Cysteine 27 Variant of the δ-Opioid Receptor Affects Amyloid Precursor Protein Processing through Altered Endocytic Trafficking

Timo Sarajärvi; Jussi T Tuusa; Annakaisa Haapasalo; Jarkko J Lackman; Raija Sormunen; Seppo Helisalmi; Johannes T Roehr; Antonio R Parrado; Petra Mäkinen; Lars Bertram; Hilkka Soininen; Rudolph E Tanzi; Ulla E Petäjä-Repo; Mikko Hiltunen

doi:10.1128/MCB.05015-11

. 2011 Jun;31(11):2326–2340. doi: 10.1128/MCB.05015-11

Cysteine 27 Variant of the δ-Opioid Receptor Affects Amyloid Precursor Protein Processing through Altered Endocytic Trafficking ^{^▿}

Timo Sarajärvi ^1,^†, Jussi T Tuusa ^2,^†, Annakaisa Haapasalo ¹, Jarkko J Lackman ², Raija Sormunen ³, Seppo Helisalmi ¹, Johannes T Roehr ⁵, Antonio R Parrado ⁴, Petra Mäkinen ¹, Lars Bertram ⁵, Hilkka Soininen ¹, Rudolph E Tanzi ⁴, Ulla E Petäjä-Repo ^2,^†, Mikko Hiltunen ^1,^*

PMCID: PMC3133236 PMID: 21464208

Abstract

Agonist-induced activation of the δ-opioid receptor (δOR) was recently shown to augment β- and γ-secretase activities, which increased the production of β-amyloid peptide (Aβ), known to accumulate in the brain tissues of Alzheimer's disease (AD) patients. Previously, the δOR variant with a phenylalanine at position 27 (δOR-Phe27) exhibited more efficient receptor maturation and higher stability at the cell surface than did the less common cysteine (δOR-Cys27) variant. For this study, we expressed these variants in human SH-SY5Y and HEK293 cells expressing exogenous or endogenous amyloid precursor protein (APP) and assessed the effects on APP processing. Expression of δOR-Cys27, but not δOR-Phe27, resulted in a robust accumulation of the APP C83 C-terminal fragment and the APP intracellular domain, while the total soluble APP and, particularly, the β-amyloid 40 levels were decreased. These changes upon δOR-Cys27 expression coincided with decreased localization of APP C-terminal fragments in late endosomes and lysosomes. Importantly, a long-term treatment with a subset of δOR-specific ligands or a c-Src tyrosine kinase inhibitor suppressed the δOR-Cys27-induced APP phenotype. These data suggest that an increased constitutive internalization and/or concurrent signaling of the δOR-Cys27 variant affects APP processing through altered endocytic trafficking of APP.

INTRODUCTION

Alzheimer's disease (AD) is the most common neurodegenerative disorder in the aging population. It is neuropathologically characterized by well-known hallmarks, such as extracellular amyloid plaques and intraneuronal neurofibrillary tangles, composed of β-amyloid peptide (Aβ) and hyperphosphorylated tau, respectively. Aβ is generated from the amyloid precursor protein (APP) after sequential cleavages by β (BACE1)- and γ-secretases. It is a well-established fact that the molecular mechanisms underlying AD pathogenesis involve alterations in APP processing which lead to increased Aβ production or, alternatively, decreased enzymatic degradation and clearance of Aβ (39). To facilitate the design of novel intervention approaches for AD, it is important to identify and functionally characterize genetic alterations which play a role in AD pathogenesis. A plausible candidate in this context is the OPRD1 gene, encoding the δ-opioid receptor (δOR), which was recently shown to form a complex with β- and γ-secretases (28, 40). Following agonist-induced activation, δOR mediates coendocytic sorting of this complex to late endosomes and lysosomes (LEL) (28, 40), in which compartments Aβ production primarily takes place. Conversely, β- and γ-secretase activities as well as Aβ levels were found to be significantly reduced in transgenic APP/PS1ΔE9 mice (overexpressing human APP with the Swedish mutation together with human presenilin-1 harboring the exon 9 deletion) treated with a selective nonpeptide antagonist for δOR (40). These results suggest that the amyloidogenic processing of APP is enhanced upon δOR activation and that the selective antagonist-mediated modulation of δOR may provide a novel treatment strategy against AD.

The δOR is a G protein-coupled receptor (GPCR) with a typical seven-transmembrane helix (7TM) topology (44). It has been implicated to have a role in the presynaptic modulation of synaptic function and in the regulation of pain and mood (6, 48). Furthermore, assessments of postmortem brain samples have revealed that opioid receptors are differentially affected in distinct brain regions in AD patients (25). The only nonsynonymous single nucleotide polymorphism (SNP) in the coding region of OPRD1, T80G (rs1042114), leading to the Phe27Cys amino acid substitution, has been reported to associate with opioid dependence (47), while another study found no association (46). Currently, no study has specifically investigated whether or not OPRD1 is genetically associated with a risk for AD (5) or other neurodegenerative diseases. Like other GPCRs, δORs not only act as monomers but can form homomeric and heteromeric complexes with other opioid receptor subtypes as well as with other GPCRs, creating new receptors with novel pharmacological properties (42). This underscores the assumption that the penetration of a putative disease-predisposing alteration may not have an additive character but may be strongly influenced by genetic and environmental interactions.

We have previously shown that cysteine at position 27 affects the maturation and subcellular localization of δOR in nonneuronal cells (20, 32). More specifically, the δOR-Cys27 variant showed a decreased mature/precursor receptor ratio, which is related to the retention of receptor precursors in the endoplasmic reticulum and enhanced turnover of mature cell surface receptors. Based on this, it was proposed that the δOR-Cys27 variant may cause a gain-of-function phenotype with possible pathophysiological consequences due to the intracellular accumulation of the receptor (20). Here, we characterized the δOR-Phe27 and δOR-Cys27 expression phenotypes in human SH-SY5Y and HEK293 cells stably overexpressing exogenous or endogenous APP and set out to assess whether these δOR variants could differentially affect APP processing. Moreover, we wanted to elucidate the genetic role of OPRD1 in AD by assessing the risk effect of T80G variation (rs1042114) in both case-control and familial AD sample sets. Results from the present study demonstrate that the δOR-Cys27 variant affects APP processing through altered endocytic trafficking.

MATERIALS AND METHODS

DNA constructs.

The Myc-δOR-Flag-pFT-SMMF and HA-δOR-pcDNA5/FRT/TO constructs, encoding the human δOR-Cys27 and δOR-Phe27 variants have been described previously (20, 21). All δOR constructs contained a cleavable hemagglutinin (HA) signal peptide and either an N-terminal Myc tag and a C-terminal Flag tag or an N-terminal HA tag with a native C terminus. The enhanced yellow fluorescent protein (EYFP)-Golgi compartment, enhanced green fluorescent protein (EGFP)-Rab7 (a marker for LEL), EGFP-Rab9 (a marker for lysosomes), and EGFP-Rab11 (a marker for recycling endosomes) constructs have been described previously (9).

Cell culture and treatments.

HEK293-AP-APP cells overexpressing alkaline phosphatase (AP) and APP695 fusion protein were grown as described previously (22). The AP ectodomain was fused to the N terminus of full-length APP695, lacking a signal peptide (27). SH-SY5Y human neuroblastoma cells overexpressing the APP751 isoform (SH-SY5Y-APP751) were grown as described previously (36). These cells express approximately 5 times more APP751 and Aβ than naive SH-SY5Y cells (reference 36 and unpublished observations). Plasmid constructs were transfected into the SH-SY5Y-APP751 and HEK293-AP-APP cells using the Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer's instructions. The preparation and maintenance of stable inducible HEK293 (HEK293_i) cell lines with inducible expression of the epitope-tagged δOR-Cys27 and δOR-Phe27 variants have been described previously (20, 21). Receptor expression was induced for 24 to 48 h by adding 0.5 μg/ml tetracycline (Invitrogen) to the culture medium. Opioid ligands, naltrexone, naltriben, ICI-174,864, and SNC-80 (all from Tocris) and bisindolylmaleimide I (Bis I; a protein kinase C [PKC] inhibitor), PP2 (a c-Src kinase inhibitor), and phorbol-12-myristate-13 acetate (PMA; a PKC activator) were added at the concentrations and for the incubation periods indicated in the figure legends. For metabolic labeling, the cells were cultivated in a media lacking cysteine and methionine and supplemented with 10% fetal calf serum, 100 units/ml penicillin, 0.1 mg/ml streptomycin, 0.5 μg/ml tetracycline, and 100 μCi/ml EXPRE³⁵S³⁵S labeling mix (Perkin-Elmer). Fresh label (50 μCi/ml) was added after a 24-h incubation for 60 min to ensure efficient labeling of newly synthesized proteins.

Preparation of protein extracts and Western blot analysis.

Total proteins were extracted from cells using transmembrane protein extraction reagent (TPER) buffer (Pierce), which contains an EDTA-free protease inhibitor cocktail (Thermo Scientific). Alternatively, cells were lysed in DDM buffer (0.5% n-dodecyl-β-d-maltoside [DDM; Alexis], 25 mM Tris-HCl [pH 7.5], 140 mM NaCl, 2 mM EDTA, 20 mM N-ethylmaleimide, 0.5 mM phenylmethanesulfonyl fluoride, 2 mM 1,10-phenantroline, 2 μg/ml aprotinin, 5 μg/ml leupeptin, 5 μg/ml trypsin inhibitor, 10 μg/ml benzamidine) by a brief 5-s sonication (see Fig. 1B, 3A, 4A, and 5 to 7). The DDM-solubilized samples were further incubated for 60 min at +4°C with constant rotation, and the insoluble material was removed by 11,000 × g centrifugation for 30 min. Membrane extracts were prepared as described previously (41) (see Fig. 4B and C). After protein quantification using either bicinchoninic acid (BCA) or DC protein assays (Pierce or Bio-Rad, respectively), samples of 10 to 30 μg of total protein lysates were subjected to 4 to 12% Bis-Tris polyacrylamide gel electrophoresis (PAGE; Invitrogen) or to 10 or 14% SDS-PAGE (Bio-Rad) and subsequently blotted onto Immun-Blot (Bio-Rad) or Immobilon P (Millipore) polyvinylidene fluoride membranes. Primary antibodies against c-Myc (4A6 [Millipore] and A-14 and 9E10 [Santa Cruz]), FLAG (M2; Sigma), APP's C terminus (A8717; Sigma), APP's N terminus (MAB348, clone 22C11; Millipore), secreted N-terminal APP fragments (sAPPα) (6E10; Biosite), β-actin (AC-15; Sigma), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (ab8245; Abcam) were used for immunoblotting. Antibody A8717 has previously been shown to specifically detect the APP C-terminal fragments (CTFs) C83, C89, C99, and APP intracellular domain (AICD), while 22C11 and 6E10 are antibodies commonly used to detect total sAPP (sAPPtot) and sAPPα, respectively (12, 16, 36). After incubation with the appropriate horseradish peroxidase-conjugated secondary antibodies (GE Healthcare), an enhanced-chemiluminescence (ECL) substrate (Amersham Biosciences/GE Healthcare) was applied to the membranes and protein bands were detected with an ImageQuant RT ECL camera (GE Healthcare) or with ECL hyperfilm (GE Healthcare) scanned with a Umax Powerlook 1120 color scanner using the Image Master 2D Platinum 6.0 software. The band intensities were quantified using Quantity One (Bio-Rad) or Image J 1.41 software.

