Address correspondence and reprint requests to Jacques Hugon, Department of Neurology, CHU Poitiers, rue de la Miletrie, 86000 Poitiers, France. E-mail: email@example.com
In Alzheimer's disease, neuropathological hallmarks include the accumulation of β-amyloid peptides (Aβ) in senile plaques, phosphorylated tau in neurofibrillary tangles and neuronal death. Aβ is the major aetiological agent according to the amyloid cascade hypothesis. Translational control includes phosphorylation of the kinases mammalian target of rapamycin (mTOR) and p70S6k which modulate cell growth, proliferation and autophagy. It is mainly part of an anti-apoptotic cellular signalling. In this study, we analysed modifications of mTOR/p70S6k signalling in cellular and transgenic models of Alzheimer's disease, as well as in lymphocytes of patients and control individuals. Aβ 1–42 produced a rapid and persistent down-regulation of mTOR/p70S6k phosphorylation in murine neuroblastoma cells associated with caspase 3 activation. Using western blottings, we found that phosphorylated forms of mTOR and p70S6k are decreased in the cortex but not in the cerebellum (devoid of plaques) of double APP/PS1 transgenic mice compared with control mice. These results were confirmed by immunohistochemical methods. Finally, the expression of phosphorylated p70S6k was significantly reduced in lymphocytes of Alzheimer's patients, and levels of phosphorylated p70S6k were statistically correlated with Mini Mental Status Examination (MMSE) scores. Taken together, these findings demonstrate that the mainly anti-apoptotic mTOR/p70S6k signalling is altered in cellular and transgenic models of Alzheimer's disease and in peripheral cells of patients, and could contribute to the pathogenesis of the disease.
In eukaryotes, protein translation includes three consecutive phases: initiation, elongation and termination. The initiation phase corresponds to processes associated with the connection between mRNA and ribosomes. The elongation phase includes the links between amino acids at the ribosomal level, and is followed by the termination phase. These three phases are highly regulated by proteins, called translation factors, that can interact directly with mRNAs. In the initiation phase, two major factors are involved: eukaryotic initiation factor 2 (eIF2) and eukaryotic initiation factor 4E (eIF4E).
The availability of eIF4E is linked to the binding of specific proteins called 4E-BPs. When these proteins are not phosphorylated, they have a great affinity for eIF4E, which is unable to bind to mRNAs, leading to a reduction of translation. These proteins are mainly phosphorylated by a kinase called mTOR (mammalian target of rapamycin) or FKBP12-rapamicyn associated protein (FRAP) (Schmelzle and Hall 2000; Raught et al. 2001). mTOR can also phosphorylate kinase p70S6k that stimulates protein synthesis (Dufner and Thomas 1999). The regulation of mTOR activity is important for the availability of eIF4E. mTOR is activated by the phosphatidyl Inositol 3 Kinase (PI3K) and Akt pathways. The mTOR pathway is also activated by IGF1 (Clemens 2001). mTOR signalling is physiologically active and allows protein and ribosome synthesis. Many studies have shown that cellular stresses can increase the binding of 4E-BPs to the eIF4E factor and can reduce protein translation (Clemens 2001). Recent studies have shown that two tumour suppressor genes, TSC1 and TSC2, and their corresponding proteins are able to regulate mTOR phosphorylation and binding to protein 14-3-3 (Gao et al. 2002). In addition, reduction of the eIF4E level is not enough to induce cellular apoptosis (Luo et al. 1994) and double-stranded RNA dependent protein kinase (PKR) can dephosphorylate the protein's 4EBPs through the activation of phosphatase 2A, leading to cell apoptosis (Xu and Williams 2000; Jeffrey et al. 2002). Recent electrophysiological results have also demonstrated that mTOR could play a role in neuronal plasticity and in the process of learning and memory. In brain slices treated with the mTOR inhibitor rapamycin, the authors observed a decrease in the late phase of long term potentiation (LTP) induced by synaptic stimulation or brain derived neurotrophic factor (BDNF) exposure (Nguyen 2002; Tang et al. 2002).
