Cholesteryl Glucoside Stimulates Activation of Protein Kinase B/Akt in the Motor Neuron-Derived NSC34 Cell Line

Abstract

Steryl glycosides and related compounds are commonly found in the environment and have been associated with neurodegenerative changes in vulnerable individuals. However, their mechanisms of action in mammalian cells have not been well investigated. In the present study the effects of cholesterol glucoside (CG), a variant form of steryl glycoside, was investigated in the motor neuron-derived NSC34 cell line. Prolonged treatment with CG was found to induce cell death in a dose- and time-dependent manner. However, transient exposure of CG preconditioned NSC34 cells for stress from serum deprivation. To study the signaling pathways activated by CG, we employed the Kinetworks™ KPSS 1.3 Phospho-site Screen to track the phosphorylation level of at least 35 diverse signaling proteins. The survival protein kinase B (PKB/Akt) displayed a 2-fold increase in phosphorylation at its Ser-473 activation site following CG stimulation. Akt signaling was important for conferring cytoprotection against serum deprivation-induced stress. Inhibition of phosphatidylinositol 3-kinase (PI3K), which indirectly triggers Akt stimulation, completely abolished CG preconditioning against serum deprivation. Our findings revealed that there may be a PI3K-independent pathway which also mediated Akt Ser-473 phosphorylation. Improved understanding of the mechanisms of action of CG should provide insights to the how other members of the steryl glycoside family induce toxicity in the mouse model of ALS-PDC, and how cells respond to these toxins.

Introduction

Steryl glycosides (SGs) are membrane lipids that are commonly found in the environment. This class of molecule can be made by most plants, some fungi, some types of bacteria, and can be expressed under certain circumstances in various mammalian cell cultures [1,2]. Members of the SG family are generically characterized by a carbohydrate unit linked to the hydroxyl group of the tetracyclic backbone. The nature of the sugar moiety can vary greatly and can be glucose, xylose, galactose, fructose, amongst other monosaccharides. Typical well-known sterols sharing this basic structure include cholesterol, campesterol, stigmasterol, brassicasterol, and sitosterol (reviewed in [2]).

Despite the relative abundance of the class of molecule in the environment, their biological roles remain unclear. SGs are one of the major glycolipid constituents of the plant cell [3,4]. Peng et al. [5] demonstrated that sterol-β-glucosides act as primers for de novo cellulose biosynthesis. Kunimoto and colleagues [2,6] suggested that SGs play an essential role in stress signal transduction cascades. Endogenous synthesis of SGs have been documented in human fibroblasts under heat stress [2], and the slime mold Physarum polycephalum was also found to rapidly induce SGs during episodes of nutrient starvation [6]. Bacterial synthesis of SGs appear to serve an entirely different purpose, namely to maintain pathogenicity. For example, HP0421 is an enzyme responsible for SG synthesis in Helicobacter pylori and is essential for them to escape host immune cells [7]. Certain bacterial pathogens, including H. pylori, have been associated with various age-related neurological conditions. These pathogens commonly have SGs as a major constituent of their lipid profiles (reviewed in [2]). SGs are also abundant in the seeds of the Cycas micronesica, the indigenous cycad tree of Guam [8,9]. Epidemiological and experimental studies have shown a strong dietary link between cycad seeds and the prevalence of the Guamanian type of amyotrophic lateral sclerosis (ALS) and parkinsonism dementia complex (PDC) [10-13]. Our laboratory previously reported that dietary exposures to washed cycad seeds in mice can induce the spectrum of neurological deficits that resemble ALS-PDC features, which include neuronal losses in the various parts of the central nervous system along with behavioral deficits [12, 14-16]. Furthermore, we isolated three sterol β–D–glucosides from washed cycad seeds that had toxic in vitro properties [17]. The isolated β-sterol-D-glucosides were identified as campesterol glucoside, stigmasterol glucoside, and β-sitosterol-D-3-glucoside (BSSG). Several experiments in our laboratory demonstrated that these phytosteryl glycosides were capable of inducing depolarizing field potentials and inducing cell death in rat cortical slice cultures [17]. These findings were further supported by in vivo studies from mice exposed to BSSG in their daily diet, which showed significant motor neuron loss and microgliosis which were accompanied by motor behavioral deficits [18]. These in vitro and in vivo data using SGs support the possibility that these molecules in cycad seeds may be involved in the etiopathogenesis of the Guamanian ALS. For this reason, we were interested in studying the effects on SGs on motor neurons. Particularly, we were interested in understanding the signal transduction pathways activated by these toxins in motor neurons. To pursue these studies, we have applied cholesteryl glucoside (CG), a variant form of the cycad-derived SGs, to a motor neuron-derived cell line, NSC-34 cells [19]. In the following, we demonstrated that prolonged exposure of CG is toxic and can lead to a substantial reduction in NSC34 cell viability. However, transient CG exposure induced activation of the survival protein kinase B, (PKB/Akt), which promoted cytoprotection against serum deprivation stress. These data support the notion that as means of self-rescuing, cells would activate survival pathways in the presence of CG-related toxins.

