Calcium-Sensing Receptors of Human Astrocyte-Neuron Teams: Amyloid-β-Driven Mediators and Therapeutic Targets of Alzheimer’s Disease

Abstract

It is generally assumed that the neuropathology of sporadic (late-onset or nonfamilial) Alzheimer’s disease (AD) is driven by the overproduction and spreading of first Amyloid-β_x-42 (Aβ₄₂) and later hyperphosphorylated (hp)-Tau oligomeric “infectious seeds”. Hitherto, only neurons were held to make and spread both oligomer types; astrocytes would just remove debris. However, we have recently shown that exogenous fibrillar or soluble Aβ peptides specifically bind and activate the Ca²⁺-sensing receptors (CaSRs) of untransformed human cortical adult astrocytes and postnatal neurons cultured in vitro driving them to produce, accrue, and secrete surplus endogenous Aβ₄₂. While the Aβ-exposed neurons start dying, astrocytes survive and keep oversecreting Aβ₄₂, nitric oxide (NO), and vascular endothelial growth factor (VEGF)-A. Thus astrocytes help neurons’ demise. Moreover, we have found that a highly selective allosteric CaSR agonist (“calcimimetic”), NPS R-568, mimics the just mentioned neurotoxic actions triggered by Aβ●CaSR signaling. Contrariwise, and most important, NPS 2143, a highly selective allosteric CaSR antagonist (“calcilytic”), fully suppresses all the Aβ●CaSR signaling-driven noxious actions. Altogether our findings suggest that the progression of AD neuropathology is promoted by unceasingly repeating cycles of accruing exogenous Aβ₄₂ oligomers interacting with the CaSRs of swelling numbers of astrocyte-neuron teams thereby recruiting them to overrelease additional Aβ₄₂ oligomers, VEGF-A, and NO. Calcilytics would beneficially break such Aβ/CaSR-driven vicious cycles and hence halt or at least slow the otherwise unstoppable spreading of AD neuropathology

PREAMBLE

Previous reviews from our Group have covered either the origins and diffusion of the “infectious Aβ and Tau seeds” advancing Alzheimer’s disease (AD) or more restricted views on the links between the calcium-sensing receptor (CaSR) and the overproduction of nitric oxide (NO) by cortical normofunctioning adult human astrocytes (NAHAs) exposed to mixtures of microglial proinflammatory cytokines and/or to Amyloid (A)β peptides (Aβs) [1-4]. The present work is based on recently gained evidences [5,6] revealing the manifold critical roles the Aβ-binding CaSRs of human astrocyte-neuron teams play to promote AD progression and the possibility of effectively interfering with the pathological consequences of Aβ●CaSR signaling through the administration of selective allosteric CaSR antagonists (“calcilytics”).

ALZHEIMER’S DISEASE SLOW CIRCUIT-BREAKING MARCH THROUGH THE BRAIN

Late-onset (sporadic or nonfamilial) Alzheimer’s disease (AD) patients (about 95% of all AD cases) are older than 65 years of age, whereas the infrequent younger cases (e.g., in their 40’s) suffer from the early onset (or familial, i.e. genetic) form of AD [7,8]. AD most likely subclinically starts its ∼30-year-long march through the brain in neurons of the parahippocampal region and the layer II of entorhinal cortex [8-10]. The neuropathology begins when such neurons groups increasingly lose their ability to rid themselves (via the action of proteases such as neprilysin or expulsion into the extracellular space and out of the brain) of the non-toxic Amyloid-β_x-42 (Aβ₄₂) peptide monomers they produce during their physiological activity [11]. The accumulating Aβ₄₂ monomers rapidly aggregate into neurotoxic oligomers [12]. The Aβ₄₂ oligomers start the full cytopathological cycle leading to the production of as well neurotoxic hyperphosphorylated (hp)-Tau oligomers, which impair axonal transport and cause synaptic failure by damaging dendritic spines [13,14]. Alternatively, the hp-Tau oligomers can by themselves trigger the hp-Tau production and/or hp-Tau “infection” [9,15]. Thus, the long march of the AD neuropathology begins when Aβ₄₂ and hp-Tau oligomers, both “AD infectious seeds”, are released from “infected” entorhinal neurons and disseminated to “infect” neurons of the dentate gyrus and the CA3 region of the memory-encoding and -retrieving hippocampal formation (reviewed in [1]). The neuropathology driven by the two kinds of “AD infectious seeds” slowly yet progressively spreads disconnecting neuronal circuits along the pathways going to the brain’s upper areas of the cognition machinery. It leaves in its wake neurons jammed by neurofibrillary tangles (NFTs) and vulnerable to being further damaged if not killed by proinflammatory cytokines released from a microglia activated by a trail of extracellular Aβ₄₂ fibrillar deposits and their associated oligomers and of “ghost” NFTs [8,14,15]. Because of the limbic sites of origin and subsequent upward trajectory of the AD neuropathology, it is a person’s increasingly failing memory and growing mental confusion that eventually attract the clinician’s attention.