EIF2 can fix guanosine triphosphate (GTP) and interact with tRNA carrying methionin. The complex formed can bind to the small 40S subunit of the ribosome, a phase in which eIF4E plays also a major role. The availability of these two initiation factors is indispensable for the onset of protein synthesis. The control of eIF2 is linked to the state of phosphorylation on the α subunit on serine 51. Upon phosphorylation, eIF2α is unable to bind to tRNA and translation is stopped. Four protein kinases are able to phosphorylate eIF2α: PKR, PERK (PKR-like endoplasmic reticulum kinase), HRI (Heme-regulated kinase) and GCN2 (amino acid-regulated kinase). PKR is a ubiquitous protein involved in viral infections, but it is also activated during different cellular stresses such as toxic stress, endoplasmic reticulum (ER) stress, intracellular calcium stress and lack of trophic factors (Williams 1999). ER stress also leads to the activation of PERK and the unfolded protein response (Harding et al. 1999). Several studies have revealed increased PKR expression with ageing (Ladiges et al. 2000), in neurodegenerative diseases (Peel et al. 2001; Chang et al. 2002b), or in the activation of eIF2α in cerebral ischaemia-reperfusion, hypoxia or zinc toxicity (Frerichs et al. 1998; Alirezaei et al. 1999; DeGraciaet al. 1999). It has been demonstrated that these kinases play an important role in the process of cellular apoptosis by interacting with protein translation and apoptotic factors.
An early study revealed that mRNA translation was disturbed in the brain of Alzheimer's disease (AD) patients (Langstrom et al. 1989). Inhibition of ribosomal translation could be responsible for modifications of gene expression detected in affected brain regions. An increased level of elongation factor 2 was also noted in AD brains (Johnson et al. 1992). We have recently demonstrated an accumulation of phosphorylated eIF2α in degenerating neurones in AD brains associated with increased staining of activated PKR (Chang et al. 2002a). This finding was confirmed by another group (Ferrer 2002). In addition, we have shown in vitro that Aβ could induce activation of PKR and the phosphorylation of eIF2α (Chang et al. 2002b). PKR activation was dependent upon caspase 3 and caspase 8 triggering. A reduction in protein synthesis has been described during ageing (Rattan 1996), as well as a decrease in eIF2α and eIF2B and an increase in PKR expression (Kimball et al. 1992; Ladiges et al. 2000). Very few studies have been carried out so far on the involvement of the mTOR/p70S6k signalling pathway following Aβ exposure in cell cultures and in AD models. The results of the present study show that there is a decrease in mTOR/p70S6k signalling in neural cells treated in vitro with Aβ, in the cortex of transgenic mice overexpressing APP and Presenilin 1 mutations, and in lymphocytes of patients suffering from AD. These findings demonstrate that translational control could represent a new molecular target for future therapeutic interventions concerning Alzheimer's disease.
Materials and methods
All-trans retinoic acid (RA), peptides Aβ 1–42 and Aβ 40–1, paraformaldehyde (PFA), Histopaque 1077, Triton X-100, sodium fluoride, phenylmethylsulfonyl fluoride, protease and phosphatase inhibitors, dithiothreitol (DTT) and Caspase 3 substrate Ac-DEVD-pNA were obtained from Sigma (Saint Quentin Fallavier, France). Pentobarbital was obtained from Ceva Santé Animale (Libourne France). Primary antibodies and secondary anti-rabbit IgG antibody conjugated with horseradish peroxidase were purchased from Cell Signalling Ozyme (Saint Quentin Yvelines, France) except for anti-β-tubulin antibody and anti-β-actin antibody which were from Sigma. The phospho-mTOR antibody was recently used as a marker for mTOR activation (Bolster et al. 2002). Secondary anti-mouse IgG antibody conjugated with horseradish peroxidase was obtained from Amersham (Orsay, France). All reagent-grade chemicals for buffers were obtained from VWR International (Strasbourg, France).
