Reelin in Alzheimer’s Disease,Increased Levels but Impaired Signaling: When More is Less

INTRODUCTION

Alzheimer’s disease (AD) is the best-recognized form of dementia, associated with a characteristic loss of memory and cognitive functions, and with a prevalence that increases with age. The pathological hallmarks of AD include extensive synaptic and neuronal loss, astrogliosis, characteristic extracellular deposition of the amyloid-β peptide (Aβ) [1], and the formation of intracellular neurofibrillary tangles of the abnormally hyperphosphorylated microtubule-associated protein, tau [2]. It is now generally accepted that Aβ and hyperphosphorylated tau (P-tau) are key pathological effectors of AD and consequently, much effort has focused on understanding their toxicity in AD. Nevertheless, how these proteins interact, how they affect the expression of other key brain proteins, and, ultimately, how they drive synaptic loss and neurodegeneration are among the fundamental questions that remain to be resolved.

The abnormal aggregation of Aβ into plaques in the brain is one of the most specific pathological features of AD. The Aβ peptide is generated by processing a larger type I transmembrane spanning glycoprotein, the amyloid-β protein precursor (AβPP), through the successive action of proteolytic secretases. Sequential processing of AβPP begins with the activity of either α- or β-secretase, followed by γ-secretase cleavage. The AβPP amyloidogenic pathway is driven by β- and γ-secretase, which leads to the formation of the 40–42 aminoacids Alzheimer-associated Aβ peptide; whereas when AβPP molecules are first cleaved within the Aβ domain by α-secretase, the generation of the Aβ peptide is precluded. The existence in human cerebrospinal fluid (CSF) of several shorter Aβ isoforms has been explained by an alternative AβPP processing involving concerted cleavage of AβPP by α- and β-secretase [3]. Moreover, a new AβPP processing pathway has been recently demonstrated, mediated in part by the membrane-bound matrix metalloproteinases MT5-MMP, referred as η-secretase, and which generates proteolytic fragments capable of inhibiting neuronal activity within the hippocampus [4].

Tau hyperphosphorylation and aggregation is an aspect of AD that is shared with another group of neurological disorders, the tauopathies, which are also frequently associated with dementia [5]. AD can be considered to be unique among tauopathies as it involves the formation of Aβ plaques as well as neurofibrillary tangles. Thus, it may be particular interesting to decipher the mechanisms and proteins involved in the crosstalk between Aβ and P-tau.

Reelin is a large signaling protein that regulates the migration of neurons during brain development and it is essential for the correct organization, development and plasticity of the cerebral cortex [6]. However, Reelin also influences synaptic neurotransmission, plasticity, and memory in the adult brain [7–10]. The Reelin signaling pathway involves a cascade of intracytoplasmatic events that ultimately limits the extent to which the tau protein is phosphorylated. This signaling is initiated by the binding of Reelin to its receptors, the apolipoprotein E receptor 2 (ApoER2) or the very-low-density liporeceptor (VLDLR) [11, 12]. Reelin binding induces the cleavage of its receptors through the sequential processing of α- and γ-secretases, which, as indicated above, also process AβPP [13–15]. Binding of Reelin to its receptor relays signals into the cell via the Dab1 adapter (Disabled-1) [16–18]. Reelin induces Dab1 tyrosine phosphorylation by Src-family kinases (SFKs) [19], which triggers an intracellular phosphorylation cascade that involves phosphatidylinositol 3-kinase (PI3K) and protein kinase B/Akt [20], and that also triggers the mitogen-activated protein kinase/extracellular signal-regulated kinase (ERK) pathway [21]. The intracellular Reelin pathway ultimately inhibits glycogen synthase kinase-3β (GSK3β), the main tau kinase [22–24], which prevents tau hyperphosphorylation [20]. Interestingly, the cyclin-dependent kinase 5 (CDK5) that also phosphorylates tau [25] can phosphorylate Dab1 independently of Reelin signaling [26]. CDK5 also participates in the positioning of cortical neurons in the developing mouse brain [27], suggesting a cross-talk between Reelin signaling and the CDK5 cascade.

Reelin signaling also regulates the activity of NMDA receptors (NMDARs) through a mechanism that requires Dab1 and the activation of SFKs [28]. NMDARs are glutamate receptors and their ion channel activity is involved in controlling synaptic plasticity and memory. Reelin augments the Ca^2 + entry through NMDARs, leading to enhanced phosphorylation and translocation of the CREB transcription factor (cAMP-response element binding protein), which ultimately modulates learning and memory (see Fig. 1A for an overview of the proteins involved with the Reelin signaling pathway).

