Increased prion protein processing and expression of metabotropic glutamate receptor 1 in a mouse model of Alzheimer's disease

Authors

  • Valeriy G. Ostapchenko,

    1. Robarts Research Institute, Department of Physiology and Pharmacology, University of Western Ontario, London, ON, Canada
    Search for more papers by this author
    • These authors contributed equally to this work.
  • Flavio H. Beraldo,

    1. Robarts Research Institute, Department of Physiology and Pharmacology, University of Western Ontario, London, ON, Canada
    Search for more papers by this author
    • These authors contributed equally to this work.
  • Andre L. S. Guimarães,

    1. Robarts Research Institute, Department of Physiology and Pharmacology, University of Western Ontario, London, ON, Canada
    2. Universidade Estadual de Montes Claros, Montes Claros, MG, Brazil
    Search for more papers by this author
  • Sanju Mishra,

    1. Robarts Research Institute, Department of Physiology and Pharmacology, University of Western Ontario, London, ON, Canada
    Search for more papers by this author
  • Monica Guzman,

    1. Robarts Research Institute, Department of Physiology and Pharmacology, University of Western Ontario, London, ON, Canada
    Search for more papers by this author
  • Jue Fan,

    1. Robarts Research Institute, Department of Physiology and Pharmacology, University of Western Ontario, London, ON, Canada
    Search for more papers by this author
  • Vilma R. Martins,

    1. International Research Center, A. C. Camargo Cancer Center and National Institute for Translational Neuroscience, São Paulo, SP, Brazil
    Search for more papers by this author
  • Vania F. Prado,

    Corresponding author
    1. Robarts Research Institute, Department of Physiology and Pharmacology, University of Western Ontario, London, ON, Canada
    2. Department of Anatomy and Cell Biology, University of Western Ontario, London, ON, Canada
    • Address correspondence and reprint requests to Dr Marco A. M. Prado or Dr Vania F. Prado, Robarts Research Institute, University of Western Ontario, 100 Perth Drive, London, ON N6A5K8, Canada. E-mails: mprado@robarts.ca or vprado@robarts.ca

    Search for more papers by this author
  • Marco A. M. Prado

    Corresponding author
    1. Robarts Research Institute, Department of Physiology and Pharmacology, University of Western Ontario, London, ON, Canada
    2. Department of Anatomy and Cell Biology, University of Western Ontario, London, ON, Canada
    • Address correspondence and reprint requests to Dr Marco A. M. Prado or Dr Vania F. Prado, Robarts Research Institute, University of Western Ontario, 100 Perth Drive, London, ON N6A5K8, Canada. E-mails: mprado@robarts.ca or vprado@robarts.ca

    Search for more papers by this author

Abstract

Prion protein (PrPC), a glycosylphosphatidylinositol-anchored protein corrupted in prion diseases, has been shown recently to interact with group I metabotropic glutamate receptors (mGluRs). Moreover, both PrPC and mGluRs were proposed to function as putative receptors for β-amyloid in Alzheimer's disease. PrPC can be processed in neurons via α or β-cleavage to produce PrPC fragments that are neuroprotective or toxic, respectively. We found PrPC α-cleavage to be 2–3 times higher in the cortex of APPswe/PS1dE9 mice, a mouse model of Alzheimer's disease. A similar age-dependent increase was observed for PrPC β-cleavage. Moreover, we observed considerable age-dependent increase in cortical expression of mGluR1, but not mGluR5. Exposure of cortical neuronal cultures to β-amyloid oligomers upregulated mGluR1 and PrPC α-cleavage, while activation of group I mGluRs increased PrPC shedding from the membrane, likely due to increased levels of a disintegrin and metalloprotease10, a key disintegrin for PrPC shedding. Interestingly, a similar increase in a disintegrin and metalloprotease10 was detected in the cortex of 9-month-old APPswe/PS1dE9 animals. Our experiments reveal novel and complex processing of PrPC in connection with mGluR overexpression that seems to be triggered by β-amyloid peptides.

image

Prion protein (PrPC) and metabotropic glutamate receptors (mGluR) are implicated in Alzheimer's disease (AD). We found age-dependent increase in PrPC processing, ADAM10 and mGluR1 levels in AD mouse model. These changes could be reproduced in cultured cortical neurons treated with Aβ peptide. Our findings suggest that increased levels of Aβ can trigger compensatory responses that may affect neuronal toxicity.

