Alzheimer's disease (AD) is the most common cause of dementia which leads to a marked impairment of cognition and learning and contributes to a reduced lifespan (Blennow et al. 2006; Katzman 1986; Selkoe 2002). AD is pathologically characterized by the progressive accumulation of amyloid plaques in the brain and blood vessels and neurofibrillary tangles in neurons (Glenner and Wong 1984). Amyloid-β protein (Aβ) is generated by sequential cleavage of amyloid precursor protein (APP) by β- and γ-secretase in subcellular compartments. The trans-Golgi network and endosomes have been identified as major sites for β-secretase activity on APP processing (Skovronsky et al. 2000). Aβ interacts with and activates several membrane receptors, including TrkA and p75NTR (Bulbarelli et al. 2009), APP itself (Verdier et al. 2004), advanced glycation end products receptors (Yan et al. 2009), insulin receptor (Townsend et al. 2007), acetylcholine receptors(Parri and Dineley 2010), contributing to the pathogenesis of AD potentially through MAPK, NF-ĸB, Jun N-terminal kinase (JNK), and other pathways. Aβ up-regulates the expression of the p75NTR through insulin-like growth factor 1 receptor (IGF-1R) phosphorylation in SH-SY5Y human neuroblastoma cells and AD transgenic mice (Chakravarthy et al. 2010, 2012; Ito et al. 2012). In our previous studies, we also observed that the expression of p75NTR in the brain of swedish-amyloid precursor protein (APPswedish) transgenic mice is increased, illustrating Aβ as a crucial factor in the regulation of p75NTR over-expression in Alzheimer's disease (Wang et al. 2011). It has been reported that Aβ up-regulates the expression of β-site APP-cleaving enzyme 1 (BACE1) through JNK pathway (Guglielmotto et al. 2011). Sortilin is a Golgi sorting protein and a member of the family of vacuolar protein sorting 10 protein (VPS10P) domain receptors and the co-receptor for proneurotrophins in different cells (Nykjaer et al. 2004; Nakamura et al. 2007; Jansen et al. 2007). Sortilin plays an important role in the regulation of neuronal viability and function (Petersen et al. 1997; Nykjaer and Willnow 2012) and is expressed in neurons of the CNS and PNS, and also in non-neural tissues like liver and fat (Petersen et al. 1997; Morris et al. 1998; Sarret et al. 2003; Kjolby et al. 2010). As a process of aging, proNGF and p75NTR are up-regulated and activate the sortilin receptor to promote neuronal cell death (Nykjaer et al. 2004; Jansen et al. 2007). Sortilin interacts with BACE1 and regulates the trafficking of BACE1, the enzyme responsible for the amyloidergic cleavage of APP, and results in overproduction of Aβ (Finan et al. 2011). Surprisingly, sortilin also directly interacts with APP (Gustafsen et al. 2013) and negatively regulates APP lipid raft targeting (Yang et al. 2013). Furthermore, sortilin interacts with p75NTR to induce neuronal apoptosis and increase AD severity (Skeldal et al. 2012). However, the role of sortilin in the pathogenesis of AD remains unclear and controversial, thus this study has sort to understand its role in AD development.
Sortilin, a Golgi sorting protein and a member of the VPS10P family, is the co-receptor for proneurotrophins, regulates protein trafficking, targets proteins to lysosomes, and regulates low density lipoprotein metabolism. The aim of this study was to investigate the expression and regulation of sortilin in Alzheimer's disease (AD). A significantly increased level of sortilin was found in human AD brain and in the brains of 6-month-old swedish-amyloid precursor protein/PS1dE9 transgenic mice. Aβ42 enhanced the protein and mRNA expression levels of sortilin in a dose- and time-dependent manner in SH-SY5Y cells, but had no effect on sorLA. In addition, proBDNF also significantly increased the protein and mRNA expression of sortilin in these cells. The recombinant extracellular domain of p75NTR (P75ECD-FC), or the antibody against the extracellular domain of p75NTR, blocked the up-regulation of sortilin induced by Amyloid-β protein (Aβ), suggesting that Aβ42 increased the expression level of sortilin and mRNA in SH-SY5Y via the p75NTR receptor. Inhibition of ROCK, but not Jun N-terminal kinase, suppressed constitutive and Aβ42-induced expression of sortilin. In conclusion, this study shows that sortilin expression is increased in the AD brain in human and mice and that Aβ42 oligomer increases sortilin gene and protein expression through p75NTR and RhoA signaling pathways, suggesting a potential physiological interaction of Aβ42 and sortilin in Alzheimer's disease.
