Address correspondence and reprint requests to Sung Sup Park and Ki-Sun Kwon, Aging Research Center, Korea Research Institute of Bioscience and Biotechnology (KRIBB), 125 Gwahak-ro, Yuseong-gu, Daejeon 305-806, Korea. E-mails: firstname.lastname@example.org; email@example.com
The role of phosphoinositide 3-kinase (PI3K) in oxidative glutamate toxicity is not clear. Here, we investigate its role in HT22 mouse hippocampal cells and primary cortical neuronal cultures, showing that inhibitors of PI3K, LY294002, and wortmannin suppress extracellular hydrogen peroxide (H2O2) generation and increase cell survival during glutamate toxicity in HT22 cells. The mitogen-activated protein kinase kinase (MEK) inhibitor U0126 also reduced glutamate-induced H2O2 generation and inhibited phosphorylation of extracellular signal-regulated kinase (ERK) 1/2. LY294002 was seen to abolish phosphorylation of both ERK1/2 and Akt. A small interfering RNA (siRNA) study showed that PI3Kβ and PI3Kγ, rather than PI3Kα and PI3Kδ, contribute to glutamate-induced H2O2 generation and cell death. PI3Kγ knockdown also inhibited glutamate-induced ERK1/2 phosphorylation, whereas transfection with the constitutively active form of human PI3Kγ (PI3Kγ-CAAX) triggered MEK1/2 and ERK1/2 phosphorylation and H2O2 generation without glutamate exposure. This H2O2 generation was reduced by inhibition of MEK. Transfection with kinase-dead 3-phosphoinositide-dependent protein kinase 1 (PDK1-KD) reduced glutamate-induced ERK1/2 phosphorylation and H2O2 generation. Accordingly, cotransfection of cells with PDK1-KD and PI3Kγ-CAAX suppressed PI3Kγ-CAAX-triggered ERK1/2 phosphorylation and H2O2 generation. These results suggest that activation of PI3Kγ induces ERK1/2 phosphorylation, leading to extracellular H2O2 generation via PDK1 in oxidative glutamate toxicity.
Glutamate induces extracellular H2O2 generation by NADPH oxidase 4 (Nox4), leading to cell death in neurons. Our experiments, using siRNAs and chemical inhibitors, showed the PI3Kγ-PDK1-MEK axis but not Akt1, mTOR, or S6K to be involved in the glutamate-induced H2O2 generation and the subsequent toxicity in neurons.
sodium dodecyl sulfate–polyacrylamide gel electrophoresis
small interfering RNA
Deregulation of the neuronal glutamatergic system leads to glutamate-induced neuronal cell death and is involved in the pathogenesis of neurodegenerative diseases (Choi 1992; Coyle and Puttfarcken 1993; Lipton and Rosenberg 1994; Olney et al. 1998). Prolonged exposure to glutamate results in the inhibition of cystine uptake, which lowers intracellular glutathione levels, leading to oxidative toxicity (Albrecht et al. 2010). Oxidative glutamate toxicity is mediated by cystine/glutamate transporters in immature cortical neurons and ionotropic glutamate receptor-deficient neuronal cell lines such as HT22 hippocampal cells (Murphy et al. 1989; Davis and Maher 1994).
Reactive oxygen species (ROS) have been studied in oxidative glutamate toxicity. Glutathione depletion is followed by the accumulation of intracellular peroxides (Tan et al. 1998) and the activation of a cGMP-dependent Ca2+ channel (Li et al. 1997). Glutamate-induced cytotoxicity has been associated with mitochondrial hyperpolarization and increased ROS production (Liu and Schubert 2009; Kumari et al. 2012). Most ROS are produced by the mitochondrial electron-transport chain. The involvement of 12-lipoxygenase, NADPH oxidase 4, and monoamine metabolism have been reported in oxidative glutamate toxicity (Maher and Davis 1996; Li et al. 1997; Ha et al. 2010). The interaction between 12/15-lipoxygenase and mitochondria is a source of ROS in HT22 cells (Pallast et al. 2009).
