Edited by: Thomas Deller, Goethe-University, Germany
Reviewed by: Eckart Förster, Ruhr University Bochum, Germany; Claus Pietrzik, University of Mainz, Germany
*Correspondence: Joachim Herz
†These authors have contributed equally to this work.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution and reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
The biological fates of the key initiator of Alzheimer’s disease (AD), the amyloid precursor protein (APP), and a family of lipoprotein receptors, the low-density lipoprotein (LDL) receptor-related proteins (LRPs) and their molecular roles in the neurodegenerative disease process are inseparably interwoven. Not only does APP bind tightly to the extracellular domains (ECDs) of several members of the LRP group, their intracellular portions are also connected through scaffolds like the one established by FE65 proteins and through interactions with adaptor proteins such as X11/Mint and Dab1. Moreover, the ECDs of APP and LRPs share common ligands, most notably Reelin, a regulator of neuronal migration during embryonic development and modulator of synaptic transmission in the adult brain, and Agrin, another signaling protein which is essential for the formation and maintenance of the neuromuscular junction (NMJ) and which likely also has critical, though at this time less well defined, roles for the regulation of central synapses. Furthermore, the major independent risk factors for AD, Apolipoprotein (Apo) E and ApoJ/Clusterin, are lipoprotein ligands for LRPs. Receptors and ligands mutually influence their intracellular trafficking and thereby the functions and abilities of neurons and the blood-brain-barrier to turn over and remove the pathological product of APP, the amyloid-β peptide. This article will review and summarize the molecular mechanisms that are shared by APP and LRPs and discuss their relative contributions to AD.
Besides the important role in lipid metabolism, members of the low-density lipoprotein (LDL) receptor family take part in a broad range of pre- and post-developmental functions in brain and play key roles in the pathogenesis of Alzheimer’s disease (AD). Much like the amyloid precursor protein (APP), members of the LDL receptor family are type-I membrane receptors with single-pass transmembrane (TM) domains that can be endocytosed, proteolytically processed and participate in a variety of protein interactions both inside and outside of the cell, including direct interactions with APP (May et al.,
In the peripheral and central nervous system, lipoprotein receptors and APP interact to control developmental processes and synaptic function. These lipoprotein receptors are highly conserved—at least as far back in evolution as
The smaller receptors within the LDL receptor family, Ldlr, Vldlr and Apoer2, contain only one EGF-precursor domain and have a juxtamembraneous domain rich in serine and threonine residues, which serve as sites for O-linked glycosylation (Kingsley et al.,
Additional somewhat distant members are Lrp5 and Lrp6 as well as the Sortilin-related receptor with LDLR class A repeats (SorLA; Figure
Despite the high degree of homology between the receptors and the overlapping expression pattern and function, the majority of these receptors are indispensable for survival or proper brain function. In fact, deletion of Lrp1 (Herz et al.,
One percent of all AD cases are early onset (EOAD) generally manifesting from mutations in APP or APP processing genes and leading to increased production of the toxic APP cleavage product, amyloid β (Aβ). The other 99% of cases are late-onset AD (LOAD) with increased Aβ-levels and deposition that are apparently independent from EOAD-like mutations in APP/APP processing genes. Instead, the leading cause in LOAD appears to be an imbalance between Aβ production and clearance from the brain (Weller et al.,
Aside from age, the most important risk modifier for developing LOAD is ApoE (Corder et al.,
The brain is the most cholesterol-rich organ, containing approximately 25%–30% of the body’s total cholesterol (Dietschy and Turley,
The cholesterol metabolism link to AD pathogenesis is further supported by additional genome-wide association studies that implicate other apolipoproteins and their receptors as AD risk factors. In addition to ApoE, a variety of SNPs in ApoJ/Clusterin from several populations are associated with LOAD (Harold et al.,
With diverse functions including gathering nutrients and clearing toxic, useless debris from the extracellular space, as well as mediating intracellular trafficking/signaling and even transcription, the indispensable nature of many of the lipoprotein receptors is not surprising. Most of these receptors play some part in APP processing or clearance of Aβ, affecting the balance between Aβ-production and clearance. Understanding how these lipoprotein receptors and their ligands influence the homeostasis of Aβ production/clearance individually, as well as in unison, will prove crucial for not only elucidating mechanisms of AD pathogenesis, but also the design of potential therapeutic interventions to counteract the disease. In this chapter, we will focus on lipoprotein receptors and their role in AD pathogenesis through regulating APP processing and Aβ clearance.