Fig. 1. — δOR-Cys27 and δOR-Phe27 variants are differentially expressed and localized in SH-SY5Y-APP751 cells. (A) Western blot analysis of total protein lysates shows that N- and C-terminally tagged δOR variants (Myc-δOR-Flag) differ significantly in their mature/precursor receptor ratios when transiently overexpressed in SH-SY5Y-APP751 cells (48-h transfection). **, P < 0.01, n = 3. Numbers at the right of the blots in panels A and B are molecular masses in kilodaltons. (B) Immunoprecipitation of the δOR variants. Aliquots of total protein lysates were immunoprecipitated with the anti-FLAG M2 antibody and analyzed by Western blotting with the anti-c-Myc A-14 antibody. (C) Confocal microscope images illustrating the differential subcellular localizations of the δOR variants in SH-SY5Y-APP751 cells transfected with δOR-Phe27 or δOR-Cys27 (Myc-δOR-Flag plasmids; anti-c-Myc = green). (Left) The δOR-Phe27 variant is localized at the plasma membrane (arrows) and in juxtanuclear compartments (arrowheads). (Right) The δOR-Cys27 variant shows a strong intracellular localization (arrows). Bar, 10 μm.

Fig. 3. — APP processing in HEK293-AP-APP cells overexpressing the δOR variants and subcellular localization in SH-SY5Y-APP751 cells. (A) Western blot analysis of anti-FLAG M2-immunoprecipitated δOR variants in HEK293-AP-APP cells (Myc-δOR-Flag plasmids; 48-h transfection) shows a difference in the mature/precursor receptor ratios between the variants. (B) Western blot analysis of total protein lysates shows a statistically significant increase in the GAPDH-normalized APP C83 levels in HEK293-AP-APP cells transiently transfected with the δOR-Cys27 variant. *, P < 0.05, n = 3. Numbers at the right of the blots in panels A and B are molecular masses (in kilodaltons). (C) Confocal microscope images of untransfected SH-SY5Y-APP751 cells show the subcellular localization of the full-length APP using an N-terminal anti-APP antibody (22C11 = green) and both the full-length APP and APP CTFs using a C-terminal anti-APP antibody (A8717 = red). Bar, 10 μm.

Fig. 4. — APP C-terminal fragments and δOR-Phe27 colocalize in SH-SY5Y-APP751 cells without direct APP-δOR interaction. (A) Confocal microscope analysis of APP (anti-APP C terminus = red) subcellular localization in SH-SY5Y-APP751 cells transiently transfected with δOR-Phe27 (anti-c-Myc = green). The two proteins colocalized (yellow) in compact juxtanuclear structures (arrowhead), with some colocalization at the plasma membrane (open arrowheads). The lower panels illustrate a 4-fold magnification of cells indicated by arrows in the upper panels. Bar, 10 μm. (B) Stably transfected tetracycline-inducible HEK293_i-δOR-Cys27 cells were induced to express the receptor for 24 h in the presence of [³⁵S]methionine/cysteine. Sequential immunoprecipitation with anti-FLAG M2 and C-terminal anti-APP antibodies was performed from mixed cellular extracts, and samples were analyzed by SDS-PAGE and fluorography. For the “Cis” samples, membrane extracts were mixed from labeled induced cells and nonlabeled uninduced cells. The “Trans” samples were prepared reciprocally. The third lane represents a control sample from the anti-APP antibody immunoprecipitation. *, an unknown 150-kDa protein. (C) Aliquots of the anti-FLAG M2 antibody immunoprecipitates were analyzed by Western blotting to show receptor recovery after the first immunoprecipitation step. Numbers at the right of the blots in panels B and C are molecular masses (in kilodaltons).

Fig. 5. — Opioid receptor ligands suppress the δOR-Cys27 expression-induced increase in the steady-state APP C83 levels. HEK293_i-δOR-Cys27 (A and B) and HEK293_i-δOR-Phe27 (A) cells were induced or not induced for 48 h in the absence or presence of opioid receptor ligands (10 μM naltrexone [NTX], 10 μM naltriben [NTB], 5 μM ICI-174,864, 10 μM SNC-80). The medium was replaced with the same supplements 24 h before the harvesting. Total protein lysates were analyzed by Western blotting. The intensities of the full-length APP (APP-fl) and APP C83 bands in panel B were quantified and normalized to β-actin. **, P < 0.01; *, P < 0.05; n = 3. Numbers at the right of the blots in panels A and B are molecular masses (in kilodaltons).

Fig. 7. — Inhibition of protein kinase C mimics the δOR-Cys27 expression-induced alteration in APP processing. (A) HEK293_i-δOR-Cys27 cells were treated with the PKC activator PMA (1 μM) for 4 h or the PKC inhibitor Bis I (10 μM) for 24 h or mock treated for 24 h. Total protein lysates were analyzed by Western blotting. Numbers at the right of the blots are molecular masses (in kilodaltons). (B) Bis I-induced upregulation of APP C83 levels is concentration dependent. Cells were treated with the indicated concentrations of Bis I and analyzed as described for panel A. (C) Inhibition of constitutive PKC activity induces a dispersal of the juxtanuclear APP CTF accumulation. SH-SY5Y-APP751 cells were treated with 1 μM Bis I for 4 h or left untreated. Cells were immunostained with the anti-APP C-terminal antibody and imaged by confocal microscopy. Bar, 10 μm.

Immunoprecipitation.

Whole-cell extracts containing equal amounts of protein were supplemented with 0.1% bovine serum albumin (BSA) and precleared with mouse IgG agarose. The precleared samples were immunoprecipitated with anti-FLAG M2 antibody affinity gel (Sigma). They were washed once with DDM buffer, twice with the same buffer containing 300 mM NaCl, twice with DDM buffer, and finally three times with DDM buffer containing 0.1%, instead of 0.5%, DDM. Bound proteins were eluted with 200 μg/ml of the FLAG peptide (Sigma) (see Fig. 1B and 3A). In the extract-mixing experiments (see Fig. 4B and C), the anti-FLAG M2 antibody affinity gel was used for the first immunoprecipitation step. After an overnight incubation, the resin was washed five times with DDM buffer, and the bound proteins were eluted with 1% SDS, 25 mM Tris-HCl (pH 7.5) at +95°C for 5 min. The eluted samples were diluted 10-fold with DDM buffer supplemented with 0.1% BSA and used for a second immunoprecipitation with the anti-APP-CTF antibody (A8717) coupled to protein A-Sepharose (GE Healthcare), followed by washes as described above and elution with the SDS-PAGE sample buffer. Aliquots of immunoprecipitates and of whole-cell extracts containing equal amounts of protein were separated by 10% SDS-PAGE and analyzed by Western blotting or fluorography as described previously (41).

Secreted APP and Aβ measurements.

sAPPα (secreted N-terminal APP fragments) and sAPPtot (= sAPPα + sAPPβ) levels were detected from cell culture media using Western blot analysis with the 6E10 (Signet Laboratories) and 22C11 (Mab348; Millipore) antibodies, respectively. β-Amyloid 40 (Aβ40) levels in the SH-SY5Y-APP751 cell culture media were determined by using the human Aβ40 enzyme-linked immunosorbent assay (ELISA) kit (the Genetics Company) as described previously (36). sAP-APP measurements from HEK293 AP-APP cells were performed as previously described (12). sAPP, sAP-APP, and Aβ40 levels were subsequently normalized to the total protein levels, which were determined from the lysates of the same cells from which the cell culture media were initially collected. Cell culture media were not concentrated or diluted at any steps during the sAPP or Aβ40 measurements.

α-Secretase activity assay.

An α-secretase activity kit (catalog no. FP001; R&D Systems) was used to measure the α-secretase activity in total protein lysates according to the manufacturer's instructions by using α-secretase-specific substrate peptides conjugated to the EDANS and DABCYL reporter molecules. Cleavage of these peptides leads to the release of a fluorescent signal, which is proportional to the α-secretase activity. After a 2-h incubation at +37°C, the emitted light (510 nm) was detected on a fluorescence microplate reader after excitation at 355 nm.

In vitro AICD generation assay.

For the in vitro AICD generation assay, we used a protocol similar to that used by Pinnix et al. (34). Briefly, SH-SY5Y-APP751 cells were plated at a density of 200,000 cells/cm². The cells were scraped into buffer A (50 mM HEPES, 150 mM NaCl, 5 mM 1,10-phenanthroline, pH 7.4) and homogenized by passing them through a 25G5/8 needle 10 times. The homogenates were centrifuged at 10,000 × g for 15 min. The resulting membrane fractions (P10) were washed once with buffer A and centrifuged at 10,000 × g. Total protein concentrations were measured in the P10 fraction, and the same amounts of protein in each sample were incubated in 30 μl of buffer B (50 mM HEPES, 150 mM NaCl, 5 mM 1,10-phenanthroline, pH 7.0, and protease inhibitor cocktail; Roche) for 2 h on ice (as a negative control) or at +37°C to allow the release of AICD. The samples were then centrifuged at 10,000 × g for 15 min. The supernatants and solubilized pellets were analyzed by Western blotting using anti-APP-CTF antibody (A8717) to detect APP C83 and AICD.

Flow cytometry.

Cell surface receptors expressed in stably transfected HEK293_i cells after a 48-h induction were analyzed by flow cytometry as described previously (33) using 1 μg/ml anti-c-Myc 9E10 antibody and 1 μg/ml phycoerythrin-conjugated secondary antibody (BD Biosciences). To assess constitutive internalization, a protocol by Markkanen and Petäjä-Repo (24) was applied. Briefly, cell surface receptors after a 24-h induction of stably transfected HEK293_i cells were labeled on 12-well plates (2 × 10⁶ cells/well) with the anti-c-Myc 9E10 antibody (1 μg/ml), washed, and chased for 4 h at +37°C. The c-Src inhibitor PP2 (4 μM) was added together with the antibody and was maintained throughout the incubations. The controls were treated with a vehicle (dimethyl sulfoxide [DMSO]). The remaining antibody-labeled receptors at the cell surface were stained with the phycoerythrin-conjugated secondary antibody (2 μg/ml). The fluorescence of live cells was measured with a Becton Dickinson FACSCalibur flow cytometer and analyzed with the CellQuestPro 6.0 software as described previously (24).

Confocal microscopy.

