|Home | About | Journals | Submit | Contact Us | Français|
Low-density lipoprotein receptor-related protein-1 (LRP1) is the main cell surface receptor involved in brain and systemic clearance of the Alzheimer's disease (AD) toxin amyloid-beta (Aβ). In plasma, a soluble form of LRP1 (sLRP1) is the major transport protein for peripheral Aβ. LRP1 in brain endothelium and mural cells mediates Aβ efflux from brain by providing a transport mechanism for A across the blood-brain barrier (BBB). sLRP1 maintains a plasma ‘sink’ activity for Aβ through binding of peripheral Aβ which in turn inhibits re-entry of free plasma Aβ into the brain. LRP1 in the liver mediates systemic clearance of Aβ. In AD, LRP1 expression at the BBB is reduced and Aβ binding to circulating sLRP1 is compromised by oxidation. Cell surface LRP1 and circulating sLRP1 represent druggable targets which can be therapeutically modified to restore the physiological mechanisms of brain Aβ homeostasis. In this review, we discuss how increasing LRP1 expression at the BBB and liver with lifestyle changes, statins, plant-based active principles and/or gene therapy on one hand, and how replacing dysfunctional plasma sLRP1 on the other regulate Aβ clearance from brain ultimately controlling the onset and/or progression of AD.
The low-density lipoprotein receptor (LDLR)-related protein 1 (LRP1/CD91/α2-macroglobulin receptor), a member of the LDL receptor family, is a multifunctional scavenger, cargo transporter and signaling receptor (Dieckmann et al., 2010; Boucher & Herz, 2011). The gene encoding LRP1 is located on chromosome 12q13-q14 (Herz et al., 1988). LRP1 is ubiquitously expressed in many cell types, including brain endothelium, neurons, smooth muscle cells, astrocytes, macrophages, fibroblasts and hepatocytes (Moestrup et al., 1992). There are two forms of LRP1; cell surface bound (LRP1) and a truncated soluble form (sLRP1). sLRP1 is normally present in plasma (Quinn et al., 1997; Grimsley et al., 1999; Sagare et al., 2007; Sagare et al., 2011a), and under physiological conditions is found only in trace amounts in the cerebrospinal fluid (CSF) in humans (Qiu et al., 2001; Liu et al., 2009). LRP1 and sLRP interact with a diverse array of ligands (Lillis et al., 2008; Zlokovic et al., 2010; Boucher & Herz, 2011).
LRP1 is one of the largest receptors (600 kDa) in the LDL receptor family (Herz & Strickland, 2001; Dieckmann et al., 2010). It is synthesized as a single polypeptide chain, cleaved by the endopeptidase furin in the trans-Golgi compartment. The resulting extracellular heavy α-chain (515 kDa) remains non-covalently coupled to the extracellular region of the transmembrane and cytoplasmic light β-chain (85 kDa) (Fig. 1). LRP1 expression is controlled at both transcriptional and translational levels (Tamaki et al., 2007; Ceschin et al., 2009; Selvais et al., 2011). Cell surface levels of LRP1 are controlled by proteolytic shedding of its ectodomain (Quinn et al., 1999; May et al., 2002; Von Arnim et al., 2005; Polavarapu et al., 2007; Liu et al., 2009; Selvais et al., 2011).
LRP1, like all LDL receptors, is composed of a modular structural organization that includes cysteine-rich complement-type repeats (CRs), epidermal growth factor (EGF) precursor repeats and β-propeller (YWTD) domains (Lillis et al., 2008). The EGF precursor consists of two cysteine-rich EGF repeats, a YWTD repeat and another EGF-like repeat (Lillis et al., 2008). There are four ligand-binding domains in the extracellular region (clusters I-IV), consisting of 2, 8, 10 and 11 CRs, respectively (Obermoeller-McCormick et al., 2001; Meijer et al., 2007). Clusters II and IV are the major ligand binding regions interacting with approximately fifty structurally diverse ligands (Table 1) including: apolipoprotein E (apoE), α2-macroglobulin (α2M), tissue plasminogen activator (tPA), proteinase-inhibitors, blood coagulation factors (e.g., factor VIII), receptor-associated protein (RAP) (Hussain et al., 1999; Neels et al., 1999; Herz 2001; Herz & Strickland, 2001; Croy et al., 2003; Lillis et al., 2008; Meijer et al., 2007; Herz et al., 2009; Boucher & Herz, 2011), Alzheimer's amyloid-Aβ (Aβ) (Deane et al., 2004a; Sagare et al., 2007), prion protein (Parkyn et al., 2008; Jen et al., 2010) and aprotinin (Demeule et al., 2008a; 2008b). Thus, LRP1 plays a major role in the transport and metabolism of macromolecules especially cholesterol associated with apoE-containing lipoproteins, and clearance of proteases, protease inhibitors, matrix proteins, viruses and toxins, such as Aβ and prion.