In transgenic mice models that lack Reelin or that dampen its activity, or in mice carrying mutations that prevent Reelin-dependent Dab1 tyrosine phosphorylation, tau phosphorylation is enhanced [12, 30]. The reduction in Reelin in transgenic mice accelerates Aβ deposition, which ultimately causes synaptic dysfuntion [30]. Likewise, Reelin knock-out mice are more sensitive to synaptic suppression induced by Aβ, and they show memory and learning disabilities [31, 32].

In the light of this, we have demonstrated that in vitro treatment with Aβ impairs Reelin signaling [33] and augments intracellular Reelin [34]. These effects on Reelin have been explored in mice over-expressing Aβ as well as in the brains of AD patients, albeit producing contradictory results. Increases in Aβ have been associated with both a depletion [35–37] and an increase in Reelin levels [34, 39].

In summary, Reelin has emerged as a signaling protein that is associated with both Aβ and tau hyperphosphorylation. Moreover, evidence that will be summarized here links elements of its signaling pathway to proteins that have been strongly implicated in AD. The goal of this article is to review the information available regarding Reelin expression in the AD brain, summarizing our recent findings as to how Aβ affects Reelin levels and its glycosylation. In addition, we discuss how ApoER2 processing may impair Reelin signaling. We propose that while Reelin expression is enhanced in the AD brain, Aβ hinders its biological activity and impairs Reelin signaling, which ultimately influences tau hyperphosphorylation, as well as memory and leaning.

REELIN SIGNALING PROTEINS, ALTERNATIVE LIGANDS, AND THEIR INTERACTIONS WITH AβPP

Previous studies indicated that AβPP can interact with Reelin [36, 40] and, as already mentioned, an interaction between Aβ and Reelin has recently been demonstrated in vitro [31]. Reelin receptors can also physically interact with AβPP (e.g., ApoER2) and in fact, AβPP is involved in the clustering of Reelin receptors that occurs after Reelin binding [41–43]. AβPP trafficking and processing is also modulated by the interaction with ApoER2, the intracellular adaptor Dab1 and the intracellular adaptor protein Fe65, which also interact with Reelin receptors [44–48]. Indeed, boosting ApoER2 expression enhances the association of AβPP with lipid rafts and increases Aβ production [40]. Accordingly, a direct role for ApoER2 on amyloid deposition and neurodegeneration in AD has been suggested [49]. There is a further evidence of a relationship between ApoER2 and AβPP, such as the ability of ApoER2 to recruit the c-Jun N-terminal kinase (JNK)-interacting proteins (JIP)-1 and -2 [50, 51], which act as scaffolds that allow the cytoplasmic domains of AβPP to interact with functional intracellular molecules [52]. The receptor-associated protein (RAP), a specialized chaperone for members of the low-density lipoprotein receptor family [53], also interacts with Aβ peptides and it promotes their cellular internalization [54]. Alternative ligands of Reelin receptors like F-spondin [55], and co-receptors like β1 integrin [56], have also been proposed to regulate the shedding of AβPP ectodomains [40, 57].

Not only is there considerable evidence that Reelin and components of its signaling pathway can influence AβPP processing but conversely, AβPP processing may also influence Reelin signaling. AβPP cytoplasmic domains in the cytosol bind to Dab1 and retain it in the cytoplasm. Thus, enhancing the cytoplasmic accumulation of AβPP prevents Dab1 from reaching the plasma membrane, and it promotes the sequestering of tyrosine phosphorylated Dab1, both of which disrupt Reelin signaling [58].

Reelin receptors can also bind apolipoprotein E (ApoE), which competes for the receptors [11, 12]. Interestingly, ApoE, and in particular the ApoE4 isoform, reduces the neuronal surface expression of ApoER2 and the ability of Reelin to prevent LTP suppression, which may accelerate the onset of dementia and neuronal degeneration [59]. Furthermore, the levels of ApoER2 are specifically affected by ApoE4 in vivo [60]. The ApoE4 variant is the most prevalent genetic risk factor for sporadic AD and one of the strongest competitors of Reelin binding to ApoER2 [11]. Thus, the differential effects of ApoE isoforms on Reelin signaling require further study. Binding of ApoE to ApoER2 independently of Reelin triggers the endocytosis of AβPP and leads to Aβ production. The ApoE4 isoform triggers more Aβ production than the ApoE2 or ApoE3 isoforms [61]. Clusterin, also known as apolipoprotein J, also plays an important role in the pathogenesis of AD [62, 63] and it acts as a ligand of Reelin receptors, potentially triggering its signaling cascade through the same pathway [64].