Materials and Methods

Reagents

Buffer ingredients, cell culture media, and water-soluble cholesterol were purchased from Sigma-Aldrich Canada Ltd (Mississauga, Ontario, Canada). Cholesterol-β-D-glucoside (FW=548.8 g/mol) was synthesized under contract by Dr. Stephen Withers (Department of Chemistry, University of British Columbia). MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) was purchased Promega Corporation (Madison, WI, USA). SP600125 (anthra[1-9-cd]pyrazol-6(2H)-one) and SB230580 (4-(4-fluorophenyl)-2-(4-methylsulfinylphenyl)-5-(4-pyridyl)1H-imidazole) were purchased from EMD Biosciences (San Diego, CA, USA). U0126 (1,4-diamino-2,3-dicyano-1,4-bis[2-aminophenylthio] butadiene), LY294002 ((2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one) were purchased from Cell Signaling Technology Inc. (Danvers, MA, USA). Antibodies against phospho-Akt Ser-473, pan specific Akt, phospho-ERK1/2, pan specific ERK1/2, phospho-JNK1/2, and phospho-p38 MAPK were purchased from Cell Signaling Technology Inc. (Danvers, MA, USA).

Neuroblastoma × Spinal Cord 34 (NSC34) Cell Line Culture

The mouse-derived neuroblastoma × spinal cord cell line, NSC34 cells were cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 2 mM L-glutamine and 10% (v/v) fetal bovine serum. Cells were maintained at 37°C in a humidified 5% CO₂ atmosphere and sub-cultured at 90% confluency. For every experiment, NSC34 cells were seeded at a density of 1500 cells/mm² onto collagen coated plates and were allowed to grow for another 5 days before treatment. Cell culture medium was replaced every 3-4 days. Collagen coating was achieved by pre-incubating culture plates with 0.1% collagen solution (500 μl/mm²) for 3 h at room temperature followed by washing once with sterile distilled water.

CG Stock Solution and Kinase Inhibitor Preparation

Cholesterol glucoside (CG) stock solutions were prepared in 100% dimethyl sulfoxide (DMSO) at concentrations of 30 mM and 50 mM. The final concentrations of CG were diluted in DMEM to give the highest experimental concentration under 0.5% DMSO. Increasing doses of DMSO up to 0.5% were previously demonstrated to have no toxic effects on cell viability [2]. Appropriate DMSO vehicle solutions were made for the kinase assay studies.

Cell Viability Assays

Cell viability was determined using the MTT assay. This method is based on the ability of viable cells to convert a tetrozolium salt (MTT) to a colored formazan product. Treated cells were replenished with fresh media containing MTT (1 μg/ml) and allowing reduction of MTT to occur at 37°C for 4 h. The media were removed and replaced by 100% DMSO to solublize the formazan product. Optical density at 540 nm was measured using an ELx808™ absorbance microplate reader (Biotek Instrument). Viability values for treated cells were expressed a percentage of control cells, the latter being defined as 100%.

SDS-Page and Immunoblot Analysis

NSC34 cells were washed once with ice cold phosphate buffered saline and harvested in lysis buffer containing 0.5% Triton X-100, 2 mM EGTA, 5 mM EDTA, 20 mM MOPS, 200 mM sodium orthovanadate, 25 mM β-glycerophosphate, 20 mM sodium pyrophosphate, 30 mM sodium fluoride, 1 mM phenylmethanesulphonylfluoride, and 1 complete mini protease inhibitor cocktail tablet (Roche Diagnostics). Lysed cells were subjected to sonication for 15 seconds on ice followed by centrifugation at 12,000 × g for 30 minutes to remove impurities such as nucleic acids and lipids. A bicinchoninic acid protein determination kit was used to determine the concentration of the protein lysate by following the manufacturer's instructions (Pierce Biotechnology Inc). Twenty μg of proteins were subjected to electrophoresis followed by electrical transfer onto a nitrocellulose membrane.