Sortilin is the co-receptor of p75NTR which signals the cell death induced by Aβ and proneurotrophins. We found that sortilin is increased in the AD brain and up-regulated by Aβ and pro-brain-derived neurotrophic factor (proBDNF). Aβ-induced upregulation of sortilin is mediated by p75NTR and the down-streaming RhoA-ROCK signaling pathway. The Aβ/Sortilinp/75NTR signaling may play a role in the pathogenesis of AD.
amyloid precursor protein
β-site APP-cleaving enzyme 1
brain-derived neurotrophic factor
methyl thiazol tetrazolium bromide
phosphate buffer saline
radio immune precipitation assay
Materials and methods
Cell culture and treatment
SH-SY5Y and SH-SY5YAPP695 human neuroblastoma cells were provided by Prof. Richard Lewis (Institute of Molecular Biosciences, University of Queensland, Australia) and Prof. Nigel Hooper (Institute of Molecular & Cellular Biology, University of Leeds). Cells were grown in Dulbecco's modified Eagle's medium (Invitrogen, Mulgrave, Australia) supplemented with 10% fetal bovine serum and 2 mM l-glutamine and 1% penicillin/streptomycin and incubated at 37°C in a humidified atmosphere of 95% air and 5% CO2. After overnight seeding of SH-SY5Y cells in 6-well plates (106/well) (Invitrogen), the SH-SY5Y cells were treated with varying doses (0.1, 0.25, 0.5, 1, 2.5, 5, 10 μM) of oligomer form of Aβ42 and scrambled Aβ42 over a time-course (3, 6, 12, 24 h). To see whether sortilin up-regulation occurs through the p75NTR receptor, SH-SY5Y cells were incubated with 10 μg/mL human p75ECD-FC recombinant peptide (Wang et al. 2011) and 10 μg/mL rabbit polyclonal anti-human p75NTR-ECD antibody (generous gift from Prof. Moses Chao, Department of Cell Biology, Skirball Institute, New York, USA) in conjunction with 1 μM Aβ42 for 24 h. Human and rabbit IgG (Sigma-Aldrich, St Louis, MO, USA) were used as negative control. To quantify the level of sortilin expression in presence of mature and premature form of neurotrophoin, SH-SY5Y cells were incubated with 50 ng/mL of proBDNF (Virovek, USA) a mutant recombinant protein which is resistant against proBDNF processing enzymes (Fan et al. 2008) and brain-derived neurotrophic factor (BDNF) (Amgen, Thousand Oaks, CA, USA) for 24 h. To investigate which signaling pathway plays a role in up-regulation of sortilin by Aβ42, ROCK inhibitor (5, 10 μM; Y27632; Sigma-Aldrich), and JNK inhibitor (2 μM; SP600125; Sigma-Aldrich) (Guglielmotto et al. 2011) were used to block the RhoA and JNK signaling pathways, respectively.