ASTROCYTE-NEURON TEAMWORKS AND AD

The conventional view of AD pathology is a “neuronocentric” one, in which the neurons are the producers, transmitters, and at the same time the victims of the Aβ₄₂ and hp-Tau toxic oligomeric “seeds” [16]. But what about the astrocytes, of which in the AD-targeted regions there are least as many as, or maybe even several fold more than, neurons [17-19]? Astrocytes have traditionally been regarded as neuron-supporting cells and/or janitorial bystanders working with the activated microglia to sweep up neuronal debris [20]. But by experimentally using cultures of functionally normal adult human astrocytes (NAHAs) isolated from the cerebral cortex we have recently gained evidences that send a new message, that is “along with the neurons, the Aβ oligomer-exposed astrocytes become makers and spreaders of neurotoxic “AD infectious seeds” thereby remarkably advancing the transmission of the AD neuropathology” [5,21] (Fig. 1).

Fig. (1).

The interactions between exogenous Aβs and the CaSRs located on the plasma membrane of astrocyte-neuron teams advance the extracellular release and spread of Aβ₄₂ and hp-Tau oligomers, the “AD infectious seeds”. Two neurons and one astrocyte of the same team are schematically depicted here. The neuropathology begins when for aging-related causes Aβ₄₂ monomers accumulate in the extracellular space, oligomerize, and bind the CaSRs inserted in the plasma membranes of both cell types (1). The engendered Aβ●CaSR signaling induces the intracellular accrual (not shown) and oversecretion of de novo produced Aβ₄₂ monomers from both neurons and astrocytes (2). The Aβ₄₂ monomers oligomerize, spread, bind, and activate the CaSRs of a further neuron of the team (3). By doing this the Aβ₄₂ oligomers cause the additional release of surplus Aβ₄₂ moieties from this and other neurons (not shown). These vicious cycles can be unceasingly repeated and hence recruit ever-increasing numbers of astrocyte-neuron teams and thus inexorably advance the progression of AD. Hp-Tau oligomers formation is also triggered in the Aβ₄₂-exposed neurons by still unclear mechanisms that might be at least partially driven by pathological Aβ●CaSR signaling (4). Secreted hp-Tau oligomers are next either taken up by the secreting cells or transferred to contiguous neurons (and maybe astrocytes too) to hinder microtubule functions and help destroy synaptic spines (5). Once produced, hp-Tau oligomers are also capable of an independent self-induction and spreading. A highly selective allosteric CaSR antagonist (calcilytic) like NPS 2143 can completely suppress the manifold noxious effects driven by pathological Aβ₄₂●CaSR signaling both in neurons and in astrocytes thereby restoring conditions close if not identical to physiological ones [5,6]. Other relevant effects of the pathological Aβ₄₂●CaSR signaling, like the surplus production and secretion of NO and VEGF-A from the astrocytes, which are also suppressed by calcilytic NPS 2143 [6,38], and the extracellular accrual of Aβ₄₂ oligomers, which activate microglia, damage oligodendrocytes, and cause cerebral amyloid angiopathy, have been omitted from the picture for the sake of clarity.

It is well established that each astrocyte can embrace several “client” neurons with its mobile processes to form a working team [22-24]. For example, a single human astrocyte domain in the hippocampus can contain from 27 x 10⁴ to 2.0 x 10⁶ synapses vs. the 2.0 x 10⁴ - 12 x 10⁴ synapses of a rodent astrocyte domain [25-28]. Thus, human cortical astrocytes have from 1.4-to-16.7-fold more chances to directly trade with their neuron partners a variety of factors influencing, amongst other processes, memory formation and retrieval. And to add to the impact of astrocytes on neuronal networks, several astrocytes may embrace the synapses of a single neuron. Under normal conditions, astrocytes induce synapses and stabilize their neuron partners’ synapses by sweeping up glutamate and K⁺ spillovers and by modulating signal transmission via so-called “gliotransmitters” [22-24,29-31]. Regional astrocyte heterogeneity may be crucial to refine neural circuits postnatally, e.g., via the release of Semaphorin 3a (Sema3a) in the spinal cord [30,31]. An important consequence of this astrocyte-neuron physical and functional embrace is the astrocytes’ ability to promote or reduce the release of neurotransmitters into the synapses they envelop with the Ca²⁺ they expel or take up during their Ca²⁺ waves [32]. Moreover, the working astrocyte-neuron teams are fueled with glucose and oxygen when signalers secreted from the astrocytes’ vessel-touching end-feet open the arteriolar “taps” and increase the local blood flow [33].