The nitrocellulose membrane was blocked in 1% BSA (w/v) in Tris-buffered saline containing 0.1% Tween-20 (v/v) (TBST) for 1 h at room temperature. The membrane was then probed with various commercially available primary antibodies diluted in 1% BSA TBST for 18 h at 4°C with gentle shaking. Incubation with secondary peroxidase-conjugated anti-mouse or anti-rabbit antibodies was performed at room temperature for 1 h. Immunoblots were developed using the ECL system (Pierce Biotechnology Inc.) and quantified with the NIH Image J 1.37 software. Kinase activity was indirectly determined using immunoblot analysis with phospho-site specific antibodies that recognized activation sites.

Kinetworks™ KPSS 1.3 Phosphosite Analysis

NSC34 cells were pretreated with 100 μM of CG for 1 h, washed with serum free media and cultured in serum free media for 4 h. The cells were harvested as described above and diluted to a final protein concentration of 1 μg/μl. The samples were then boiled and sent to Kinexus Bioinformatics Corp., Vancouver, BC. for the Kinetworks™ KPSS 1.3 Phospho-Site multi-immunoblotting analyses. The KPSS 1.3 screen tracks the phosphorylation levels of more than 35 protein kinases and their substrates. For a full listing of phospho-proteins tracked in the KPSS 1.3 screen, please refer to the company's website at www.kinexus.ca.

Statistics and Data Analyses

Statistical significance was determined by one way ANOVA followed by Tukey's honest post hoc test or an unpaired Student's t-test if the data were normally distributed and the variances were homogenous. Statistical significance was defined as p<0.05. All data are presented as mean±S.E.M. from experiments performed at least in triplicate.

Results

CG Precondtioning for Serum Deprivation Stress

The cytotoxicity of CG was first established for the NSC34 cells. CG treatment reduced NSC34 cell viability in a dose- and time-dependent manner (Two way ANOVA, Concentration: F_4,217=52.3 p<0.001; Duration: F_3,217=33.3 p<0.001). Significant cell death was observed only after 24 h of CG treatment in the 50-250 μM range (Fig.1A). Treatment with cholesterol only (without the glucose moiety) did not have a significant effect on NSC34 cell viability over increasing dose and time (Fig.1B). To investigate whether transient CG treatment could confer neuroprotection against serum deprivation in NSC34 cells, we developed the preconditioning model as indicated in Fig 2A. Serum deprivation alone for 24 h induced approximately 25% cell death, but this effect was prevented by pretreating cells with 25-100 μM CG for 1 hour (One way ANOVA, F_5,30=4.78 p<0.05) (Fig. 2B). Prolonged CG treatment did not protect NSC34 cells from cell death induced by serum deprivation (unpaired Student's t-test p<0.05) (Fig. 2C)

Figure 1. Time and concentration-dependent effect of CG on NSC34 cell viability.

(A) NSC34 cells were treated with increasing concentrations of CG for up to 7 days and cell viability was monitored using the MTT reduction assay. A two-way ANOVA (duration: F_3,217=33.3 p<0.001 and concentration: F_4,217=52.3 p<0.001) followed by Tukey's honest post hoc (*p<0.05) indicated significant decrease in cell viability with increasing concentrations and duration of CG treatment. (B) Cholesterol without the glucose moiety did not show any significant effect on cell viability. Cell viability is expressed as a percentage of control cells without any treatment (defined as 100%) and represented as mean±S.E.M (n=12).

Figure 2. Transient cholesteryl glucoside (CG) exposure preconditions NSC34 cells for serum deprivation-induced cell death.

(A) CG preconditioning model where NSC34 cells are treated with CG under normal condition in complete cell culture medium for 1 h. CG was removed from the cell culture medium and the NSC34 cells were deprived of serum for 24 h, followed by monitoring cell viability. (B) CG exposure for 1 h to NSC34 cells prior to serum deprivation protected against serum deprivation induced cell death. One way ANOVA followed by Tukey's honest post hoc showed significant increase in cell survival in NSC34 cells pretreated with 25-100 μM of CG prior to serum deprivation (*p<0.05). (C) Simultaneous CG treatment with serum deprivation for the indicated times did not protect against serum deprivation-induced cell death. Instead, a greater reduction in cell number was observed when CG was exposed to serum-deprived cells (unpaired Student's t test * p<0.05). Cell number is represented as a percentage of control cells (untreated and not serum deprived). Data are represented as the mean±S.E.M.