For confocal microscope analysis, SH-SY5Y-APP751 cells were plated on sterile coverslips coated with 100 μg/ml poly-d-lysine (Sigma) and transfected with the δOR-Cys27 or δOR-Phe27 construct. Untransfected cells were used as controls. To inhibit PKC activity, the cells were treated with 1 μM Bis I for 4 h. The cells were fixed in 4% paraformaldehyde for 15 min at room temperature. Next, the cells were incubated in the blocking and permeabilization buffer containing 0.5 or 5% BSA and 0.1% Triton X-100 (BHD Laboratory Supplies) in phosphate-buffered saline for 30 to 45 min and stained with primary antibodies against c-Myc (4A6, 1:200) and the APP C terminus (A8717, 1:250 or 1:1,000). For APP subcellular-localization experiments, SH-SY5Y-APP751 cells were plated on chamber slides (Lab-Tek) and transfected with the EYFP-Golgi compartment, EGFP-Rab7 (a marker for LEL), EGFP-Rab9 (a marker for lysosomes), or EGFP-Rab11 (a marker for recycling endosomes) construct (9). The fixed and permeabilized cells were stained with primary antibodies against the APP C terminus (A8717, 1:1,000), the APP N terminus (22C11, 1:100), calnexin (A4, 1:100, a marker for endoplasmic reticulum; BD Biosciences), the transferrin receptor (13-6800, 1:200, a marker for plasma membrane and early endosomes; Zymed Laboratories), and lysobisphoshatidic acid (LBPA; 6C4, 1:100, a marker for multivesicular bodies [MVBs]; Echelon Biosciences). Alexa Fluor 488 goat anti-mouse or Alexa Fluor 568 goat anti-rabbit (1:250 or 1:500; Invitrogen), Cy5 goat anti-rabbit (1:1,000; Jackson ImmunoResearch Laboratories), and Cy3 sheep anti-mouse (C-2181, 1:500; Sigma) were used as secondary antibodies. Single optical z-sections were obtained with a Nikon Eclipse TE300 microscope together with the UltraVIEW laser scanning confocal unit (Perkin Elmer) or, alternatively, with a Carl Zeiss LSM700 or LSM510 confocal laser scanning microscope. The images were processed with the Adobe Photoshop CS4 software (version 11.0).

Electron microscopy.

For electron microscopy, induced (24 h) or mock-induced HEK293_i-hδOR-Cys27 cells were rinsed with phosphate-buffered saline and fixed with 4% paraformaldehyde in 0.1 M phosphate buffer containing 2.5% sucrose. The cryosectioning, immunolabeling, and imaging were done essentially as described earlier (1), with minor modifications. Antibodies against LBPA (6C4, 1:200) and the APP C terminus (A8717, 1:700) were used for immunodetection, followed by incubations with protein A-conjugated 5-nm and 10-nm gold particles, respectively.

Patients and controls.

All AD cases and controls originated from eastern Finland and were examined at the Department of Neurology, Kuopio University Hospital. The study cohort consisted of 533 AD patients (mean age at onset of AD, 71.2 ± 7.0 years; range, 43 to 89 years; 69% were women) and 671 age-matched healthy control subjects (mean age at the time of neuropsychological examination, 69.2 ± 6.1 years; range, 37 to 87 years; 59.8% were women). All patients fulfilled the NINCDS-ADRDA criteria for probable AD (26). Control subjects had no signs of dementia according to the interview and neuropsychological testing. There were 95 AD patients and 155 controls with an onset age or age at examination of ≤65 years. These early-onset AD patients did not show conclusive evidence of autosomal dominant transmission (19), and there were no APP or PSEN1/2 mutations. The Kuopio University Hospital ethics committee approved the study. The NIMH Genetics Initiative Study sample contained 1,376 individuals from 410 families, of which all members were of self-reported European ancestry.

SNP genotyping.

Four tag SNPs (rs678849, rs760588, rs4654323, and rs529520) were selected from the Applied Biosystems database (Human Pre-Designed Assays) using the SNPbrowser 3.0 software (Table 1). SNP rs1042114 was designed with the Custom TaqMan SNP genotyping assay design tool (Applied Biosystems). The selected SNPs were part of a single haplotype block, according to the HapMap database (http://www.hapmap.org/), in a Caucasian population. DNA samples were isolated from blood with the BloodPrep chemistry kit (Applied Biosystems), and SNP genotyping was performed by using TaqMan allelic discrimination assays (Applied Biosystems). The TaqMan PCRs were done on an MJ Research Peltier thermal cycler (model PTC-200), and the fluorescence was detected on an ABI Prism 7000 sequence detector (Applied Biosystems). The PCR program was the following: one cycle at +50°C for 2 min to activate uracil-N-glycosylase, which was added to prevent carryover contamination, and +95°C for 10 min to activate the AmpliTaq Gold polymerase and then 39 cycles of +92°C for 15 s, denaturation at +60°C for 1 min, annealing, and extension. APOE genotyping was performed as previously described (19). Genotyping of the NIMH Genetics Initiative Study sample set was completed with the GeneChip human mapping 500K array set from Affymetrix (4).

Table 1.

Summary of allele and genotype frequencies of OPRD1 SNPs among Finnish AD patients and controls

SNP (location [kb])^a	Allele	Allele frequency in:		P value^b	Genotype	Genotype frequency in:		P value^b
SNP (location [kb])^a	Allele	Controls	AD patients	P value^b	Genotype	Controls	AD patients	P value^b
rs1042114 (0)		n = 1,324	n = 1,002			n = 662	n = 501
	T	0.875	0.865	0.47	GG	0.024	0.012	0.07
	G	0.125	0.135		TG	0.201	0.246
					TT	0.775	0.743
rs678849 (6.2)		n = 1,340	n = 1,062			n = 670	n = 531
	T	0.534	0.518	0.44	TT	0.209	0.271	0.54
	C	0.466	0.482		TC	0.515	0.493
					CC	0.276	0.235
rs760588 (23.6)		n = 1,340	n = 1,064			n = 670	n = 532
	A	0.678	0.675	0.88	AA	0.460	0.438	0.36
	G	0.322	0.325		AG	0.436	0.474
					GG	0.104	0.088
rs4654323 (30)		n = 1,340	n = 1,062			n = 670	n = 531
	G	0.701	0.699	0.88	TT	0.090	0.079	0.60
	T	0.299	0.301		TG	0.418	0.444
					GG	0.493	0.476
rs529520 (36)		n = 1,342	n = 1,066			n = 671	n = 533
	C	0.569	0.540	0.17	CC	0.306	0.287	0.24
	A	0.431	0.460		AC	0.526	0.507
					AA	0.168	0.206

Open in a new tab

Locations of SNPs are indicated in the 5′-to-3′ orientation with respect to SNP rs1042114.

Allele and genotype frequencies were compared using a two-sided Pearson χ² test. All the studied SNPs were in Hardy-Weinberg equilibrium in both cases and controls (P > 0.05). The marginal association at rs1042114 was confirmed in an independent family-based data set (TG versus GG plus TT; P = 0.021, OR = 1.62 [95% CI, 1.00 to 2.64]) (see the text for more details).

Imputation analysis.

Markers for the NIMH families were generated as part of a genome-wide association study (GWAS), as previously described (5). Since SNP rs1042114 was not directly genotyped on the Affymetrix 500K SNP array, we inferred genotypes at this site via imputation using IMPUTE v2.0 (23). As reference panels, we used the precompiled HapMap 3 and 1000 Genomes Project “CEU+TSI” panels from the IMPUTE website (obtained on 2 June 2010). These contain HapMap 3 data (from release 2 of February 2009) and 1000 Genomes Project data from Pilot 1 genotypes (released in August 2009) for autosomal SNPs. Only genotypes called with 0.9 or greater posterior call probability were called or coded as missing if the posterior call probability fell below 0.9. Manual regenotyping of markers imputed via this procedure yielded genotype-wide consistency rates of ∼99% in a similar project using imputed genotype data from the same NIMH families (based on 12 SNPs) (L. Bertram and J. T. Roehr, unpublished observations).

Statistical analyses.

A Hardy-Weinberg equilibrium of SNPs in AD patient and control cohorts was calculated with the GenePop program (http://genepop.curtin.edu.au/genepop/genepop_op1.html). The SPSS statistical software for Windows (version 14.0.1; SPSS, Inc.) was used to assess differences in allele and genotype frequencies by applying two-sided Pearson χ² and Fisher exact tests. Kaplan-Meier survival analyses were performed to assess the age-of-onset effects. Multivariate logistic regression analysis was applied to assess age, gender, and APOE ε4-adjusted risk effects. Pairwise LD values and estimation of haplotypes and haplotype blocks were performed with the HaploView program (http://www.broadinstitute.org/scientific-community/science/programs/medical-and-population-genetics/haploview/haploview). The GraphPad Prism 4.01 software was used for one-way analysis of variance (ANOVA), followed by Bonferroni's posttest, in experiments with more than one variable. In experiments with only one variable, paired t tests, an independent-sample t test (equal variances assumed), or a Mann-Whitney U test (equal variances not assumed) were used to test statistical significance between the sample groups. Association analyses in the NIMH GWAS data set were analyzed using PBAT v3.6 as previously described (4). Effect sizes in the family data were estimated in SAS v9.2 using conditional logistic regression stratified on family as described previously (37). Meta-analyses were based on random-effect models and were performed in R programming language, version 2.10.0, using rmeta, version 2.16. All values are reported as means ± standard deviations (SD). The level of statistical significance was set to a P of <0.05.

RESULTS

Expression of the δOR-Cys27 variant leads to APP C83 accumulation and decreased Aβ levels in human SH-SY5Y cells overexpressing APP and in HEK293 cells with endogenous or overexpressed APP.

Several lines of recent evidence suggest that δOR contributes to AD pathogenesis by affecting Aβ generation in vitro and in vivo (28, 40). However, the evidence has so far focused on the agonist-induced receptor internalization, and the consequent coendocytic trafficking, of β- and γ-secretases, while constitutive agonist-independent effects of the receptor on APP processing are still elusive. We have recently shown that in nonneuronal cells, the less common δOR-Cys27 variant shows significantly decreased receptor maturation efficiency, increased intracellular accumulation, and reduced stability at the cell surface compared to the δOR-Phe27 variant, although the two variants are indistinguishable in their pharmacological properties (20). Keeping in mind these findings, we wanted to elucidate whether the δOR-Cys27 and δOR-Phe27 variants differentially affect APP processing. To characterize the δOR expression phenotypes, we first transiently overexpressed N- and C-terminally tagged δOR-Phe27 and δOR-Cys27 in human SH-SY5Y neuroblastoma cells stably overexpressing the APP751 isoform (SH-SY5Y-APP751) (Fig. 1). As with the HEK293 and CHO cells used previously (20) (see also Fig. 5A), the δOR variants differed significantly in their mature/precursor receptor ratios when expressed in SH-SY5Y-APP751 cells (Fig. 1A and B). The δOR-Phe27 variant showed an approximately 2-fold-higher mature/precursor receptor ratio than the δOR-Cys27 variant, suggesting less efficient maturation for the δOR-Cys27 variant and/or higher turnover of its mature form. A similar difference in the mature/precursor receptor ratios was detected following transient transfection of the variants in HEK293 cells stably overexpressing APP fused to alkaline phosphatase (AP; HEK293-AP-APP) (see Fig. 3A). To assess whether the δOR-Cys27 variant was more prominently observed in intracellular compartments, we studied receptor subcellular localization in the δOR-Cys27- and δOR-Phe27-overexpressing SH-SY5Y-APP751 cells using confocal microscopy (Fig. 1C). Staining with the anti-c-Myc antibody recognizing the N-terminal Myc tag in δOR-Cys27 and δOR-Phe27 showed a markedly augmented intracellular localization of the δOR-Cys27 variant compared to the localization of δOR-Phe27. δOR-Phe27 was more abundantly localized at the plasma membrane and intracellularly in compact juxtanuclear foci.