Abbreviations used
AD

Alzheimer's disease

β-amyloid peptide

APP

amyloid precursor protein

DHPG

dihydroxyphenylglycine

mGluR

metabotropic glutamate receptor

PrPC

cellular prion protein

RIPA

radioimmunoprecipitation assay

TBS

tris-buffered saline

WT

wild type

Alzheimer's disease (AD) is the most common form of dementia. Its pathological hallmarks include accumulation of proteinaceous aggregates (senile plaques) formed by peptides derived from amyloid precursor protein (APP) processing and neurofibrillary intracellular tangles, because of the hyperphosphorylation of the microtubule-associated protein Tau (Selkoe 2011). Recent evidence has provided support for a role of soluble oligomeric forms of β-amyloid (Aβ) peptides as a potential toxin in AD (Lambert et al. 1998; Klein 2002; Walsh and Selkoe 2007; Ferreira and Klein 2011), by interacting with distinct synaptic components to affect synaptic transmission (Barry et al. 2011). Among possible targets, NMDA receptors (Snyder et al. 2005; Decker et al. 2010), α7 nicotinic acetylcholine receptors (Wang et al. 2000; Magdesian et al. 2005; Bomfim et al. 2012), the insulin receptor (Xie et al. 2002; De Felice et al. 2009; Zhao et al. 2009; De Felice 2013), and the prion protein (PrPC) (Lauren et al. 2009; Caetano et al. 2011; Freir et al. 2011) have received considerable attention in the last few years.

Of all the potential receptors for Aβ oligomers, PrPC is one of the most enigmatic. Though its corruption by misfolding in prion diseases has been thoroughly supported (Prusiner 1998), understanding of its biological functions is still a challenge. Recent studies identified PrPC as an Aβ oligomer receptor (Lauren et al. 2009). This interaction has been confirmed by several other laboratories (Chen et al. 2010; Kessels et al. 2010; Caetano et al. 2011), albeit the pathological significance of Aβ oligomer interaction with PrPC has been the subject of some controversy (Balducci et al. 2010; Benilova and De Strooper 2010; Calella et al. 2010). Two N-terminally located charged clusters in the PrPC sequence have been identified as Aβ oligomer binding sites (Lauren et al. 2009; Chen et al. 2010). Removal or antibody-mediated blocking of these sites significantly decreased Aβ oligomer binding and the corresponding toxic effect in neurons (Chen et al. 2010; Chung et al. 2010; Freir et al. 2011).

PrPC undergoes proteolytic processing that resembles that of APP. Constitutive α-cleavage of PrPC releases the N1 peptide (PrPC residues 23–110), and the C-terminal counterpart (C1 peptide) remains glycosylphosphatidylinositol-anchored (Mange et al. 2004). α-cleavage depends on ‘a disintegrin and metalloprotease’ (ADAM) family of proteases, with the potential participation of ADAM10 and ADAM17 (Vincent et al. 2001), during the later stages of protein secretion (Walmsley et al. 2009). Of note, α-cleavage eliminates Aβ oligomer binding sites, both of which reside within the N1 peptide. Coincidentally, the released N1 peptide possesses neuroprotective properties (Guillot-Sestier et al. 2009), particularly by reducing Aβ oligomer-induced cell death (Guillot-Sestier et al. 2012). β-cleavage of PrPC produces the peptides N2 and C2, by cutting PrPC in the vicinity of residue 90 (Mange et al. 2004). β-cleavage has been originally detected in prion-infected brains (Chen et al. 1995) and has also been observed under oxidative stress conditions (McMahon et al. 2001). In addition, ADAM10 has been shown to shed PrPC off the membrane at the residue 229 (Taylor et al. 2009).

Studies on PrPC functions have uncovered a number of ligands with potential physiological significance (Zanata et al. 2002; Lopes et al. 2005; Caetano et al. 2008; Linden et al. 2008; Beland and Roucou 2012). These studies led to the hypothesis that PrPC works as an extracellular scaffold, organizing distinct synaptic components involved with neuronal signaling (Martins et al. 2002; Linden et al. 2008). We have found previously that some of the signaling activity of PrPC is mediated by group I metabotropic glutamate receptor family (mGluR1 and mGluR5) (Beraldo et al. 2011). This pathway can be modulated in AD because of changes in glutamate signaling (Li et al. 2009), particularly, due to Aβ activation of group 1 mGluRs (Renner et al. 2010).