Cell cultures and Aβ treatment
Neuro-2a murine neuroblastoma cell lines were obtained from the American Type Culture Collection (ATCC). Neuro-2a cells were cultured with Minimal Essential Medium (MEM) (Gibco-BRL, Cergy Pontoise, France), supplemented with 10% fetal bovine serum (FBS) and 1% antibiotics [50 U/mL penicillin + 50 µg/mL streptomycin (Gibco-BRL)]. The cells were differentiated into mature neurones by incubating with 20 µm RA for 2 days. Cells were maintained in a humidified 5% CO2 atmosphere at 37°C.
Cells were treated with 20 µm Aβ (1–42) or (40–1) from 30 min to 24 h in serum-free medium. The peptides were incubated at 37°C from 24 to 48 h to form aggregates prior to use. After treatment, neural cells were lysed in ice-cold lysis buffer containing protease and phosphatase inhibitor cocktail. The lysate was sonicated and then centrifuged at 15 000 g for 15 min at 4°C. The protein content of the supernatant fluid was measured using a protein assay kit (BioRad, Marnes la Coquette, France); 20 µg protein (from the supernatant fluid) per lane were used for mTOR analysis, 10 µg for all other proteins and 60 µg for the caspase activity assay.
Caspase 3 assay
In a reaction volume of 100 µL, lysate buffer supernatant fluid was added to assay buffer [312.5 mm HEPES (pH 7.5), 31.25% w/v sucrose, 0.3125% w/v CHAPS] with 100 mm DTT and the caspase-3 substrate Ac-DEVD-pNA (200 µm), and incubated at 37°C. Release of p-nitroaniline was monitored by recording the OD405nm at 15 min intervals.
Western blot analysis
Protein (20 µg per sample) was separated on 6% or 10% Tris-glycine polyacrylamide gels and transferred to polyvinylidene difluoride (PVDF) membranes (Immobilon-P, Millipore, Bedford, MA, USA). Immunoblots were blocked for 2 h in Tris-buffered saline Tween-20 (TBST, 20 mm Tris-HCl, 150 mm NaCl, pH 7.5, 0.05% Tween 20) containing 5% non-fat milk and 0.21% sodium fluoride. The blots were incubated with primary antibody in blocking buffer at 4°C overnight. Membranes were washed twice with TBST and then incubated with the secondary antibody : peroxidase-conjugated anti-rabbit IgG (1 : 1000 dilution) for 1 h at 25°C. Membranes were washed again and then developed with Enhanced Chemiluminescence (ECL) plus western blotting (Amersham Biosciences), followed by apposition of the membrane to autoradiographic films (Hyperfilm ECL, Amersham Biosciences). After two washes with TBST, membranes were probed with a monoclonal antibody against tubulin (1 : 1000, Sigma) or β-actin (1 : 1000, Sigma) for 2 h at 25°C. They were then washed with TBST, incubated with anti-mouse peroxidase-conjugated secondary antibody for 1 h and developed. Band intensities were analysed with a gel documentation system (BioRad). mTor and p70S6k expressions were adjusted to tubulin expression. The protein levels were expressed as densitometry and percentage of controls.
The different transgenic mouse lines were generated at the Research Centre of Paris, Sanofi Aventis, in Neurodegenerative Disease Group (Vitry sur Seine, France). All the mice were female and all experiments were performed in compliance with, and following approval of the Aventis Animal Care and Use Committee, in accordance with the standards of the Guide for the Care and Use of Laboratory Animals (CNRS ILAR), and with respect to French and European Community rules. The genetic background of the mice was mixed CBA/C57BL/6.
The following groups of mice were examined: 24-month-old control mice [n = 5, body weight (BW) = 36.14 ± 4.49 g]; 24-month-old transgenic mice (n = 5, BW = 28.78 ± 1.84 g) expressing human mutant APP751 (APP HE) (carrying the Swedish and London mutations KM670/671NL and V717l, Thy1 promoter); 12-month-old control mice (C57BL/6, n = 8, BW = 34.58 ± 2.22 g); 12-month-old transgenic mice expressing single human mutant presenilin-1 (APP/PS1 WT/HE) (PS-1 M146L, HMG promoter, n = 6, BW = 34.06 ± 3.05 g), 12-month-old transgenic mice expressing human mutant APP751 SL and human mutant presenilin-1 M146L (APP/PS1 HE/HE, n = 4, BW = 28.85 ± 0.48 g). APP/PS-1 double-transgenic mice were generated by crossing PS-1 (HMG PS-1 M146L) homozygous mice with hemizygous APP (Thy1 APP751 SL) transgenic mice. A detailed description of these transgenic mice has been given previously (Wirths et al. 2001; Blanchard et al. 2003). APP mice were analysed at 24 months of age and APP-PS1 mice were assessed at 12 months of age because they revealed comparable levels of amyloid plaques in the brain at these respective ages.