Preparation of oligomer form of Aβ42
In this study, the oligomer form of Aβ42 was used. Synthetic Aβ42 and scrambled Aβ42 were purchased from American Peptide (Sunnyvale, CA, USA) and prepared following the protocols described previously. In brief, the Aβ42 peptide was dissolved in 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP, Sigma-Aldrich) at 1 mg/mL concentration, and aliquoted. The HFIP was allowed to evaporate in the fume hood and the resulting clear peptide film was dried under vacuum overnight and pellets stored at −20°C until use. For oligomerization of Aβ42, dried pellet was re-suspended in Dulbecco's modified Eagle's medium to a final concentration of 100 μM and incubated at 4°C for 24 h. The oligomer form of Aβ42 was tested by western blot before use (Dahlgren et al. 2002; Wang et al. 2011).
Transgenic AD mice
Double APPswe/PS1dE9 transgenic mice (2xTg-AD) on C57BL/6j background were obtained from The Jackson Laboratory. These transgenic mice bear a chimeric mouse/human (Mo/Hu) APP695 with mutations linked to familial AD (KM 593/594 NL) and human PS1 carrying the exon-9-deleted variant associated with familial AD (PS1dE9) in one locus under the control of a prion promoter element (Jankowsky et al. 2001). Before use, genotyping of APPswe/PS1dE9 transgenic mice was performed by PCR. The animals were kept under standardized barrier breeding conditions with free access to water and food. All procedures involving animals were approved by the Animal Welfare Committee of the University of South Australia and undertaken according to the guidelines of the National Health and Medical Research Council of Australia. The use of genetically modified animals was approved by the Biosafety Committee of the University of South Australia.
The brain tissue of human
Human brain tissue was obtained from the South Australia Brain Bank at Flinders University. Post-mortem brain tissue from four normal female subjects (mean age 85.6 years) and four female patients (mean age 81 years) neuropathologically diagnosed with AD was snap-frozen and stored at −80°C. Temporal cortical samples were homogenized in a radio immune precipitation assay (RIPA) buffer containing 2 mM phenyl methane sulfonyl fluoride and protease inhibitors (Roche). Homogenate was centrifuged at 2500 g for 10 min at 4°C. The protein concentration of the lysates was determined using BCA protein assay kit (Thermo Scientific, Rockford, USA). Supernatants containing 30 μg of protein were subjected to SDS-PAGE for western blot analysis. All procedures involving human subjects/patients were approved by the Flinders Clinical Research Ethics Committee.
Protein extraction from cell line and the brain tissue of mice
Total protein from cell lines and brain tissue of mice brain was obtained to subject to western blotting and real-time PCR. Briefly, culture medium was removed and cells were gently washed twice with chilled phosphate buffer saline (PBS1X) and lysed by suspending in the RIPA buffer (50 mM Tris, 150 mM NaCl, 1 mM EDTA, 0.5% Triton X-100, 0.5% Sodium deoxycholate, pH7.5) containing a protease inhibitor cocktail. 6-month-old APPSwe/PS1dE9 and control mice were killed by overdosing with pentobarbital and the cortex of brain was extracted and homogenised in RIPA buffer containing protease inhibitors cocktail. The cells and brain lysates were sonicated and then centrifuged at 16 000 g for 10 min at 4°C and total protein concentration of supernatants were determined using BCA protein assay kit (Thermo Scientific, Rockford, USA). All lysates were stored at −80°C until use.
Western blot analysis
Protein extracts (30 μg) from cell lines and the brain of mice and human were separated by 4–12% precast SDS-PAGE gels (Bio-Rad, USA) which were run at 80–120 V using Tris-glycine running buffer and transferred to nitrocellulose membrane (0.2 mm pore size, GE Healthcare, Uppsala, CA, USA). Membrane was then blocked for 1 h at 25°C in PBS containing 5% skim milk and incubated with rabbit anti-sortilin (Osenses, Adelaide, Australia), rabbit polyclonal anti-sortilin (Abcam, Cambridge, MA, USA), sheep anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Osenses, Australia), and mouse monoclonal anti β-actin (Sigma-Aldrich), mouse monoclonal anti-Aβ42 (6E10) (Covance, Dedham, MA, USA), mouse monoclonal anti-RhoA-GTP (New-East Biosciences, Malvern, PA, USA) and rabbit polyclonal anti- total Rho (New-East Biosciences) in 2% skim milk in PBS for overnight, respectively, followed by incubation with the horseradish peroxidase-linked secondary antibodies (Sigma-Aldrich) in 1 : 3000 in 2% skim milk in phosphate buffered saline tween-20 for 1 h at 25°C. Membranes were developed by Image Quant LAS 4000 (GE Healthcare Bio Science AB, Uppsala, CA, USA) and Image J software (Research Service Branch; National Institute of Health, http://rsbweb.nih.gov/ij/index.html) was used for quantitative analysis.