While LRP1 has been regarded mainly as a receptor which internalizes its ligands and directs them to the lysosomes for proteolytic degradation, it has been also reported that LRP1 transports several ligands transcellularly across the blood-brain barrier (BBB) including Aβ (Shibata et al., 2000; Deane et al., 2004a; Davis et al., 2004; Wu et al., 2005; Cirrito et al., 2005; Bell et al., 2007), RAP (Pan et al., 2004), tPA (Benchenane et al., 2005), lipid free and lipidated apoE2 and apoE3, apoE2 and apoE3 complexes with Aβ (Deane et al., 2008) and a family of Kunitz domain-derived peptides (Demeule et al., 2008a; 2008b; Bertrand et al., 2011). These findings suggest that LRP1 may play a key role in regulating the exchange of several ligands between the brain and the blood.
The cytoplasmic domain of LRP1 contains two NPxY, one YxxL motifs and two di-leucine motifs (Li et al., 2001) (Fig. 1). It has been suggested that the YxxL motif and distal di-leucine repeats may be associated with a rapid endocytotic rate of LRP1 (Li et al., 2001; Deane et al., 2004a; Deane et al., 2008). The cytoplasmic tail is phosphorylated on several serine, threonine or tyrosine residues (Bu et al., 1998; Li et al., 2001; van der Geer, 2002). It interacts with a number of adaptor proteins including Shc, Disabled-1, JIP1, PSD-95, CED-6/GULP and Fe65 (Table 2), involved in directing cellular trafficking and transmembrane signaling (Trommsdorff et al., 1998; Gotthardt et al., 2000; Herz et al., 2009; Boucher & Herz, 2011). The NPxY motifs serve as docking sites for the adaptor proteins, Shc, Disabled-1, JIP1, PSD-95, CED-6/GULP Snx17 and Fe65. Phosphorylation of the cytoplasmic domain regulates interaction of adaptor proteins to the NPxY motifs and endocytosis (Li et al., 2001). Thus, LRP1 has a dual role as a rapid cargo endocytotic cellular transporter and a transmembrane cell signaling receptor.
While some genetic studies have suggested that LRP1 is linked to AD and cerebral amyloid angiopathy (CAA) (Kang et al., 1997; Lambert et al., 1998; Wavrant-DeVrieze et al., 1999; Christoforidis et al., 2005), others have shown weak or no such association (Beffert et al., 1999; Bertram et al., 2000; Pritchard et al., 2005; Harold et al., 2009; Lambert et al., 2009; Bertram et al., 2010; Chalmers et al., 2010; Natunen et al., 2012). Earlier studies have demonstrated that LRP1 and many of its ligands are deposited in senile plaques (Rebeck et al., 1995; Arelin et al., 2002). In addition to regulating Aβ clearance from brain (Shibata et al., 2000; Deane et al., 2004a), it has been shown that the LRP1 cytoplasmic C-terminal domain interacts with APP's (Aβ-precursor protein) cytoplasmic domain via FE65, an LRP1 adaptor protein, which in turn influences the APP processing and Aβ generation (Pietrzik et al., 2004; Waldron et al.,2008). It has also been demonstrated that LRP1 in neurons mediates Aβ cellular uptake and possibly retention in the brain via its ligands α2M and apoE (Narita et al., 1997; Qiu et al., 1999; DeMattos et al., 2004; Zerbinatti et al., 2004; Zerbinatti & Bu, 2005; Deane et al., 2008). However, the exact implications of these findings for the development of Aβ pathology and cognitive decline remain unclear.
Aβ concentration in brain interstitial fluid (ISF) is regulated by its rate of production from APP (Selkoe, 2001a; Cirrito et al., 2003) and clearance from brain (Mawuenyega et al., 2010) and across the BBB mainly via LRP1 (Shibata et al., 2000; Deane et al., 2004a). In addition, influx of circulating Aβ into certain brain regions expressing the receptor for advanced glycation end products (RAGE) in brain endothelium may contribute to Aβ levels in brain (Deane et al., 2003; Dries et al., 2012). Aβ clearance is modulated by apoE in an isoform-specific manner (Zlokovic, 1996; Martel et al., 1997; Tanzi et al., 2004; Moir & Tanzi, 2005; Deane et al., 2008) and apoJ (Zlokovic et al., 1996; Calero et al., 2000; Bell et al., 2007). The enzymatic degradation of Aβ may also contribute to its clearance from brain (Selkoe, 2001b; Saido & Leissring, 2012). In contrast to the capillaries in peripheral organs allowing free exchanges of many molecules between the blood and the ISF (Mann et al., 1985), the BBB is normally impermeable to small polar molecules, peptides and proteins (Zloković et al., 1985; Zlokovic & Apuzzo, 1997). Therefore, transport of Aβ across the BBB requires a specialized transport system. Aβ is transported from blood to brain across the BBB of several species that have been studied including guinea-pigs, mice, rats and primates (Zlokovic et al., 1993; Ghilardi et al., 1996; Martel et al., 1996; Poduslo et al., 1997; Mackic et al., 1998b; Mackic et al., 2002; Ujiie et al., 2003; Eisele et al., 2010).