In summary, these multiple links between Reelin signaling and AβPP/Aβ strongly suggest that the progression of AD could influence Reelin signaling. Conversely, Reelin and elements of its signaling pathway might contribute to the neuronal dysfunction associated with AD and neurodegeneration. However, there is no consensus as to how Reelin levels are affected in AD, which complicates the interpretation of the role of this signaling protein in the disease.

THE GENETIC OF REELIN SIGNALING PROTEINS IN AD

The link between Reelin with AD is also suggested by the association of some Reelin single-nucleotide polymorphisms (SNPs) with the pathogenesis of AD [39, 65–68]. Moreover, a possible association with AD has also been proposed for genetic polymorphisms of both the Reelin receptors, ApoER2 and VLDLR [69–71], although this remains to be confirmed [72–74]. Recently, Dab1 was also proposed as a novel candidate gene for AD, although the strength of the contribution of Dab1 may differ among populations [75]. In addition, transcriptomics studies connected synaptic dysfunction with Reelin signaling in AD [75].

ASSESSMENT OF REELIN LEVELS IN AD

Before reviewing the existing data obtained from human samples, it is important to consider relevant biochemical and methodological issues related to the study of Reelin in brain extracts. Mouse Reelin is a large, 3461 amino acid secreted protein [77] with, like other extracellular molecules, many potential glycosylation sites. In fact, full-length Reelin is predicted to be  385 kDa in size and the higher molecular weight protein often detected in electrophoretic analysis ( 420 kDa) is thought to be the result of its glycosylation [78, 79]. Reelin undergoes proteolytic cleavage after interaction with its receptor [80–82], specifically at the sites referred to as N-terminal and C-terminal [83], resulting in the formation of several fragments whose relative abundance differs in distinct tissues [84]. The exact N-terminal and C-terminal cleavage sites were recently identified, together with candidate proteases [82, 85] (see also Fig. 2A). Reelin can also be processed through the activity of extracellular matrix metalloproteinases independently of its reaction with its receptors, which interestingly generates the same proteolytic fragments [86–89]. The full mechanism and significance of the specific proteolytic cleavage of Reelin remains elusive [90], although the proteolytic fragments of Reelin may fine tune Reelin signaling through ApoER2 binding [80, 91]. In fact, impaired Reelin processing has been associated with pathological progression in a mouse model of temporal lobe epilepsy [92]. Reelin forms homodimers in the human brain and plasma [93–95], these being the active forms that bind to receptors and initiate the Dab1-phosphorylation pathway [42]. Interestingly, truncated fragments of Reelin can form large assemblies and bind to receptors, but they do not efficiently induce the tyrosine phosphorylation of Dab1 [94].

Full-length Reelin (420 kDa) and two Reelin fragments (310 and 180 kDa) are detected in brain extracts and CSF with commonly used N-terminal antibodies, the smallest 180 kDa fragment being the most abundant (Fig. 2B). Accordingly, it is possible to measure the full-length Reelin protein and Reelin fragments at the same time, thereby evaluating the possible alterations in Reelin signaling associated with certain pathological conditions. However, in many studies of Reelin in the brain of AD patients only the abundant 180 kDa Reelin fragment has been analyzed.