The mitogen-activated protein kinase kinase (MEK)-specific inhibitor U0126 has been shown to profoundly protect HT22 cells against the oxidative stress induced by glutamate (Satoh et al. 2000). U0126-mediated protection has been associated with a reduction in elevated intracellular Ca2+ levels that leads to cell death in a neuronal cell line and primary cortical neuron cultures (Stanciu et al. 2000). However, the phosphoinositide 3-kinase (PI3K)-Akt pathway is thought to serve a pro-survival signaling role in neurons (Datta et al. 1999). Fibroblast growth factor 1 (FGF1) protection against glutamate toxicity is dependent on GSK3β inactivation via a pathway involving activation of the PI3K-Akt cascades in rat primary neuronal and HT22 cells (Hashimoto et al. 2002). Recently, the transient inhibition of PI3K has been shown to protect immature primary cortical neurons from oxidative toxicity (Levinthal and DeFranco 2004). We have reported that the activation of extracellular signal-regulated kinase (ERK) 1/2 is involved in glutamate-induced extracellular H2O2 generation (Ha et al. 2010). U0126 inhibited ERK1/2 phosphorylation and H2O2 generation in HT22 cells. Here, we investigated whether PI3K induces MEK1/2 activation and induces H2O2 generation. The contribution of the PI3K isotypes to H2O2 generation and cell survival was assessed. ERK1/2 phosphorylation by the PI3K isotype was analyzed via the expression of constitutively active PI3K isoform and siRNA directed against them. Furthermore, we identified a possible connection between PI3K activation and ERK phosphorylation.
Materials and methods
Plasmids and reagents
l-glutamate, 3-[4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), LY294002, wortmannin, AS605240, U0126, and anti-HA-tag antibody were purchased from Sigma-Aldrich (St. Louis, MO, USA). Dulbecco's modified Eagle's medium, fetal bovine serum, Amplex Red® H2O2 assay kit, Lipofectamine 2000, small interfering RNAs (siRNAs), and TRIzol antibody were purchased from Invitrogen (Carlsbad, CA, USA). Anti-phospho-Akt (p-Akt), anti-phospho-ERK1/2 (p-ERK1/2), and anti-MEK1/2 antibodies were obtained from Cell Signaling Technology (Danvers, MA, USA). Anti-Akt, anti-ERK1/2, anti-phospho-MEK1/2, and anti-β-actin antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). DNA primers were purchased from NeoProbe (Daejeon, Korea). The myc-tagged human wild type (PDK1-WT) and kinase-dead PDK1 (PDK1-KD) were kindly provided by Dr H. Ha (Chungbuk National University, Cheongju, Korea). An empty plasmid was used as a mock control. To generate a wild-type PI3Kγ (HA-PI3Kγ-WT) construct, we performed PCR using full-length human PI3Kγ cDNA as the template. First, the forward primer 5′-CCAGATTACGCTCGGTCGACCGAGCTGGAGAACTATAA-3′ and the reverse primer 5′- CTGCGCTAGCTTAGGCTGAATGTTTCTCTCC-3′ were used for amplification; then, the forward extension primer 5′-CACCATGTACCCATACGATGTTCCAGATTACGCTCGGTC-3′ was used for 5′-extension. To generate a constitutively active PI3Kγ (HA-PI3Kγ-CAAX), the forward extension primer and the reverse CAAX-tag primer 5′-ATGCTAGCTTAGGAGAGCACACACTTGCAGGCTGAATGTTTCTCTCC-3′ were used. PCR products were cloned in pLenti6-V5/D-TOPO (Invitrogen) according to the manufacturer's protocol. The clone was confirmed by nucleotide sequencing of both sides.
Cell culture and viability assay
HT22 mouse hippocampal cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and passaged with trypsinization at less than 50% confluence as described previously (Ha et al. 2010). The survival of HT22 cells was assessed using the MTT assay. Briefly, cells (5 x 103) seeded in 96-well plates were incubated at 37°C in a humidified atmosphere of 5% CO2 for 24 h and then exposed to 5 mM glutamate (unless otherwise indicated) in the presence or absence of inhibitors. MTT reduction in the cells was determined 24 h after glutamate exposure by measuring the absorption at 570 nm on a microplate reader (Model 680, Bio-Rad, Hercules, CA, USA). The data for each experiment are presented as values relative to the control treated with phosphate buffer saline.
