LDLR-related protein 10 (LRP10) regulates amyloid precursor protein (APP) trafficking and processing: evidence for a role in Alzheimer's disease

Abstract

Background: The Aβ peptide that accumulates in Alzheimer's disease (AD) is derived from amyloid precursor protein (APP) following proteolysis by β- and γ-secretases. Substantial evidence indicates that alterations in APP trafficking within the secretory and endocytic pathways directly impact the interaction of APP with these secretases and subsequent Aβ production. Various members of the low-density lipoprotein receptor (LDLR) family have been reported to play a role in APP trafficking and processing and are important risk factors in AD. We recently characterized a distinct member of the LDLR family called LDLR-related protein 10 (LRP10) that shuttles between the trans-Golgi Network (TGN), plasma membrane (PM), and endosomes. Here we investigated whether LRP10 participates in APP intracellular trafficking and Aβ production.

Results: In this report, we provide evidence that LRP10 is a functional APP receptor involved in APP trafficking and processing. LRP10 interacts directly with the ectodomain of APP and colocalizes with APP at the TGN. Increased expression of LRP10 in human neuroblastoma SH-SY5Y cells induces the accumulation of mature APP in the Golgi and reduces its presence at the cell surface and its processing into Aβ, while knockdown of LRP10 expression increases Aβ production. Mutations of key motifs responsible for the recycling of LRP10 to the TGN results in the aberrant redistribution of APP with LRP10 to early endosomes and a concomitant increase in APP β-cleavage into Aβ. Furthermore, expression of LRP10 is significantly lower in the post-mortem brain tissues of AD patients, supporting a possible role for LRP10 in AD.

Conclusions: The present study identified LRP10 as a novel APP sorting receptor that protects APP from amyloidogenic processing, suggesting that a decrease in LRP10 function may contribute to the pathogenesis of Alzheimer's disease.

Figure 1

LRP10 interacts with APP. (A)In vivo interaction of LRP10-HA and GFP-APP₆₉₅ proteins. Lysates of HEK cells transfected with HA-tagged LRP10 and GFP or GFP-tagged APP were immunoprecipitated with anti-HA and then immunoblotted with anti-HA or anti-GFP antibody to detect LRP10 and GFP, respectively. (B) LRP10-HA interacts with endogenous APP. Lysates of HEK cells transfected with pcDNA3-HA or LRP10-HA were immunoprecipitated with anti-APP antibody and then immunoblotted with anti-HA antibody.

Figure 2

Interaction of the ectodomains of LRP10 and APP. (A) Schematic representation of the LRP10 FLAG-tagged deletion mutants used to determine the binding domain of APP. The structural elements of LRP10 are depicted, including (from amino to carboxyl terminus) the CUB domains, LDLRA repeat, and transmembrane (TM) domain. CD, Cytoplasmic domain; ED, Ectodomain. (B) The interaction between LRP10 and APP does not depend on the cytoplasmic domain of LRP10. Lysates of HEK cells transfected with APP₆₉₅ and pcDNA3-FLAG, FLAG-tagged LRP10^ΔCD, or FLAG-tagged LRP10^ΔED were immunoprecipitated with anti-APP or anti-FLAG and were immunoblotted with anti-APP or anti-FLAG antibody. (C) Schematic representation of the GST-APP deletion mutants used to determine the binding domain of LRP10. The structural elements of APP are depicted, including the cysteine, acidic, carbohydrate, amyloid beta, and cytoplasmic domains. (D)In vitro interaction of LRP10 with the ectodomain of APP. The APP deletion mutants shown in (C) and the GST protein (10 μg each) were immobilized on glutathione beads and were incubated with in vitro translated ³⁵ S-labeled LRP10. Bound proteins were separated by SDS-PAGE and were detected by autoradiography. GST proteins were detected by coomassie staining. Input equaled 2.5% of the total in vitro-translated product.