Signaling Pathways Underlying CG Preconditioning in NSC34 Cells

The previous experiment clearly demonstrated that CG preconditioned NSC34 cells for serum deprivation. To investigate potential signaling pathways involved in CG preconditioning, the Kinetworks™ KPSS-1.3 phospho-site screen was employed to track the phosphorylation status of at least 35 different proteins at specific phosphorylation sites. Examples of the multi-immunoblot images generated from NSC34 cells pretreated or not with CG are shown in Fig. 3. All the known phosphorylation sites detected by the KPSS 1.3 screen are marked with a red arrow (Fig. 3A). Only those phospho-sites that demonstrated a percentage change from control (%CFC) in signal intensity greater than 25% after CG pretreatment were indicated by dotted circles (Fig. 3B) and their percentage change were shown in Fig. 3C. There were various unidentified immunoreactive protein bands that were also detected by the phospho-site screen. These unidentified protein bands were arbitrary labeled A to E, and the %CFC for each unclassified protein were indicated below the arbitrary names on the multi-immunoblot in Fig. 3B.

Figure 3. Kinetworks™ KPSS-1.3 phospho-site multi-immunoblot analysis of CG treated and serum deprived NSC34 cell lysates.

NSC34 cells were pretreated with 100 μM of CG for 1 h and serum deprived for 4 h. Control cells were serum deprived for 4 h only. Cell lysates were subjected to Kinetworks™ KPSS 1.3 phospho-site multi-immunoblot analysis to screen for broad phosphorylation changes during CG preconditioning. (A) A multi-immunoblot of serum deprived NSC34 cells with the tracked phospho-proteins at the indicated phosphorylation sites. The arrows marked the locations of the tracked phospho-proteins bands. (B) A multi-immunoblot of NSC34 cells pretreated with 100 μM and serum deprived for 4 h. Circled bands indicate those target phospho-sites where the observed percent change from control (%CFC) was 25% or greater. Protein bands that are labeled A to E show significant changes in band intensity but are currently not identified. The %CFC for these proteins is presented under the corresponding letter.

NSC34 cells were pretreated with 100 μM of CG for 1 h and serum deprived for 4 h. Control cells were serum deprived for 4 h only. Cell lysates were subjected to Kinetworks™ KPSS 1.3 phospho-site multi-immunoblot analysis to screen for broad phosphorylation changes during CG preconditioning. (C) The intensities (counts per minute) of the immunoblot enhanced chemiluminence signals for target phospho-proteins were quantified for serum deprived NSC34 cells pretreated with 100 μM CG (black bars) or without pretreatment (white bars). The data shown in the figure include only those target phospho-sites where the observed percent change from control (%CFC) was 25% or higher.

Previous studies indicated that the PI3K/Akt pathway is important in modulating preconditioning against various stress stimuli [21-24]. For this reason, it was of particular interest to investigate the role of the PI3K/Akt pathway in CG preconditioning for serum deprivation in NSC34 cells. These cells were preconditioned with CG as previously described and serum deprived for up to 8 h. The activity state of Akt was determined using a phospho-site specific antibody for Akt Ser-473 activation site. Phosphorylation at Ser-473 was significantly upregulated by CG stimulation (group: F_1,16=4.63 p<0.05; time F_3,16=15.7 p<0.001). For further analysis, an unpaired Student's t-test indicated that Akt Ser-473 phosphorylation was significantly increased 0 and 2 h after serum deprivation following CG pretreatment (Fig. 4A). Inhibition of the PI3K/Akt pathway with the PI3K inhibitor LY294002 (30 μM) during CG preconditioning prevented CG's cytoprotective effect (unpaired Student's t-test p<0.05).

Figure 4. PI3K/Akt pathway involved in CG preconditioning.

(A) NSC34 cells were either pretreated with 100 μM of CG for 1 h (black bars) or replenished with complete media without CG (white bars), followed by serum deprivation for up to 8 h. Akt activation was monitored by Western blot analysis using a phospho-site specific antibody. Two way ANOVA indicated significant changes in Akt (Time F_3,16=15.7 p<0.001; Group F_1,16=4.63 p<0.05). For follow-up analysis of the significant changes indicated above, an unpaired Student's t-test was used to indicate changes between CG pretreated and untreated groups (* p<0.05). (B) NSC34 cells were treated 30 μM of LY294002 (PI3K inhibitor) during CG preconditioning (100 μM for 1 h), and followed by serum deprivation for 24 h. Transient exposure of CG preconditioned NSC34 cells against cell death induced by serum deprivation. Inhibition with LY294002 to suppress the PI3K/Akt pathway prevented the cytoprotective effect of CG preconditioning against serum deprivation (unpaired Student's t-test *p<0.05). Data are represented as mean±S.E.M.