Several studies have reported that LRP1 transports Aβ from brain to blood (Shibata et al., 2000; Banks et al., 2003; Deane et al., 2004a; Ito et al., 2006; Bell et al., 2007; Deane et al., 2008; Bell et al., 2009) and is the main brain Aβ clearance receptor (Table 3). In contrast, RAGE mediates transport of blood-borne Aβ into brain and is associated with oxidant stress and pro-inflammatory response (Mackic et al., 1998a; Deane et al., 2003; Deane et al., 2012). RAGE expression has been found to be increased in brain endothelial cells and vascular smooth muscle cells in animal models of aging as well as in AD patients (Yan et al., 1996; Deane et al., 2003; Donahue et al.,2006; Miller et al., 2008; Silverberg et al., 2010b), whereas LRP1 expression is decreased both at the BBB and in cerebral arterial vascular smooth muscle cells (VSMC) (Shibata et al., 2000; Bading et al., 2002; Deane et al., 2004a; Donahue et al.,2006; Herring et al., 2008; Bell et al., 2009; Silverberg et al., 2010a). These changes in Aβ key transport receptors, therefore, favor accumulation of brain Aβ during AD pathogenesis.
Continuous removal of Aβ from brain, blood and the entire organism is essential for preventing its accumulation in the brain (Zlokovic et al., 2000; Zlokovic et al., 2010; Zlokovic, 2011). As illustrated in Fig. 2, LRP1 plays a key role in the three-step serial clearance mechanism mediating Aβ elimination from the brain under physiological conditions. Step 1. Binding of Aβ to LRP1 at the abluminal brain endothelial cell membrane initiates rapid Aβ clearance across the BBB into the blood in vivo (Shibata et al., 2000; Deane et al., 2004a; Cirrito et al., 2005; Ito et al., 2006; Sagare et al., 2007; Bell et al., 2007; Deane et al., 2008; Bell et al., 2009). Human Aβ injected into different brain regions in mice emerges intact in murine plasma confirming its elimination from the brain (Shiiki et al., 2004; Bell et al., 2007). Step 2. Circulating plasma sLRP1 binds to and sequesters free Aβ in plasma providing a key endogenous peripheral Aβ ‘sink’ promoting continuous removal of Aβ from brain (Sagare et al., 2007; Sehgal et al., 2012). In humans and mice, sLRP1 normally binds > 70% of circulating Aβ preventing free Aβ access to the brain (Sagare et al., 2007). In AD patients and AD transgenic mice, however, increased oxidation of sLRP1 decreases binding affinity for Aβ (Sagare et al., 2007; Sagare et al., 2011) resulting in elevations in free Aβ40 and Aβ42 in plasma which may then lead to increased transport into the brain via RAGE (Deane et al., 2003; Ujiie et al., 2003; Donahue et al., 2006; Miller et al., 2008; Deane et al., 2012). Step 3. LRP1 localized to hepatic cells binds to and systemically clears circulating Aβ (Tamaki et al., 2006). Reduced hepatic LRP1 levels are associated with decreased peripheral Aβ clearance in aged rats (Tamaki et al., 2006; Tamaki et al., 2007). In aged Squirrel monkeys, Aβ systemic clearance is also reduced and associated with increased Aβ levels in the brain (Mackic et al., 1998b; Mackic et al., 2002). In addition to the liver, sLRP1-Aβ complexes and free Aβ are eliminated through the kidneys (Sagare et al., 2007). Recently, the liver was also shown to be a major source of Aβ and can regulate brain Aβ levels (Sutcliffe et al., 2011; Sagare et al., 2011b).
Besides endothelial cells, LRP1 is also expressed in other cell types of the neurovascular unit including pericytes (Bell et al., 2012), VSMC (Bell et al., 2009) and astrocytes (Koistinaho et al., 2004). It has been suggested that LRP1 in VSMC (Bell et al., 2009) and astrocytes (Koistinaho et al., 2004) participates in Aβ clearance. Whether LRP1 in pericytes can clear Aβ in vivo is presently not known. However, the role of LRP1 in maintaining the integrity of the BBB has been recently demonstrated (Bell et al., 2012). Therefore, one would expect that drugs that can increase LRP1 expression at the BBB will also promote LRP1-mediated Aβ clearance by other cell types in brain and improve the BBB integrity, assuming that these drugs can penetrate the BBB and reach their cellular targets in brain expressing LRP1.