Brain tissue preparation
For western blot analysis, mice were anaesthetized with pentobarbital (40 mg/kg, i.p.) and killed. Brains were removed and cortex and cerebellum dissected. These cerebral regions were homogenized by 10 up-and-down strokes of a pre-chilled Teflon-glass homogenizer in 20 volumes of lysis buffer (25 mm Tris-HCl, 150 mm NaCl, 1 mm EDTA, pH 7.4) and supplemented with 50 mm sodium fluoride, 1 mm phenylmethylsulfonyl fluoride, protease and phosphatase inhibitor cocktails (50 µL/g tissue and 10 µL/mL lysis buffer, respectively). Lysates were sonicated and centrifuged at 15 000 g for 15 min at 4°C. The resulting supernatant fluids were collected for BioRad protein assay and western blot analysis as described above.
The brain tissue preparation for immunohistochemistry required an intracardial perfusion with Phosphate-Buffered Saline (PBS) (154 mm NaCl, 1.54 mm KH2PO4, 2.7 mm Na2HPO4.7H2O, pH 7.2) followed by 4% (v/v) paraformaldehyde (PFA) in PBS. Brains were removed and post-fixed for 24 h in 4% PFA at 4°C. They were then rinsed in PBS, dehydrated, and embedded in paraffin for sectioning (4 µm in thickness).
Sections were de-waxed for 2 × 10 min with Histosol (Shandon, Cergy Pontoise, France) and rehydrated by 5 min incubations in each of 100, 95 and 75% ethanol. Sections were rinsed twice for 5 min in ultra-high quality water (H2OUHQ) and washed twice in PBS (2 × 5 min). To unmask antigens, the slides were immersed in citrate buffer (10 mm sodium citrate, pH 6) and microwaved for 2 × 10 min (with 650 W). For amyloid peptide (Aβ) immunostaining, sections were incubated in 10 mm sodium citrate buffer, pH 9, for 40 min in a 98°C water bath. Slides were allowed to reach room temperature before two 5 min incubations in H2OUHQ. For rabbit polyclonal primary antibody anti-phospho-mTOR (Ser2448), dilution 1/20, the EnVision System (DakoCytomation, Trappes, France) was used. This protocol includes four steps: (i) a 5 min peroxidase block to quench endogenous peroxidase activity; (ii) an incubation for 30 min at room temperature with diluted primary antibody in TBST (50 mm Tris-HCl, 150 mm NaCl pH 7,6 with 0.1% Tween) containing 1% of bovine serum albumin; (iii) an incubation for 30 min at room temperature with the peroxidase-labelled polymer; (iv) staining (8 min incubation) with 3,3′-diaminobenzidine (DAB) substrate-chromogen which results in a brown-coloured precipitate at the antigen site. The monoclonal mouse primary antibody anti-human β-amyloid (1–40/42) C-terminal, dilution 1/25 (Calbiochem, Strasbourg, France), was immunostained with the DakoCytomation DAKO ARK system. The diluted primary antibody was labelled using the biotinylation reagent, a monovalent biotinylated anti-mouse immunoglobulin. Tissue sections were incubated for 15 min at room temperature (RT) with the biotin-labelled primary antibody and then incubated, also for 15 min at RT, with a streptavidin-alcaline phosphatase conjugate. Staining was completed with fast red substrate-chromogen, resulting in a red end product at the site of the target antigen. Mayer's haematoxylin or methyl green was eventually used as counterstain.