CG Stimulates AKT/PKB Activation

In the previous experiment, CG preconditioning resulted in elevated Akt phosphorylation and paralleled cytoprotection against serum deprivation-induced cell death, making it likely that one of the effects of CG was to stimulate Akt activation. To test this hypothesis, NSC34 cells were stimulated with CG and Akt activation was assessed as previously described. Stimulation of NSC34 cells with 250 μM CG produced a marked increase in the phosphorylation level of Akt Ser-473 in time- and concentration-dependent manners (Fig. 5A and 5B). Treatment with 0.3% DMSO used as a control did not have any noticeable effect on Akt Ser-473 phosphorylation. To further characterize the effect of CG stimulation on Akt Ser-473 phosphorylation, NSC34 cells were serum-deprived for 12 h and stimulated with DMEM containing 10% fetal bovine serum. When these serum-deprived NSC34 cells were later stimulated with serum, Akt was phosphorylated in a time-dependent manner (Fig. 5C). However, in the presence of 250 μM of CG, serum-stimulation of Akt phosphorylation was ablated (unpaired Student's t-test p<0.05) (Fig. 5C, black bars).

Figure 5. Cholesteryl glucoside (CG) induce Akt phosphorylation.

(A) CG treatment induced rapid activation of Akt (One way ANOVA F_5,29=5.18 p<0.01). (B) CG-induced Akt activation is also concentration-dependent. 250 μM of CG treatment for 5 min can trigger more than 2-fold Akt activation (One way ANOVA F_5,30=2.63 p<0.05). Treatment with 0.3% DMSO demonstrated that the DMSO in the highest CG concentration is not involved in activating Akt. Tukey's honest post hoc showed significant changes compared to control cells without treatment (* p<0.05). Significant changes of percentage from controls (%CFC) are indicated on top of each bar. (C) NSC34 cells were serum deprived for 12 h and then stimulated with 10% serum in the presence or absence of 250 μM CG for the indicated times. CG treatment suppresses serum stimulation of Akt phosphorylation (unpaired Student's t-test *p<0.05). (D) NSC34 cells maintained in complete media were treated with either 30 μM of LY294002 (PI3K inhibitor) or 200 nM of rapamycin (mTOR inhibitor) for 1 h. The inhibitors were then removed and the cells were incubated with complete media in the presence of absence of 250 μM of CG for 10 min. As determined before, CG treatment can induce phosphorylation of Akt^Ser-473 phosphorylation under basal condition or following inhibition of the PI3K pathway with LY294002 (unpaired Student's t-test *p<0.05). mTOR is a downstream target of Akt that was previously found to phosphorylate Akt at Ser-473. However, inhibition of mTOR with rapamycin did not have any effect on Akt Ser-473 phosphorylation. The data are represented as the mean±S.E.M from at least three independent experiments.

Akt activation is primarily mediated through the PI3K/PDK1 pathway. However, the mammalian target of rapamycin (mTOR) has been speculated to catalyze the phosphorylation of Akt at Ser-473, which is required for full activation [25-27]. For this reason, we investigated whether CG induced Akt phosphorylation via PI3K/PDK1 or mTOR. NSC34 cells were pretreated with the PI3K inhibitor LY294002 (30 μM) and the mTOR inhibitor rapamycin (200 nM) prior to CG stimulation. Inhibition of the PI3K/Akt pathway followed by stimulation with 250 μM CG for 10 min resulted in a three-fold increase in Akt phosphorylation (unpaired Student's t-test p<0.01; LY294002 + CG vs LY294002 only) (Fig. 5D). CG stimulation of Akt Ser-473 phosphorylation in the presence of LY294002 indicated an alternative mechanism for Akt activation independent of PI3K. Inhibition of mTOR by rapamycin treatment, had no effect on basal phosphorylation, but impeded the action of CG on Akt Ser-473 phosphorylation (Fig. 5D).

CG and the MAPK Signaling

The Erk1/2 signaling pathway was believed to promote cell survival but has been found to be aberrantly regulated in various neuropathological conditions [28]. Erk1 and Erk2 activities appeared to be differentially regulated by CG treatment. Erk1 phosphorylation (Fig. 6A) was suppressed by 250 μM CG stimulation from 2 to 32 min (One way ANOVA F_5,30=10.2 p<0.001; Tukey's honest post hoc p<0.05). Erk2 phosphorylation was not significantly affected by CG stimulation.

Figure 6. Cholesteryl glucoside (CG) stimulation does not affect MAPK signaling.