Next, we determined the effects of the δOR-Phe27Cys variation on APP processing in SH-SY5Y-APP751 cells (Fig. 2). Transient overexpression of the N- and C-terminally tagged δOR-Cys27 variant resulted in a robust increase (∼3-fold) in GAPDH-normalized APP C83 C-terminal fragment (CTF) levels compared to levels in cells transfected with δOR-Phe27 or the control plasmid (Fig. 2A). Importantly, the APP C99 CTF levels were not significantly affected, suggesting that δOR-Cys27 overexpression does not lead to a general enhancement of APP processing. Coinciding with the augmented APP C83 levels, we also observed an accumulation of the APP intracellular domain (AICD) in the δOR-Cys27-overexpressing cells, while the levels of immature APP (APPim), mature APP (APPm), and total APP (APPtot) remained unchanged. In order to assess whether the observed increase in the APP C83 levels was related to an augmented α-secretase activity, we performed an activity assay from membrane protein fractions of the δOR-Cys27-, δOR-Phe27-, and vector-transfected SH-SY5Y-APP751 cells (Fig. 2A) but found no changes in the α-secretase activity. The total-protein-normalized, secreted APPα (sAPPα) levels in the conditioned cell media were not increased after δOR-Cys27 transfection either (Fig. 2B). In contrast, total-protein-normalized Aβ40 levels were significantly decreased in the δOR-Cys27-transfected cell medium on average by 50%, in parallel with moderately decreased total sAPP (sAPPtot = sAPPα + sAPPβ) levels (Fig. 2B). The in vitro AICD generation assay, however, did not reveal changes in the γ-secretase activity in SH-SY5Y-APP751 cells after δOR transfection (Fig. 2C). Transient overexpression of the δOR variants in HEK293-AP-APP cells resulted in the accumulation of APP C83 and AICD without affecting sAPPtot levels specifically in cells overexpressing δOR-Cys27 in a manner similar to that in SH-SY5Y-APP751 cells (Fig. 3 B). Importantly, a similar alteration in APP processing upon δOR-Cys27 overexpression was observed when the δOR variants with an N-terminal hemagglutinin (HA) tag and a native untagged C terminus were transiently overexpressed in SH-SY5Y-APP751 cells (data not shown). Furthermore, analogous changes in the processing of endogenous APP were detected in tetracycline-inducible HEK293_i cells stably overexpressing the δOR-Cys27 variant, but not the δOR-Phe27 one (HEK293_i-δOR-Cys27 and HEK293_i-δOR-Phe27, respectively [20]) (see also Fig. 5A).

In addition to revealing the alteration in APP processing in δOR-Cys27- but not in δOR-Phe27-expressing cells, confocal microscopy analysis of SH-SY5Y-APP751 cells transiently transfected with δOR-Phe27 revealed that APP colocalized predominantly in compact juxtanuclear structures with the δOR-Phe27 variant (Fig. 4 A and see Fig. 8A), indicating that APP and δOR are in a close proximity to each other. This led us to investigate whether there is a direct, and possibly a variant-specific, physical interaction between APP and δOR. We performed protein coimmunoprecipitation assays with HEK293_i-δOR-Cys27 and HEK293_i-δOR-Phe27 cells. Western blot analysis showed minor amounts of APP to coimmunoprecipitate with both variants (data not shown). However, this was not evident for the newly synthesized proteins in a metabolic labeling experiment. Only a very minor coimmunoprecipitation of APP with δOR was found after a prolonged film exposure (data not shown), and this also took place under postlysis conditions (Fig. 4B and C). Thus, the observed modest coimmunoprecipitation of δOR and APP was likely to be a postlysis artifact. Although the δOR variants and APP do not seem to form a stable complex, it is still possible that these two proteins transiently interact with each other via an additional protein(s). This interaction could be mediated, e.g., by the γ-secretase complex, which has been found to independently interact with both APP and δOR (28, 40).

Fig. 8. — Overexpression of δOR-Cys27 reduces the localization of APP C-terminal fragments in late endosomes and lysosomes. (A) Confocal microscopy of APP subcellular localization in untransfected (UNT) and transiently transfected SH-SY5Y-APP751 cells expressing δOR-Phe27 or δOR-Cys27. APP (anti-APP C terminus = red) localizes in compact juxtanuclear structures (arrow) and at or near the plasma membrane in the δOR-Phe27-overexpressing cells or in untransfected cells. The δOR-Phe27 variant (anti-c-Myc = green) is predominantly localized near or at the plasma membrane but shows colocalization with APP (yellow) in the juxtanuclear structures. The δOR-Cys27 variant is localized mostly intracellularly, and it rarely colocalizes with APP in juxtanuclear structures (arrow). (B) Characterization of the APP-positive juxtanuclear structures. APP (anti-APP C terminus = red) strongly colocalizes with markers of the LEL and Golgi compartments (EGFP-Rab7 and EYFP-Golgi, respectively; green) in untransfected SH-SY5Y-APP751 cells. Bar, 10 μm. (C) Quantification of the numbers of δOR-Phe27- or δOR-Cys27-transfected cells with APP-positive LEL/Golgi compartments. *, P < 0.05, n = 3.

In summary, these results indicate that the expression of the δOR-Cys27 variant, but not the δOR-Phe27 one, leads to the accumulation of APP C83 and AICD, but not APP C99, in SH-SY5Y and HEK293 cells expressing either endogenous or overexpressed APP. In parallel with this, the secreted Aβ40 levels were decreased in the δOR-Cys27-overexpressing cells, while the sAPPα levels remained unchanged. These changes were not related to altered α- or γ-secretase activities, nor were there any obvious indications that a direct δOR-APP interaction in a variant-specific manner plays a role in the altered APP processing. Interestingly enough, the APP processing phenotype in the δOR-Cys27 variant-expressing cells is intriguingly similar to that observed previously with APP under pH-neutralized conditions (45), suggesting that these two conditions may share common mechanistic features.

δOR-Cys27 variant-induced changes in APP processing can be suppressed by a long-term treatment with δOR-specific ligands via a mechanism involving altered trafficking of the receptor.

To investigate the possibility that the observed changes in APP processing in the δOR-Cys27-expressing cells relate to the inefficient maturation of this δOR variant, we assessed whether enhancement of receptor maturation suppresses the increase in the APP C83 levels (Fig. 5). In order to enhance δOR maturation, the HEK293_i-hδOR-Cys27 cells were treated with membrane-permeable δOR ligands that are known to act as pharmacological chaperones by facilitating the maturation and cell surface delivery of newly synthesized receptors (21, 31). Interestingly, a 48-h treatment with saturating concentrations of a few, but not all, of the tested receptor ligands suppressed the δOR-Cys27-induced increase in the β-actin-normalized APP C83 levels (Fig. 5B). Importantly, this occurred in a manner that was not directly related to their ability to enhance receptor maturation. The two tested antagonists, naltrexone and naltriben, and the inverse agonist ICI-174,864 all showed efficient pharmacological chaperoning; i.e., they assisted the folding of the receptor, as was evident from the increased mature/precursor receptor ratio. However, only naltrexone and ICI-174,864 significantly suppressed the increase in the APP C83 levels (Fig. 5B). SNC-80, a membrane-permeable δOR agonist, showed a moderate, but not statistically significant, suppression of the APP C83 levels. SNC-80 had a chaperoning effect, as the steady-state amount of receptor precursors was decreased; however, as expected, it also caused a strong downregulation of mature receptors (Fig. 5B). The effect of a peptidic membrane-impermeable ligand, Leu-enkephalin, that shows no pharmacological chaperoning activity was similar to the effect of SNC-80 (data not shown). The δOR ligands did not significantly change the amount of the full-length APP when the receptor was overexpressed (Fig. 5B). The maturation efficiency of APP and the levels of full-length APP and the APP CTF were not affected by the ligands in the absence of receptor expression (data not shown). These data thus demonstrate that the alteration in APP processing induced by δOR-Cys27 expression can be suppressed by long-term treatment with a subset of receptor-specific ligands. However, this is not related to inefficient maturation of the δOR per se. Rather, the receptor variant-specific effects are more likely related to other differences between the two δOR variants, namely, to the altered cell surface stability and trafficking properties of the δOR-Cys27 variant (20).

Altered APP processing is coupled to changes in the endocytic trafficking of the δOR-Cys27 variant.

The observed APP processing phenotype in the δOR-Cys27-overexpressing cells appears to relate to the altered behavior of mature receptors that have reached the cell surface. Thus, we next examined whether specific signal transduction pathways linked to δOR might be involved. We reasoned that constitutive internalization of the δOR-Cys27 variant (20) might lead to the activation of signaling pathways in the absence of agonist-mediated receptor activation. We therefore screened the effects of inhibitors of several GPCR signaling molecules on APP processing in HEK293_i-δOR-Cys27 cells. Pertussis toxin, a reagent that uncouples the receptor from the G_i protein, U73122, an inhibitor of phospholipase C, or H-89, an inhibitor of protein kinase A, showed no discernible effects (data not shown). In contrast, PP2, a c-Src tyrosine kinase inhibitor, strongly suppressed the δOR-Cys27 expression-induced increase in the APP C83 levels (Fig. 6 A). Furthermore, it increased the amount of δORs at the cell surface by about 30% (Fig. 6B). A prolonged incubation of the δOR-Cys27-overexpressing cells with PP2 caused a moderate decrease in the expression of full-length APP (Fig. 6A). However, the maturation efficiency of APP was unchanged. Importantly, the PP2-mediated suppression of the elevated APP C83 levels in δOR-Cys27-expressing cells was still significant when the amount of APP C83 was normalized to the amount of full-length APP (data not shown).

Fig. 6. — c-Src inhibition suppresses the δOR-Cys27 expression-induced elevation of APP C83 levels and constitutive receptor internalization. (A) HEK293_i-δOR-Cys27 cells were induced or not induced for 48 h, and the c-Src-inhibitor PP2 was added or not added for the last 24 h as indicated. Total protein lysates were analyzed by Western blotting. ***, P < 0.001, n = 4. Numbers at the right of the blots are molecular masses (in kilodaltons). (B) HEK293_i-δOR-Cys27 cells were induced and treated with PP2 (5 μM) as described for panel A. Cell surface receptors were labeled with the anti-c-Myc 9E10 antibody followed by the phycoerythrin-conjugated secondary antibody. The fluorescence intensity of live cells was measured by flow cytometry. A representative histogram (n = 3) is shown. Curves shaded black are results with no receptor expression induced, gray curves reflect results with the vehicle control, and black curves reflect results with PP2. (C) HEK293_i-δOR-Cys27 cells were induced for 24 h, and cell surface receptors were labeled for 30 min at +37°C with the anti-c-Myc 9E10 antibody in the absence or presence of 4 μM PP2. After being washed, cells were harvested immediately or chased for 4 h at +37°C in the continued absence or presence of PP2, and the remaining antibody-labeled receptors at the cell surface were stained with the phycoerythrin-conjugated secondary antibody and analyzed by flow cytometry. *, P < 0.05, n = 3.

c-Src is implicated in agonist-mediated δOR desensitization, internalization, and recycling (2, 13), and it constitutively phosphorylates δOR (3). We therefore examined whether inhibition of c-Src activity by PP2 in HEK293_i-δOR-Cys27 cells inhibits constitutive receptor internalization. Indeed, in the flow cytometer internalization assay, a significantly higher number of receptors remained at the cell surface in PP2-treated cells than in controls following a 4-h treatment (Fig. 6C). Thus, it is likely that the observed PP2-induced increase in the steady-state amount of the δOR-Cys27 variant at the cell surface (Fig. 6B) was due to a decrease in constitutive receptor internalization. Together, these results suggest that constitutive internalization of the δOR-Cys27 variant and/or a closely coupled c-Src-dependent process plays a part in the altered processing of APP in cells overexpressing δOR-Cys27.