Selection of patients and cognitive tests
Twenty-six patients (18 female, 8 male) (mean age: 76.3 ± 6.5 years, range: 60–89 years) suffering from AD were selected from the department of Neurology (Poitiers University Hospital), according to the NINCDS-ADRDA clinical criteria, and included in the study along with 28 control subjects (21 female, 7 male) (mean age: 77.4 ± 7.2 years, range: 64–94 years) without neurological disorders. All patients and controls gave their written informed consent for the study. All control individuals had a Mini Mental Status Examination (MMSE) above 25. AD patients had a MMSE below 26. Blood samples (20 mL) were drawn at the same time as MMSE was performed. This study was approved by the regional ethics committee of the Region Poitou-Charentes (France) (CCPPRB).
Preparation of lymphocyte cells
Peripheral venous blood was obtained from healthy volunteer donors and Alzheimer patients. Mononuclear cells were isolated using Ficoll-Histopaque 1077 (Sigma, Saint Quentin Fallavier, France) density gradient centrifugation at 700 g for 20 min. The lymphocyte cells obtained were then washed in RPMI 1640 (Gibco-BRL) and re-suspended in 200 µL lysis buffer [50 mm Tris-HCl, 50 mm NaCl, pH 6.8, 1% Triton X-100, 1% (v/v) protease inhibitor cocktail, 1% (v/v) phosphatase inhibitor cocktail]. After homogenization, lysates were sonicated for 2 min and centrifuged at 15 000 g for 15 min at 4°C. The supernatant fluid was analysed for protein content using a protein assay kit (BioRad). Aliquots of 2 µg protein/µL were frozen at − 80°C until further analysis. As internal control, we used the same protein sample obtained from murine neuroblastoma cells (Neuro-2a).
Results are expressed as means with SEM. Data for multiple variable comparisons were analysed by one-way analysis of variance (anova) or the non-parametric Kruskal–Wallis test. For comparison of significance, Bonferroni's test or Dunns's test was used as a post hoc test according to the statistical program GraphPad Instat. For comparison of significance between two groups, Student's t-test was conducted. The level of significance was p < 0.05. The Pearson correlation and linear regression between p70S6k levels and MMSE scores was also analysed.
Caspase 3 activation induced by 1–42 Aβ
Figure 1a shows the results of caspase 3 activation in Neuro-2a cells after exposure to Aβ. After 8 h, there is a significant increase in active caspase 3 in treated cells compared with control cells, indicating that apoptosis pathways are triggered by Aβ in Neuro-2a cells. The activity of caspase 3 is augmented (174% of control) 4 h after Aβ exposure. After 8 h, the activity is 267% of controls and at the same time, inactivation of mTOR is around 70% of controls in Neuro-2a cells.
Inactivation of mTOR in neural cells induced by Aβ peptide 1–42
Expression of mTOR was examined using polyclonal antibodies against its total and phosphorylated forms. Figure 1b shows a marked decrease in mTOR phosphorylation when the differentiated murine neuroblastoma cells were treated with 20 µm aggregated Aβ 1–42 for various times. The ratio between the phosphorylated and the total forms is already reduced (Fig. 1c) 30 min after the beginning of exposure. This inactivation augments with time and is around 75% in Neuro-2a cells after 24 h (*p < 0.05 compared with the respective control). As a control, the non-toxic homologue Aβ 40–1 did not enhance mTOR inactivation in cells.
Western blots also depicted the effects of Aβ 1–42 exposure on phosphorylated (Thr389 controlled by mTOR signalling) p70S6k expression in neural cells (Figs 2a and b). Analysis revealed that Aβ 1–42 induced a progressive reduction of the ratio between the phosphorylated and the total forms of p70S6k. This inactivation is significant after 24 h (*p < 0.05 compared with the respective control). As a control, the reverse peptide Aβ 40–1 did not induce p70S6k inactivation in neural cells.