Figure 3

LRP10 colocalizes with APP and modulates its intracellular distribution. HeLa cells were transfected with GFP-APP and control vector pcDNA3 (A, D), wild-type LRP10-HA, (LRP10^wt-HA; B, E), or HA-tagged-LRP10 in which two DXXLL motifs (that bind the clathrin adaptors GGAs and AP1/2) in the cytoplasmic tail were mutated (LRP10^2DXXAA-HA, C, F). Cells were fixed, permeabilized, and immunostained with anti-GFP (green), anti-HA (red), and anti-TGN46 or EEA1 (blue) antibodies. The labeled cells were examined by confocal fluorescence microscopy. (A, D) In control pcDNA3 cells, APP-GFP was detected mainly in the juxtanuclear region and surrounding vesicles where it partially overlapped with TGN46 (A, inset, arrowheads) and EEA1 (D, inset, arrowheads). (B, E) LRP10^wt-HA was detected in the juxtanuclear region and surrounding vesicles and partially overlapped with TGN46 (B) and EEA1 (E). In these cells, GFP-tagged APP was detected mainly in the Golgi region, where it partially overlapped with LRP10^wt-HA (B). The merged images show a partial overlap between LRP10^wt-HA and GFP-APP in the Golgi cisternae labeled by TGN46 (B, inset, arrowheads) and surrounding endosomes labeled by EEA1 (E, inset, arrowheads). (C, F) HA-tagged LRP10^2DXXAA was redistributed to the plasma membrane (arrows) and peripheral early endosomes (F, inset). In these cells, GFP-APP was also detected on the plasma membrane (F, inset, arrow), Golgi (C, inset) and in peripheral endosomes labeled by EEA1 (F, inset), where it colocalized with LRP10^2DXXAA-HA (F, arrowheads, inset). Scale bar, 10 μm.

Figure 4

Comparison of the intracellular trafficking pathways of LRP10^wtand LRP10^2DXXAA. The internalization of LRP10 was evaluated in HeLa cells transfected with LRP10^wt or LRP10^2DXXAA tagged with an extracellular FLAG epitope. The cells were pre-incubated on ice to arrest endocytosis, and LRP10 molecules exposed on the cell surface were immunolabeled with antiserum directed against FLAG at 4°C. The cells were then washed and incubated at 37°C to allow internalization. Endocytosis and TGN targeting of cell surface-labeled LRP10 were evaluated after 0, 5, 10, 60, and 120-min chases at 37°C. The cells were then fixed, permeabilized, and immunostained with anti-TGN46 (green) or anti-EEA1 (blue) antibodies. The labeled cells were examined by confocal fluorescence microscopy. In cells expressing FLAG-LRP10^wt, low level of FLAG-LRP10^wt was observed at the cell surface at time 0 (A, inset, arrows). Internalized LRP10^wt (int-LRP10^wt) was observed in vesicles near the plasma membrane following a 5-min (C) and a 10-min (E) chase where it colocalized with EEA1 (C, E, insets, arrowheads). After a 60-min chase, int-LRP10^wt localized in the Golgi region and partially colocalized with the TGN marker TGN46 (G, inset, arrowheads). This colocalization was higher after a 120-min chase (I, inset, arrowhead). In cells expressing FLAG-LRP10^2DXXAA, large amounts of LRP10^2DXXAA were detected on the cell surface at time 0 (B, arrows). After a 5-min chase, int-LRP10^2DXXAA was still distributed at the plasma membrane (D, arrow) or partially colocalized with EEA1 in vesicles near the PM (D, inset, arrowheads). After a 10-min chase, int-LRP10^2DXXAA was observed in EEA1-labeled vesicles (F, inset, arrowheads). However, after a 60-min chase (H), and even a 120-min chase (J), int-LRP10^2DXXAA was still mainly observed in early endosomes (H, J, insets, arrowheads). Scale bar, 10 μm.