But this constructive astrocyte-neuronal teamwork in the circuits of the cortical cognitive regions can become disruptive if aging or mutant neurons start producing, accruing, and releasing a surplus of Aβ peptides (Aβs). When neurons and the brain tissue become increasingly unable to get rid of the accumulating Aβ monomers, toxic Aβ oligomers are generated that trigger the formation, oligomerization, and release of pathological hp-Tau oligomers. Thus, neurons are likely to spill out both kinds of “AD infectious seeds” onto their astrocyte team partners [13] with consequences that are now beginning to emerge (Fig. 1).

An example of the destructive potential of an Aβ₄₂-attacked astrocyte-neuron team is provided by Talantova et al. [34]. Aβ₄₂ oligomers released from a neuron can bind to its astrocyte partner’s α-7-nicotinic acetylcholine receptors (α-7-nAChRs). Next, the signals from these receptors induce the astrocyte to exocytose the glutamate spillover from the neuronal synapses it has been sweeping up. The astrocyte-released glutamate activates the partner neuron’s extrasynaptic glutamate N-methyl-D-aspartate receptors (NMDARs). This in turn triggers intraneuronal Ca²⁺ surges inducing a cascade of events that include dysfunctional mitochondria pumping out reactive oxygen species (ROS). In sequence, ROS inflict oxidative damage that destroys neuronal synapses thereby cutting communications between the client neurons of the astrocyte’s network [14,34].

Amyotrophic lateral sclerosis (ALS) may be another example of astrocytes actually being in the driver’s seat of a major motor neuron disease. Re et al. [35] set up a model system in which astrocytes isolated from sporadic ALS patients are co-cultured with human embryonic motor neuron stem cells. They found that the ALS astrocytes secreted neurotoxic factors that killed the motor neurons via necroptosis. In other words, as stated by Pirooznia et al. [36], the motor neurons in sporadic ALS may be surrounded by killer astrocytes, which therefore should be the targets of therapeutic efforts. Thus, it seems that astrocytes might be the sole drivers of sporadic ALS, while both astrocytes and neurons are the co-drivers of AD.

In addition, astrocytes can express enough prion protein (PrP^C) to support progressive neurodegeneration in prion disease [37] and are also likely to act essential parts in neurodevelopmental diseases such as Rett syndrome and fragile X mental retardation [30].

Some of the roles astrocytes play in AD were brought to light by our demonstration that when cultured NAHAs isolated from the cerebral cortex were exposed to exogenous Aβs, both their β-secretase (BACE1) and γ-secretase (γ-S) activities were stimulated to produce increased amounts of Aβ₄₂ out of the endogenous amyloid precursor protein (APP) [5,21]. The astrocytes then intracellularly accrued and oversecreted the endogenous oligomerizing Aβ₄₂ monomers into the culture medium together with a surplus of nitric oxide (NO) and vascular endothelial growth factor (VEGF)-A [5,6,38]. If in vivo surplus Aβ oligomers, VEGF-A, and NO are released from the astrocytes’ end-feet, they will also impair the supply of oxygen and glucose by distorting and eventually destroying the local vasculature [39,40]. Thus can start the astrocyte-neuronal engine that widely disseminates cognition-destroying neurotoxic “AD infectious seeds” [13-15]. And notably, while the AD pathology is driven by the toxic secretions from both team members, it appears that human astrocytes are far stronger and abler to survive than their teams’ client neurons because they appear to be unaffected by their several neurotoxic secretions. Consequently, as the astrocytes can keep producing and releasing their neurotoxins, they become the chief murderers of the neurons [5,6].

Thus, our in vitro observations strongly suggest that Aβ₄₂ released from aging or mutant human neurons partaking in astrocyte-neuron teams in vivo would likely stimulate the teams’ astrocytes to make and secrete additional Aβ₄₂ oligomers. These would spread and recruit even more astrocyte-neuron teams to produce and secrete surplus Aβ₄₂ oligomers. The involved neurons would also synthesize and release hp-Tau oligomers (whether the astrocytes too play any part in this process remains to be ascertained). Therefore, both kinds of “AD seeds” would advance the inexorable progression of AD [5,13].

Here, the obvious question that arises is how to stop or at least significantly slow AD’s development. Ideally, inhibiting the overproduction of the pivotal Aβ₄₂ and its toxic oligomers by both the astrocytes and neurons would halt the disease. But preventing the astrocytes and their client neuron partners from oversecreting their endogenous Aβ₄₂ oligomers could also effectively hinder AD spreading and progression (Fig. 1).