Ectopic treatments that impair normal endosomal trafficking can mimic the δOR-Cys27 variant-induced changes in APP processing.

Like pH-neutralizing conditions (43), inhibition of protein kinase C (PKC) is known to alter the endocytic trafficking of internalized cell surface proteins to the perinuclear area (15). Therefore, it was interesting that in the screen of inhibitors of GPCR signaling molecules, bisindolylmaleimide I (Bis I), a specific inhibitor of all PKC isoforms, was found to change APP processing in a manner that closely mimicked the one caused by δOR-Cys27 overexpression. Exposure of uninduced HEK293_i-δOR-Cys27 cells (and a parental cell line, HEK293_i [data not shown]) to increasing concentrations of Bis I led to the accumulation of APP CTFs and AICD in a concentration-dependent manner without an increase in the sAPPα levels (Fig. 7 A and B). As expected, an increased sAPPα secretion was observed when PKC was activated by phorbol-12-myristate-13 acetate (PMA) (35). These results suggest that the δOR-Cys27 expression-induced, agonist-independent, increase in the APP C83 levels is related to disturbances in a constitutive cellular process(es) rather than to PKC-dependent or -independent induction of the nonamyloidogenic APP processing described for several other GPCRs upon agonist-mediated activation (8, 29, 30). It is conceivable that this phenotype is related to an impairment of PKC-dependent cellular trafficking. To test this, the SH-SY5Y-APP751 cells were exposed to Bis I for 4 h, followed by an analysis of APP subcellular localization by confocal microscopy using an antibody directed against the APP C terminus. APP was most intensely localized in a compact juxtanuclear cluster of vesicles in the nontreated cells, but inhibition of PKC caused a major dispersion of the APP-positive vesicles to the cell periphery (Fig. 7C). This is in line with the notion that altered processing and impaired trafficking of APP may be coupled.

Expression of the δOR-Cys27 variant leads to a decreased localization of APP C-terminal fragments in the LEL compartment of SH-SY5Y-APP751 cells.

APP C83 has previously been shown to accumulate under pH-neutralizing conditions. This was suggested to take place because of reduced degradation of C83 in the LEL compartment as well as through partial α-secretase-mediated conversion of C99 to C83 (45). Therefore, we wanted to elucidate whether overexpression of the δOR-Cys27 variant affected the subcellular localization of APP CTFs in SH-SY5Y-APP751 cells (Fig. 8). Confocal microscopy using an antibody recognizing the APP C terminus revealed that APP was localized mostly in the δOR-Phe27-expressing and untransfected cells in a compact juxtanuclear structure with a partial localization at or near the plasma membrane (Fig. 8A). In the δOR-Cys27-overexpressing cells, however, APP-positive juxtanuclear structures were rarely observed. Rather, APP was localized ubiquitously in these cells, suggesting that APP dispersed away from the juxtanuclear compartment(s) in δOR-Cys27-transfected cells (Fig. 8A). Costaining of the untransfected cells with APP N- and C-terminal antibodies revealed that the APP staining in the juxtanuclear structure originated mainly from the APP CTFs and not from the full-length form of APP (Fig. 3C).

Next, we wanted to characterize the juxtanuclear structure in which the APP CTFs predominantly localized in the untransfected and δOR-Phe27-overexpressing cells. Using markers specific for different subcellular compartments (Fig. 8B and 9), juxtanuclear APP staining was found to strongly colocalize with EGFP-Rab7, a LEL marker, in untransfected SH-SY5Y-APP751 cells. Partial colocalization was also observed with the EYFP-Golgi marker, indicating that the APP CTFs resided predominantly in the LEL and Golgi compartments. Quantification of the transfected cells containing APP-positive juxtanuclear staining revealed an average 60% decrease in the number of cells containing APP CTFs in the LEL/Golgi compartments in δOR-Cys27-transfected cells compared to δOR-Phe27-transfected ones (P < 0.05) (Fig. 8C). In cells overexpressing the δOR-Cys27 variant, APP CTFs showed a partial colocalization with the endosomal marker LAMP2 and the multivesicular body (MVB) marker lysobisphoshatidic acid (LBPA) (data not shown), but the exact nature of these compartments could not be determined by confocal microscopy. Therefore, we investigated the effect of the δOR-Cys27 variant on the targeting of APP CTFs by immunoelectron microscopy in HEK293 cells. The number of LBPA-positive vesicle clusters with or without the MVB-type outer membrane was increased in cells induced to express the δOR-Cys27 variant compared to the number in uninduced controls. Interestingly, APP CTFs were abundantly concentrated in these vesicles (Fig. 10). These vesicles may represent intermediates which have failed to mature and/or move to the LEL/Golgi compartments. Collectively, the present results suggest that overexpression of the δOR-Cys27 variant leads to a decreased localization of APP CTFs in late endosomes and lysosomes in SH-SY5Y-APP751 cells. The altered APP processing phenotype caused by δOR-Cys27 expression is thus different from the one resulting from classical nonamyloidogenic APP processing and is most likely related to constitutive δOR internalization and altered endocytic trafficking.

Fig. 9. — Characterization of APP subcellular localization in SH-SY5Y-APP751 cells utilizing subcellular markers. APP localization in subcellular compartments was analyzed by confocal microscope analysis using the C-terminal anti-APP antibody (A8717 = red) and antibodies against calnexin (endoplasmic reticulum marker), the transferrin receptor (TfR; a marker for plasma membrane and early endosomes), or LBPA (a marker for MVBs) or after cotransfection with EGFP-Rab11 (endosomal recycling compartment marker) or EGFP-Rab9 (lysosomal marker) (all green) in SH-SY5Y-APP751 cells. A partial colocalization of APP was observed with TfR and EGFP-Rab9 (yellow), whereas APP colocalization with other markers was not evident (right panels, merged). Bar, 10 μm.

Fig. 10. — δOR-Cys27 overexpression induces increased targeting of APP CTFs into LBPA-positive vesicle clusters and MVBs. HEK293_i-δOR-Cys27 cells were induced for 24 h and analyzed by electron microscopy. LBPA, an MVB marker labeled with 5-nm gold particles (white arrows), shows colocalization with APP CTFs (10-nm gold particles, black arrows) in organelles with typical MVB morphology (the outer membrane is marked with arrowheads) and also in vesicle clusters without a surrounding membrane.

Genetic analysis of OPRD1 gene polymorphisms in case-control and family-based AD cohorts.

Reinforced by the functional findings related to the δOR-Phe27Cys variation in AD patients, we finally wanted to assess in an unbiased manner whether or not genetic variations in OPRD1 affect the risk for AD among a Finnish case-control cohort consisting of, all together, ∼1,200 AD patients and controls. In addition to the T80G variation (rs1042114), which leads to the Phe27Cys substitution, four tag SNPs from the OPRD1 gene were selected for genotyping based on the haplotype block structure in the central European population (Table 1). According to the allele and genotype association analyses, only rs1042114 showed a borderline genotype association with AD (P = 0.07) (Table 1), which was driven by TG heterozygotes, who were marginally more frequent among AD cases than among controls [P = 0.07; odds ratio [OR] = 2.5; 95% confidence interval [CI], 0.94 to 6.50). Ages of onset and cerebrospinal fluid Aβ42 analyses with respect to the rs1042114 variation did not reveal statistically significant changes among the AD patients (data not shown).

To independently replicate the results observed with rs1042114, we extracted this SNP from genome-wide association study (GWAS) data sets obtained from Caucasian multiplex AD families from the National Institute of Mental Health (NIMH) Genetics Initiative Study (7). As the rs1042114 variant was not included on the Affymetrix 500K GWAS array, we inferred genotypes of this SNP by imputation via IMPUTE v2.0 using HapMap CEU and 1000 Genomes Project pilot release data. In these independent, family-based data, we also observed an overtransmission of the heterozygous TG genotype to those affected (P = 0.021). Using conditional logistic regression stratified by family, this translated into an OR of 1.62 (95% CI, 1.00 to 2.64). Combining these data with the case-control results via random-effects meta-analysis yielded an OR of 1.77 (95% CI, 1.14 to 2.74; P = 0.011). Collectively, these data suggest that the Phe27Cys variation in δOR may play a role in AD.

DISCUSSION

It was recently shown that δOR is linked to AD pathogenesis by affecting Aβ generation in vitro and in vivo (28, 40). Keeping in mind this intimate link, we assessed whether the phenylalanine-to-cysteine substitution at position 27 of δOR plays a functional role in APP processing in vitro. We have previously shown that cysteine at position 27 affects the maturation and subcellular localization of δOR in nonneuronal cells (20, 32). More specifically, the δOR-Cys27 variant showed a decreased mature/precursor receptor ratio, which was related to the retention of receptor precursors in the endoplasmic reticulum and enhanced turnover of mature cell surface receptors. Based on this, it was proposed that the δOR-Cys27 variant may cause a gain-of-function phenotype with possible pathophysiological consequences due to the intracellular accumulation of the receptor (20). Here, our experiments with human SH-SY5Y neuroblastoma cells overexpressing APP corroborated the previous data from HEK cells and demonstrated a less efficient maturation of the δOR-Cys27 variant than that of the δOR-Phe27 variant. This finding indicates that the receptor phenotypes with respect to the phenylalanine-to-cysteine substitution are similar in neuronal and nonneuronal cells. Furthermore, we observed, in parallel with δOR-Cys27 expression, a robust accumulation of APP C83 and AICD, but not APP C99, in both SH-SY5Y and HEK293 cells that was not evident in cells expressing δOR-Phe27. The similar results in cells expressing exogenous or endogenous APP confirm that the observed changes in APP processing are not caused by APP overexpression per se. Further characterization of the altered APP processing phenotype in δOR-Cys27-expressing cells revealed that the sAPPα levels in the cell culture medium were not altered but that the total sAPP (= sAPPα + sAPPβ) and, particularly, the Aβ40 levels were significantly decreased. This suggests that cellular events affecting APP processing, such as the altered trafficking of newly synthesized APP in the secretory pathway, are not the cause of the observed changes in δOR-Cys27-expressing cells. Moreover, the α-secretase activity in SH-SY5Y-APP751 cells expressing the δOR-Cys27 variant was not enhanced, which excludes the possibility that the accumulation of C83 is a consequence of increased α-secretase-mediated cleavage of APP. We also did not find evidence for a decreased γ-secretase activity in SH-SY5Y-APP751 cells expressing δOR-Cys27 that could explain the reduced Aβ40 levels. This finding emphasizes the notion that the agonist-independent effects of the δOR-Cys27 variant are different from those triggered by agonist-mediated δOR activation, which has been shown to increase γ-secretase activity (28, 40). Interestingly, though, we observed that the changes in APP processing induced by the δOR-Cys27 variant can be suppressed by a long-term treatment with a subset of δOR ligands. This may occur via a mechanism which involves altered trafficking and stability, but not inefficient maturation of the receptor. However, even though the altered APP processing was not associated with the intracellular accumulation of δOR-Cys27 precursors, we cannot exclude the possibility that the latter phenotype has other adverse effects in the pathogenesis of AD.