Western blotting analysis of MAPK activation in CG-treated NSC34 cells using phospho-site specific antibodies (A) CG treatment induced suppression of Erk1 phosphorylation from 2 to 32 min (One way ANOVA F_5,30=10.2 p<0.0001) but (B) Erk2 phosphorylation was not significantly affected. (C) NSC34 cells treated under the same condition as previously described were assessed for JNK1/2 and p38 MAPK activation. The blots were stripped and probed with anti-actin to confirm equal protein loading in each lane. Both JNK1/2 and p38 MAPK activation are activated only several hours later after CG treatment. The data represent the mean±S.E.M of at least three independent experiments.

The activities of the stress-activated protein kinases c-Jun N-terminal protein kinases (JNK) and p38 MAPK are known to be upregulated under cytotoxic conditions [29]. In contrast to the rapid responses for Akt and Erk1/2, both the JNK and p38 MAPK activations occurred much more slowly (Fig. 6C). On the one hand, JNK phosphorylation was first observed after 8 h of CG treatment and persisted for at least 24 h. On the other hand, the phosphorylation of p38 MAPK was not evident until several hours after CG treatment. When these immunoblots were stripped and probed with an anti-actin antibody, all lanes appeared to be equally loaded.

Role of PI3K/AKT and MAPK Pathways in CG Toxicity

In view of CG modulation of the PI3K/Akt pathway and Raf1-Mek1/2-Erk1/2 pathways in NSC34 cells, we next explored the role of these kinases in CG-induced cell death using pharmacological inhibitors. As shown in Fig. 7, NSC34 cells exposed to 250 μM CG for 48 h showed over 30% reduction in cell viability (unpaired Student's t-test p<0.05). To investigate the role of Akt and Erk1/2 in CG-mediated cytotoxicity, the PI3K inhibitor LY294002 (30 μM) and the Mek1/2 inhibitor U0126 (10 μM) were treated simultaneously with CG. Inhibition of the Mek/Erk pathway neither protected nor worsened CG-induced cell death (Fig. 7A), indicating that the Mek1/2/Erk1/2 pathway was not involved in cell death mechanisms in CG-treated NSC34 cells. Inhibition of the PI3K/Akt pathway with LY294002 during CG stimulation resulted in higher cell death as compared to CG treatment alone (unpaired Student's t-test p<0.05) (Fig. 7A). Inhibitor treatment alone had no effect on cell viability as compared to control untreated cells.

Figure 7. Effect of Mek1/2, PI3K, JNK1/2, and p38 MAPK inhibition on CG-treated NSC34 cell viability.

NSC34 cells maintained in complete media were treated with 250 μM CG and various kinase inhibitors for 20 h. The cells were then maintained in complete media with or without CG for a total of 48 h. (A) 250 μM of CG treatment resulted in approximately 40% decreased in cell viability (unpaired Student's t-test **p<0.01). CG treatment in the presence of 10 μM Mek1/2 inhibitor, U0126 or 30 μM PI3K inhibitor, LY294002 also resulted in a significant decreased in cell viability when compared to inhibitor treatment alone (unpaired Student's t-test **p<0.01). Inhibition of PI3K with LY294002 exacerbated CG induced decreased in cell viability (CG alone vs. CG with LY294002; unpaired Student's t-test *p<0.05). (B) CG treatment in the presence of 10 μM p38 MAPK inhibitor, SB230580 or 10 μM JNK1/2 inhibitor, SP600125 did not prevent cell death (inhibitor alone vs. inhibitor with CG, unpaired Student's t-test *p<0.05, **p<0.01). Inhibitor treatment alone did not have any significant effect on cell viability as compared to control cells (no treatment). Data are represented as mean±S.E.M.

The JNK and p38 MAPK are activated by stress stimuli and are believed to mediate pro-apoptotic signals under cytotoxic stress [29]. To test whether the delayed activation of JNK and p38 MAPK with prolonged CG stimulation was involved in mediating cell death, the p38 MAPK inhibitor SB203580 (10 μM) and the JNK inhibitor SP600125 (10 μM) were co-treated with CG as previously described. Inhibition of p38 MAPK and JNK with these compounds had no significant effect on cell death (Fig. 7B). Inhibitor treatment alone did not have any significant effects on cell viability as compared to control untreated cells.