The long-term consequences of increased Aβ peptides and its influence on PrPC and its signaling partners, such as mGluRs, are poorly understood. Here, we used APPswe/PS1dE9, a mouse model of AD in which PrPC has been previously shown to mediate AD-related learning and memory deficits (Gimbel et al. 2010), to investigate the consequences of familial AD mutations for PrPC processing. We found increased expression of mGluR1 and augmented proteolysis of PrPC in this mouse line. Our data suggest a possible link between Aβ, mGluR1, ADAM10 expression and PrPC processing. We propose that alterations in PrPC processing in this mouse line may be part of a compensatory response with potential to regulate how neurons respond to toxic Aβ peptide species.

Material and methods

Animals

All studies were conducted in accordance with the Canadian Council of Animal Care guidelines. The procedures were approved by the University of Western Ontario Institutional Animal Care and Use Committee (protocol #2008-127). All efforts were made to minimize suffering of animals. Male APPswe/PS1dE9 transgenic mice in the C57BL6/j background were purchased from Jackson Laboratory (Bar Harbor, ME, USA). These mice carry the human APP with Swedish mutation and the DeltaE9 mutation of the human presenilin 1 gene (Jankowsky et al. 2004). Wild-type (WT) littermates were used as controls.

Immunohistochemistry

Anesthetized APPswe/PS1dE9 and WT mice were transcardially perfused with 0.9% saline followed by 4% paraformaldehyde in phosphate-buffered saline. Brains were removed and post-fixed overnight. Coronal sections (50 μm) were cut through the brain using a vibratome. Polyclonal anti-human β-Amyloid antibody (Invitrogen, Camarillo, CA, USA) was used to determine amyloid plaques in free floating sections. Aβ was retrieved by boiling the sections in 10 mM sodium citrate buffer. Non-specific binding was blocked by incubation with 2% horse serum (Gibco, Carlsbad, CA, USA) and 2% bovine serum albumin in Tris-buffered saline (TBS) with 0.3% Triton X100. Sections were then processed for immunostaining by overnight incubation at 4°C in the primary antibody (1 : 200) diluted in TBS. After rinsing two times in TBS, sections were incubated for 2.5 h at 4°C with the secondary antibody (anti-rabbit Alexa Fluor 488, Invitrogen, Camarillo, CA, USA) diluted in TBS containing 1% horse serum and 1% bovine serum albumin, rinsed three times and mounted.

(1–42) ELISA

For Aβ(1–42) quantification we used APPswe/PS1dE9 mice by 2, 6, and 9 months of age. Total amyloid from mouse cortex was extracted as described previously (Lewis et al. 2004). Briefly, the tissue was homogenized in 10 volumes of 5 M guanidine chloride, 50 mM HEPES (pH 7.3), 5 mM EDTA with protease inhibitor cocktail (Calbiochem, San Diego, CA, USA) followed by rotation at 22°C for 3 h. Homogenate was diluted 10-fold into ice-cold TBS with protease inhibitor cocktail and cleared by centrifuging at 16 000 g for 20 min at 4°C. Aβ(1–42) was detected using Invitrogen ELISA kit for human Aβ(1–42) (cat#KHB3544, Invitrogen). Final values were normalized to the amount of loaded wet tissue.

qPCR

For mRNA analysis, cortical and hippocampal samples from 6- or 9-month-old mice were homogenized, frozen in a mixture of dry ice/ethanol and kept at −80 °C until use. For real-time quantitative PCR (qPCR), experiments were performed as previously described (Guzman et al. 2011; Martins-Silva et al. 2011). Briefly, total RNA was extracted using the Aurum Total RNA kit (Bio-Rad, Hercules, CA, USA). Quantity and quality of RNA were analyzed by microfluidic method with a Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA) . First-strand cDNA was synthesized using High Capacity cDNA transcription kit (Applied Biosystem, Foster City, CA, USA). cDNA was subsequently subjected to 40 cycles of qPCR on a CFX-96 Real-Time System using the iQ SYBR GREEN SUPERMIX (Bio-Rad). For each experiment, a non-template reaction was used as a negative control. Absence of DNA contaminants was confirmed in reverse transcription-negative samples and by melting curve analysis. Relative quantification of gene expression was done with the delta-delta-Ct method using β-actin gene expression to normalize the data. Sequences of primers used are available upon request.