The above-mentioned changes in APP processing imply the possibility that the δOR-Cys27 variant influences the endocytic trafficking of APP. Confocal microscopy of δOR-Phe27-expressing and untransfected SH-SY5Y-APP751 cells showed similar subcellular localizations of APP in juxtanuclear structures confirmed as LEL and Golgi compartments. Interestingly, APP CTFs in particular were present in these compartments, whereas full-length APP resided mostly on or close to the plasma membrane. In contrast, the number of cells with APP-positive juxtanuclear LEL and Golgi compartments was significantly reduced in δOR-Cys27-expressing cells, suggesting that APP CTFs were dispersed from these compartments to localize diffusely in the cells. Since APP C83 showed a robust accumulation in the total protein extracts from δOR-Cys27-expressing cells, it is conceivable that the APP CTF-positive vesicles are consequently dispersed away from their normal localization. In this context, it should be noted that the decreased APP CTF localization in LEL is not associated with a differential interaction between the δOR variants and APP, as our coimmunoprecipitation experiments did not provide proof for a direct interaction between the δOR variants and APP. Collectively, our results suggest that in δOR-Cys27-expressing cells, the endocytic trafficking of APP CTFs to LEL is reduced, which leads to the accumulation of APP C83 due to its inefficient degradation.

δOR-Cys27 variant expression was found to result in altered APP processing that is intriguingly similar to the APP processing phenotype under pH-neutralizing conditions (45). Under those conditions in ammonium chloride-treated CHO cells, the accumulation specifically of APP C83, but not APP C99, was detected in conjunction with decreased Aβ and unaltered sAPPα levels. APP C99 was also partially converted to APP C83 through an α-secretase-mediated cleavage in CHO cells stably overexpressing APP C99 under pH-neutralizing conditions, explaining the decreased APP C99/C83 ratio (45). Consistent with this finding, it was recently shown that C99 and C89 are processed by α-secretase (ADAM10) to C83 (16). Also, AICD levels were significantly increased, suggesting that AICD was generated predominantly from the surplus of APP C83 through γ-secretase-mediated ε-site cleavage. This interpretation is in line with findings showing that AICD is generated by γ-secretase at the plasma membrane and/or early endosomes (14) but that Aβ is produced primarily in LEL (10, 17). Related to this, it was recently shown that alkalizing drugs, such as bafilomycin A₁, induces the accumulation of APP C83 and AICD in luminal MVB vesicles (43). MVBs are transport intermediates between early and late endosomes involved in receptor recycling and protein degradation processes (11). Interestingly, MVBs are also suggested to play a role in AD pathogenesis since several APP derivatives accumulate in these vesicles in AD patients (38) as well as in a transgenic mouse model of amyloidosis (18). These similarities between alkalization and δOR-Cys27 variant phenotypes in terms of altered APP processing suggest the involvement of overlapping molecular mechanisms which may be related to changes in the endocytic trafficking pathway.

Since the δOR-Cys27 variant is internalized extensively from the cell surface under agonist-independent conditions, in contrast to δOR-Phe27 (20), there may be a difference in constitutive downstream signaling between the δOR variants. We found that a c-Src tyrosine kinase inhibitor, PP2, suppressed the APP processing phenotype induced by the expression of the δOR-Cys27 variant in a dose-dependent manner. In conjunction with the altered APP processing phenotype, a significantly higher number of receptors resided at the plasma membrane than in the control cells and the constitutive internalization of cell surface receptors was impaired following c-Src inhibition. This is an important finding from a mechanistic point of view since c-Src has been shown to participate in agonist-induced δOR internalization and recycling (2, 13). As PP2 does not discriminate between tyrosine kinases in the Src-family, it is also possible that other Src-related kinases, like Fyn, could be involved. Interestingly, we also observed that the specific PKC inhibitor Bis I dose dependently induced an APP processing phenotype similar to that of the δOR-Cys27 variant and caused the accumulation of APP C83 and AICD. Furthermore, PKC inhibition resulted in the dispersion of APP CTF-positive vesicles into the cell periphery in a manner analogous to that of δOR-Cys27-expressing cells. All together, our results suggest that c-Src and possibly also PKC activities play a role in the regulation of molecular events underlying the altered endocytic trafficking of APP CTFs in cells expressing the δOR-Cys27 variant (Fig. 11). Furthermore, the results allow us to conclude that the altered APP processing caused by δOR-Cys27 expression is different from the classical nonamyloidogenic APP processing pathway and that it is related to constitutive receptor internalization and altered endocytic trafficking.

Fig. 11. — Potential mechanism for the δOR-Cys27 variant-induced alteration in APP processing. APP can be processed via the nonamyloidogenic (I) or the amyloidogenic (II) pathway. γ-Secretase-induced intramembranous cleavage in the juxtanuclear late endosome/lysosome (LEL) compartment (III) is preceded by ectodomain shedding by α-secretase at the cell surface (I) or by β-secretase in the endosomal compartment (II). Agonist-mediated activation of δOR (IV) has previously been shown to enhance the amyloidogenic route by coendocytic targeting of β- and γ-secretases (40), leading to the increased production of Aβ. Expression of the δOR-Cys27 variant (V), which displays enhanced constitutive internalization (20), inhibits the trafficking of APP CTFs to LEL and causes APP C83 accumulation in the dispersed MVB-like compartments (VI). The changed processing and targeting phenotype of APP in δOR-Cys27-expressing cells is identical to those caused by PKC inhibition or ectopic alkalization (45), thus suggesting similar or common underlying molecular mechanisms. A long-term treatment with δOR-specific antagonists/inverse agonists (VII) or with a c-Src tyrosine kinase inhibitor (VIII) suppresses the accumulation of δOR-Cys27 variant-induced APP C83. This indicates that the constitutive agonist-independent internalization and/or related signaling of δOR-Cys27 results in the altered trafficking and processing of APP. In contrast, the δOR-Phe27 variant, which harbors a higher intrinsic stability at the cell surface (IX) (20), does not induce a similar process. EE, early endosomes.

The fact that the δOR-Cys27 variant robustly affected the endocytic trafficking of APP prompted us to elucidate in an unbiased manner also the genetic variants of OPRD1 in two independent AD cohorts. Interestingly, our data suggest that the rs1042114 variation may play a role in AD since the meta-analysis of combined case-control and familial data indicated a statistically significant risk effect for TG heterozygotes. Since the T→G allele change at the rs1042114 site leads to the phenylalanine-to-cysteine substitution in the N-terminal domain of δOR, it is possible that this variation itself rather than some other proximal alteration may have a functional relevance in AD, which is in LD with rs1042114. In this context, however, it should be noted that additional replication studies of different sample cohorts will be needed before making any firm conclusions regarding the δOR-Phe27Cys variation in AD.

Our results have significant relevance in the context of AD pathogenesis, as the intervention approaches focusing on the formation and trafficking of the GPCR/β- and γ-secretase complex have been suggested as a novel strategy against AD (28, 40). Since ∼25% of AD patients are heterozygous for the δOR-Phe27Cys variation, it is possible that these patients would respond to, e.g., δOR antagonist/agonist treatments differently from δOR-Phe27 homozygotes. These possibilities emphasize the importance of determining the genetic background of the δOR-Phe27Cys variation on a case-by-case basis. Moreover, we do not yet know the functional role of δOR heteromers composed of the two δOR variants, or even other opioid receptor subtypes, in receptor signaling and in other events, such as APP processing in vitro and in vivo. Taken together, our results underscore the importance of taking into account the genetic profile of the δOR-Phe27Cys variation when developing δOR-related therapeutic strategies against AD.

ACKNOWLEDGMENTS

This study was supported by strategic funding from the University of Eastern Finland and by grants from the Health Research Council of the Academy of Finland (grant 14133 to A.H. and grant 127199 to H.S., M.H., and U.E.P.-R.), the Sigrid Juselius Foundation (to U.E.P.-R. and M.H.), the Kuopio University Hospital (EVO grant 5772708 to H.S.), the Nordic Centre of Excellence in Neurodegeneration (to H.S. and M.H.), the Finnish Cultural Foundation (to J.T.T.), the Cure Alzheimer's Fund and NIA (to R.E.T.), and the Finnish Funding Agency for Technology and Innovation (grant 70048/09 to H.S.).

We are grateful to Minna Männikkö for her help in the initial estimation of SNP rs1042114 allele frequencies in a control population and to Sirpa Kellokumpu for preparation of the electron microscopic samples.

Footnotes

^▿

Published ahead of print on 4 April 2011.