Therefore, we set out to answer this question through the experimental use of NAHAs and human cortical postnatal (HCN-1A) neurons. This allowed us to find a way to stop the exogenous Aβ-treated NAHAs and neurons from oversecreting their endogenous Aβ₄₂ and concurrently the NAHAs from releasing toxic amounts of NO and VEGF-A [5,6,38]. The clue that led us to this discovery was the knowledge that Aβ oligomers bind several membrane receptors harbored by both astrocytes and neurons, namely those for ApoE, insulin, NMDA, cellular prion protein, advanced glycation end products, and in addition the p75 neurotropin receptor, α-7-nAChR, frizzled receptor, formyl peptide receptor-like 1, and the CaSR (see [5] for references). The CaSR was of particular interest to us, as we had previously shown that it is involved in the overproduction of NO by NAHAs exposed to a mixture of proinflammatory cytokines or to Aβ_25-35,an established proxy of Aβ₄₂ [5,41]. But before proceeding further it is worth recalling here the salient features of the CaSR to get a better glimpse of the complex roles it plays in AD.

STRUCTURE, SIGNALING, AGONISTS AND ANTAGONISTS OF THE CaSR

Structure

The CaSR belongs to family 3 of G protein-coupled receptors (GPCRs), also named seven transmembrane spanning receptors (7TMRs), and plays regulatory roles in Ca²⁺ homeostasis and cellular signaling [42]. It was first cloned from the bovine and later the human parathyroid gland [43,44]. In humans, the CaSR’s structure is comprised of an aminoterminal extracellular domain (ECD; 612 amino acids (aa)), seven α-helical transmembrane domains (TMD with TM1-TM7; 250 aa), which are typical of the GPCRs, and an intracellular domain with a carboxyterminal tail (ICD; 216 aa). A cysteine rich domain (CysRD) links the ECD with TM1 and is important for signals transmission to the TMD upon binding of a ligand (Fig. 2). The ECD’s surface is glycosylated and can form CaSR●CaSR homeodimers via a disulphide-link involving Cys¹²⁹ and Cys¹³¹ [45,46] or CaSR●mGluR1α heterodimers. The conformational arrangement of each monomer, determined by molecular modeling, indicates that the CaSR’s ECD has a bilobed Venus flytrap (VFT)-like structure. The orthosteric Ca²⁺-binding site is located between the two lobes of the VFT [46] (Fig. 2). The cleft formed by the orthosteric Ca²⁺-binding site is thought to be open when the agonist is missing and closed after binding the Ca²⁺ or any other orthosteric (type I) agonist. A second putative Ca²⁺-binding site is placed in the TMD, because an ECD-lacking CaSR still responds to a Ca²⁺ signal [46]. A binding site for aromatic L-amino acids is located near the orthosteric Ca²⁺-binding site. Several putative binding pockets are located in the seven extracellular loops of the TMD [47]. Ca²⁺ binding triggers changes in the conformational structure of the TMD and ICD that allow the ICD’s carboxyterminal tail to interact with various G proteins (e.g., G_11α, G_i/o, G_qα), which in turn mediate the activation/inhibition of manifold signaling pathways [48].

Fig. (2).

A CaSR homeodimer inserted in the plasma membrane. Each extracellular domain (ECD) has a bilobed Venus fly-trap (VFT; enclosing between its lobes the orthosteric binding site of the Ca²⁺ or of divalent or trivalent cations, aminoglycoside antibiotics and polyamines. A cysteine-rich domain (CysRD) connects the VFT with the first (TM1) of the seven transmembrane helices, being important for signal transmission. Aromatic L-α-amino acids, calcimimetics, and calcilytics are among the allosteric (type II) CaSR modulators and bind overlapping but not identical sites of the receptor, i.e. calcilytics between TM3 and TM5, and calcilytics and calcimimetics between TM6 and TM7. The last transmembrane helix (TM7) is joined to the intracellular domain (ICD). Ligand binding changes the CaSR conformation (not shown), and causes the interaction of the ICD’s tail with various G-proteins (not shown), which then activate/inhibit several signaling pathways (only a few of them are depicted here; see the text for more details and relevant references). By phosphorylating the ICD tail Ca²⁺/diacyl glycerol (DAG)-activated protein kinase Cs (PKCs) attenuate the signaling of the ligand bound CaSR. AA, arachidonic acid; Akt, protein kinase B; cAMP, cyclic AMP; p-CREB, phosphorylated cAMP response element-binding protein; ER, endoplasmic reticulum; IP3, inositol triphosphate; p38, JNK, and ERK are members of the mitogen-activated protein kinases (MAPKs); PA, phospholipase A.