Western blot analysis

Cortical and hippocampal samples were homogenized in ice-cold radioimmunoprecipitation assay buffer (50 mM Tris-HCl, pH 8, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, and 1% sodium dodecyl sulfate) with protease and phosphatase inhibitor cocktails. Samples were centrifuged at 16 000 g for 15 min. Protein concentration was evaluated by the Bradford method and 20–60 μg of protein were used for immunoblotting analysis. After sodium dodecyl sulfate–polyacrylamide gel electrophoresis [4–12% PAGEr® Gold Precast Gels (Lonza, Rockland, ME, USA) or 10% Tris-Tricine (prepared according to a standard protocol)] proteins were transferred onto polyvinylidene difluoride membranes. Immunoblotting was done using 8H4 (1 : 2500) (Abcam, Cambridge, MA, USA) , anti-ADAM10 (1 : 2000), anti-mGluR1 (1 : 2000), or anti-mGluR5 (1 : 2000) (Millipore, Billerica, MA, USA) antibodies.

PrPC processing

PrPC processing in mouse tissues was analyzed as described elsewhere (Walmsley et al. 2009). Namely, brain region extracts prepared in radioimmunoprecipitation assay buffer containing protease inhibitor cocktail were diluted in PNGase F buffer containing 0.1% sodium dodecyl sulfate and 50 mM 2-mercaptoethanol and heated at 95 °C for 5 min. After that, samples were incubated with PNGase F (Glyco, 1 mU per 30 μg of total protein) for 2 h at 37°C and then analyzed by immunoblotting with 8H4 antibody. C1 (18 kDa) and C2 (20-22 kDa) fragments were quantified by densitometry, relative to the total amount of PrPC.

Cortical neuronal cultures

Primary cultures of cortical neurons were obtained from 17-day-old mouse embryos as described previously (Vincent et al. 1996). Briefly, the cortex was first incubated with trypsin (0.25%) for 20 min at 37°C and mechanically dissociated with a pipette in Neurobasal medium (Invitrogen) supplemented with 10% of fetal bovine serum. Cells were then plated at a density of 2 × 106 cells per 35-mm dish pre-coated with poly-l-Lysine (10 mg/mL) (Sigma) and kept at 37°C and 5% CO2. Dihydroxyphenylglycine (DHPG, 50 μM), 1 μM Aβ oligomers- prepared as described previously (De Felice et al. 2009; Caetano et al. 2011)- and/or 100 μM LY 367385 (mGluR1 antagonist, Tocris Bioscience, Southampton, UK) were added to 8 DIV neurons. After 8–24 h of incubation, conditioned media and cells were separated and used for protein analyses.

Calcium signaling

Primary cortical neurons, obtained as described above, were loaded with 5 μM of intracellular Ca2+ probe Fluo-4 AM (Invitrogen) for 30 min at 37°C in neurobasal medium supplemented with 1 mM CaCl2. Cells were washed three times with HBSS (Invitrogen) and kept in Krebs buffer (124 mM NaCl, 4 mM KCl, 25 mM HEPES, 1.2 mM MgSO4, and 10 mM glucose) supplemented with 2 mM CaCl2. Data acquisition was performed using a confocal microscope LSM510 (Carl Zeiss Microscopy GmbH, Jena, Germany) with excitation at 488 nm and emission at 505–530 nm. Fluorescence was normalized as F1/F0 (F1, maximal fluorescence and F0, basal fluorescence). For each condition, three different neuronal cultures and 30–50 cells were analyzed.

Immunoprecipitation of secreted PrPC peptides

Conditioned media from cultured neurons was collected and cleared from cell fragments by centrifugation for 10 min at 1000 g at 4°C. After pre-clearing with protein G sepharose (10 μL per 1 mL of medium), conditioned media was incubated with SAF-32 antibodies (against N-terminal part of PrPC, 1 : 500, Cayman Chemical, Ann Arbor, MI, USA) and 10 μL of protein G sepharose overnight at 4°C. After spinning down, beads were washed with Tris-buffered saline with 0.1% Tween 20, heated in SDS–PAGE sample buffer and resolved in 10% Tris-Tricine SDS–PAGE. Proteins were analyzed by immunoblotting using SAF-32 antibodies (1:2000).

Statistical analysis

Statistical analyses were done using GraphPad Prism 5 (GraphPad, La Jolla, CA, USA). Results represent means ± SE of at least three experiments performed for each condition. Data were analyzed by one-way anova and Tukey's post hoc test or by Student's t-test as described in figure legends.