REFERENCES

1. Apaja P. M., Tuusa J. T., Pietila E. M., Rajaniemi H. J., Petaja-Repo U. E. 2006. Luteinizing hormone receptor ectodomain splice variant misroutes the full-length receptor into a subcompartment of the endoplasmic reticulum. Mol. Biol. Cell 17:2243–2255 [DOI] [PMC free article] [PubMed] [Google Scholar]
2. Archer-Lahlou E., et al. 2009. Src promotes delta opioid receptor (DOR) desensitization by interfering with receptor recycling. J. Cell. Mol. Med. 13:147–163 [DOI] [PMC free article] [PubMed] [Google Scholar]
3. Audet N., Paquin-Gobeil M., Landry-Paquet O., Schiller P. W., Pineyro G. 2005. Internalization and Src activity regulate the time course of ERK activation by delta opioid receptor ligands. J. Biol. Chem. 280:7808–7816 [DOI] [PubMed] [Google Scholar]
4. Bertram L., et al. 2008. Genome-wide association analysis reveals putative Alzheimer's disease susceptibility loci in addition to APOE. Am. J. Hum. Genet. 83:623–632 [DOI] [PMC free article] [PubMed] [Google Scholar]
5. Bertram L., McQueen M. B., Mullin K., Blacker D., Tanzi R. E. 2007. Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database. Nat. Genet. 39:17–23 [DOI] [PubMed] [Google Scholar]
6. Bie B., et al. 2010. Nerve growth factor-regulated emergence of functional delta-opioid receptors. J. Neurosci. 30:5617–5628 [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Blacker D., et al. 1997. ApoE-4 and age at onset of Alzheimer's disease: the NIMH genetics initiative. Neurology 48:139–147 [DOI] [PubMed] [Google Scholar]
8. Camden J. M., et al. 2005. P2Y2 nucleotide receptors enhance alpha-secretase-dependent amyloid precursor protein processing. J. Biol. Chem. 280:18696–18702 [DOI] [PubMed] [Google Scholar]
9. Choudhury A., et al. 2002. Rab proteins mediate Golgi transport of caveola-internalized glycosphingolipids and correct lipid trafficking in Niemann-Pick C cells. J. Clin. Invest. 109:1541–1550 [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Golde T. E., Estus S., Younkin L. H., Selkoe D. J., Younkin S. G. 1992. Processing of the amyloid protein precursor to potentially amyloidogenic derivatives. Science 255:728–730 [DOI] [PubMed] [Google Scholar]
11. Gruenberg J., Stenmark H. 2004. The biogenesis of multivesicular endosomes. Nat. Rev. Mol. Cell Biol. 5:317–323 [DOI] [PubMed] [Google Scholar]
12. Hiltunen M., et al. 2006. Ubiquilin 1 modulates amyloid precursor protein trafficking and Abeta secretion. J. Biol. Chem. 281:32240–32253 [DOI] [PubMed] [Google Scholar]
13. Hong M. H., et al. 2009. Role of Src in ligand-specific regulation of delta-opioid receptor desensitization and internalization. J. Neurochem. 108:102–114 [DOI] [PubMed] [Google Scholar]
14. Kaether C., Schmitt S., Willem M., Haass C. 2006. Amyloid precursor protein and Notch intracellular domains are generated after transport of their precursors to the cell surface. Traffic 7:408–415 [DOI] [PubMed] [Google Scholar]
15. Kermorgant S., Zicha D., Parker P. J. 2003. Protein kinase C controls microtubule-based traffic but not proteasomal degradation of c-Met. J. Biol. Chem. 278:28921–28929 [DOI] [PubMed] [Google Scholar]
16. Kuhn P. H., et al. 2010. ADAM10 is the physiologically relevant, constitutive alpha-secretase of the amyloid precursor protein in primary neurons. EMBO J. 29:3020–3032 [DOI] [PMC free article] [PubMed] [Google Scholar]
17. LaFerla F. M., Green K. N., Oddo S. 2007. Intracellular amyloid-beta in Alzheimer's disease. Nat. Rev. Neurosci. 8:499–509 [DOI] [PubMed] [Google Scholar]
18. Langui D., et al. 2004. Subcellular topography of neuronal Abeta peptide in APPxPS1 transgenic mice. Am. J. Pathol. 165:1465–1477 [DOI] [PMC free article] [PubMed] [Google Scholar]
19. Lehtovirta M., et al. 1996. Clinical and neuropsychological characteristics in familial and sporadic Alzheimer's disease: relation to apolipoprotein E polymorphism. Neurology 46:413–419 [DOI] [PubMed] [Google Scholar]
20. Leskela T. T., Markkanen P. M., Alahuhta I. A., Tuusa J. T., Petaja-Repo U. E. 2009. Phe27Cys polymorphism alters the maturation and subcellular localization of the human delta opioid receptor. Traffic 10:116–129 [DOI] [PubMed] [Google Scholar]
21. Leskela T. T., Markkanen P. M., Pietila E. M., Tuusa J. T., Petaja-Repo U. E. 2007. Opioid receptor pharmacological chaperones act by binding and stabilizing newly synthesized receptors in the endoplasmic reticulum. J. Biol. Chem. 282:23171–23183 [DOI] [PubMed] [Google Scholar]
22. Lichtenthaler S. F., et al. 2003. The cell adhesion protein P-selectin glycoprotein ligand-1 is a substrate for the aspartyl protease BACE1. J. Biol. Chem. 278:48713–48719 [DOI] [PubMed] [Google Scholar]
23. Marchini J., Howie B., Myers S., McVean G., Donnelly P. 2007. A new multipoint method for genome-wide association studies by imputation of genotypes. Nat. Genet. 39:906–913 [DOI] [PubMed] [Google Scholar]
24. Markkanen P. M., Petäjä-Repo U. E. 2008. N-glycan-mediated quality control in the endoplasmic reticulum is required for the expression of correctly folded delta-opioid receptors at the cell surface. J. Biol. Chem. 283:29086–29098 [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Mathieu-Kia A. M., Fan L. Q., Kreek M. J., Simon E. J., Hiller J. M. 2001. Mu-, delta- and kappa-opioid receptor populations are differentially altered in distinct areas of postmortem brains of Alzheimer's disease patients. Brain Res. 893:121–134 [DOI] [PubMed] [Google Scholar]
26. McKhann G., et al. 1984. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology 34:939–944 [DOI] [PubMed] [Google Scholar]
27. Neumann S., et al. 2006. Amyloid precursor-like protein 1 influences endocytosis and proteolytic processing of the amyloid precursor protein. J. Biol. Chem. 281:7583–7594 [DOI] [PubMed] [Google Scholar]
28. Ni Y., et al. 2006. Activation of beta2-adrenergic receptor stimulates gamma-secretase activity and accelerates amyloid plaque formation. Nat. Med. 12:1390–1396 [DOI] [PubMed] [Google Scholar]
29. Nitsch R. M., et al. 1993. Receptor-coupled amyloid precursor protein processing. Ann. N. Y. Acad. Sci. 695:122–127 [DOI] [PubMed] [Google Scholar]
30. Nitsch R. M., Slack B. E., Wurtman R. J., Growdon J. H. 1992. Release of Alzheimer amyloid precursor derivatives stimulated by activation of muscarinic acetylcholine receptors. Science 258:304–307 [DOI] [PubMed] [Google Scholar]
31. Petaja-Repo U. E., et al. 2002. Ligands act as pharmacological chaperones and increase the efficiency of delta opioid receptor maturation. EMBO J. 21:1628–1637 [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Petaja-Repo U. E., Hogue M., Laperriere A., Walker P., Bouvier M. 2000. Export from the endoplasmic reticulum represents the limiting step in the maturation and cell surface expression of the human delta opioid receptor. J. Biol. Chem. 275:13727–13736 [DOI] [PubMed] [Google Scholar]
33. Petaja-Repo U. E., et al. 2006. Distinct subcellular localization for constitutive and agonist-modulated palmitoylation of the human delta opioid receptor. J. Biol. Chem. 281:15780–15789 [DOI] [PubMed] [Google Scholar]
34. Pinnix I., et al. 2001. A novel gamma-secretase assay based on detection of the putative C-terminal fragment-gamma of amyloid beta protein precursor. J. Biol. Chem. 276:481–487 [DOI] [PubMed] [Google Scholar]
35. Racchi M., Mazzucchelli M., Pascale A., Sironi M., Govoni S. 2003. Role of protein kinase C alpha in the regulated secretion of the amyloid precursor protein. Mol. Psychiatry 8:209–216 [DOI] [PubMed] [Google Scholar]
36. Sarajarvi T., et al. 2009. Down-regulation of seladin-1 increases BACE1 levels and activity through enhanced GGA3 depletion during apoptosis. J. Biol. Chem. 284:34433–34443 [DOI] [PMC free article] [PubMed] [Google Scholar]
37. Schjeide B. M., et al. 2009. Assessment of Alzheimer's disease case-control associations using family-based methods. Neurogenetics 10:19–25 [DOI] [PMC free article] [PubMed] [Google Scholar]
38. Takahashi R. H., et al. 2002. Intraneuronal Alzheimer abeta42 accumulates in multivesicular bodies and is associated with synaptic pathology. Am. J. Pathol. 161:1869–1879 [DOI] [PMC free article] [PubMed] [Google Scholar]
39. Tanzi R. E., Bertram L. 2005. Twenty years of the Alzheimer's disease amyloid hypothesis: a genetic perspective. Cell 120:545–555 [DOI] [PubMed] [Google Scholar]
40. Teng L., Zhao J., Wang F., Ma L., Pei G. 2010. A GPCR/secretase complex regulates beta- and gamma-secretase specificity for Abeta production and contributes to AD pathogenesis. Cell Res. 20:138–153 [DOI] [PubMed] [Google Scholar]
41. Tuusa J. T., Markkanen P. M., Apaja P. M., Hakalahti A. E., Petaja-Repo U. E. 2007. The endoplasmic reticulum Ca2+-pump SERCA2b interacts with G protein-coupled receptors and enhances their expression at the cell surface. J. Mol. Biol. 371:622–638 [DOI] [PubMed] [Google Scholar]
42. van Rijn R. M., Whistler J. L., Waldhoer M. 2010. Opioid-receptor-heteromer-specific trafficking and pharmacology. Curr. Opin. Pharmacol. 10:73–79 [DOI] [PMC free article] [PubMed] [Google Scholar]
43. Vingtdeux V., et al. 2007. Alkalizing drugs induce accumulation of amyloid precursor protein by-products in luminal vesicles of multivesicular bodies. J. Biol. Chem. 282:18197–18205 [DOI] [PubMed] [Google Scholar]
44. Waldhoer M., Bartlett S. E., Whistler J. L. 2004. Opioid receptors. Annu. Rev. Biochem. 73:953–990 [DOI] [PubMed] [Google Scholar]
45. Waldron E., et al. 2008. Increased AICD generation does not result in increased nuclear translocation or activation of target gene transcription. Exp. Cell Res. 314:2419–2433 [DOI] [PubMed] [Google Scholar]
46. Xuei X., et al. 2007. The opioid system in alcohol and drug dependence: family-based association study. Am. J. Med. Genet. B Neuropsychiatr. Genet. 144B:877–884 [DOI] [PubMed] [Google Scholar]
47. Zhang H., Kranzler H. R., Yang B. Z., Luo X., Gelernter J. 2008. The OPRD1 and OPRK1 loci in alcohol or drug dependence: OPRD1 variation modulates substance dependence risk. Mol. Psychiatry 13:531–543 [DOI] [PMC free article] [PubMed] [Google Scholar]
48. Zhang Z., Pan Z. Z. 2010. Synaptic mechanism for functional synergism between delta- and mu-opioid receptors. J. Neurosci. 30:4735–4745 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B1] 1. Apaja P. M., Tuusa J. T., Pietila E. M., Rajaniemi H. J., Petaja-Repo U. E. 2006. Luteinizing hormone receptor ectodomain splice variant misroutes the full-length receptor into a subcompartment of the endoplasmic reticulum. Mol. Biol. Cell 17:2243–2255 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B2] 2. Archer-Lahlou E., et al. 2009. Src promotes delta opioid receptor (DOR) desensitization by interfering with receptor recycling. J. Cell. Mol. Med. 13:147–163 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B3] 3. Audet N., Paquin-Gobeil M., Landry-Paquet O., Schiller P. W., Pineyro G. 2005. Internalization and Src activity regulate the time course of ERK activation by delta opioid receptor ligands. J. Biol. Chem. 280:7808–7816 [DOI] [PubMed] [Google Scholar]

[B4] 4. Bertram L., et al. 2008. Genome-wide association analysis reveals putative Alzheimer's disease susceptibility loci in addition to APOE. Am. J. Hum. Genet. 83:623–632 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B5] 5. Bertram L., McQueen M. B., Mullin K., Blacker D., Tanzi R. E. 2007. Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database. Nat. Genet. 39:17–23 [DOI] [PubMed] [Google Scholar]

[B6] 6. Bie B., et al. 2010. Nerve growth factor-regulated emergence of functional delta-opioid receptors. J. Neurosci. 30:5617–5628 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B7] 7. Blacker D., et al. 1997. ApoE-4 and age at onset of Alzheimer's disease: the NIMH genetics initiative. Neurology 48:139–147 [DOI] [PubMed] [Google Scholar]

[B8] 8. Camden J. M., et al. 2005. P2Y2 nucleotide receptors enhance alpha-secretase-dependent amyloid precursor protein processing. J. Biol. Chem. 280:18696–18702 [DOI] [PubMed] [Google Scholar]

[B9] 9. Choudhury A., et al. 2002. Rab proteins mediate Golgi transport of caveola-internalized glycosphingolipids and correct lipid trafficking in Niemann-Pick C cells. J. Clin. Invest. 109:1541–1550 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B10] 10. Golde T. E., Estus S., Younkin L. H., Selkoe D. J., Younkin S. G. 1992. Processing of the amyloid protein precursor to potentially amyloidogenic derivatives. Science 255:728–730 [DOI] [PubMed] [Google Scholar]

[B11] 11. Gruenberg J., Stenmark H. 2004. The biogenesis of multivesicular endosomes. Nat. Rev. Mol. Cell Biol. 5:317–323 [DOI] [PubMed] [Google Scholar]