The development of therapies that target brain Aβ to slow the progression of AD continues to be challenging (Hardy, 2009; Rosenberg, 2011; McKhann, 2011). There is an urgent need to explore new therapeutic approaches and to test these in clinical trials before disease onset (Rosenberg, 2011). Restoration or enhancement of normal LRP1-driven Aβ control mechanisms may offer new opportunities to reduce and/or prevent Aβ accumulation and the associated cognitive decline. As suggested by LRP1's three-step homeostatic control of brain Aβ levels, potential points of intervention are: (1) restoration and/or enhancement of BBB LRP1 expression, (2) replacement of dysfunctional sLRP1, and (3) restoration and/or enhancement of hepatic LRP1 expression.
Approaches directed at increasing LRP1 activity at the BBB and liver and/or replacing sLRP1 peripheral binding of Aβ that we discuss below are all designed to promote Aβ clearance from brain, which conceptually is similar to Aβ clearance therapy mediated by anti-Aβ antibodies (Deane et al., 2009; Mucke et al., 2009). But, some antibodies cross the BBB and can activate microglia response contributing to amyloid clearance, which is not a feature of LRP1-based therapy. Whether this aspect of antibody action contributes to microhemorrhages observed with some anti-Aβ antibodies (Deane et al., 2009), in contrast to LRP1-based approaches which do not cause hemorrhages (Sagare et al., 2007), is not clear at present. One would also expect that γ- and β-secretase inhibitors, by reducing the Aβ load in brain, will act synergistically with LRP1-based strategies directed at increasing Aβ clearance. However, the interaction of LRP1-based therapy with β-secretase inhibitors might be more complex because β-secretase also cleaves sLRP1 (von Arnim et al., 2005), which on one hand would increase the BBB and cellular Aβ clearance by increasing LRP1 levels, while on the other hand might reduce Aβ peripheral sink by lowering soluble sLRP1
LRP1 expression at the BBB is reduced during normal aging and in AD (Kang et al. 2000; Shibata et al., 2000; Bading et al., 2002; Deane et al., 2004a; Donahue et al., 2006; Silverberg et al., 2010a). Therefore, normalization or enhancement of LRP1 levels at the BBB should increase the clearance of Aβ from brain and restore control of brain Aβ levels. Increased BBB LRP1 expression may be achieved by: (a) lifestyle changes (e.g., diet, exercise and enriched environment), (b) pharmacological agents (e.g., statins, plant-based active principles), and (c) gene therapy.
Cerebral oxidative stress is increased in AD (Smith et al., 1991; Montine et al., 2002; Lovell & Markesbery, 2007), and could render BBB LRP1 dysfunctional as a result of oxidation and/or reduced proteosomal degradation (Deane et al., 2004a). Therefore, enhancing the body's natural anti-oxidant defense mechanisms, with antioxidant nutrients, such as polyphenol-rich foods (Williams & Spencer, 2012; Hwang et al., 2012), vitamin A, C and E, could potentially protect LRP1 from oxidative damage, and thus reduce associated declines in Aβ clearance and cognitive function. In rodents, antioxidants supplemented in the diet including vitamins (Sung et al., 2004), grape seed extract (Liu et al., 2011), wine (Wang et al., 2006; Ho et al., 2009), pomegranate juice (Hartman et al., 2006), blackberries (Shukitt-Hale et al., 2009), wild blueberry extract (Papandreou et al., 2009), green tea (Chan et al., 2006; Assunção et al., 2011) or extra virgin olive oil (Farr et al., 2012) improved learning acquisition and memory retention. Some preliminary studies suggest that diet supplemented with polyphenols improves memory in older adults (Krikorian et al., 2010a; Krikorian et al., 2010b). A higher intake of vitamin C and E (Masaki et al., 2000; Engelhart et al., 2002; Zandi et al., 2004) and vitamin E (Morris et al., 2002) have been shown in some studies to be associated with a reduced risk of AD. However, other studies in different populations have failed to show this association (Sano et al., 1997; Luchsinger et al., 2003; Petersen et al., 2005; Galasko et al., 2012) and the collective results from clinical trials remain equivocal.