mTOR and p70S6k expression in transgenic mice
Antibody against p70S6k phosphorylated at Thr389 showed two positive bands corresponding to phosphorylated p70/p85 S6 kinases that are the cytosolic and nuclear isoforms, respectively. In mouse homogenates, we analysed only the signal band at 70 kDa (Fig. 4). There was no difference in the activation of mTOR and p70S6k (corresponding to the ratio between phosphorylated and total forms of the kinases) between transgenic mice expressing human mutant APP751 and control mice in cortex homogenates (Figs 3a and 4a). In cerebellum homogenates, an increased level of phosphorylated mTOR and p70S6k was observed but was only significant for p70S6k (increase: 34%, p < 0.01) (Figs 4b and d). A significant decrease in mTOR activation was observed in the cortex (decrease: 34%) of APP/PS-1 double-transgenic mice expressing human mutant APP751 SL and human mutant presenilin-1 M146L compared with 12-month-old control mice (Figs 3a and c). As shown in Fig. 5, APP/PS-1 double-transgenic mice developed in the cortical regions extracellular Aβ deposits (Fig. 5a) with very intensively Aβ-immunostained neurones (Fig. 5b). A reduction in phosphorylated mTOR immunostaining was observed in cortical neurones of APP/PS-1 double-transgenic mice (Fig. 5d) compared with cortical neurones of control mice showing mainly cytoplasmic localizations (Fig. 5c). In the cortex of APP/PS-1 double-transgenic mice, the inactivation of mTOR detected by western blots was associated with a significant inactivation of p70S6k (decrease: 31%, Figs 4a and c). On the other hand, a large increase in mTOR activation was observed in cerebellum homogenates of APP/PS-1 double-transgenic mice (increase: 250% of the ratio phospho-mTOR/mTOR, p < 0.01) (Figs 3b and d) with a non-significant increase in the ratio phospho-p70S6k/p70S6k (Figs 4b and d). Morphological results confirmed that phospho-mTOR immunostaining was clearly marked in the cerebellum of APP/PS1 transgenic mice (Fig. 5f).
Expression and phosphorylation of p70S6 kinase in lymphocytes of Alzheimer patients
The expression and phosphorylation of p70S6 kinase (Thr389) was analysed by western blotting in lymphocytes of 28 control individuals and 26 Alzheimer patients (Figs 6a, b and c). Phosphorylation site (Thr389) is the specific site of p70S6k phosphorylation by kinase mTOR. Data are expressed as a percentage of internal control taken from neuroblastoma cells. Individual results and means are depicted in Fig. 6(b) and indicate that p70S6k activation, measured by the ratio phospho-p70S6k/p70S6k, was significantly decreased (51%, p < 0.0005) in lymphocytes of Alzheimer patients compared with control individuals (AD patients: 241.8 ± 30.2, controls: 490.2 ± 52.6). The sensitivity of the results is 88.4% and the specificity is 64.2%. This clear-cut decrease was due to a significant reduction in the phosphorylated form of p70S6k. The scattergram (Fig. 6c) shows that the mean value of phospho-p70S6k (Thr389) expression alone was significantly decreased by 74% in lymphocytes of Alzheimer patients compared with control lymphocytes (AD patients: 49.2 ± 10.4; controls: 188.0 ± 17.2; p < 0.0005).
Correlation between p70S6 kinase activation in lymphocytes of Alzheimer patients and MMSE scores
Correlation and regression analyses were carried out to investigate the relations between p70S6k activation in lymphocytes of Alzheimer patients and MMSE scores. The phospho-p70S6k/p70S6k ratio measured in Alzheimer patients lymphocytes showed a significant correlation with MMSE scores (r = 0.726; p < 0.0001; Fig. 7).
Our results demonstrate that Aβ exposure induces a sustained reduction in mTOR/p70S6k signalling in neuroblastoma cell cultures, marked by a progressive decrease in phosphorylated mTOR and phosphorylated p70S6k. In addition, mTOR signalling is reduced in the cortex but not in the cerebellum of APP/PS-1 mutant transgenic mice and in lymphocytes of AD patients compared with control individuals. Taken together, these findings suggest that the inactivation of mTOR/p70S6k pathway, which regulates cell growth and proliferation, protein translation and autophagy and is thought to transfer anti-apoptotic properties, could be a factor contributing to the degenerative process detected in the brains of AD patients.