Discussion

Brief exposures to CG within physiological levels can trigger protective intracellular signaling, but prolonged exposures are cytotoxic (Fig. 1A). These findings indicated that cells may have intrinsic resistance mechanisms to prevent cell death mediated by CG and possibly other stress stimuli. The aims of the current study included an examination of whether CG treatment could prevent cell death induced by serum deprivation and the mechanisms underlying CG preconditioning. Brief exposure of exogenous CG was found to trigger survival pathways to counteract cell death induced by serum deprivation (Fig. 2). Using a broad phospho-site immunoblotting screen, we identified various protein kinases that may be involved in CG preconditioning (Fig. 3). As in other types of preconditioning conditions [21-24], Akt activity was upregulated during CG preconditioning (Fig. 4), further supporting the notion that the PI3K/Akt pathway plays a central role in cell survival. Moreover, inhibition of the PI3K/Akt pathway completely abolished the cytoprotection induced by CG preconditioning (Fig. 4B).

Preconditioning as an Intrinsic Property of the Cell

Murray and colleagues [30] were first to describe ischemic preconditioning, where the myocardium subjected to brief episodes of ischemia developed resistance to subsequent ischemic episodes. The ischemic preconditioning phenomenon extended beyond cardiac tissues, and was documented in other organs including the brain [31-34]. Preconditioning was later observed in cell culture preparations. Moreover, the concept has also been extended to preconditioning triggered by non-ischemic stresses, such as by reactive oxygen species, hypoxia, stretch, and various chemicals [33, 35-37]. For example, Han and colleagues [22] found that transient hydrogen peroxide exposure preconditioned rat fibroblast L-cells for subsequent hydrogen peroxide exposures. This finding was extended to two other cell lines, the human embryonic kidney cell line (HEK293 cells) and a rat ventricular cell line (H9c2) [22]. Hypoxia also preconditioned the PC12 rat pheochromocytoma cell line to serum deprivation and various pharmacological agents that induced apoptosis [21]. Bishop et al. [38] reported that motor neurons developed adaptive resistance to nitric oxide (NO) treatment. They observed that sublethal treatment of NO to NSC34 cells triggered adaptive mechanisms that counteracted cytotoxic NO levels [38]. Taken together, these results indicated that preconditioning is an intrinsic adaptive mechanism for cell survival.

CG Preconditioning in NSC 34 Cells

CG has been proposed to function as a mediator in the early stages of stress response [40]. CG synthesis can be rapidly induced after exposure of slime molds and human fibroblast to heat shock [1,6,39,40]. Moreover, rats subjected to cold restraint stress-induced gastric ulcer had more than a 2-fold increase in serum CG levels immediately after stress [40]. The elevated serum CG levels returned to basal levels after one hour post stress. Furthermore, oral administration of CG prior to cold stress effectively prevented ulcer formation [40]. The mechanism for CG removal has not been determined. Kunimoto and colleagues [40] argued that rapid induction and rapid removal of this glycolipid have anti-stress effects that protects against cold stress-induced ulcers. This finding parallels our observation that brief episodes of CG stimulation promote survival, whereas prolonged treatment is cytotoxic. Specifically, we showed that brief CG treatment triggered activation of a survival kinase, which protected against cell death induced by serum deprivation. However, continuous CG exposure resulted in more cell death than serum deprivation alone (Fig. 5).

The physiological concentration of CG in mammals remained undetermined. In our study, we used 250 μM of CG for preconditioning since this concentration was the required amount for prevention of cold stress-induced ulcers from rats [40]. In view of the possibility that CG could be a lipid stress mediator, it is interesting to speculate that the basal level of CG fluctuates as the organism experiences insults in its daily activities. A sudden increase in CG would trigger mechanisms for its removal. An inability to remove CG might result in deleterious damage to the organism. Our results indicate that CG is toxic to NSC34, but that it could also evoke intrinsic survival mechanisms to counteract other cytotoxic stimuli.

Signaling Pathways Involved in CG Preconditioning

The Akt pathway mediates survival signaling in cells and the phospho-site screen indicated the greatest increase in Akt Ser-473 phosphorylation during CG preconditioning. Akt is an important anti-apoptotic protein kinase, whose signaling mechanisms suppresses cell death triggered by various cytotoxic stress stimuli [21-23]. Furthermore, the loss of Akt signaling has been correlated with neuronal death in various models of neurodegeneration. For example, the loss of Akt signaling has been found to be the underlying cause of cell death in motor neurons deprived of trophic support [41]. Furthermore, an early decrease in PI3K/Akt signaling was found in presymptomatic stages of a mouse model of familial ALS [42].