Physical activity bolsters cerebral angiogenesis in rodents (Isaacs et al., 1992; Swain et al., 2003; Pereira et al., 2007; Latimer et al., 2011) and human subjects (Pereira et al., 2007) and may delay cognitive decline by reducing cerebrovascular risk, including the contribution of small vessel disease to dementia (Ahlskog et al., 2011). In chronic conditions, such as AD, aberrant angiogenesis reduces vascular density thereby leading to reductions in both capillary perfusion and Aβ clearance contributing to neurodegeneration (Wu et al; 2005; Zlokovic et al., 2005). Physical activity and cognitive stimulation in the form of enriched environment consisting of various objects (tunnels, balls, soft materials and locomotive items, such as wooden ramp, ladders, stairs and running wheel) resulted in significant reduction in cerebral amyloid-β deposits in AD transgenic mice (Lazarov et al., 2005; Adlard et al., 2005; Ambree et al., 2006; Herring et al., 2008). Mice subjected to enriched environment for 4 weeks also showed up-regulation of pro-angiogenic genes, increased cerebral vessel density, along with decreased cerebral levels of RAGE and increased LRP1 levels which would favor a net clearance of Aβ from brain (Herring et al., 2008). Increased cerebrovascular density would lead to a larger surface area for transport of molecules across the BBB. Recent studies have shown beneficial effects of physical activity on cognition in older adults (Lautenschlager et al., 2008; Middleton et al., 2008; Baker et al., 2010; Ahlskog et al., 2011; Graff-Radford, 2011; Tseng et al., 2011).
Cholesterol is an essential component of the cell membrane and an important precursor for both steroids and bile acids. Hypercholesterolemia is a major risk factor of vascular disease (Kannel et al., 1971). Some studies have indicated that elevated serum cholesterol levels might increase the risk of AD (Notkola et al., 1998; Kivipelto et al., 2001; Schneider & Simons, 2012). Systemic and central cholesterol levels maybe controlled independently since the BBB restricts the transport of apolipoprotein-containing cholesterol from blood to brain (Martel et al., 1997) and from brain to blood (Bell et al., 2007; Deane et al., 2008). Statins are cholesterol lowering drugs that reversibly inhibit the rate-limiting enzyme, hydroxymethylglutaryl-coenzyme A (HMGCoA) reductase, in cholesterol biosynthesis (Goldstein & Brown, 1990). There are several statins, including simvastatin, atorvastatin, fluvastatin, cerivastatin, pitavastatin, pravastatin and rosuvastatin, each of which has differences in physio-chemical properties (Illingworth et al., 2001). The lipophilic statins, such as simvastatin and atorvastatin, are more likely to enter the brain than the hydrophilic statins, such as pravastatin and rosuvastatin (Vuletic et al., 2006; Sierra et al., 2011).
Retrospective case studies have suggested that statins reduce the risk of AD (Jick et al., 2000; Wolozin et al., 2007), but prospective studies and clinical trials have produced inconsistent data (Di Paolo & Kim, 2011) with some showing reduced risk (Wolozin et al., 2000; Cramer et al., 2008; Sparks et al., 2008; Haag et al., 2009) and others showing no change (Rea et al., 2005; Zandi et al., 2005; McGuinness et al., 2009; Benito-Leon et al., 2010; Feldman et al., 2010). These conflicting data may be due to a number of factors, including the type of statin used, duration of the treatment, baseline cholesterol plasma levels, age at which the statin therapy was started and whether a given statin can enter the brain and influence central (e.g., CSF) cholesterol levels. Interestingly, there were no differences in the reduction in risk of AD produced by simvastatin (BBB permeable) and pravastatin (BBB impermeable) (Haag et al., 2009). In addition to lowering cholesterol, statins possess multiple cholesterol-independent beneficial effects (Evans et al., 2009). For example, they increase endothelial nitric oxide (NO) production and blood flow by upregulating NO-synthase (Laufs et al., 1998), increase tPA levels (Essig et al., 1998) and reduce levels of the potent vasoconstrictor endothelin-1 (Hernandez-Perera et al., 1998). Collectively, these changes could increase cerebral blood flow which could revitalize the brain and help mitigate neuronal dysfunction (Bell et al., 2009; Zlokovic, 2011).
Statins increase the expression of LDL receptors and LRP1 in organs such as the liver (Rudling et al., 1992; Moon et al., 2011a). However, there is little work on the role of statins on LRP1 expression at the BBB. Our earlier studies have shown that treatment of human brain endothelial cells with simvastatin or lovastatin increased expression of LRP1 (Deane et al., 2004b). In a transgenic mouse model of AD (APP23), treatment with fluvastatin orally for 4 weeks increased LRP1 expression levels at the BBB leading to increased Aβ clearance and reduced accumulation (Shinohara et al., 2010). While these are promising data further work is needed to study the role of statins on LRP1 expression at the BBB and brain in AD.
In a recent study, Probucol, a non-statin cholesterol-lowering drug was found to prevent cognitive impairment induced by Aβ peptide in mice (Santos et al., 2012). Probucol was originally synthesized as an antioxidant (Yamamoto, 2008) but later found to have cholesterol lowering properties (Buckley et al., 1989). It is of note that Probucol administration in aged rats for 30 days resulted in a significant increase in hippocampal LRP1 expression (Champagne et al., 2003).