Results

APPswe/PS1dE9 mice have a progressive increase in Aβ(1–42) peptides followed by impairment in learning and memory (Savonenko et al. 2005; Reiserer et al. 2007). We used ELISA assays to quantify Aβ(1–42) in APPswe/PS1dE9 brain extracts from 2, 6, and 9-month-old mice and confirmed age-related Aβ(1–42) peptide accumulation, which seem to reach a steady state by 6 months (Fig. 1a). Using immuno labelling, we also confirmed the presence of amyloid plaques in cortex and dentate gyros of 9-month-old APPswe/PS1dE9. No immunoreactivity was observed in WT mice at the same age (Fig. 1b).

Figure 1.

Analysis of beta-amyloid content in mouse brain. (a) ELISA analysis of human Aβ(1–42) levels in the brain of APPswe/PS1dE9 mice. (b) Detection of Aβ(1–42)-positive plaques in the hippocampus and cortex of wild-type and APPswe/PS1dE9 mice as described in Methods. Results are presented as means + SE of three animals for each group analyzed and compared by one-way anova and Tukey's post hoc test. ***p < 0.001.

Next, we investigated PrPC expression and processing in APPswe/PS1dE9 mice. In previous experiments, expression of PrPC in this mouse model has been determined for the whole brain in 12 month-old mice (Gimbel et al. 2010). Here, we examined the hippocampus and cortex individually, the two brain regions that are most affected in this transgenic mouse line, at two different ages, 6 and 9 months. These ages were chosen based on accumulation of Aβ peptides in this mouse line. We found that in the hippocampus, PrPC mRNA and protein expression levels did not differ between APPswe/PS1dE9 and WT mice at 6 or 9 months (Fig. 2a–c). In the cortex, PrPC mRNA levels were also not changed in APPswe/PS1dE9 mice compared to WT controls (Fig. 3a). However, APPswe/PS1dE9 mice showed PrPC protein levels slightly increased at 9 months of age (p < 0.05) when compared to control mice (Fig. 3b and c).

Figure 2.

Cellular prion protein (PrPC) expression and processing in hippocampus from APPswe/PS1dE9 and wild-type mice. (a) PrPC mRNA expression. (b) Representative immunoblotting of PrPC expression and processing. (c) PrPC protein expression quantification. (d) PrPC α-cleavage quantification. (e) PrPC β-cleavage quantification. PrPC C2-peptide appeared as two bands around 22 kDa in Tris-Tricine sodium dodecyl sulfate–polyacrylamide gel electrophoresis gels. mRNA levels were quantified by qPCR and actin was used to normalize the data. Protein levels were quantified using actin as a loading control. Results are presented as means + SE of three to eight animals for each group compared by Student's t-test. *p < 0.05.

Figure 3.

Cellular prion protein (PrPC) expression and processing in cortex from APPswe/PS1dE9 and wild-type mice. (a) PrPC mRNA expression. (b) Representative immunoblotting of PrPC expression and processing. (c) PrPC protein expression quantification. (d) PrPC α-cleavage quantification. (e) PrPC β-cleavage quantification. mRNA levels were quantified by qPCR and actin was used to normalize the data. Protein levels were quantified using actin as a loading control. Results are presented as means + SE of three to eight animals for each group compared by Student's t-test. *p < 0.05, **p < 0.01.

We then compared PrPC processing in hippocampal and cortical tissues of wild type and APPswe/PS1dE9 mice. Hippocampal extracts showed no changes in α-processing and a slight decrease in PrPC β-processing in 6-month-old transgenic mice (p < 0.05, compared to WT), but this change was not observed consistently at 9 months (Fig. 2d and e). In contrast, PrPC cleavage products were significantly increased in cortical fractions of transgenic mouse brains (compared to WT), with a 2.5-fold increase in C1 levels observed both at 6 (p < 0.01) and at 9 (p < 0.05) months of age and a twofold increase in C2 levels (p < 0.05) at 9 months of age (Fig. 3d and e). It is worth noting that C2 was presented generally as a double band, while in a few previous studies a single smeared band at this size (~21 kDa) was observed (Chen et al. 1995; McMahon et al. 2001). β-cleavage, unlike α-cleavage, is more vaguely assigned to the vicinity of residue 90, enriched with basic residues, which may allow for production of PrPC peptides with different mobility. We assumed that these two bands reflect different sites of cleavage of PrPC products and therefore used both of them for C2 quantification. Increase in C2 levels has been proposed to indicate augmented oxidative stress (McMahon et al. 2001), which is likely to occur because of glutamatergic signaling abnormalities (Li et al. 2009; Renner et al. 2010). Importantly, our recent work has demonstrated that mGluR1 and mGluR5 can form a functional complex with PrPC that is involved in transducing extracellular activation of PrPC to intracellular signaling (Beraldo et al. 2011). These observations led us to analyze the expression of the PrPC partners, mGluR1 and mGluR5, to determine if they present age-related alterations in this mouse model. qPCR analysis showed that in the hippocampus, mGluR5 mRNA expression levels were decreased in 9-month old (p < 0.05), but not in 6-month-old APPswe/PS1dE9 mice when compared to controls (Fig. 4a). This decrease did not result in changes in mGluR5 hippocampal protein levels (Fig. 4b). We also analyzed tissues from 2-month-old mice, and detected no significant differences in mGluR5 mRNA or protein levels (data not shown). In the cortex, there was no significant difference in mGluR5 expression, both at the mRNA or protein levels when transgenic and control mice were investigated (Fig. 4c and d).