[B12] 12. Hiltunen M., et al. 2006. Ubiquilin 1 modulates amyloid precursor protein trafficking and Abeta secretion. J. Biol. Chem. 281:32240–32253 [DOI] [PubMed] [Google Scholar]

[B13] 13. Hong M. H., et al. 2009. Role of Src in ligand-specific regulation of delta-opioid receptor desensitization and internalization. J. Neurochem. 108:102–114 [DOI] [PubMed] [Google Scholar]

[B14] 14. Kaether C., Schmitt S., Willem M., Haass C. 2006. Amyloid precursor protein and Notch intracellular domains are generated after transport of their precursors to the cell surface. Traffic 7:408–415 [DOI] [PubMed] [Google Scholar]

[B15] 15. Kermorgant S., Zicha D., Parker P. J. 2003. Protein kinase C controls microtubule-based traffic but not proteasomal degradation of c-Met. J. Biol. Chem. 278:28921–28929 [DOI] [PubMed] [Google Scholar]

[B16] 16. Kuhn P. H., et al. 2010. ADAM10 is the physiologically relevant, constitutive alpha-secretase of the amyloid precursor protein in primary neurons. EMBO J. 29:3020–3032 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B17] 17. LaFerla F. M., Green K. N., Oddo S. 2007. Intracellular amyloid-beta in Alzheimer's disease. Nat. Rev. Neurosci. 8:499–509 [DOI] [PubMed] [Google Scholar]

[B18] 18. Langui D., et al. 2004. Subcellular topography of neuronal Abeta peptide in APPxPS1 transgenic mice. Am. J. Pathol. 165:1465–1477 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B19] 19. Lehtovirta M., et al. 1996. Clinical and neuropsychological characteristics in familial and sporadic Alzheimer's disease: relation to apolipoprotein E polymorphism. Neurology 46:413–419 [DOI] [PubMed] [Google Scholar]

[B20] 20. Leskela T. T., Markkanen P. M., Alahuhta I. A., Tuusa J. T., Petaja-Repo U. E. 2009. Phe27Cys polymorphism alters the maturation and subcellular localization of the human delta opioid receptor. Traffic 10:116–129 [DOI] [PubMed] [Google Scholar]

[B21] 21. Leskela T. T., Markkanen P. M., Pietila E. M., Tuusa J. T., Petaja-Repo U. E. 2007. Opioid receptor pharmacological chaperones act by binding and stabilizing newly synthesized receptors in the endoplasmic reticulum. J. Biol. Chem. 282:23171–23183 [DOI] [PubMed] [Google Scholar]

[B22] 22. Lichtenthaler S. F., et al. 2003. The cell adhesion protein P-selectin glycoprotein ligand-1 is a substrate for the aspartyl protease BACE1. J. Biol. Chem. 278:48713–48719 [DOI] [PubMed] [Google Scholar]

[B23] 23. Marchini J., Howie B., Myers S., McVean G., Donnelly P. 2007. A new multipoint method for genome-wide association studies by imputation of genotypes. Nat. Genet. 39:906–913 [DOI] [PubMed] [Google Scholar]

[B24] 24. Markkanen P. M., Petäjä-Repo U. E. 2008. N-glycan-mediated quality control in the endoplasmic reticulum is required for the expression of correctly folded delta-opioid receptors at the cell surface. J. Biol. Chem. 283:29086–29098 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B25] 25. Mathieu-Kia A. M., Fan L. Q., Kreek M. J., Simon E. J., Hiller J. M. 2001. Mu-, delta- and kappa-opioid receptor populations are differentially altered in distinct areas of postmortem brains of Alzheimer's disease patients. Brain Res. 893:121–134 [DOI] [PubMed] [Google Scholar]

[B26] 26. McKhann G., et al. 1984. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology 34:939–944 [DOI] [PubMed] [Google Scholar]

[B27] 27. Neumann S., et al. 2006. Amyloid precursor-like protein 1 influences endocytosis and proteolytic processing of the amyloid precursor protein. J. Biol. Chem. 281:7583–7594 [DOI] [PubMed] [Google Scholar]

[B28] 28. Ni Y., et al. 2006. Activation of beta2-adrenergic receptor stimulates gamma-secretase activity and accelerates amyloid plaque formation. Nat. Med. 12:1390–1396 [DOI] [PubMed] [Google Scholar]

[B29] 29. Nitsch R. M., et al. 1993. Receptor-coupled amyloid precursor protein processing. Ann. N. Y. Acad. Sci. 695:122–127 [DOI] [PubMed] [Google Scholar]

[B30] 30. Nitsch R. M., Slack B. E., Wurtman R. J., Growdon J. H. 1992. Release of Alzheimer amyloid precursor derivatives stimulated by activation of muscarinic acetylcholine receptors. Science 258:304–307 [DOI] [PubMed] [Google Scholar]

[B31] 31. Petaja-Repo U. E., et al. 2002. Ligands act as pharmacological chaperones and increase the efficiency of delta opioid receptor maturation. EMBO J. 21:1628–1637 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B32] 32. Petaja-Repo U. E., Hogue M., Laperriere A., Walker P., Bouvier M. 2000. Export from the endoplasmic reticulum represents the limiting step in the maturation and cell surface expression of the human delta opioid receptor. J. Biol. Chem. 275:13727–13736 [DOI] [PubMed] [Google Scholar]

[B33] 33. Petaja-Repo U. E., et al. 2006. Distinct subcellular localization for constitutive and agonist-modulated palmitoylation of the human delta opioid receptor. J. Biol. Chem. 281:15780–15789 [DOI] [PubMed] [Google Scholar]

[B34] 34. Pinnix I., et al. 2001. A novel gamma-secretase assay based on detection of the putative C-terminal fragment-gamma of amyloid beta protein precursor. J. Biol. Chem. 276:481–487 [DOI] [PubMed] [Google Scholar]

[B35] 35. Racchi M., Mazzucchelli M., Pascale A., Sironi M., Govoni S. 2003. Role of protein kinase C alpha in the regulated secretion of the amyloid precursor protein. Mol. Psychiatry 8:209–216 [DOI] [PubMed] [Google Scholar]

[B36] 36. Sarajarvi T., et al. 2009. Down-regulation of seladin-1 increases BACE1 levels and activity through enhanced GGA3 depletion during apoptosis. J. Biol. Chem. 284:34433–34443 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B37] 37. Schjeide B. M., et al. 2009. Assessment of Alzheimer's disease case-control associations using family-based methods. Neurogenetics 10:19–25 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B38] 38. Takahashi R. H., et al. 2002. Intraneuronal Alzheimer abeta42 accumulates in multivesicular bodies and is associated with synaptic pathology. Am. J. Pathol. 161:1869–1879 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B39] 39. Tanzi R. E., Bertram L. 2005. Twenty years of the Alzheimer's disease amyloid hypothesis: a genetic perspective. Cell 120:545–555 [DOI] [PubMed] [Google Scholar]

[B40] 40. Teng L., Zhao J., Wang F., Ma L., Pei G. 2010. A GPCR/secretase complex regulates beta- and gamma-secretase specificity for Abeta production and contributes to AD pathogenesis. Cell Res. 20:138–153 [DOI] [PubMed] [Google Scholar]

[B41] 41. Tuusa J. T., Markkanen P. M., Apaja P. M., Hakalahti A. E., Petaja-Repo U. E. 2007. The endoplasmic reticulum Ca2+-pump SERCA2b interacts with G protein-coupled receptors and enhances their expression at the cell surface. J. Mol. Biol. 371:622–638 [DOI] [PubMed] [Google Scholar]

[B42] 42. van Rijn R. M., Whistler J. L., Waldhoer M. 2010. Opioid-receptor-heteromer-specific trafficking and pharmacology. Curr. Opin. Pharmacol. 10:73–79 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B43] 43. Vingtdeux V., et al. 2007. Alkalizing drugs induce accumulation of amyloid precursor protein by-products in luminal vesicles of multivesicular bodies. J. Biol. Chem. 282:18197–18205 [DOI] [PubMed] [Google Scholar]

[B44] 44. Waldhoer M., Bartlett S. E., Whistler J. L. 2004. Opioid receptors. Annu. Rev. Biochem. 73:953–990 [DOI] [PubMed] [Google Scholar]

[B45] 45. Waldron E., et al. 2008. Increased AICD generation does not result in increased nuclear translocation or activation of target gene transcription. Exp. Cell Res. 314:2419–2433 [DOI] [PubMed] [Google Scholar]

[B46] 46. Xuei X., et al. 2007. The opioid system in alcohol and drug dependence: family-based association study. Am. J. Med. Genet. B Neuropsychiatr. Genet. 144B:877–884 [DOI] [PubMed] [Google Scholar]

[B47] 47. Zhang H., Kranzler H. R., Yang B. Z., Luo X., Gelernter J. 2008. The OPRD1 and OPRK1 loci in alcohol or drug dependence: OPRD1 variation modulates substance dependence risk. Mol. Psychiatry 13:531–543 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B48] 48. Zhang Z., Pan Z. Z. 2010. Synaptic mechanism for functional synergism between delta- and mu-opioid receptors. J. Neurosci. 30:4735–4745 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Cysteine 27 Variant of the δ-Opioid Receptor Affects Amyloid Precursor Protein Processing through Altered Endocytic Trafficking ▿

Timo Sarajärvi

Jussi T Tuusa

Annakaisa Haapasalo

Jarkko J Lackman

Raija Sormunen

Seppo Helisalmi

Johannes T Roehr

Antonio R Parrado

Petra Mäkinen

Lars Bertram

Hilkka Soininen

Rudolph E Tanzi

Ulla E Petäjä-Repo

Mikko Hiltunen

Abstract

INTRODUCTION

MATERIALS AND METHODS

DNA constructs.

Cell culture and treatments.

Preparation of protein extracts and Western blot analysis.

Fig. 1.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 7.

Immunoprecipitation.

Secreted APP and Aβ measurements.

α-Secretase activity assay.

In vitro AICD generation assay.

Flow cytometry.

Confocal microscopy.

Electron microscopy.

Patients and controls.

SNP genotyping.

Table 1.

Imputation analysis.

Statistical analyses.

RESULTS

Expression of the δOR-Cys27 variant leads to APP C83 accumulation and decreased Aβ levels in human SH-SY5Y cells overexpressing APP and in HEK293 cells with endogenous or overexpressed APP.

Fig. 2.

Fig. 8.

δOR-Cys27 variant-induced changes in APP processing can be suppressed by a long-term treatment with δOR-specific ligands via a mechanism involving altered trafficking of the receptor.

Altered APP processing is coupled to changes in the endocytic trafficking of the δOR-Cys27 variant.

Fig. 6.

Ectopic treatments that impair normal endosomal trafficking can mimic the δOR-Cys27 variant-induced changes in APP processing.

Expression of the δOR-Cys27 variant leads to a decreased localization of APP C-terminal fragments in the LEL compartment of SH-SY5Y-APP751 cells.

Fig. 9.

Fig. 10.

Genetic analysis of OPRD1 gene polymorphisms in case-control and family-based AD cohorts.

DISCUSSION

Fig. 11.

ACKNOWLEDGMENTS

Footnotes

REFERENCES

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Cysteine 27 Variant of the δ-Opioid Receptor Affects Amyloid Precursor Protein Processing through Altered Endocytic Trafficking ^{^▿}