Signaling

CaSR signaling is triggered and modulated by several factors and the kind(s) of response(s) it evokes depend(s) on the cell type considered and the pathways involved. Different ligands besides Ca²⁺ trigger CaSR’s activity, such as divalent and trivalent cations (e.g., Mg²⁺, Sr²⁺, La³⁺, Gd³⁺), polycations (e.g., polylysine, polyarginine), polyamines (spermine, spermidine, protamine), and aminoglycoside antibiotics (tobramycin, neomycin, gentamycin) [49-51]. These type I agonists bind the orthosteric Ca²⁺-binding site even in concert with the Ca²⁺ itself. The levels of these agonists vary in the tissues expressing the CaSR; hence the signaling differs accordingly. In tissues where the Ca²⁺ abounds, like the bone, CaSR’s activity is affected by the constitutive presence of high Ca²⁺ levels. On the other hand, the gastrointestinal mucosa cells are exposed to changeable Ca²⁺ concentrations depending on the variable composition of meals and their content of other cations [52,53]. Multivalent cationic proteins like Aβs also bind the CaSR [6,38,54,55] (Fig. 3B). Moreover, CaSR’s signaling activity is modulated by the extracellular pH and ionic strength, both of which alter the EC₅₀ for Ca²⁺, as they do in the kidney tubules [56,57]. CaSR signaling is affected also by factors modifying its level of mRNA and protein expression at a given Ca²⁺ concentration, e.g., cellular proliferation or mitotic quiescence, and agents like interleukin (IL)-1β, IL-6, vitamin D3, and Ca²⁺ itself [39,58-60]. On the other hand, stromal derived factor (SDF)-1 and macrophage chemotactic protein (MCP)-1 and some ligands (see below) regulate CaSR’s traffic from the endoplasmic reticulum to the plasma membrane [61]. Various studies have identified a panoply of G-protein-mediated CaSR signal transducing pathways besides the TRPC6-encoded Ca²⁺ channels. They include: (A) the inhibition of adenylyl cyclase preventing the synthesis of cyclic AMP (cAMP) [62]; (B) the activation of lipid kinases including several phospholipases with the production of (i) inositol triphosphate (IP3) eliciting the release of Ca²⁺ from intracellular stores [48]; and of (ii) diacyl glycerol (DAG) activating conventional protein kinase C (PKC) isoforms that phosphorylate at Ser¹³³ the CREB transcription factor [63] and CaSR’s Thr⁸⁸⁸ in the ICD that has been recognized as a critical negative regulator of CaSR signaling [64]; and (C) the stimulation of other protein kinases, e.g., AKT, and filamin-regulated mitogen-activated protein kinases (MAPKs), including MEK, ERK, and JNK [41,55,56]. In turn, the various second messengers involved activate downstream-placed signaling cascades, which are likely to amplify the noxious effects of a pathological Aβ●CaSR signaling [41,48,65] (Fig. 2).

Fig. (3).