There is some conflicting data emerging from studies on human AD brains and CSF (see Table 1). Reasons for the discrepancies among the studies are not clear, but one source of confusion and variability in published data maybe due to methodological factors influencing the measurement of Reelin. We first reported by western blots a significant increase of the highly abundant 180 kDa Reelin fragment in CSF from AD patients, compared to healthy individuals [96]. At that time, the limited sensitivity of the assay did not allow to estimate the amount of full-length Reelin in CSF. In a later study, there was no correlation of the CSF-Reelin 180 kDa fragment with the neurological disease state, probably due to the large subject-to-subject variation and/or the storage of samples at -20°C [97], an analytical condition that was later characterized, among others, to affect the assessment of Reelin levels [38, 95]. In this regard, other pre-analytical and analytical factors affect the estimation of Reelin, including freeze-thaw cycles, and sample boiling prior to SDS-PAGE, a step in the preparation of samples for western blot analysis [38, 95]. For all our following studies, we employed a 3 min denaturation at 98°C to prepare samples for western blots, avoiding freeze-thaw cycles, information that is not always available in other studies. In our next study increases in 180 kDa Reelin and full-length were found in AD CSF [38]. When brain extracts were studied, an increase in Reelin protein (full-length and fragments) and Reelin mRNA was corroborated in the AD frontal cortex, a regions targeted by the disease, while in the cerebellum from the same cases, a relatively neglected area of the AD brain, remained unaffected [38]. In the same study, intriguingly, the glycosylation pattern of Reelin appears affected in AD CSF. In a later report, the levels of Reelin 180 kDa fragment in the CSF from the AD group were not statistically different from those in the controls, although the alterations in glycosylation were confirmed [34]. Interestingly, in cultured cells treated with Aβ₄₂ peptide the levels of Reelin resulted increased and furthermore, with an altered pattern of glycosylation, resulting in changes similar to those observed in the AD brain and CSF [34]. We also found large increases in Reelin protein and mRNA in the brain of humans with Down syndrome, in which the genomic trisomy provokes AβPP over-expression [34]. Increased Reelin expression has been also possibly associated with the specific vulnerability of neurons to AD [98]. Finally, a decrease in Reelin protein levels was detected in several brain areas of AD subjects, although these changes did not reach significance at the transcriptional levels [37]. In this study, proteins were obtained after mRNA extraction with TRIzol reagent, which may influence Reelin determination. Morphometric analyses in brain supported the depletion of Reelin, but, interestingly, Reelin-positive deposits were only found in AD patients whereas control subjects remained free of Reelin accumulations [37].

Other groups have addressed the study of Reelin in AD analyzing its expression in the Cajal-Retzius cells, which express high levels of Reelin in the developing cerebral cortex [99]. Nonetheless, in the adult brain, Reelin expression in the cerebral cortex is mostly restricted to subsets of GABAergic interneurons, and remains low in Cajal-Retzius cells [99]. Anyhow, the studies of Reelin affectation in AD Cajal-Retzius cells were inconclusive. A first report indicated that Reelin-positive Cajal-Retzius cells are preserved in the AD entorhinal cortex [100], while a further study suggested that a decline in the number of Cajal-Retzius cells in the temporal isocortex was associated with AD [101].

Moreover, in terms of morphometry, considerable differences in staining have been reported with several anti-Reelin antibodies, which may be particularly relevant to the cell types with less Reelin [102]. The specific preparation of tissue for immunohistochemical procedures may also influence the detection of Reelin immunoreactivity [102, 103].

We have discussed above that methodological factor can contribute to the commented discrepancies among the reports; although, in our studies, we also found large inter-subject variability in Reelin protein and mRNA levels, particularly in AD cases. AD cases are typically classified with respect to their Braak and Braak stage based in neurofibrillary tangles and neuropil threads which exhibit a characteristic distribution pattern, and not in Aβ plaques which vary widely not only within architectonic units but also from one individual to another [104]. Indeed, in addition to inherent biological variability, sporadic AD is a disease of complex etiology, and amyloid accumulation is not an invariable occurrence being absent in a sizable proportion of AD patients [105]. Thus, since Reelin expression is mainly influenced by Aβ, it appears as possible that discrepancies between different studies with reduced sample size also reflect variability in amyloid load in the brain.

In transgenic AD models, where intrinsic variability is low and differences can only be attributed to other variables like Aβ dosage or the age of the animals, controversial results have also been obtained. In an AβPP transgenic mouse that carries the Swedish and Indiana familial AD mutations, less Reelin protein has been observed in the brain, together with either a decrease in mRNA expression [35] or with no evident transcriptional change [37]. Other studies in AD transgenic models indicate an early decrease in Reelin content, at least part of which corresponds to the 180-kDa Reelin fragments and not the full-length Reelin [106–108]. In Tg2576 AD mice that carry the AβPP Swedish mutation, a decrease was evident in the 180-kDa Reelin fragment but not for the 310-kDa fragment, although the study was performed on a small cohort of three animals in which the levels of full-length Reelin were not reported [36]. However, we found increased brain Reelin in Tg2576 AD mice [34]. Reelin accumulation has been described in aged Ts65Dn mice that possess three copies of the segment of mouse chromosome 16 containing AβPP (orthologous to the region of human chromosome 21 that is responsible for the phenotype of Down syndrome) [109]. By contrast, other microscopic analyses of transgenic mice revealed no major changes in the distribution of cells expressing Reelin in the hippocampus compared to the wild type controls [110, 111].