mTOR is known to phosphorylate p70S6k at Thr389, but other kinases such as ERK1/2 are able to phosphorylate p70S6k at Thr421/Ser424 (Dennis et al. 1998). In addition, a recent report showed that p70S6k could be activated not by a linear pathway from mTOR but by two separate pathways triggered by mTOR and PI3 kinase (Raught et al. 2001). This could explain the different activation profile observed for mTOR and p70S6k dephosphorylation. This dephosphorylation is not very rapid and complete as it was observed after energy depletion in another cell culture system (Inoki et al. 2003). This difference is possibly linked to the slow toxicity induced by Aβ. We have previously demonstrated that Aβ 1–42 activates PKR, a kinase known to down-regulate protein translation, acting prior to mTOR in translational control and leading to phosphorylation of the eukaryotic initiation factor 2α (eIF2α) and cellular apoptosis (Suen et al. 2003). Increased PKR signalling and reduced mTOR/p70S6k signalling are the combined consequences of Aβ 1–42 exposure in neural cells. Previous reports have revealed an increase in intracellular calcium following Aβ exposure (Mattson and Chan 2003). The role of calcium in PKR activation was also demonstrated in our previous study (Suen et al. 2003) and could be involved in the transient rise of p70S6k phosphorylation observed at 2 h because this enzyme is known to be sensitive to variations in intracellular calcium (Hannan et al. 2003). It is now known that mTOR is an oncogenic protein participating in cell growth and proliferation but with probable anti- and pro-apoptotic properties that have been recently reviewed (Castedo et al. 2002). In our neural cell system, the reduction in mTOR phosphorylation occurred as early as 30 min to 2 h after Aβ exposure and was amplified over the following 24 h. This finding could contribute to the neurotoxic and apoptotic properties of Aβ previously described in neural cells (Chang et al. 2002b).
The immunohistochemical results showed a large number of Aβ−positive plaques in different regions of the central nervous system including the cortex, as well as an increase in expression of Aβ inside neurones. As previously described, almost no neuronal degeneration was detected in the brains of the transgenic mice at the opposite of what was in the brain of AD patients. The changes in mTOR/p70S6k signalling were only detected in APP/PS1 transgenic mice and not in transgenic mice expressing only human mutant APP. Since both types of transgenic mice expressed Aβ plaques, it is possible to postulate that the augmented expression of Aβ can only be partly responsible for the reduced mTOR/p70S6k signalling in the same regions. Presenilin (Culvenor et al. 1997) and mTOR (Drenan et al. 2004) have been localized in the ER–Golgi network and mutated PS1 could contribute, together with Aβ accumulation, to the reduced activity of mTOR/p70S6k in the brains of these APP/PS1 mice. A recent report has also noted that mutated human PS1 down-regulates the PI3K/Akt pathway which controls mTOR activity (Baki et al. 2004). We have observed a significant modification of mTOR phosphorylation in the cerebellum of transgenic mice overexpressing the double APP/PS1 human mutations. In addition, using histological methods, we did not observe plaques or overt positive Aβ immunostaining in the cerebellum of these animals; this observation concurs with previously published data assessing the same mice (Blanchard et al. 2003). The increased levels of phosphorylated mTOR observed in the cerebellum with western blots in APP-PS1 mice is associated with a lack of Aβ accumulation. The reasons for these biochemical and histological findings are not known but it can be assumed that specific molecular pathways in the cerebellum might prevent major extracellular and intracellular Aβ accumulations. Further studies will be necessary to determine whether these two observations are linked and are implicated in specific neuronal susceptibility of the cortical versus the cerebellar areas. In AD brains, only diffuse plaques and very few senile plaques have been described in the cerebellum of patients with a faint morphological reaction in their microenvironment (Joachim et al. 1989). A recent report has revealed that positive staining for phosphorylated p70S6k on two sites (Thr389, Thr421/Ser424) was detected in neurones of AD patients with increased brain levels. Only activated p70S6k expression at the Thr421/Ser424 site was correlated with levels of total tau and phosphorylated tau (An et al. 2003). The phosphorylation of p70S6k on Thr421/Ser424 is linked to ERK1/ERK2 triggering and not to mTOR activation (Dufner and Thomas 1999). The molecular mechanisms leading to increased expression of phosphorylated p70S6k on Thr389 in specific AD neurones could be comparable with those leading to the transient increased activity induced in vitro by Aβ exposure in our cell culture system. Whether or not the staining of activated p70S6k (Thr389) in AD brains corresponds to specific modulation of mTOR signalling activity or to another contributory factor in these neurones remains to be determined.