Figure 5

LRP10^2DXXAAinhibits the retrograde transport of APP from the endosome to the TGN. Internalization of APP was evaluated in HeLa cells transfected with APP₆₉₅ and pcDNA3 (A, D), HA-tagged LRP10^wt(B, E), or LRP10^2DXXAA(C, F). The cells were pre-incubated on ice to arrest endocytosis, and APP molecules exposed on the cell surface were immunolabeled with antiserum directed against APP (α-22 C11) at 4°C. Endocytosis and TGN targeting of APP were evaluated after 60-min chase periods at 37°C. Cells were then fixed, permeabilized, and immunostained with anti-HA (red) and anti-TGN46 or anti-EEA1 (blue) antibodies. The labeled cells were examined by confocal fluorescence microscopy. In cells expressing pcDNA3 and LRP10^wt-HA, internalized APP (int-APP, green) was localized in the Golgi region and partially colocalized with the TGN marker TGN46 (A, B, arrowheads). However, in cells expressing LRP10^2DXXAA-HA, internalized APP (int-APP) was distributed mainly in vesicles labeled with the endosomal marker EEA1 and LRP10^2DXXAA-HA (F, arrowheads). Arrowheads indicate structures in which APP colocalized with TGN46 or EEA1 while arrows indicate structures in which APP and TGN46 or EEA1 did not colocalize. Scale bar, 10 μm.

Figure 6

LRP10 affect the trafficking of APP in neuronal SH-SY5Y cells. (A-F) LRP10 colocalized and altered the distribution of endogenous APP. SH-SY5Y cells were stably transfected with empty pcDNA3 vector (A, D) or high levels of HA-tagged-LRP10^wt (B, E) or -LRP10^2DXXAA (C, F). Cells were fixed, permeabilized, and immunostained with anti-APP (green), anti-HA (red), and anti-TGN46 or anti-EEA1 (blue) antibodies. The labeled cells were examined by confocal fluorescence microscopy. (A, D) In control cells (pcDNA3), endogenous APP was detected mainly in the Golgi region where it partially overlapped with TGN46 (A, inset arrowheads) and surrounding vesicles where it partially overlapped with EEA1 (D, inset, arrowheads). (B, E) LRP10^wt-HA was detected in the TGN (B, inset) and surrounding EEA1-labeled endosomes (E, inset). Endogenous APP was mainly detected in the Golgi region, where it partially overlapped with LRP10^wt-HA (B, inset). The merged image shows a clear overlap between LRP10^wt-HA and endogenous APP in the TGN (B, arrowheads, insets). (C, F) LRP10^2DXXAA-HA was redistributed mainly to peripheral EEA1-labeled early endosomes (F, inset). Endogenous APP was also detected in peripheral EEA1-labeled endosomes (F, inset), where it colocalized with HA-tagged LRP10^2DXXAA (F, arrowheads, inset). Scale bar, 10 μm. (G) The level of cell surface APP was altered by the expression of LRP10. SH-SY5Y cells stably expressing pcDNA3 vector or high levels of HA-tagged-LRP10^wt or -LRP10^2DXXAA were surface-biotinylated with sulfo-NHS-SS-biotin. Biotinylated proteins were precipitated with neutravidin, and samples (Surface proteins) were analyzed by immunoblotting with antisera directed against HA or APP. The total cell lysate (Total proteins) was also analyzed to assess APP and LRP10 expression levels. The three APP variants (APP₆₉₅, APP₇₅₁, and APP₇₇₀) were detected by the anti-APP antibodies. The asterisk indicates an accumulation of mature APP (glycosylated APP_751/770) in the presence of LRP10^wt. Actin was used as a loading control. (H) The maturation of APP was altered in SH-SY5Y cells expressing LRP10^wt. SH-SY5Y control cells or cells stably expressing LRP10^wt or LRP10^2DXXAA were pulse-labeled with [³⁵ S]methionine for 5 min and chased at 37 °C for the indicated time. Radiolabeled APP was immunoprecipitated from the cell extracts and was analyzed by SDS-PAGE and autoradiography. The three APP variants (APP₆₉₅, APP₇₅₁, and APP₇₇₀) are indicated by arrowheadss. Accumulations of mature APP (*, indicates mature APP_751/770; +, indicates mature APP₆₉₅) were observed in cells expressing LRP10^wt. Graph of the ratio of mature (m) APP versus immature (im) APP in the indicated SH-SY5Y stable clones as determined by densitometric scanning of pulse-chase experiments exemplified in the autoradiogram above. The results indicated that there is a delayed turnover of the mature, fully glycosylated APP variants when they are co-expressed with LRP10^wt. Results are expressed as means ± SD (n = 3). *, p < 0.05; **, p < 0.005 (compared to control cells).