Figure 4.

mGluR5 expression in hippocampus and cortex from APPswe/PS1dE9 and wild-type mice. (a) mGluR5 mRNA expression in hippocampus. (b) mGluR5 protein expression in hippocampus. (c) mGluR5 mRNA expression in cortex. (d) mGluR5 protein expression in cortex. mRNA levels were quantified by qPCR and actin was used to normalize the data. Protein levels were quantified using actin as a loading control. Results are presented as means + SE of three to nine animals for each group compared by Student's t-test. *p < 0.05.

Expression levels of mGluR1 in the hippocampus (mRNA and protein) were not statistically different between APPswe/PS1dE9 and WT mice in both ages, although at 9 months we detected a clear trend for increased levels of mGluR1 protein (Fig. 5a and b). In contrast, in the cortex there was a robust increase in mGluR1 mRNA levels in 9-month-old APPswe/PS1dE9 mice compared to controls (p < 0.05), which was accompanied by a threefold increase (p < 0.01) in mGluR1 protein levels (Fig. 5c and d).

Figure 5.

mGluR1 expression in APPswe/PS1dE9 and wild-type mice. (a) mGluR1 mRNA expression in hippocampus. (b) mGluR1 protein expression in hippocampus. (c) mGluR1 mRNA expression in cortex. (d) mGluR1 protein expression in cortex. mRNA levels were quantified by qPCR and actin was used to normalize the data. Protein levels were quantified using actin as a loading control. Results are presented as means + SE of three to eight animals for each group compared by Student's t-test. *p < 0.05, **p < 0.01.

To start dissecting the mechanisms involved with increased PrPC processing, and to determine potential links between mGluR1 expression and PrPC processing, we used cultured cortical neurons. Aβ oligomers, which are thought to induce detrimental signals in neurons (Ferreira and Klein 2011), were used to reproduce the increased levels of Aβ peptides in mice. Eight-hour treatment of neurons with Aβ oligomers did not produce significant changes in PrPC processing or mGluR1 expression (not shown). In contrast, 24-h exposure to Aβ oligomers induced moderate, but significant, increases in both mGluR1 and C1 (Fig. 6a and b, p < 0.05). Interestingly, in the presence of mGluR1 antagonist LY 367385, Aβ oligomers failed to increase C1 levels, although mGluR1 expression was still increased, suggesting a potential link between α-processing of PrPC and Aβ modulation of mGluR activity. In addition, Aβ oligomer treatment increased expression of ADAM10 (p < 0.05), an enzyme that has been implicated in both shedding and α-processing of PrPC in cultured cortical neurons (Vincent et al. 2001; Altmeppen et al. 2011). Aβ oligomers can attack neurons by interacting with distinct receptors and synaptic proteins (Ferreira and Klein 2011), thus, in other to understand potential consequences of mGluR1 activity for PrPC processing we treated cortical neurons with DHPG, an mGluR1/5 agonist. Treatment with DHPG increased levels of intracellular Ca2+ and also augmented expression levels of ADAM10 (p < 0.05), but α-processing of PrPC was unaltered (Fig. 6c–e). This result suggests that increased α-processing of PrPC induced by Aβ oligomers may depend on signaling activities that add to mGluR activation. Noteworthy, PrPC shedding from neurons was increased robustly by DHPG (p < 0.05), which agrees with the increased levels of ADAM10 we observed. Of note, basal level of α-processing in neuronal cultures was one order of magnitude lower than that in adult mouse brains, which could contribute to our inability to detect small levels of PrPC α-processing in this condition.

Figure 6.