Akt signaling is believed to have a protective role in response to neuronal injury by phosphorylating and inhibiting downstream regulators of apoptosis [43]. Furthermore, increased Akt activity parallels various preconditioning conditions. For example, Akt is activated during oxidative [22] and hypoxic [21], preconditioning against various various stress stimuli. Inhibition of Akt by blocking its upstream activator, PI3K prevented this cytoprotective effect. Inhibition of PI3K during CG preconditioning also abolished the cytoprotective effect further demonstrating that the PI3K/Akt pathway has a survival role.

Akt activation may be an intrinsic mechanism for resistance to CG stimulation. However, the mechanism that mediates Akt activation during cellular preconditioning has yet to be determined. The finding that CG stimulated Akt phosphorylation after LY294002 treatment indicated an alternative pathway for Akt activation. Many researchers believe that mTOR may be the same enzyme as PDK2, which phosphorylates Akt specifically at Ser-473 [26]. We reported that rapamycin treatment alleviated CG's ability to induce Akt Ser-473 phosphorylation, but do not affect basal phosphorylation of Akt. mTOR usually forms a complex with two associating molecules, Rictor and Raptor. Recent work demonstrated that the mTOR-Rictor complex is rapamycin-insensitive, in contrast to the mTOR-Raptor complex which is rapamycin-sensitive and is believed to be the main upstream PDK2 kinase that mediates Akt Ser-473 phosphorylation [44,45]. In NSC34 cells, inhibition with rapamycin did not block basal Akt Ser-473 phosphorylation, thus the observed effect in the current studies is due primarily to the mTOR-Rictor complex. Since rapamycin prevented CG-induced Akt Ser-473 phosphorylation, it would be interesting to speculate that CG treatment resulted in a shift from the mTOR-Raptor to the mTOR-Rictor complex. Determining whether or not CG-stimulated Akt Ser-473 phosphorylation depends on mTOR-Raptor activity will require further studies. Furthermore, it is possible that this is the alternative pathway for CG stimulation of Akt Ser-473 phosphorylation after LY294002 treatment.

CG suppressed serum stimulation of Akt in NSC34 cells contradicted the finding that CG-induced Akt activation mediated cytoprotection. It is possible that serum deprivation may have altered the normal physiology of NSC34 cells resulting in a differential response to CG stimulation. In any case, PI3K was found to be important in Akt activation and delayed the toxic effect of CG on NSC34 cell viability. Inhibition of PI3K during CG treatment resulted in further reduction of cell viability than CG treatment alone. At this concentration and duration of exposure, LY294002 alone did not have any effect on NSC34 cell viability. However, when the cells were treated with LY294002 alone for 48 h, cell viability was significantly decreased.

In contrast, Erk1/2 signaling had negligible effects on NSC34 cell viability during CG stimulation. Pharmacological inhibitors that interfere with Erk1/2 activation did not worsen cell viability during CG treatment. NSC34 cells exposed to CG showed a selective inhibition of Erk1 phosphorylation at its activation sites, but not Erk2 phosphorylation at the analogous sites. The mechanism for Erk1 inhibition is unclear, but Erk1 appears to be more sensitive to CG treatment. The stress-activated protein kinases, JNK1/2 and p38 MAPK are activated later during CG treatment. JNK and p38 MAPK are not involved in promoting CG-induced cell death, since suppressing their activities did not protect against CG toxicity. It is possible that activation of JNK and p38 MAPK may be secondary events caused by other cytopathological conditions downstream of CG stimulation.

The precise activity of various signal transduction pathways remains elusive during the progression of human ALS-PDC. We believe that at each disease stage of ALS-PDC, neurons are constantly striving for the molecular balances between survival and death signal transduction pathways [46]. These balances between neuronal survival and death might be mediated in part by the competing activation of different cell signaling pathways. Sometimes cell-survival pathways, which might be mediated by PKB/Akt, predominate in response to a neurotoxic stimuli and the neurons survives. At other times cell death pathways prevail during a toxic stimulation and neurodegeneration occurs. Our laboratory identified a family of putative neurotoxic steryl glysocides that may be involved in the etiopathogenesis of ALS-PDC and related disorders (for review, see [2]). It remains unclear the exact binding site(s) of CG and how these steryl glycosides induce cell death. Our current study demonstrated differential effects of CG: Chronic exposure was neurotoxic; transient CG exposure could protect NSC34 cells against serum deprivation by mechanisms involving PI3K/Akt signaling. In contrast, MAPK signaling pathways had a negligible role in CG toxicity. Future studies will focus on studying biochemical pathways specifically involved in CG-mediated cell death. A better understanding of the molecular mechanism underlying CG toxicity may provide insights to the toxic actions of steryl glycosides and the role they may play in neurological disease etiology.