Ldlr, the founding member of the LDL receptor family, is ubiquitously expressed throughout the body, where it plays a key role in regulating cholesterol homeostasis (reviewed in Go and Mani,
While impaired Ldlr function in humans leads to elevated plasma cholesterol and premature cardiovascular disease due to reduced uptake of cholesterol-rich LDLs (Hobbs et al.,
As cholesterol metabolism is linked to AD and regulated by Ldlr, Ldlr knockout mice have been used as a model organism to study the interplay between cholesterol and Aβ-deposition in several studies. While Ldlr has no known direct or indirect interaction with APP or APP processing, Ldlr binds to Aβ and mediates its clearance by degradation in astrocytes, but does not alter APP processing (Kim et al.,
The second receptor identified in the LDL receptor family, Lrp1 (Herz et al.,
Global Lrp1 knockout mice are embryonically lethal (Herz et al.,
Lrp1 directly interacts with APP extracellularly and regulates the localization and processing of APP (Kounnas et al.,
The extracellular interaction of Lrp1 and APP only occurs with APP isoforms containing the Kunitz protease inhibitor (KPI) domain and promotes the internalization of APP (Kounnas et al.,
Intracellular interactions with APP and Lrp1 also appear important in modulating the amyloidogenic processing of APP. Both Fe65 and Dab1 interact with Lrp1 NPxY motifs and modify intracellular signal transduction (Trommsdorff et al.,
Despite promoting neuronal Aβ production, Lrp1 participates in Aβ clearance (reviewed in Kanekiyo and Bu,
Importantly, both the Vldlr- and Lrp1-mediated Aβ clearance mechanisms via the BBB are differentially slowed down by ApoE-isoforms: ApoE4 > ApoE2 or ApoE3 (Deane et al.,
Lrp1b is very similar to Lrp1 in overall structure and sequence (~59% identical). Where Lrp1b differs most from Lrp1 is an extra LBR in the ECD and a 33 amino acid insert in the ICD (Liu et al.,
Mutations in Lrp1b are associated with multiple different types of cancer (Liu et al.,
Lrp1b binds to fibrinogen and ApoE carrying proteins (Haas et al.,
Both Apoer2 and Vldlr are quite similar in size and domain composition to Ldlr (Figure
Apoer2 and Vldlr are almost exclusively expressed in the brain where they act as receptors not only for ApoE but also for the neuromodulator Reelin (D’Arcangelo et al.,
The signaling initiated by Reelin binding to Apoer2 and Vldlr plays essential roles during the development of the CNS and neuronal function through adulthood (Förster et al.,
Apoer2 and Vldlr double knockout leads to a phenotype comparable to Reelin or Dab1 deficiency: mice develop strong ataxia, a smaller cerebellum, and defective lamination of cerebellum, cortex and hippocampus (Trommsdorff et al.,
Cortical Cajal-Retzius cells die out after birth and the amount of hippocampal Cajal-Retzius cells dramatically thins out later during postnatal hippocampal maturation (Chowdhury et al.,
Apoer2 and Vldlr bind Reelin and cluster together resulting in the phosphorylation of Dab1 and Src-kinase-mediated phosphorylation of NR2 subunits of the NMDA receptor (Hiesberger et al.,
There are several lines of evidence that implicate Reelin signaling as protective against AD pathogenesis. First, Reelin-signaling can counteract Aβ-induced synaptic suppression (Durakoglugil et al.,
Both Apoer2 and Vldlr interact with APP-binding proteins and influence the amyloidogenic processing of APP (reviewed Hoe and Rebeck,
Ligand binding to Apoer2 induces homotypic clustering as well as clustering with other receptors, including APP (Divekar et al.,
F-spondin is an extracellular ligand for both Apoer2 (Hoe et al.,
Like Lrp1, the NPxY domain of Apoer2 binds the cytosolic adaptor protein Fe65. While Lrp1 and Fe65 enhance Aβ production, Fe65 increases the interaction of APP and Apoer2 and decreases APP processing by stabilizing them at the cell surface (Hoe et al.,
Apoer2 directly interacts with APP extracellularly (Fuentealba et al.,
Lrp2 is structurally very similar to Lrp1b and one of the most studied lipoprotein receptors in conjunction with AD. Similar to Lrp1, Lrp2 undergoes proteolytic processing to release the ECD followed by γ-secretase cleavage to release the ICD (Zou et al.,
During neural tube formation and forebrain development Lrp2 is required for the dorsal to ventral gradient of the bone morphogenic protein 4 (BMP4) and sonic hedgehog (Shh). Lrp2 mediates endocytosis of Bmp4 for degradation and Bmp4 levels are increased in Lrp2-deficient mice (Spoelgen et al.,
Lrp2-deficient mice die shortly after birth due to respiratory insufficiency. Lrp2 function is critical during neural tube formation, as it acts to organize Shh-mediated forebrain development during neurulation (Christ et al.,
In the adult brain, Lrp2, facilitated by its ligand ApoJ/Clusterin, mediates Aβ clearance from the CSF (Hammad et al.,
Lrp2 expression decreases with age, which goes along with a reduced clearance rate of Aβ (Carro et al.,
One of the shorter members of the LDL receptor family, Lrp4, is critical for survival in that LRP4 knockout mice die after birth due to defects in the neuromuscular junction (NMJ; Weatherbee et al.,