Plant-based active principles have played a vital role in modern drug development. In a recent report, oral administration of an extract from the root of Withania somnifera was shown to enhance LRP1 expression in brain microvessels and liver and increase the levels of circulating sLRP1 in plasma resulting in reversal of AD pathology in APP/PS1 AD transgenic mice (Sehgal et al., 2012; Dries et al., 2012).
While the BBB is also a major obstacle to the effective delivery of gene transfer vectors into the brain, restoration of age-dependent or disease-associated reductions in LRP1 expression at the BBB may be accomplished through specific targeting of cerebrovascular endothelial cells. Endothelial cells are the cellular constituents of the BBB and possess two opposing plasma membranes: one facing the vascular lumen and is exposed to blood and one facing brain parenchyma and is exposed to brain interstitial fluid. Therefore, when targeting cerebrovascular endothelial cells, gene transfer vectors do not need to cross the BBB given direct plasma exposure when administered systemically. Effective gene transfer can be achieved with either non-viral or viral systems. However, viral based systems are usually more effective in mediating cell entry and transfer of genes to the target cell and are in many instances preferred (Davidson et al., 2000). Adeno-associated virus (AAV) is frequently used as a vector for gene therapy for achieving long-term gene expression in both dividing and non-dividing cells (Grieger & Samulski, 2012). AAV also lacks pathogenicity as it is unable to autonomously replicate without a helper virus (Grieger & Samulski, 2012). The capsid amino acid sequence for AAV serotypes are different, except 1 and 6 (Daya & Berns, 2008), and may contribute to host cell specificity. Of the common serotypes, AAV-2 has been developed for transduction of brain endothelial cells by phage panning (Chen et al., 2009), and was used in neurological studies in mouse models of lysosomal storage disease (Fu et al., 2002; Chen et al., 2009) and Parkinson disease (Bankiewicz et al., 1998). Recently, AAV-9 with a lower prevalence of neutralizing antibodies in humans and more efficient gene transfer capacity in vivo has been developed for endothelial cell-directed gene transfer (Varadi et al., 2011).
Targeting the gene transfer vector to the BBB could be direct or indirect. In the direct approach, vector targeting is mediated by a small peptide that has been inserted into the viral capsid sequence to modify viral tropism, as used for endothelial cells (Stachler & Bartlett, 2006; White et al., 2004). This could be enhanced by identifying molecular signature epitopes in the cerebral endothelial cells of AD brains and presenting these epitopes on the capsid of AAV to enhance site-specific distribution after intravenous injections (Chen et al., 2009). In lysosomal storage disease, specific epitopes have been identified suggesting unique vascular signature imparted by the disease (Liu et al., 2005; Chen et al., 2009). The indirect targeting involves associating molecules that interact with both viral surface and the specific cell surface receptor such as antibodies (Bartlett et al., 2000) and biotin (Ponnazhagan et al., 2002). In summary, it may be possible to use AAV-2 carrying the cDNA of LRP1 or its smaller fragments to restore reduced LRP1 expression in brain vascular endothelial cells in AD.
In plasma, sLRP1 is the main carrier of plasma Aβ since it binds 70-90% of Aβ in normal individuals (Sagare et al., 2007). In AD patients, sLRP1 levels are reduced by about 30% and dysfunctional due to its oxidation (Sagare et al., 2007). Consequently, there is reduced Aβ binding to sLRP1 resulting in an increase of free Aβ levels in plasma (Sagare et al., 2007; Sagare et al., 2011a) which could re-enter brain via RAGE (Martel et al., 1996; Deane et al., 2004a, 2004b). In mild cognitive impairment (MCI), patients who progress to AD (MCI-AD) and AD patients have greater levels of oxidized sLRP1 and free Aβ in plasma. Furthermore, these elevations correlate with established biomarkers such as increased CSF tau/Aβ42 ratio and reduced MMSE scores (Sagare et al., 2011a). In a mouse model of AD (6 months of age), treatment with low levels (40 g/kg) of recombinant wild type ligand binding domain IV of LRP1 (LRPIV), a truncated form of endogenous LRP1 which consists of a single binding domain with high affinity for Aβ for 3 months reduced brain Aβ levels, improved cerebral blood flow (CBF) responses to brain stimulation and enhanced learning and memory without altering plasma levels of other LRP1 ligands such as apoE, tPA, and MMP9 (Sagare et al., 2007). Treatment with LRPIV did not affect levels of LDLR and LRP1 in brain, liver or kidneys, nor did it affect the levels of phosphorylated LRP1 in brain and liver (Sagare et al., 2007). LRPIV did not enter CSF (Sagare et al., 2007) and, therefore, it was efficacious systemically. These data demonstrated that recombinant LRPIV holds potential as a replacement therapy for AD and should be explored in other pre-clinical models. To further enhance specificity for binding Aβ in relative isolation of other plasma ligands, site-directed mutagenesis was performed with LRPIV (Zlokovic et al., 2009). A lead mutant LRPIV has been identified and selected as a result of its preferential binding to Aβ40 and Aβ42. More specifically, this LRP1 fragment binds Aβ42 with 3-fold greater affinity than the LRPIV used in the earlier studies, while binding other LRP1 ligands with reduced affinity (Sagare et al., 2007). This mutant LRPIV may, therefore, represent a more effective sLRP1 replacement therapy for MCI-AD and AD patients than endogenous sLRP1.