Real-time quantitative RT-PCR
SH-SY5Y and SH-SY5YAPP695 were seeded in 6-well plates (106/well) (Invitrogen) and after treatment, culture medium was removed and cells were gently washed twice with chilled phosphate buffer saline (PBS) and cells lysed in lysis buffer (QIAGEN, Doncaster, Australia). Total RNA (400–600 ng/well) was extracted using an optimized RNA extraction protocol based on the RNeasy Mini Kit Isolation System (Qiagen Ltd.) according to the manufacturer's protocol and immediately afterward, cDNA was synthesized using iScripTM cDNA Synthesis Kit (Bio-Rad Hercules, CA, USA) according to the manufacturer's instructions. As a negative control, a cDNA synthesis reaction was also carried out without the addition of reverse transcriptase. All PCR primers (Table 1) were obtained from GeneWorks (Hindmarsh, Australia) and were designed using human DNA sequences and NCBI PRIMER BLAST software (Bethesda, MD, USA), whereby forward and reverse primers were designed in different exons, 200–300 bp in length, CG content 40–50%, and Tm 2 at 55–60. For quantitative RT-PCR of genes of interest, a reaction volume of 10 mL was found adequate using iTaq™ Universal SYBR® Green supermix (Bio-rad). For each cDNA sample, quantitative PCR assays were run on a CFX Connect Real-Time PCR System (Bio-Rad) in duplicate. From the amplification curves, relative expression was calculated using the comparative Ct (2−ΔCt) method, with GAPDH serving as the endogenous control.
|Gene||Forward primer||Reverse primer|
Cell viability assay
The vitality of the SH-SY5Y cell line treated with Aβ42 oligomer of was assessed by Methyl Thiazoly Blue Tetrazolium Bromide (MTT) Assay kit (Sigma) following the manufacturer's instructions. Briefly, the SH-SY5Y cells were cultured at a density 2 × 104 cells per well in growth medium for 12 h in 96-well plates, and then exposed to different concentrations of Aβ42 oligomer (0.1, 0.25, 0.5, 1, 2.5, 5, 10 μM) for 24 h. 10 μL of MTT solution (5 mg/mL) was added to each well and incubated at 37°C for 4 h. Absorbance was recorded at 490 nm with a micro-plate reader WALLAC 1420 (PerkinElmer, Waltham, MA, USA). In all cases, test substances were added containing medium alone to determine whether they interfered with the assay. The data are expressed as mean percent viable cells compared to respective control cultures.
All statistical analyses were performed with Statistical Package for the Social Sciences (spss) software (version 18.0; IBM, Armonk, NY, USA). Data are expressed as mean ± SEM and significance was achieved when p < 0.05. Variables between groups were determined by student's t-test and one-way anova. When significance (p < 0.05) was achieved, a post hoc analysis of groups was performed using a Tukey's post test.
Sortilin expression is up-regulated in brain of human AD and APPswe/PS1dE9 mice
As sortilin interacts with APP and BACE1, it is likely that sortilin plays an important role in AD. To examine functions of sortilin in AD, we first investigated whether the level of sortilin expression in AD brain and in the brain of APPswe/PS1dE9 transgenic mice is altered. A pilot study of n = 4 AD brain and n = 4 controls was performed and it illustrates that the expression of sortilin in human post-mortem AD brain cortex was significantly increased compared with control brain (p < 0.05) (Fig. 1a and b). We also examined the level of sortilin in the brain of 6-month-old double APPswe/PS1dE9 transgenic (2xTg-AD) mice. These results show that sortilin protein in the brain tissue of APPswe/PS1dE9 mice was increased compared with C57BL/6j (WT) (**p < 0.01) (Fig. 1c and d).