Effect of Aβ and dihydroxyphenylglycine (DHPG) treatments in embryonic cortical cultures. (a) Representative WB images for PNGase-treated Cellular prion protein (PrPC) (flPrPC for full-length PrPC, C1 for PrPC C1 peptide), mGluR1 and a disintegrin and metalloprotease (ADAM)10 in cultures treated with 1 μM Aβ and/or 100 μM LY 367385. (b) Quantitative analysis of (a). Results are presented as means + SE of at least four independent experiments analyzed with one-way anova and post hoc Tukey's test. *p < 0.05 (c) Effect of 50 μM DHPG treatment on intracellular calcium levels detected by fluo 4 dye. Thapsigargin (THG) was added in the end of each experiment as a positive control of intracellular calcium release. (d) Representative WB images for PNGase-treated PrPC, ADAM10, and shedded PrPC in cultures treated with 50 μM DHPG. (e) Quantitative analysis of (d). Results are presented as means + SE of at least five independent experiments analyzed with paired t-test. *p < 0.05.

ADAM10 has been previously found to be involved in numerous neuroprotective events including PrPC α-cleavage and shedding (Vincent et al. 2001), and APP α-cleavage (Lichtenthaler 2011). Therefore, we tested whether ADAM10 expression was altered in APPswe/PS1dE9 mice. At 6 months of age, there was no significant difference in the expression level of ADAM10 in both hippocampus and cortex of transgenic mice. However, in 9-month-old animals ADAM10 expression was consistently higher in APPswe/PS1dE9 cortices (p < 0.05, compared to controls) with a similar trend in hippocampal tissues (Fig. 7). Of note, ADAM10 protein expression pattern correlates with that of mGluR1 suggesting a potential relationship between mGluR1 increased expression and increased levels of ADAM10.

Figure 7.

A disintegrin and metalloprotease (ADAM)10 expression in hippocampus and cortex from APPswe/PS1dE9 and wild-type mice. ADAM10 in hippocampal (a) and cortical (b) extracts. Protein levels were quantified using actin as a loading control. Results are presented as means + SE of three to nine animals for each group compared by Student's t-test. *p < 0.05.

Our experiments reveal novel and complex processing of PrPC in connection with mGluR over-expression that seems to be triggered by β-amyloid peptides.

Discussion

The results presented in this study reveal novel forms of regulation of PrPC, and some of its associated proteins, in APPswe/PS1dE9 mice The complex processing of PrPC presented here seems to be triggered by β-amyloid peptides and has the potential to mediate, at least in part, how Aβ peptides influence neuronal function. PrPC is present in cells both as full-length glycosylphosphatidylinositol-anchored 209-amino acid protein, and also as N-terminally truncated forms. Processing of PrPC occurs constitutively in the later secretory pathway by three pathways: 1. α-cleavage at the PrPC residue 111. 2. shedding at the PrPC C-terminus. 3. reactive oxigen species-mediated β-cleavage in the vicinity of residue 90 (Mange et al. 2004). During α-cleavage, the N-terminal peptide N1 is released from PrPC and the C1 peptide that remains at the cell membrane lacks Aβ oligomer binding sites. Hence, the increase in α-processing, with generation of N1 and C1 peptides that we observed in APPswe/PS1dE9 mice suggests the possibility of age and Aβ-dependent changes in the amount of prion protein capable of transducing Aβ signaling. Importantly, recent work indicated that C1 may protect mice against prion diseases, by acting as a dominant-negative PrPC isoform (Westergard et al. 2011).

β-cleavage of PrPC was first detected by the presence of its C-terminal product C2 in prion-infected brains (Chen et al. 1995) and subsequently this process was associated with reactive oxygen species (McMahon et al. 2001). C2 formation is probably related to changes in lysosomal activity (Dron et al. 2010), which for amyloid-related diseases may be connected with failure of lysosomes to clear aggregated proteins (Lindner and Demarez 2009). Thus, the increased accumulation of C2 fragments observed at 9 months in the cortex of APPswe/PS1dE9 mice may reflect distinct processed such as lysosomal failure to clear Aβ aggregates and/or increased reactive oxygen species induced by stress, which is a common feature in AD mouse models (Garcia-Alloza et al. 2009). Hence, increased C2 levels are not a specific feature of AD or prion diseases and may reflect neuronal stress levels. Notably, the C2 fragment of PrPC preserves one of the Aβ oligomer binding sites, therefore up-regulation of C2 production may not change Aβ-induced signaling. Mice expressing the truncated prion protein PrP90–231, corresponding to the C2 fragment, showed increased susceptibility to scrapie infection, with shorter lag-phase after inoculation (Fischer et al. 1996). Consequently, higher levels of C2 peptide might have a role in the increased susceptibility of AD mouse models with Swedish APP mutation to prion infection (Morales et al. 2010).