Figure 7

LRP10 overexpression alters APP processing in SH-SY5Y cells. APP cleavage products in SH-SY5Y stable clones expressing pcDNA3 vector alone (Ctl) or HA-tagged LRP10^wt or LRP10 trafficking mutant LRP10^2DXXAA. (A) Representative Western blot of total sAPP secreted in the media and of β-CTF and total APP in the cell lysates of the indicated SH-SY5Y stable clones. Actin was used as loading control. (B) AlphaLISA quantitative analysis of sAPPα, sAPPβ, and Aβ₄₀ in the media of the SH-SY5Y stable clones expressing pcDNA3 vector alone (Ctl), high level of HA-tagged LRP10^wt (LRP10^wt) or low and high levels of LRP10 trafficking mutant (LRP10^2DXXAA). Results are expressed as means ± SD (n = 3). *, p < 0.01; **, p < 0.005; ***, p < 0.001 (compared with control cells).

Figure 8

LRP10 levels in healthy and AD brains. Analysis of LRP10 protein and mRNA levels in healthy and AD brains. (A) LRP10 protein expression in the frontal cortex and hippocampus of healthy (Control) and Alzheimer’s disease (AD) patients was compared by Western blotting. Representative data showing lysates subjected to SDS-PAGE and immunoblotted with antisera directed against LRP10 and neuron-specific Class III β-tubulin (TUJ1). (B) Densitometric analysis of Western blots, such as those shown in (A), normalized to the signal of TUJ1. Results are expressed as means ± SD (n ≥ 3). *, p < 0.01; **, p < 0.005; ***, p < 0.001 (compared with healthy patients). (C) LRP10 mRNA levels in the frontal cortex of healthy (CTL) and AD patients were compared by qRT-PCR. Total mRNA was reverse transcribed, and the levels of LRP10 cDNA were analyzed by qPCR with SYBR Green and were expressed relative to the endogenous control (RPL13) using the comparative CT method. Results are expressed as means ± SD (n = 5 samples, in duplicate). The difference between CTL and AD was not significant (p = 0.4).

Figure 9

Model for the role of LRP10 in APP trafficking. LRP10 is transported from the Golgi to the plasma membrane in the secretory pathway independent of the DXXLL motifs. LRP10^wt is rapidly internalized from the cell surface into early endosomes where it is efficiently recycled back to the Golgi. LRP10 may also traffic directly from the TGN to the endosomes. LRP10^2DXXAA, which is unable to associate with GGAs and AP1/2, is slowly internalized from the plasma membrane and accumulates in the early endosomes, since it is incapable of recycling to the TGN. APP is constitutively transported to the plasma membrane and internalized in endosomes. Interaction of APP with LRP10^wt prolonged its presence in the TGN. LRP10 could trap APP in the TGN (1), reducing the amount of APP transported to the cell surface for non-amyloidogenic processing by α- and γ-secretases. In addition, LRP10 may shuttle APP from early endosome back to the TGN (2), reducing its amyloidogenic processing by β- and γ-secretases. LRP10^2DXXAA traps APP in endosomes (3), resulting in enhanced accessibility to amyloidogenic secretases and thus processing into Aβ.