The CaSRs of GFAP-expressing NAHAs form specific complexes with soluble Aβ_25-35 moieties visualized via the in situ PLA (isPLA) method. (A). Untransformed proliferatively quiescent NAHAs set in culture strongly express the GFAP marker. WB. GFAP Western immunoblotting was carried out according to Chiarini et al. [108] on NAHAs protein lysates using an anti-human GFAP rabbit polyclonal antiserum followed by an alkaline phosphatase (AP)-conjugated secondary antibody. Lane 1, molecular weight markers. Lane 2, specific GFAP protein band (53 kDa). IHC. GFAP immunohistochemistry on NAHAS using an anti-human GFAP mouse IgG1 followed by an AP-conjugated secondary antibody (for technical details see [108]). (B) Imaging of specific Aβ_25-35●CaSR complexes via the in situ Proximity Ligation Assay (isPLA) according to Söderberg et al. [105] and 3D digital renderings. Proliferatively quiescent NAHAs were exposed for 1 h at 37 °C to soluble biotinylated Aβ_25-35 (5.0 mM, Anaspec, CA, USA) dissolved in complete growth medium prior to fixation in 4 %v/v paraformaldehyde (PFA). Notably, after fixation NAHAs were not permeabilized, which restricted the antigen-antibody interactions to the outside of their plasma membrane. A mouse anti-CaSR monoclonal and a rabbit anti-biotin polyclonal were the two primary antibodies. The PLUS and MINUS isPLA probes were donkey anti-rabbit IgG (heavy + light chains) and donkey anti-mouse IgG (heavy + light chains), respectively (Olink Bioscience, Uppsala, Sweden). Once challenged with antibodies and isPLA probes, the samples were sequentially incubated with the Duolink Hybridization Solution, Ligation Mix, and Amplification Mix (all from Olink Bioscience) to produce single-stranded rolling-circle amplification products that were hybridized to oligonucleotide probes labeled with the red fluorophore Tye624 (λ_ex = 594 nm and λ_em = 624 nm) by incubating with the Duolink Detection Reagents (Olink Bioscience). Next, nuclear DNA was stained with 1.0 µg ml⁻¹ 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI, Sigma-Aldrich). Pictures were acquired under a Leica TCS SP5 AOBS Laser confocal microscope (Leica-Microsystems, Wezlar, Germany). A 40X/1.25 NA oil-immersion objective (HCX PL APO 40x 1.25 OIL UV, Leica-Microsystem) was used. The Maximum Intensity Projection (MIP) renderingsof original isPLA pictures were obtained using the Image Pro Plus 3D Viewer (Image-Pro Plus ^TM, version 7.0, Media Cybernetics, Bethesda, MD). (I) Untreated (i.e. not exposed to Aβ_25-35) NAHAS show no specific isPLA signal or background. Only the DAPI-stained nuclei are detectable. (II) Aβ_25-35-incubated NAHAs exhibit isPLA signals (red dots) corresponding to specific Aβ_25-35●CaSR complexes that are in part aggregating in patches. Nuclear DNA is blue (DAPI). (III) A higher magnification of Aβ_25-35-exposed adult human cortical astrocytes showing specific Aβ_25-35●CaSR complexes (in red) visualized via the isPLA method. Nuclear DNA is blue (DAPI). The complexes have started aggregating in patches, an event preceding their internalization. Some of the patches have been purposedly marked by differently colored arrows to allow the identification of the same patches in a different projection shown in panel (IV). (IV) The 3D MIP-rendering of the picture shown in panel (III) seen in an oblique lateral projection. The colored arrows indicate the same specific Aβ_25-35●CaSR complexes (in red) as in panel (III). The location of the Aβ_25-35●CaSR complexes at the periphery of the cytoplasm can be appreciated here somewhat better than in the top-bottom 2D image in panel (III).

CaSR’s Allosteric Agonists (Calcimimetics) and Antagonists (Calcilytics)

The allosteric or type II and agonists and antagonists bind the CaSR at sites different from the orthosteric one and sensitize the receptor to activation by the Ca²⁺ in opposite directions. The best characterized classes of type II CaSR agonists are aromatic L-amino acids and two highly selective and CaSR-specific synthetic polyalkylamines, i.e. NPS R-568 and Cinacalcet HCl, which were named “calcimimetics” because their effects mimic those elicited by extracellular Ca²⁺ [66,67]. Conversely, highly selective synthetic allosteric antagonists or “calcilytics”, i.e. NPS 89636 and NPS 2143, diminish the CaSR response to Ca²⁺ and other type I agonists [66,67]. Indeed, calcilytics, which increase PTH secretion, and calcimimetics, which decrease PTH secretion, were first developed by NPS Pharmaceuticals [66,67]. As revealed by the results of point mutation studies, these quite interesting pharmacological agents bind to partially overlapping but not identical sites of the CaSR, the calcilytics between TM3 and TM5, and both the calcimimetics and calcilytics between TM6 and TM7 [68,69]. Calcimimetics are right now used to treat primary or secondary hyperparathyroidism conditions and to rescue loss-of function CaSR mutants. Conversely, calcilytics were initially meant (but till now not clinically used) to treat osteoporosis and may mitigate the effects of gain-of-function CaSR mutants [66-73]. Finally, it is worth recalling that the CaSR exhibits several distinct conformational states, each of which is induced and stabilized by a different ligand, allosteric agonists and antagonists included, and is linked to a particular set of intracellular signaling pathways—a property defined “ligand-biased signaling” [74]. And, it is the specific cell type considered that determines the preferential activation of a particular set of CaSR signaling pathways by the same ligand [75].