The expression level of sortilin in SH-SY5YAPP695 is higher than in SH-SY5Y
To further examine sortilin expression in AD models, we tested whether SH-SY5Y human neuroblastoma cells express sortilin and that they are a suitable model to investigate the effects of Aβ42 on sortilin expression in an in vitro study. 30 μg/well of total SH-SY5Y cell protein of was used on a western blot and sortilin protein was detected by rabbit anti-sortilin antibody (Fig. 2a). Recent studies have highlighted the potential importance of soluble oligomeric forms of Aβ42 in AD (Walsh and Selkoe 2007; Reddy et al. 2010; Glabe 2008; Kim et al. 2003). As Aβ42 is more likely to form oligomers than Aβ40, and is more critical in AD (Iwatsubo et al. 1994; Jarrett et al. 1993), we used an Aβ42 oligomer in this study for any further experiments. To check the oligomer formation after preparation and before use, 0.2 μg of Aβ42 oligomer was subjected to western blot and detected by anti-Aβ antibody (6E10) (Fig. 2b). To determine whether there is any difference in the expression level of sortilin mRNA, and protein in SH-SY5Y and SH-SY5YAPP695, we applied quantitative real-time PCR and western blot to analyze, respectively, mRNA and protein level of sortilin in normal and transgenic cell line of human neuroblastoma cells. We found a significant increase in mRNA and protein levels of sortilin in SH-SY5YAPP695 compared with SH-SY5Y (p < 0.05, p < 0.01; Fig. 2c–e).
A dose-dependent effect of Aβ42 oligomers on sortilin expression in SH-SY5Y cells
On the basis of the data above, we hypothesized that Aβ42 may regulate the expression of sortilin. To test the hypothesis, we treated SH-SY5Y cells with different concentrations of Aβ42 oligomers for 24 h and examined the gene and protein expression of sortilin. Aβ42 oligomer up-regulated the expression of sortilin in SH-SY5Y human neuroblastoma in a dose-dependent manner (p < 0.05, p < 0.01; Fig. 3a and b) up to 1 μM. Quantitative real-time PCR showed that Aβ42 oligomer enhanced the mRNA level of sortilin (p < 0.05; Fig. 3c).
MTT assay was conducted to check cell viability in the presence of different concentrations of Aβ42, and observed a dose-dependent toxicity of Aβ42 on cell viability in SH-SY5Y cells (p < 0.05; Fig. 3d). Based on this finding, 1 μM of Aβ42 was found to be optimal and was used in the following experiments described herein.
As Aβ mediates neurotoxicity via the p75NTR receptor (Coulson et al. 2009), it is likely that other p75NTR ligands also regulate sortilin expression. To understand whether neurotrophins and proneurotrophins alter the protein and gene expression of sortilin, SH-SY5Y cells were incubated with 50 ng/mL of proBDNF and BDNF for 24 h for western blot and quantitative real-time PCR (Fig. 3e–g). The results showed that proBDNF but not mature BDNF significantly elevated the levels of both protein (Fig. 3e–f) and mRNA (Fig. 3g) of sortilin.