One of the major functions of the liver is to regulate the levels of multiple diverse plasma proteins through controlling both biosynthesis and clearance (Charlton, 1996). LRP1 is highly expressed in hepatocytes serving as a cargo transporter of many macromolecules, including chylomicron remnants (Willnow et al., 1994; Mahley & Huang, 2007), α2 macroglobulin (Willnow et al., 1995; Poller et al., 1995), serine protease inhibitors (serpin)-enzyme complex (Kounnas et al., 1996; Maekawa & Pollefsen, 1996) and factor VIII (Lenting et al., 1999; Bovenschen et al., 2003). Interestingly, conditional deletion of LRP1 from liver reduced the plasma levels of HDL-containing cholesterol compared to wild-type mice (Basford et al., 2011). Recently, the liver was shown to be a major source of peripheral Aβ and that inhibition of peripheral Aβ production with Gleevec reduced plasma and brain Aβ levels (Sutcliffe et al., 2011; Sagare et al., 2011b). The liver is also a major organ for rapid peripheral clearance of Aβ (Tamaki et al., 2006) and LRPIV-Aβ complexes (Sagare et al., 2007). Hepatic LRP1 levels have been shown to be reduced in aging rats, and this is associated with decreased peripheral Aβ clearance (Tamaki et al., 2006; Tamaki et al., 2007). Increased Aβ level in brain is associated with reduced systemic Aβ clearance in aging Squirrel monkeys (Mackic et al., 1998b; 2002). It is also possible that the age-dependent reduction in liver LRP1 expression may contribute to systemic accumulation of macromolecules normally cleared by LRP1 that could lead to atherosclerotic lesions (Espirito Santo et al., 2004). Thus, restoring normal LRP1 levels in liver is essential for peripheral clearance of Aβ and the sink action for removal of brain Aβ. Potential therapies to increase LRP1 expression levels in liver include a) statins, b) insulin, and c) expression of recombinant LRP1 fragments such as LRPIV by gene therapy.
The liver is the major organ for the synthesis of peripheral apolipoprotein-associated cholesterol and, therefore, targeted by statins to reduce systemic cholesterol levels and consequently decreasing the risk of cardiovascular disease or stroke (Hiki et al., 2009). In the liver, statins not only decrease the synthesis of cholesterol but also increase the expression levels of LDLR, which increases the uptake and clearance of systemic apolipoprotein-associated cholesterol. However, studies on the role of statins on LRP1 expression in this organ are limited. In one study, rats treated orally with atorvastatin (20 mg/kg, daily) for 6 weeks starting at 20 weeks of age had increased levels of LRP1 and, as expected, LDLR in the liver (Moon et al., 2011a). Identifying statins that can increase the LRP1 expression levels at the BBB and in liver could provide rapid and effective clearance of brain Aβ Further work is needed to extend these studies using different statins, doses, duration of treatment, different starting ages of treatment, different models of AD and in different species.
Epidemiological studies have suggested an association between insulin-resistant type II diabetes mellitus and AD (Matsuzaki et al., 2010). However, the mechanism for this is unclear. Since insulin is lower in type II diabetes, its role in regulating LRP1 in liver was explored. In rats, insulin treatment increased LRP1 levels in the liver by facilitating its translocation to the hepatic plasma membrane which enhanced LRP1-dependent Aβ clearance by the liver (Tamaki et al., 2007). If these findings could be confirmed in humans, then insulin treatment would potentially benefit a subgroup of AD patients.
One of the goals would be to selectively enhance LRP1-dependent Aβ clearance by the liver but no other LRP1 ligands in plasma, so as to minimize the risk of potential adverse effects. Similar to the approach suggested for targeting cerebrovascular cells in the BBB, including gene transfer vector to express mutant LRPIV (Zlokovic et al., 2009) may selectively increase the clearance of peripheral Aβ via the liver without significant changes in plasma levels of other LRP1 ligands. The most promising gene transfer vector system for the liver is the AAV vectors, which have been successfully used for many liver disorders including hypercholesterolemia (Lebherz et al., 2004; Van Craeyveld et al., 2011). Using a vector dose of 1012 genome copies (gc) carrying the human LDLR cDNA and the liver specific promoter thyroxin binding globulin (TBG), the transduction efficiency was 25-fold greater for AAV-8 compared to AAV-2 in LDLR-/- mice (Lebherz et al., 2004). AAV-8 achieved a stable transduction, high levels of LDLR expression in liver, without toxic effects and normalized plasma lipid profile. AAV-8 is highly effective in liver transduction in several species including mice, dogs and non-human primates, and the expression of green fluorescent protein is not affected by the type of promoters (Bell et al., 2011). At present, AAV-8 has the most favorable features for clinical applications of liver-directed gene transfer (Wang, 2011). It may be possible to use AAV-8 to carry the cDNA of a mutant LRPIV and a liver-restricted promoter to restore normal function of LRP1 in the liver of AD patients. While this approach appears feasible, it needs to be tested in mouse models of AD for its efficacy in controlling brain Aβ levels.