HUMAN ASTROCYTES, CaSR, AND AD

In recent years, a growing body of novel evidences has led to realize that human cortical astrocytes substantially differ from rodent ones. The former are much bulkier, emit 10-fold more primary processes, exhibit novel morphological subtypes, e.g., the interlaminar one, control broader synaptic domains, and are capable of performing more complex and intense functional tasks, e.g., faster Ca²⁺ waves propagation, than the latter [22,24,76-83]. The new evolutionary features acquired by the astrocytes have significantly impacted on human brain physiology. This is indirectly confirmed by recent findings of Han et al. [84], who showed that human astrocytes engrafted on the brains of mice increase the learning abilities and activity-dependent plasticity of the animals. Obviously, astrocytes’ evolutionary changes have also impacted on human neurodegenerative diseases, AD included. Perhaps, this is the reason why animal AD models fail to fully reproduce the human disease and why pharmacological findings gained in animal AD models cannot be successfully translated to human clinical settings [85-89]. Various authors had previously shown that astrocytes are able to engulf and degrade exogenously accruing Aβs [90,91]. However, more recent findings prove that astrocytes can also significantly contribute to the Aβs overload of an AD brain. In fact, unstimulated (control) adult human cortical astrocytes exhibit a discrete level of activity of β-site APP-cleaving enzyme 1/β-secretase (BACE1/β-S) and γ-secretase (γ-S), and under chronic stress or during AD or after an exposure to exogenous Aβs both these enzymatic activities surge remarkably thereby leading to the de novo production of larger amounts of Aβs [5,21,92].

From a neuropathological standpoint, a diffuse astrogliosis is detected in both AD-model animals [92] and postmortem human AD brains [93]. Generally, the astrocytes become hypertrophied, keep their spatial domains, and overexpress S100β and GFAP proteins while partially losing their complement of glutamate metabolizing enzymes [80,83,92]. Notably, an activation of the human astrocytes can be detected via Positron Emission Tomography (PET) after administering the inflammation-revealing tracer (11)C-D-deprenyl to mild cognitive impairment (MCI) patients, being more intense in them than in patients at later AD stages and in healthy individuals [94]. Alterations in astrocyte signaling as revealed by intercellular Ca²⁺ waves and synchronous hyperactivity were reported to occur in transgenic AD-model animals [95]. In addition, Aβ-exposed astrocytes exhibited increases in their intracellular Ca²⁺ levels, yet at variance with neurons they did not die [5,96]. Hence, as a Ca²⁺ dyshomeostasis befalls in the activated astrocytes of AD brains (reviewed in [97-99]), what role(s) would play the CaSRs jutting from the plasma membranes of the astrocytes and of their client neurons?

The CaSR is expressed ubiquitously in the brain, though more intensely in some regions, e.g., the hippocampus, than in others [100]. In primary cultures of rat embryo brains, neurons and oligodendrocyte progenitor cells expressed CaSR’s mRNA more intensely than astrocytes did. In addition, CaSR expression tended to decline from postnatal to adult age likely in relation to oligodendrocyte and astrocyte differentiation [101]. Chattopadhyay et al. [102] were the first to show that a functional CaSR is expressed in cultured human embryo astrocytes, besides human astrocytoma and meningioma cells. On their own part, Dal Prà et al. [41] showed that NAHAs cultured from surgical left-overs of the cerebral cortex also express a functional CaSR, at lower levels while proliferating and at higher ones while mitotically quiescent, levels that were little affected by the actual Ca²⁺ concentration in the medium.

In tissues like the brain, where the cells are not involved in the maintenance of systemic Ca²⁺ homeostasis, the extracellular Ca²⁺ acts as the first messenger to regulate through the CaSR a variety of cellular functions; (i) during the developmental stages, in which controls proliferation, migration, and differentiation of oligodendrocytes and neurons; (ii) in the postnatal life, in which modulates neurotransmission and synaptic plasticity [51,82,103]; and (iii) in the course of diseases affecting the central nervous system [104].