Primary neuron culture and cytotoxicity
Mouse primary cortical neuronal cultures were prepared from albino ICR mice (Daehan, Daejeon, Korea) at E15. The cortices were dissociated as previously described (Ha et al. 2010). After dissociation, the neurons were plated on 96- or 6-well plates (Nunc, Roskilde, Denmark) that were coated with laminin (4 μg/mL, Sigma-Aldrich) and poly-d-lysine (100 μg/mL, Sigma-Aldrich), and the cells were cultured in neurobasal medium supplemented with B-27. Cytosine arabinoside (10 μM, Invitrogen) was added to the cultures 24 h after plating, thus inhibiting the proliferation of non-neuronal cells. One third of the culture medium was replaced with fresh medium without cytosine arabinoside every 3 days, and the neurons of div 14–16 were used for the experiments. To determine the cytotoxicity, we measured the cellular release of lactate dehydrogenase into the culture medium (1 × 105 cells per well in 96-well plates) using a commercial assay kit according to the manufacturer's protocol (Promega, Madison, WI, USA). The data indicate the percentage of lactate dehydrogenase release relative to the release for the control that was treated with lysis buffer.
Transient transfection and western blotting
The indicated plasmid DNA was transfected into HT22 cells with Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer's instructions. Cells were transferred in 96-well or 6-well plates and incubated for 24 h before glutamate exposure in the absence or presence of inhibitors. After the indicated time of incubation, cells were lysed, and soluble proteins (10 μg) were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and blotted onto a polyvinylidene difluoride membrane. The membranes were incubated with antibodies against Akt, p-Akt (Thr308 or Ser473), ERK, p-ERK1/2 (Thr202/Tyr204), myc-tag, and β-actin. Next, the membranes were incubated with a horseradish peroxidase-conjugated secondary antibody. After washing, the membranes were developed with an enhanced chemiluminescence system (Thermo Scientific, Rockford, IL, USA) according to the manufacturer's instructions.
Extracellular H2O2 measurement
We measured H2O2 release from cells using the Amplex Red® H2O2 assay kit (Ha et al. 2010). Briefly, cells were seeded in 96-well plates (5 × 103 cells per well) and incubated for 24 h, then cells were exposed to glutamate alone or glutamate in combination with the indicated inhibitors for 6 h in 100 μl of Krebs-Ringer phosphate buffer (KRPG; 145 mM NaCl, 5.7 mM NaH2PO4, 4.9 mM KCl, 0.5 mM CaCl2, 1.2 mM MgSO4, and 5.5 mM glucose, pH 7.3). The culture medium (50 μL) was transferred to an equal volume of KRPG containing 50 μM Amplex Red reagent and 0.2 U/mL horseradish peroxidase. After 10 min incubation at 25°C, the fluorescence was measured using the Fusion α microplate fluorometer (PerkinElmer, Santa Clara, CA, USA). The excitation and emission wavelengths were 545 nm and 590 nm, respectively. The background (KRPG alone) fluorescence was subtracted from each value. The H2O2 generation data are presented as values relative to glutamate alone (unless otherwise indicated).
Knockdown of PI3K expression with siRNAs
HT22 cells were transfected with siRNAs using 150 pM siRNA and Lipofectamine 2000 according to the manufacturer's (Invitrogen) protocol. Isotype-specific siRNAs corresponding to murine PI3Kα-1 901-921 (GCCCAACUUGAUGCUGAUGGC), PI3Kα-2 (946–966, GCCGAUUGAUAGCUUCACCAU), PI3Kβ-1 (1091–1110, UGGAGUAUGUGUUUGGCGAU), PI3Kβ-2 (1324–1344, CCUUUCCAAAUUACCUUGGUU), PI3Kγ-1 (1492–1511, CAACAAGUCCUCUGCCAAAG), PI3Kγ-2 (1561-1580, UAUCAAAAUCAAAGACUUGC), PI3Kδ-1 (1531-1551, UCG CUU GGG CCA ACCU CAU GCU), and PI3Kδ-2 (1544–1564, CCUCAUGCUAUUCGACUACAA) were used. Six hours after transfection, the medium was replaced with fresh medium and incubated at 37°C. The cells were transferred to 96-well or 6-well plates 24 after transfection and cultured for another 24 h before exposure to glutamate. The H2O2 levels and cell survival were measured 6 and 24 h after glutamate exposure, respectively. Stealth RNAi negative control duplexes (scrambled siRNA, Invitrogen) were used as controls (SC).