Time-dependent effect of Aβ42 oligomers on sortilin expression in SH-SY5Y cells
To test the effect of Aβ42 time-dependent sortilin gene and protein expression, SH-SY5Y cells incubated with 1 μM Aβ42 over a time course of 3, 6, 12, 24 h and the cell lysate was subjected to western blot and real-time PCR for quantification of sortilin protein and gene expression, respectively. Interestingly, 12 and 24 h incubation with 1 μM Aβ42 significantly up-regulated the protein and gene expression of sortilin in SH-SY5Y cells, however, a dramatic over-expression of sortilin mRNA was observed after 6 h incubation with Aβ42 (p < 0.05, p < 0.01; Fig. 4a–c). As sorLA is another member of the VPS10P family of receptors and has been linked to the development of AD (Rogaeva et al. 2007; Gustafsen et al. 2013), we investigated whether Aβ42 enhances the gene expression of sorLA in SH-SY5Y. These cells were incubated with 1 μM Aβ42 for 3, 6, 12, 24 h, but Aβ42 oligomers did not elevate the gene expression of sorLA (Fig. 4d).
Aβ42 enhances sortilin expression via p75NTR receptor in SH-SY5Y cells
To elucidate whether p75NTR is involved in Aβ-induced sortilin over-expression, we blocked p75NTR with 10 μg/mL p75ECD-FC recombinant peptide or 10 μg/mL anti-p75ECD antibody for 24 h in conjunction with 1 μM Aβ42. Interestingly, Aβ42-induced sortilin protein and mRNA over-expression appears to occur via the p75NTR receptor, as the p75ECD-FC and anti-p75ECD blocked the gene and protein up-regulation of sortilin in the presence of Aβ42 (p < 0.05, p < 0.01; Fig. 5a–c).
Aβ42 up-regulates sortilin expression through the RhoA signaling pathway in SH-SY5Y cells
Several signaling pathways are involved in the signal transduction of p75NTR. Aβ and proneurotrophins can activate JNK and RhoA via the p75NTR receptor (Sun et al. 2012; Skeldal et al. 2011; Yao et al. 2005). Next, we asked whether RhoA is involved in sortilin over-expression by Aβ42. Y27632 (a well characterized ROCK inhibitor) was used to inhibit ROCK. When 5 and 10 μM Y27632 were incubated with SH-SY5Y cells for 24 h, 10 μM Y27632 resulted in a significant down-regulation of endogenous sortilin protein (p < 0.05; Fig. 6 a and b). This indicates that RhoA is involved in sortilin regulation, however, to elucidate whether Y27632 reduces Aβ-induced sortilin up-regulation, SH-SY5Y cells were treated with 1 μM Aβ42 and 10 Y27632 for 24 h. It appeared that Y27632 suppressed sortilin expression induced by Aβ42 (p < 0.05, p < 0.01; Fig. 6c and d). We also observed that 1 μM Aβ42 significantly increased the ratio of active RhoA-GTP relative to total Rho (p < 0.05; Fig. 6e and f).
Aβ42 does not activate the over-expression of sortilin receptor through JNK signaling pathway in SH-SY5Y cells
The c-JNK signaling pathway is one potential mechanism regulating the transcriptional activity of Aβ42 and mediates the up-regulation of essential proteins such as BACE1 in AD (Yao et al. 2005; Kadowaki et al. 2005; Ma et al. 2009; Ramin et al. 2011; Guglielmotto et al. 2011). Therefore, we speculated that JNK may play a role in the regulation of sortilin expression in the presence of Aβ42. To investigate this, a JNK inhibitor (SP600125, 2 μM) (Guglielmotto et al. 2011) was incubated with SH-SY5Y cells for 24 h and the cell lysate collected to assess sortilin protein expression. SP600125 did not alter the endogenous expression of sortilin (Fig. 7a and b). We also tested if SP600125 prevents sortilin up-regulation in the presence of Aβ42. Thus, SH-SY5Y cells were incubated together with 1 μM Aβ42 and 2 μM SP600125 for 24 h, however, SP600125 did not block the over-expression of sortilin protein in the presence of Aβ42 (Fig. 7c and d).