Increased glutamate signaling has been observed under AD-related conditions, because of inhibition of glutamate uptake by Aβ oligomers (Li et al. 2009). Interestingly, glutamate exposure-related oxidative stress have also been shown to specifically up-regulate the production of mGluR1 in cortical neuronal cultures (Sagara and Schubert 1998). In addition to finding that APPswe/PS1dE9 mice have age-dependent increased levels of mGluR1, we also found that treatment of cortical cultures with Aβ oligomers for 24 h increased mGluR1 expression, suggesting a potential mechanism by which mGluR1 may be increased in this mouse model. The increased levels of mGluR1 we detected contrast with results obtained in studies investigating human AD brains, which showed down-regulation of mGluR signaling pathway (Albasanz et al. 2005; Parameshwaran et al. 2008). The difference observed may be because of the fact that AD mouse models reflect early pre-clinical phases of the disease (Ashe and Zahs 2010; Ferretti et al. 2011), while the human brains investigated likely represent an advanced disease phase.

Although the molecular mechanisms responsible for the selective up-regulation of mGluR1 are not understood, dysregulation of this signaling pathway has the potential to cause several distinct outcomes. For example, mGluRs are involved in regulating synaptic plasticity (Luscher and Huber 2010; Cosgrove et al. 2011), which is affected in AD. Particular roles of group I mGluRs in AD are somewhat controversial as they have been shown to have both excitotoxic and neuroprotective effects under stress conditions. mGluR1 excitotoxicity occurs under post-ischemic condition (Pellegrini-Giampietro 2003). Abnormal activation of group I and II mGluRs upon Aβ-mediated inhibition of glutamate transporters seems to increase LTD (Li et al. 2009). On the other hand, activation of group I mGluRs seems to protect neurons from a number of stress factors, including increases in nitric oxide (Vincent and Maiese 2000), hydrogen peroxide or platelet-activating factor (Zhu et al. 2004), glucose starvation, and cystine deprivation (Sagara and Schubert 1998; Zhu et al. 2004), as well as NMDA-induced excitotoxicity (Blaabjerg et al. 2003). Here, we found that blocking mGluR1 activity prevents C1 increase in cortical neurons treated with Aβ oligomers, arguing for a possible role of this receptor in modulating PrPC α-cleavage. Our previous experiments have indicated that mGluR1 interacts with PrPC (Beraldo et al. 2011); hence, this receptor may influence PrPC processing either directly or through signaling. Indeed, it is possible that mGluR1 and PrPC could mediate some of the effects of Aβ oligomers as a protein complex.

Remarkably, in cultured cortical neurons, activation of mGluRs seems to cause up-regulation of ADAM10. This enzyme participates in APP processing, but also it can modulate PrPC processing. Increased levels of ADAM10 correlated with increased PrPC shedding in neuronal cultures, which agrees with recent studies showing that ADAM10 may be responsible for shedding, but it is not required for α-processing of PrPC (Altmeppen et al. 2011). In agreement with neuronal culture experiments, APPswe/PS1dE9 mice also showed an increase in ADAM10 levels. Although the mechanism of ADAM10 action in PrPC processing is still unknown, PrPC (Fevrier et al. 2004) and ADAM10 are secreted by exosomes and ADAM10 is able to cleave proteins inside these organelles (Gutwein et al. 2005; Stoeck et al. 2006). Whether there is contribution of exosomal release of PrPC fragments remains to be determined. It would be important to investigate whether shedding or processing of PrPC is also increased in early phases of disease in individuals affected with familial AD.

In summary, we showed that the APPswe/PS1dE9 mouse model of AD exhibits altered PrPC processing and a selective increase in cortical mGluR1 expression as well as up-regulated levels of ADAM10. These changes may reflect a compensatory mechanism that could regulate for Aβ-mediated toxicity in the cortex of APPswe/PS1dE9 AD mouse model.

Acknowledgements

The authors declare no conflict of interest. This study was supported by The Alzheimer's Association (USA), CIHR (MOP 93651 and 89919), Canadian Foundation for Innovation (CFI), Ontario Research Fund (ORF) and PrioNet-Canada.

Ancillary