RNA isolation and semiquantitative RT-PCR
Total RNA was isolated from HT22 cells 48 h after transfection using TRIzol reagent (Invitrogen) according to the manufacturer's instructions. A total of 5 μg RNA was applied for first-strand cDNA synthesis using the first-strand cDNA synthesis kit (Solgent, Daejeon, Korea). The following primers were used: PI3Kα, 5′-TGACACAGGATTTTCTTGATTGTG-3′ and TCAAAAGATTGTAGTTCTGGCATT-3′; PI3Kβ, 5′-GCCAGACTCGCTGAGAATCTAT-3′ and 5′-TCTTTTTATATAGAACTCCCACT-3′; PI3Kγ, 5′-ACCCTGGTGATCGAGAAATG-3′ and 5′-AGCATCCGCAGGTCCAAGCC-3′; and PI3Kδ, 5′-CAGCCAGAATTGAAGTCCTCTT-3′ and 5′-CGTCCTTGCTCACACTTTTCGAG-3′. β-Actin RNA was amplified (5′-GTGGGCCGCCCTAGGCACCA-3′ and 5′-CTCTTTGATGTCACGCACGATTTC-3′) as a control. PCR was performed with Platinum Taq DNA polymerase (Invitrogen) using the following parameters: an initial 5 min denaturation step at 95°C followed by 35 cycles of 94°C, 62°C, and 72°C of 30 sec at each step and a final 10 min extension at 72°C. The PCR-amplified fragments were separated on 2% agarose gels (Invitrogen) and identified by sequencing. β-Actin was amplified as a control.
Quantitative data are expressed as the means ± SD. Statistical comparisons between experimental groups were made using the two-tailed, unpaired Student's t-test. Probability values of p <0.05 were considered significant.
PI3K inhibitors protect cells from oxidative glutamate toxicity
Transient PI3K inhibition is known to protect immature primary cortical neurons (DIV 3–4) from oxidative glutamate toxicity (Levinthal and DeFranco 2004). First, to determine whether PI3K can be activated by prolonged exposure to glutamate in HT22, we monitored Akt phosphorylation after cells were treated with 5 mM glutamate in the presence or absence of the inhibitors LY294002 or wortmannin. Glutamate apparently induced Akt phosphorylation, and this phosphorylation was significantly suppressed by LY294002 or wortmannin in a concentration-dependent manner (Fig. 1a). The total Akt expression was not changed by glutamate. Next, to evaluate the contribution of PI3K activation to extracellular H2O2 generation, we measured the released H2O2 using the Amplex Red H2O2 assay kit after cells were treated with glutamate in the presence of inhibitors. LY294002 and wortmannin suppressed H2O2 generation in dose-dependent manners (Fig. 1b). H2O2 generation in the presence of inhibitors (10 μM LY294002 or 1 μM wortmannin) was reduced to 44% of the levels observed in controls treated with dimethylsulfoxide. Cell survival was in inverse proportion to increases in H2O2 generation. Cell survival increased from 16% to 65–66% in the presence of 10 μM LY294002 or 1 μM wortmannin (Fig. 1c). These results, along with the finding that the removal of H2O2 from the extracellular space rescues HT22 cells from glutamate toxicity (Ha et al. 2010), suggest that PI3K inhibitors protect HT22 cells from oxidative glutamate toxicity through suppression of extracellular H2O2 generation.