We analysed the expression of phosphorylated p70S6k (Thr389) and total p70S6k in lymphocytes of patients and control individuals. There was a clear reduction in the mean level of the ratio between phosphorylated p70S6k versus total p70S6k. This difference was also detected for phosphorylated p70S6k alone in AD patients as compared with control subjects. This finding demonstrates that peripheral cells are marked by a decrease in the initiation phase of protein translation in patients suffering from AD. Many previous studies have demonstrated various biological disturbances in lymphocytes and other peripheral cells in AD patients, including increased cytosolic free calcium levels (Gibson et al. 1987; Adunsky et al. 1991; Eckert et al. 1997; Palotas et al. 2002), an impairment of mitogenic proliferation (Stieler et al. 2001), an enhanced sensitivity to apoptosis (Eckert et al. 2001) and the presence of increased reactive oxygen species (Mecocci et al. 2002). In line with the results of the present study, several of these observations could be either the cause or the consequence of reduced translational control. Recent data have revealed that lymphocytes of AD patients are less sensitive to the action of rapamycin, a potent antagonist of mTOR signalling, on cell cycle kinesis and proliferation (Nagy et al. 2002). We suggest that if the mTOR/p70S6k signalling is already down-regulated in AD lymphocytes, the availability of the remaining activated mTOR/p70S6k proteins is insufficient for proper rapamycin action. The other interesting finding is that the reduction of translational control assessed by the ratio between activated p70S6k versus total p70S6k is correlated with the cognitive decline of patients evaluated by MMSE scores. In this study, we have only included AD patients with a MMSE score ranging from 25 to 10; in these patients, memory anomalies in the first phase of the disease are very often at the forefront of clinical signs. mTOR has been experimentally associated with molecular mechanisms of long-term potentiation (LTP), and learning and memory (Cammalleri et al. 2003). In fact, rapamycin can experimentally modify these cognitive processes. If the translational control is altered in AD brains, this modification could contribute to the onset of memory impairment observed in AD. As mentioned previously, protein synthesis appears to be reduced in AD brains. This finding was also observed in experimental transient cerebral ischaemia associated with a decrease in mTOR/p70S6k signalling marked by reduced activated p70S6k expression (Mengesdorf et al. 2002).
We have previously shown that PKR levels are altered in hippocampal neurones of AD patients and are marked by increased expression of phosphorylated PKR, suggesting that protein translation is reduced in these neurones (Chang et al. 2002a). It would be interesting to find out whether or not the modifications of mTOR/p70S6k signalling in peripheral lymphocytes can reflect biological neuronal disturbances associated with memory decline in AD patients, and it can be assumed that a general alteration of translational control implicates various cell types in AD models and in the cells of AD patients, including neurones and peripheral lymphocytes.
In summary, it is not yet known whether the down-regulation of mTOR/p70S6k signalling observed in lymphocytes is also present in the brains and neurones of AD patients. Cellular and transgenic AD models have revealed down-regulated mTOR/p70S6k signalling, and further studies will have to determine the contribution of this disturbed signalling to neuronal degeneration. The mTOR/p70S6k pathway, which is already at the centre of active anti-cancer research (Bjornsti and Houghton 2004), could also represent a new target for future therapeutic approaches and neuroprotection in AD, as well as in other neurodegenerative disorders.
The authors would like to thank Raymond Poncharaud and Brigitte Vidal for technical help, Claudette Pluchon, Bernard Fauconneau and Alain Piriou for useful advice, and André Brizard for scientific support. This study was supported by the CHU de Poitiers with a grant from the French Ministry of Health (PHRC) and by the University of Poitiers and the French Ministry of Education and Research, with a grant to the Research Unit EA 3808.