Concerning AD, Ye et al. [54] showed that exogenous Aβs bind and activate the CaSR causing the opening of a Ca²⁺-permeable non-selective cation channel that elicited a sustained surge of intracellular Ca²⁺ in cultures of hippocampal pyramidal neurons from wild-type mice and rats with resultant neuronal dysfunction. By contrast, exogenous Aβs could not evoke these intracellular Ca²⁺ surgesin CaSR^-/- mice [54]. In addition, exogenous Aβs activated the same cationic channel in human embryo kidney cells overexpressing the CaSR (HEK293-CaSR), but failed to do so in wild-type HEK293 cells [54]. Moreover, using a sensitive luciferase-reporter gene assay, Conley et al. [55] demonstrated that in CaSR-transfected Cos1 cells exogenous Aβ_1-42 activated CaSR signaling in a dose-dependent fashion. Recently, by applying to GFAP-expressing NAHAs in cultures (Fig. 3A) the in situ Proximity Ligation Assay (isPLA) method, which reveals through a fluorescent sharp signal the very close (≤30 nm) and definite interaction between two proteins [105], we showed that soluble Aβ oligomers do specifically bind the CaSRs at the plasma membrane [5,6] (Fig. 3B) to be subsequently endocytosed (unpublished observations). That exogenous Aβs not only bind but also activate the human astrocytes’ CaSR is supported by our findings that (i) either soluble or fibrillar Aβ₂₅₋₃₅,an established Aβ₄₂ functional surrogate [5,106], stimulates the excess production, accrual, and release of endogenous Aβ₄₂ in the NAHAs and in human cortical postnatal HCN-1A neurons in parallel with the astrocytes’ surplus synthesis and secretion of NO and VEGF-A [5,38]; (ii) a highly selective calcimimetic, NPS R-568 also stimulates the synthesis, accrual, and release of surplus Aβ₄₂ besides of NO and VEGF-A from the NAHAs thereby mimicking the effects of exogenous Aβs [5,6,38]; (iii) conversely, in the presence of exogenous Aβs, calcilytic NPS 2143 effectively blocks the oversecretion of Aβ₄₂ by the cortical astrocytes and neurons, and of NO and VEGF-A by the NAHAs [5,6,38]; and (iv) last but not least, calcilytic NPS 2143 fully preserves the viability of the Aβs-exposed human cortical postnatal HCN-1A neurons, which would otherwise have progressively died [5]. Therefore, the extracellularly accruing Aβ₄₂ oligomers do bind and activate the CaSRs of both members of the astrocyte-neuron teams. Thus, they excite the further release and spreading of a set of neurotoxins, including Aβ₄₂, NO, and VEGF-A, which would advance AD progression. Importantly, the addition of a calcilytic agent like NPS 2143 fully prevents these noxious effects driven by the pathological Aβ●CaSR signaling [5].

Interestingly, an exposure of the NAHAs to exogenous Aβ oligomers also induced by 48 hours a significant albeit transient increase in astrocytes’ total CaSR proteins; under the same respect calcimimetic NPS R-568 was ineffective, calcilytic NPS 2143 given by itself caused an early, but transient decrease of the total CaSR protein, whereas in the presence of exogenous Aβs NPS 2143 elicited an intense and persistent fall of the total CaSR levels and likely of the Aβ●CaSR signaling intensity [5]. These are examples of alterations of the CaSR life cycle according to the bound agonist/antagonist that up- or downregulate CaSR availability and consequently its signaling intensity occurring in the NAHAs. Breitwieser [107] designated this recently identified mechanism as “agonist-driven insertional signaling (ADIS)” and our just mentioned findings support her suggestion that ADIS is likely to have therapeutic relevance. In addition, together with the lysosomes the proteasome helps reduce the availability of the total CaSR and the intracellular Aβ₄₂ accrual in the NAHAs since its 20S chymotrypsin-like activity is remarkably though transiently increased by calcilytic NPS 2143 in the presence of exogenous Aβs [5].

CONCLUSIONS

Our view that the CaSRs of human astrocyte-neuron teams play specific and relevant roles in the spreading and progression of the AD neuropathology is supported not only by their just mentioned ability to form specific complexes with Aβ oligomers [6], but also by the opposite effects of CaSR’s highly selective synthetic allosteric agonists or antagonists [5,6,38]. Therefore, the suggestion of using allosteric CaSR antagonists (calcilytics) like NPS 2143 or similarly acting agents to hinder or at least remarkably slow the otherwise inexorable progression of AD stems from the experimental results gained through the use of cultured NAHAs and human cortical HCN-1A neurons [5,6,38,41]. Given the destructive effects of the disease on cortical neurons, it would be advisable to try CaSR antagonists, just like any other anti-AD candidate drug, on MCI or early post-MCI cases to assess their real therapeutic potential.

AUTHORS’ CONTRIBUTIONS

All the authors contributed to the drafting of the manuscript and to the drawing of the figures. All the authors read and approved the final manuscript.

CONFLICT OF INTEREST

The author(s) confirm that this article content has no conflict of interest.

LIST OF ABBREVIATIONS

aa

 =  amino acid

Aβ₄₂

 =  amyloid-β_x-42

Aβs

 =  amyloid-β peptides

AD

 =  Alzheimer’s disease

ALS

 =  amyotrophic lateral sclerosis

CaSR

 =  Ca²⁺-sensing receptor

ECD

 =  extracellular domain

hp

 =  hyperphosphorylated

ICD

 =  intracellular domain

MCI

 =  mild cognitive impairment

MIP

 =  Maximum Intensity Projection

NAHA

 =  normal adult human astrocyte

NMDAR

 =  N-methyl-D-aspartate receptor

α-7-nAChR

 =  α-7-nicotinic acetylcholine receptor

NO

 =  nitric oxide

ROS

 =  reactive oxygen species

TMD

 =  transmembrane domain

VEGF

 =  vascular endothelial growth factor

VFT

 =  Venus fly-trap