The MEK/ERK inhibitor U0126 has been shown to suppress glutamate-induced H2O2 generation in HT22 cells (Ha et al. 2010). LY294002 reduced glutamate-induced H2O2 generation to the same extent as diphenyleneiodonium (DPI), an NADPH oxidase inhibitor, and U0126 (Fig. 2a). To determine whether glutamate-induced H2O2 generation plays a role in MEK1/2 activation, we examined glutamate-induced MEK1/2 activation in the presence of DPI. DPI did not reduce glutamate-induced ERK1/2 phosphorylation (Fig. 2b). This result shows that ERK1/2 phosphorylation was independent of extracellular H2O2. To investigate whether PI3K activation contributes to MEK1/2 activation, we analyzed Akt and ERK1/2 phosphorylation after cells were treated with glutamate in the presence of LY294002 or U0126. Glutamate (5 mM) exposure induced the phosphorylation of Akt and ERK1/2 in the absence of inhibitors (Fig. 2c), and the ERK phosphorylation was slightly faster than the Akt phosphorylation. However, the presence of LY294002 reduced the phosphorylation of ERK1/2 and Akt, whereas U0126 blocked ERK1/2 phosphorylation alone. These results suggest that PI3K activation contributes to the activation of MEK1/2 in glutamate-induced H2O2 generation.
PI3Kγ contributes to glutamate-induced H2O2 generation
Class I PI3K catalytic subunits exist in 4 isoforms: PI3Kα, PI3Kβ, PI3Kγ, and PI3Kδ. To investigate which isoform is responsible for glutamate-induced H2O2 generation and cell survival, we measured H2O2 generation using isoform-specific knockdown cells. Semiquantitative RT-PCR showed that transfection with isoform-specific siRNA significantly suppressed expression of each PI3K isoform compared with control scrambled (SC) siRNA (Fig. 3a). Knockdown of the expression of PI3Kβ and PI3Kγ apparently reduced glutamate-induced H2O2 generation, whereas PI3Kα and PI3Kδ knockdowns decreased gene expression but negligibly altered H2O2 generation (Fig. 3b). Glutamate-induced H2O2 generation was 70% and 40% in PI3Kβ and PI3Kγ knockdown cells compared with glutamate-treated control cells (SC), respectively. Cell survival was inversely proportional to the increase in H2O2 generation (Fig. 3c). Survival in PI3Kβ and PI3Kγ knockdown cells increased to 46% and 81% compared to glutamate-treated controls (SC, 15%), respectively, whereas survival of PI3Kα and PI3Kδ knockdown cells increased insignificantly. Similar results for H2O2 generation and cytotoxicity were observed in a primary neuron culture (Fig. 3d and e). The PI3Kγ knockdown in the primary neuron culture reduced the glutamate-induced cytotoxicity by 39.2% and the H2O2 generation by 55.9%. The finding that the PI3Kγ-specific inhibitor AS605240 reduced H2O2 generation in a dose-dependent manner (Fig. 3f) confirmed the contribution of PI3Kγ to H2O2 generation. These results suggest that PI3Kγ is heavily involved in oxidative glutamate toxicity.
PI3Kγ-CAAX expression induced H2O2 generation and ERK phosphorylation
To further assess the contribution of PI3Kγ to H2O2 generation, we measured H2O2 without glutamate exposure after cells were transfected with various amounts of PI3Kγ-CAAX plasmid, which codes for a constitutively active form of PI3Kγ. The CAAX motif permanently attaches the enzyme to the membrane and places it close to its lipid substrates (Hehl et al. 2001). First, Akt was phosphorylated because of the expression of PI3Kγ-CAAX in a dose-dependent manner without glutamate treatment (Fig. 4a). The H2O2 generation was also increased in a dose-dependent manner (Fig. 4b). PI3Kγ-CAAX increased H2O2 generation up to 28% in the absence of glutamate exposure compared to glutamate exposure alone (5 mM), whereas wild-type PI3Kγ (PI3Kγ-WT) did not induce significant H2O2 generation. Glutamate exposure showed that PI3Kγ-CAAX increased glutamate-induced H2O2 generation by 21-45% compared with that for a vector control (Fig. 4c). H2O2 generation triggered by PI3Kγ-CAAX without glutamate was suppressed by 64% in the presence of U0126 (Fig. 4d). To investigate whether PI3Kγ-CAAX without glutamate induces MEK activation, ERK1/2 and MEK1/2 phosphorylation was assessed 24 h after the cells were transfected with PI3Kγ-CAAX, PI3Kγ-WT, or the vector control. In contrast to the vector control and the PI3Kγ-WT expression, PI3Kγ-CAAX expression apparently induced MEK1/2 and ERK1/2 phosphorylation without glutamate exposure (Fig. 4e). Furthermore, the finding that the PI3Kγ knockdown suppressed glutamate (2 mM)-induced ERK1/2 phosphorylation (Fig. 4f) supports the idea that PI3Kγ induces MEK activation. Similar results for the suppression of glutamate-induced ERK1/2 phosphorylation for a PI3Kγ knockdown were observed in primary neuron cultures (Fig. 4g). These results suggest that PI3Kγ contributes to MEK activation and H2O2 generation in glutamate toxicity. ERK1/2 phosphorylation was not significantly reduced by PI3Kγ knockdown after 5 mM glutamate exposure (data not shown).