We have reviewed the evidence suggesting that LRP1 has a major role in regulating the clearance of Alzheimer's Aβ from brain. Specifically, we have discussed LRP1-based potential therapeutic interventions at three steps in restoring the normal controlling mechanisms regulating brain Aβ levels. The multimodal regulatory roles of cell surface LRP1 at the BBB, circulating soluble LRP1 and LRP1 in peripheral organs such as the liver could be a major but still underexplored therapeutic target in AD.
Although most AD patients are affected by one or two vascular risk factors and/or develop vascular changes in brain (i.e., > 65% of cases over the age of 60 and > 85% of cases over 80) (Wu et al., 2005; Bell et al., 2009; Zlokovic, 2011), it remains elusive whether MCI patients who later develop vascular dementia without amyloid deposition can also develop lower activity in the LRP1 system (Sagare et al., 2011). Future studies should address this important issue. It would be interesting to determine the impact of the LRP1 system on non-Aβ related vascular pathway of AD pathogenesis. For instance, recent findings indicated that pericytes are critical for maintaining the BBB integrity (Bell et al., 2010), and that the inability of astrocyte-secreted apoE (i.e., apoE4) to interact with LRP1 on pericytes may lead to BBB breakdown and microvascular degeneration followed by secondary neuronal degeneration (Bell et al., 2012). This latter study suggested that lower expression of LRP1 in vascular cells may promote vascular dysfunction and secondary neuronal degeneration independently of Aβ clearance. More studies are needed, however, to clarify the exact role of LRP1 in vascular dementia
Some studies have shown that BBB is leakier in AD patients than in age-matched controls (Zipser et al., 2007; Farrall & Wardlaw, 2009). Exactly how this influences the LRP1 clearance system in brain is not clear at present. It is conceivable that focal BBB disruptions will allow more Aβ to enter the brain from the periphery which might add an additional burden to the LRP1 clearance mechanism at the BBB.
Pharmacological and/or gene therapy strategies have been shown to increase the levels of LRP1 at the BBB and may hold the potential to prevent and/or reduce Aβ accumulation during AD pathogenesis. Replacement therapy for the native oxidized and dysfunctional circulating sLRP1 with recombinant LRP1 clusters may help maintain the peripheral Aβ ‘sink’ activity at an early stage of the disease in patients with MCI by reducing the influx of circulating Aβ into the brain which may slow down disease progression. Oxidized sLRP1 in plasma, which does not bind peripheral Aβ, is also an important early biomarker in MCI individuals converting into AD as well as in AD patients. Finally, approaches to increase LRP1 expression in peripheral organs such as the liver pharmacologically or by gene therapy might help promote systemic Aβ elimination, which has been shown to improve peripheral Aβ-‘sink’ activity by reducing sLRP1-Aβ and free Aβ levels in plasma and promoting Aβ clearance from the brain.
As a note of caution, we would like to stress that the development of LRP1-based therapies for AD requires careful toxicity and safety monitoring of unwanted potential side effects given that LRP1 participates in multiple control systems in the body, i.e., from the cellular transport and metabolism of cholesterol to anticoagulation and inflammation. However, our preliminary findings suggest that LRP1-based therapeutics such as the recombinant LRP1 clusters (Zlokovic et al., 2009) can be tailored by genetic engineering to be more specifically directed at Aβ with minimal effects on other systems in the body. These new LRP1 variants produced by site-directed LRP1 mutagenesis may have the potential either as recombinant proteins adapted for systemic administration and/or as Aβ-selective LRP1 gene constructs adapted for tissue-specific expression in the brain, liver and/or other peripheral organs.
Findings were supported by R37 AG023084, R37 NS34467 and HL63290, the ALS Association (grant 1859) and the Zilkha family to B.V. Zlokovic and AG29481 to R. Deane. The authors thank Ethan Winkler for critical reading.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of interest statement
BVZ is the scientific founder of Socratech LLC, a start-up biotechnology company with a mission to develop new therapeutic approaches for stroke and Alzheimer's disease. APS and RD have no conflicts of interest to disclose.