Our study has uncovered a novel functional effect of Aβ42 on sortilin expression in AD. We demonstrated that Aβ42 stimulates the expression of the sortilin receptor in SH-SY5Y cells and that the level of sortilin expression in human AD brain is higher than controls. Protein expression data indicate that sortilin is also up-regulated in the brain of 6-month-old double APPswe/PS1dE9 transgenic mice when compared with aged-matched wild type controls. We also found that Aβ42 activates sortilin expression through the p75NTR/ RhoA signaling pathway.
The mammalian Vps10p sorting receptors, consisting of five members of type-I transmembrane protein (Sortilin, SorLA, and SORCS1-3), may play differential roles in the pathogenesis of AD. They are involved in mediating a variety of intracellular sorting and trafficking functions between endosome and trans-Golgi network compartments and subsequently increase the progress of neurodegenerative diseases like AD (Nyborg et al. 2006; Lane et al. 2012). For example, sorLA and SORCS1 mediate retrograde trafficking of APP from the cell surface to Golgi compartments and reduce the amyloidergic processing of APP and production of Aβ and decrease the risk of AD (Nyborg et al. 2006; Gustafsen et al. 2013; Rogaeva et al. 2007). The function of sorLA and SORCS1 on APP processing occurs via their interactions with the retromer complex (Small et al. 2005; Lane et al. 2010; Fjorback et al. 2012). Sortilin, on the other hand, plays an important role in the regulation of neuronal viability and function and it interacts with p75NTR and proNGF(Nykjaer et al. 2004) and is required for the toxic action of Aβ oligomers (Takamura et al. 2012). The functions of sortilin in APP trafficking and Aβ production are not fully understood. Sortilin interacts with BACE1, regulates its retrograde trafficking and accelerates production of Aβ42. On the other hand, sortilin binds APP, regulates dendritic and axonal targeting of APP and mediates the processing of APP toward the non-amyloidergic pathway and increases the production of sAPPα (Gustafsen et al. 2013). We recently found that sortilin interacts with APP in a head to head and tail to tail manner, in which the FLVHRY motif of sortilin interacts with the APP NPTYKFFE motif, regulating lysosomal targeting of APP and reducing production of Aβ (Yang et al. 2013).
It has been shown that the high-affinity, pro-survival nerve growth factor receptor TrkA is reduced in AD (Mufson et al. 2010; Finan et al. 2011). As sortilin interacts with APP and BACE1, it is thought that sortilin plays an important role in the progression of AD. In support of this hypothesis, we first found that the level of sortilin expression in AD brain and in the brain of 6-month-old double APPswe/PS1dE9 transgenic mice to be higher than age-matched normal brain of human and wild type mice. The finding from human AD brain, while being from a small sample size, is consistent with the results from the transgenic mice and AD cell line models. In particular, we demonstrated that Aβ42 oligomer caused up-regulation of the sortilin receptor at the gene and protein level. Interestingly, proBDNF, but not mature BDNF, also stimulates gene expression of sortilin. It appears that the role in sortilin expression is specific, as sorLA gene expression is not altered in the presence of Aβ42.