PDK1 but not Akt1 induces glutamate-induced ERK phosphorylation
PDK1 is known to bind and directly phosphorylate MEK (Sato et al. 2004). To identify whether PDK1 is involved in glutamate-induced MEK activation, the phosphorylation of ERK1/2 was determined after cells were transfected with PDK1-KD and then treated with glutamate. The phosphorylation of ERK1/2 and Akt was apparently suppressed in 2 mM glutamate, although the phosphorylation of both was slightly but significantly reduced in 5 mM glutamate (Fig. 5a). Wild-type PDK1 (PDK1-WT) did not affect Akt or ERK phosphorylation regardless of glutamate concentration (2 mM and 5 mM). Glutamate-induced H2O2 generation was reduced by PDK1-KD expression (Fig. 5b). The H2O2 generation for the PDK1-KD expression was 44% and 31% lower for the 2 mM and 5 mM glutamate treatments, respectively, than the H2O2 generation for the control vector expression. Over-expression of PDK1-WT did not induce H2O2 generation without glutamate treatment. Moreover, PDK1-KD expression suppressed the PI3Kγ-CAAX-induced ERK1/2 activation (Fig. 5c). ERK1/2 phosphorylation was induced by transfection with PI3Kγ-CAAX, and this phosphorylation was suppressed by cotransfection of cells with PDK1-KD. Furthermore, co-expression of PI3Kγ-CAAX and PDK1-KD reduced PI3Kγ-CAAX-induced H2O2 generation by 60% (Fig. 5d). These results suggest that PDK1 expression is involved in PI3Kγ-induced MEK activation and H2O2 generation. Inhibitors of Akt and its downstream effectors, such as GSK3 and mammalian target of rapamycin (mTOR), consistently did not affect glutamate-induced H2O2 generation (Figure S1a). Furthermore, the constitutively active form of Akt1 did not induce ERK1/2 phosphorylation (Figure S1b), H2O2 generation (Figure S1c), or cell death (Figure S1d) without glutamate treatment. These results imply that Akt1 is not involved in glutamate-induced H2O2 generation through ERK1/2 phosphorylation.
In this study, we investigated the contribution of PI3K to oxidative glutamate toxicity and the connection between PI3K and MEK. Persistent activation of ERK contributes to glutamate-induced oxidative toxicity in immature primary cortical neuron cultures (Stanciu et al. 2000). A similar study has shown that delayed activation of ERK accompanies glutamate-induced oxidative stress (Levinthal and DeFranco 2004). The authors of this study suggested that the intracellular ROS generated by glutamate possibly activate ERK to induce oxidative toxicity. ERK activation is downstream of 12-lipoxygenase (Stanciu et al. 2000). However, we have shown that ERK1/2 activation by glutamate is responsible for the extracellular H2O2 generation that leads to cell death in HT22 cells (Ha et al. 2010). For the acute and chronic kinetics of ERK activation corresponding to different functions within the cell (Colucci-D'Amato et al. 2003), chronic ERK phosphorylation may contribute to the neuronal death induced by prolonged exposure to glutamate. The subcellular localization of the NADPH oxidase (NOX) protein (Serrander et al. 2007) and the compartmentalization of ERK could play a role in H2O2 generation, leading to cell death.