Also in this context, although Reelin is a soluble glycoprotein, there is increasing evidence that Reelin can be found into Aβ deposits [103, 112] and accumulate as Reelin-enriched aggregates [113]. Reelin staining tends to be very pronounced in AD regions where many neurons undergo granulovacuolar degeneration [39]. However, there is no noticeable Reelin staining in the neurofibrillary tangles and neuritic plaques in AD brains [39, 100]. By contrast, Reelin is present in corpora amylacea deposits, age-related spherical bodies thought to contain a collection of neuronal breakdown products, including aggregated proteins and abnormal glycogen bodies that are more prevalent in the AD brain [105]. In transgenic mice models that express both human mutant AβPP and human mutant presenilin-1, Reelin may be located in plaque-like structures [112]. Finally, in hippocampal plaques of aged wild-type mice, Reelin and Aβ co-localize [103], a physical interaction that may also compromise Reelin solubilization and the assessment of its levels in amyloid conditions.

IMPAIRED REELIN SIGNALING IN AD

As indicated, there is increasing interest in determining whether Reelin levels are altered in the AD brain so as to shed light on the potential role of the Reelin signaling in this disease [114–118]. A reduced Reelin expression is associated with impaired signaling, resulting in cognitive decline [119, 110] and increased tau phosphorylation [11, 30]. Inhibition of Reelin signaling is sufficient to impair memory [121]. Thus, decreased Reelin levels have been described in other neuropsychiatric conditions, such as schizophrenia and other psychoses, or epilepsy [114, 123], where impaired Reelin signaling is assumed to contribute to the clinical symptoms and pathological progression. The dysregulation of tau phosphorylation has been also hypothesized as a point of convergence in the pathogenesis of AD and schizophrenia [124].

Nevertheless, in AD the correlation between estimation of Reelin levels and Reelin signaling may be altered by the influence of Aβ. We recently demonstrated that Aβ alters Reelin glycosylation and compromises its capacity to bind to ApoER2, impeding the ability to down-regulate tau phosphorylation via GSK3β kinase [33]. Changes in Reelin glycosylation have been reported in other diseases, including liver cirrhosis and rheumatic pathology [125, 126]; and interestingly, glycosylation may also regulate ApoE receptor processing [13] and thereby compromise the binding of Reelin [127]. Indeed, differential splicing and glycosylation of ApoER2 regulates its role in synaptic function and memory [128]. However, we are still awaiting a specific study of the differential splicing and glycosylation of ApoER2 in AD.

We have associated the inability of Reelin to bind to its receptor, ApoER2, with the ineffectiveness of altered Reelin glycoforms to form homodimers, the active signaling ligand. Remarkably, Reelin fragments can form large protein complexes which can be characterized by native gels [93]. Indeed, we have demonstrated that Reelin extracted from AD brains has a tendency to form larger structures than homodimers [33], also involving shorter Reelin fragments (see Fig. 2C). While Reelin and Aβ interact in vitro, whether Aβ contributes directly to this impaired Reelin homodimerization by interfering with its assembly requires further investigation. To our knowledge, the existence of abnormal Reelin complexes has been not addressed in other neurological conditions, but the existence of these complexes should be taken into account when determining total Reelin levels in enzyme-linked immunosorbent assays [129]. The altered Reelin oligomerization in AD brains is likely to produce altered Reelin signaling.

How Reelin signaling is altered in the AD brain is difficult to assess. Studying both the full-length Reelin protein and its fragments can be informative, since Reelin proteolysis parallels signaling. However, the fact that Reelin cleavage can also be exerted extracellularly by matrix metalloproteinases, independent of any receptor interaction [83, 87–89], means that the quantification of Reelin fragments is not a good indicator of Reelin activity. Furthermore, recent evidence indicates an upregulation of matrix metalloproteinases in AD, in association with amyloid disease progression [130]. Therefore, to fully test whether Reelin signaling is really affected in the AD brain, it will be important to also assess the state of downstream events, like ApoER2 processing, the generation of receptor fragments and the phosphorylation of Dab1, essential for its role in memory and learning [7, 131].