Oligomeric forms of Aβ42 play a critical role in progress of AD as Aβ42 is more likely to form oligomers rather than Aβ40 (Walsh and Selkoe 2007; Reddy et al. 2010; Glabe 2008; Kim et al. 2003). The oligomers disrupt cognitive functions in the brain and potently inhibit long-term potentiation, enhance long-term depression and reduce dendritic spine density in normal rodent hippocampus (Chromy et al. 2003; Cleary et al. 2005; Lacor et al. 2004). Interestingly Aβ oligomers activate differential gene expression in the human brain (Sebollela et al. 2012). For example, it has been shown that Aβ42 up-regulates the expression of p75NTR neurotrophin receptor in SH-SY5Y human neuroblastoma cells and AD transgenic mice through insulin-like growth factor 1 (Chakravarthy et al. 2010, 2012; Ito et al. 2012). In our previous study, we reported that the expression of p75NTR in APPswe transgenic mouse brain is higher than wild type (Wang et al. 2011). It is also suggested that Aβ42 triggers upregulation of BACE1 through the JNK Pathway and induces physiological transcriptional activation of BACE1 (Guglielmotto et al. 2011; Piccini et al. 2012). It is also known that Aβ can activate RhoA causing neurite collapse (Petratos et al. 2008), but the mediating receptor is not known. Elucidation of the receptor pathway involved in upregulation of sortilin in the presence of Aβ42 is crucial for understanding the significance of this biological phenomenon. Previous studies have shown that some gene expression is predominantly regulated by proneurotrophins via p75NTR (Nakamura et al. 2007; Mufson et al. 2010). In addition, Aβ can bind p75NTR mediating the apoptosis and degeneration of cholinergic neurons (Coulson et al. 2009). Therefore, we investigated whether Aβ42 up-regulates sortilin via p75NTR. Three pieces of evidence support the notion that p75NTR mediates the gene and protein upregulation of sortilin in the presence of Aβ42. First, the diffusible extracellular domain of p75NTR can block the up-regulation of sortilin induced by Aβ. Secondly, the antibody to the extracellular domain of p75NTR also blocks the response. Thirdly, an alternate p75NTR /sortilin ligand proBDNF, but not p75NTR/trkB ligand mature BDNF triggers the response. These data indicate that neurodegenerative ligands of p75NTR/sortilin receptors up-regulate the expression of sortilin. It is likely that such a regulatory mechanism operates in vivo in the brain of AD mice and human patients as sortilin is clearly increased in brains of AD mice. What are the down-stream signals of p75NTR responsible for the up-regulation of sortilin in response to Aβ and proBDNF? It is well known that C-JNK 3 is one of most common signaling pathways to be activated by proNTs and through sortilin/ p75NTR, resulting in neuronal cell death (Skeldal et al. 2011; Reichardt 2006; Volosin et al. 2006; Koshimizu et al. 2010). It also reported that Aβ42 induces BACE1 expression through JNK (Guglielmotto et al. 2011). Therefore, it is of interest to investigate whether the upregulation of sortilin in the presence Aβ42 occurs through JNK. To our surprise, a JNK inhibitor did not influence Aβ-induced sortilin upregulation. The neurodegenerative signals of proBDNF, proNGF and Aβ cause neurite collapse via activating p75NTR and RhoA GTPase/ROCK pathway (Sun et al. 2012; Wang et al. 2011; Petratos et al. 2008). We propose that this pathway may be involved in the regulation of sortilin expression. Consistent with this idea, we found that the ROCK inhibitor, Y27632, blocked the response induced by Aβ. In addition, we also demonstrated that Aβ42 increased the level of active RhoA-GTP, confirming the previous finding (Petratos et al. 2008). These data suggest that the upregulation of sortilin is positively regulated by Aβ, proBDNF and p75NTR via activation of the RhoA-GTPase/ROCK signal pathway.
In summary, in contrast to sorLA, the expression of sortilin is increased in the AD brain of mice and humans and in an AD cell line model. The increased expression of sortilin is regulated by Aβ oligomers by activating the RhoA GTPase/ROCK pathway through the neurotrophin receptor p75NTR. As sortilin is important in neuronal cell death induced by the proneurotrophin/ p75NTR pathway and Aβ production, this novel regulation mechanism may play a role in AD pathogenesis.
We thank Prof. Richard Lewis from the Institute of Molecular Biosciences, University of Queensland, Australia for providing SH-SY5Y cells and Prof. Nigel Hooper from Institute of Molecular & Cellular Biology, University of Leeds, UK for providing SH-SY5YAPP695 human neuroblastoma cells. We also thank Dr Benjamin Roberts from School of Pharmacy and Medical Sciences, University of South Australia for his helpful and critical suggestions. This work was supported by research grants from National Health and Medical Research Council of Australia (NHMRC 488022 and 1021408). Khalil Saadipour was supported by an International Postgraduate Research Scholarship (IPRS) from Flinders University.