Deficiency in Lrp4, MuSK, Agrin, APP and APLP2, or the intracellular scaffold Rapsyn lead to neonatal lethality, due to failure to form NMJs (Gautam et al.,
Studies at the NMJ also revealed important insights how different members of the APP-family interact (Choi et al.,
On the muscle fiber membrane, MuSK and Lrp4 form a functional receptor complex for Agrin. Upon Agrin binding to Lrp4, MuSK is phosphorylated resulting in Rapsyn-dependent focal clustering of nAChR (Shen et al.,
Similar to the lipoprotein receptor ligand Reelin, multiple functions have been described for Agrin in shaping and maintaining neuronal activity in the brain. Agrin stimulates filopodia formation to allow structural plasticity (McCroskery et al.,
Lrp4 also contributes to synaptic plasticity. Mice lacking Lrp4 or expressing a truncated Lrp4 retaining the ECD (Lrp4-ECD) in the brain have impaired hippocampal LTP and impaired memory (Gomez et al.,
In the hippocampus, besides neurons, astrocytes express functional α7-type AchRs (Shen and Yakel,
Heparan sulfate proteoglycans (HSPG) inhibit BACE1 mediated APP cleavage (Scholefield et al.,
Lrp5 and Lrp6 share 71% homology and are more distantly related members of the family. Despite encoding three LBRs and four EGF-precursor homology domains, compared to the core members, the domains appear in an inverse order with the ligand-binding domains adjacent to the TM segment rather than at the N-terminus. Additionally, their ICDs lack NPxY motifs. Both receptors have important functions in Wnt/β-catenin signaling, whereby Wnt and the Frizzled-receptors, mediate intracellular β-catenin translocation to the nucleus for transcriptional control of target gene expression (reviewed by Joiner et al.,
Lrp5 deficiency causes osteoporosis and bone fracture in mice due to reduced osteoblast proliferation and low bone mass (Kato et al.,
Wnt signaling via Lrp6 has been implicated in neuronal differentiation (Jeong et al.,
SorLA is a hybrid-type receptor, as the only member of the LDL receptor family with a Vps10p (yeast vacuolar protein sorting 10 protein) domain and six Fibronectin repeats (Figure
Defective homeostasis of SorLA and its cargo disrupts cellular function and causes AD, atherosclerosis and obesity (Caglayan et al.,
The ICD of SorLA is important for retrograde trafficking from endosomes to the trans-Golgi network (TGN) by binding to the retromer complex and anterograde trafficking by interacting with clathrin-adaptors (Jacobsen et al.,
Lrp3, Lrp10 (murine Lrp9) and Lrp12 (ST7/Mig13) share high homology (Battle et al.,
The function of APP and Aβ beyond AD is not well understood and understudied, especially in conjunction with lipoprotein receptors. Different chapters of this series discuss the physiological role of APP and its cleavage products from various physiological perspectives. APP and its trafficking and processing plays a role in neurite outgrowth and synaptogenesis, APP-deficiency decreases dendritic spine numbers and impairs LTP, which can be rescued by sAPPα but not sAPPβ (Tyan et al.,
APP trafficking and processing is controlled by a large variety of proteins, but little is known about their physiological relevance. APP interacts with numerous type-I TM receptors, many of which are lipoprotein receptors, and several other ligands, adaptor and scaffolding proteins, which together provide a protein-protein network involved in signaling, processing of various receptors, partially through endocytic pathways.
APP processing to Aβ and in particular the accumulation of the amyloidogenic Aβ42 product, either from increased production or impaired clearance, are initiating events in AD, and ApoE genotype is the most important late onset risk factor for AD. Both APP and ApoE interact with LDL receptor family members to regulate APP trafficking, processing and elimination. Therefore, it is all but certain, that LDL receptor family members play a pivotal role in the pathogenesis of AD.
As a result of the work reviewed in this article, we have learned much about the potential molecular mechanisms that these lipoprotein receptors play in AD pathogenesis, yet the relative importance of each individual event is still unclear. Continuing work on the biology of LDL receptor related genes and their ligands on the physiology of the APP processing machinery holds great promise not only to greater understanding of the disease process but also for the identification of novel and effective therapeutic approaches.
TP and CRW jointly wrote the article and designed the figures under JH guidance and JH edited the manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
This work was supported by grants from the NHLBI (R37 HL063762), the NIA (RF AG053391), the NINDS and NIA (RO1 NS093382), as well as, the Consortium for Frontotemporal Dementia Research (A108400), and the Brightfocus Foundation (A2016396S). We would like to thank Nancy Heard and Barbara Dacus for their help in preparing the figures.