The role of PI3Kγ in neurological function has been reported. PI3Kγ deficiency reduces brain infarction and neurological deficits (Jin et al. 2011). The involvement of PI3Kγ in the generation of ROS has been observed in the brain (Jin et al. 2011) and in neutrophils (Suire et al. 2006). We showed that the PI3K inhibitors LY294002 and wortmannin protect cells from glutamate toxicity by reducing extracellular H2O2 generation (Fig. 1b and c). Knockdown of PI3Kβ and PI3Kγ reduced H2O2 generation and the cellular toxicity induced by glutamate in HT22 cells (Fig. 3b and c) and primary neuron cultures (Fig. 3d and e). Inhibition of H2O2 generation by AS605240, a PI3Kγ-specific inhibitor (Fig. 3f), indicates that PI3Kγ activation is crucial in the glutamate-induced H2O2 generation that leads to cell death. We next assessed the relationship between PI3K and MEK. U0126, a MEK inhibitor, suppressed ERK1/2 phosphorylation alone, whereas LY294002, a PI3K inhibitor, reduced Akt and ERK1/2 phosphorylation (Fig. 2c). Expression of PI3Kγ-CAAX induced ERK1/2 phosphorylation in the absence of glutamate (Fig. 4e), and PI3Kγ knockdown reduced ERK1/2 phosphorylation in the presence of glutamate (Fig. 4f and g). These results indicate that PI3Kγ activation is required for MEK1/2 activation.
Many lines of evidence support the existence of cross-talk between PI3K and ERK and the potential for PI3K to act as an upstream activator of ERK (York et al. 2000; Lin et al. 2001). Our observations are consistent with the idea that the transient early inhibition of PI3K protects cortical neurons from oxidative glutamate toxicity by blocking the development of delayed ERK activation (Levinthal and DeFranco 2004). Interestingly, H2O2 generation was found to be downstream of MEK activation in our study. Glutamate-induced (Fig. 2a) and PI3Kγ-CAAX-triggered (Fig. 4d) H2O2 generation were suppressed by U0126. Furthermore, DPI inhibited glutamate-induced H2O2 generation (Fig. 2a) but did not inhibit glutamate-induced ERK1/2 phosphorylation (Fig. 2b).
Several previous studies have shown that inhibition of PI3K prevents the ERK activity induced by stimuli such as platelet-derived growth factor, epidermal growth factor, and lovastatin (Choudhury et al. 1997; Mahimainathan et al. 2005; Ghosh-Choudhury et al. 2007). PDK1 has been shown to phosphorylate MEK (Sato et al. 2004) regardless of pleckstrin homology domain. The finding that expression of PDK1-KD suppressed glutamate- and PI3Kγ-CAAX-induced ERK1/2 phosphorylation (Fig. 5a and c) demonstrates that MEK1/2 activation is PDK1-dependent. Interestingly, PDK1-WT expression neither induced H2O2 generation without glutamate treatment (Fig. 5b) nor activated ERK without PI3Kγ-CAAX (Fig. 5c). This study represents the first report showing that the PI3K-PDK1-MEK (ERK) signaling pathway produces H2O2 during oxidative glutamate toxicity. Akt1 and its effector molecules, GSK3 and mTOR, are not likely to be involved in glutamate-induced H2O2 generation (Figure S1), although glutamate stimulation results in PI3K-mediated Akt phosphorylation.
In summary, glutamate-triggered H2O2 generation and ERK1/2 phosphorylation require PI3K activation. Of the four PI3K isoforms, PI3Kγ has the strongest role in the generation of oxidative stress. PDK1 mediates PI3Kγ-induced MEK activation (ERK1/2 phosphorylation). A PI3Kγ-PDK1-MEK pathway largely contributes to oxidative glutamate toxicity. Future studies will elucidate the activation link between PI3K-ERK1/2 and NADPH oxidase resulting in H2O2 generation.
We thank Dr H. Ha of Chungbuk National University for providing the myc-tagged human wild type and kinase-dead PDK1. This research was supported by a grant from KRIBB Research Initiative Program and Biomedical Technology Development Programs 20120009082 and 20120009022. Authorship credit: acquisition and interpretation of data, JSH; design of project and drafting of manuscript, K-SK; conception of project and revising of manuscript, SSP.
Conflict of interest
The authors declare that there are no conflicts of interest.