Dab1 is not only an adaptor protein essential for the intracellular transduction of Reelin signaling but it can also be phosphorylated on serine by CDK5 in a Reelin-independent manner. Hence, Dab1 is likely to fulfill other unidentified functions [26] independent of those performed by Dab1 activated through tyrosine phosphorylation by SFKs [19]. Nevertheless, determining the tyrosine phosphorylation state of Dab1 in AD may reflect Reelin activity in the brain. To date the total levels of Dab1 are believed to be upregulated in the AD brain [132], although other studies indicated Dab1 is depleted in transgenic AD mice, with increased or unchanged levels in the AD brain [37]. Nevertheless, to date the tyrosine phosphorylation of Dab1 has yet to be studied in relation to AD.

Like Dab1, we postulate that quantifying ApoER2 fragments may reflect the efficiency of Reelin signaling in the brain. ApoER2 is apparently the dominant Reelin receptor in the human forebrain, participating in the modulation of synaptic plasticity and memory formation [79]. Reelin binding to ApoER2 instigates Reelin signaling in neurons, although it also induces the clustering and proteolytic processing of this receptor [42, 55]. The proteolytic processing of ApoER2 upon ligand binding involves the sequential action of α- and γ-secretases, similar to AβPP processing. These enzymes generate an intracellular C-terminal domain (ICD) and a soluble extracellular fragment, as described previously [13, 55]. Soluble extracellular fragments of ApoER2 encompass the entire ligand-binding domain and they may also have a dominant-negative effect on Reelin signaling [131]. Other ApoER2 ligands promote its cleavage, such as ApoE and α2-macroglobulin [55], and neurotrophins like BDNF can modulate this process [134]. Interestingly, the neuroprotective ApoE2 ligand drives the strongest accumulation of C-terminal ApoER2 fragments compared with the AD-associated ApoE4 [135], which did not produce significant accumulation of intracellular ApoER2 [55]. More recent studies indicated that clustering of ApoER2 is relatively weak with ApoE as the ligand, in the absence of lipoproteins [43].

Taken together, we speculate that impaired Reelin signaling will result in reduced processing of ApoER2 and the generation of fewer ApoER2 fragments. Thus, the quantification ApoER2 fragments in the brain and CSF appears to be a suitable read-out of Reelin signaling.

FINAL REMARKS AND CONCLUSIONS

To evaluate Reelin signaling in AD it is necessary to assess other factors involved in the cross-talk with other signaling pathways, as well as the roles of alternative ApoER2 ligands like ApoE and clusterin. The effect of ApoER2 splice variants on Reelin binding affinity [91, 136], and the complexity of Dab1 splicing [137], should also be contemplated as a means to modulate Reelin signaling [138, 139].

The mechanisms that regulate Reelin expression remain relatively unclear. We recently demonstrated a putative auto-regulatory mechanism whereby the release of the ICD fragment of ApoER2 following Reelin binding influences Reelin transcription [140]. We demonstrated that inhibiting the generation of the ApoER2 ICD by γ-secretase increased Reelin levels. In the AD brain, where Reelin binds less efficiently to ApoER2, we expect less ApoER2-ICD to be generated, which might increase Reelin expression as occurs in the AD brain [34, 38]. Interestingly, the expression of Dab1 also appears to be upregulated in the frontal cortex of the AD brain [132]. How Aβ affects Reelin activity, directly and/or by altering its glycosylation [33], requires further research. Nevertheless, the activation of Reelin signaling sets up a vicious circle whereby Reelin up-regulation may be driven by a chronic failure in Reelin signaling (Fig. 1B).

Increasing Reelin activity has been proposed as a therapeutic option for AD to protect against Aβ [31, 141]. Reelin supplementation can enhance cognitive ability and synaptic plasticity [142], and recover learning/cognitive deficits in a heterozygous “reeler” mouse [143], as well as in a mouse model for Angelman syndrome [144]. Reelin supplementation, as well as Reelin secretion by transplanted GABAergic precursor cells, may also prevent the induction of cognitive and sensory-motor gating deficits induced by phencyclidine [145, 146]. Finally, F-spondin, which also activates the Reelin pathway [147], improves spatial learning/memory in wild type mice and reduces Aβ plaque deposition in transgenic mouse models of AD [148].

In summary, since altered Reelin function is predicted to have significant effects on neural function, it is also plausible that acute activation of the Reelin pathway may represent a new therapeutic strategy to ameliorate the cognitive decline associated with AD pathology. A specific description of the effects of Reelin signaling in AD and greater efforts to increase our understanding of the interaction between Aβ and Reelin are fundamental to define potential therapeutic interventions.