Astroglial Calcium Signaling in Aging and Alzheimer's Disease

Abstract

Astrocytes are the homeostatic and protective cells of the central nervous system (CNS). In neurological diseases, astrocytes undergo complex changes, which are subclassified into (1) reactive astrogliosis, an evolutionary conserved defensive rearrangement of cellular phenotype aimed at neuroprotection; (2) pathological remodeling, when astrocytes acquire new features driving pathology; and (3) astrodegeneration, which is manifested by astroglial atrophy and loss of homeostatic functions. In aging brains as well as in the brains affected by Alzheimer's disease (AD), astrocytes acquire both atrophic and reactive phenotypes in a region- and disease-stage-dependent manner. Prevalence of atrophy overreactivity, observed in certain brain regions and in terminal stages of the disease, arguably facilitates the development of neurological deficits. Astrocytes exhibit ionic excitability mediated by changes in intracellular concentration of ions, most importantly of Ca²⁺ and Na⁺, with intracellular ion dynamics triggered by the activity of neural networks. AD astrocytes associated with senile plaques demonstrate Ca²⁺ hyperactivity in the form of aberrant Ca²⁺ oscillations and pathological long-range Ca²⁺ waves. Astroglial Ca²⁺ signaling originating from Ca²⁺ release from the endoplasmic reticulum is a key factor in initiating astrogliotic response; deficient Ca²⁺ signaling toolkits observed in entorhinal and prefrontal cortices of AD model animals may account for vulnerability of these regions to the pathology.

THE BRAIN AGING

The cellular mechanisms of aging in general, and of brain aging in particular, remain controversial; many hypotheses have been proposed, while no unifying theory has been agreed upon. Brain senescence is associated with a progressive loss of function with deficits in learning and memory, and yet, surprisingly, the healthy brain withstands the process of aging remarkably better than other organs and systems. Indeed, the physical fitness of youth cannot be matched even at middle age, when the cognitive capabilities are reaching their zenith. Indeed, brain cognitive capacity lasts well into advanced age. This bright scenario of extended cognitive capacity, however, is not universal because aging is the chief risk factor for the development of neurodegenerative pathologies that invariably end in senile dementia characterized by severe cognitive impairments. Thus, at least from the presentation point of view, the brain aging appears in two guises: the physiological (with preserved cognition) and pathological (with cognitive deficits).

What determines the fateful divergence between preserved cognition and decline? There are obviously the environmental factors (such as lifestyle, diet, education or physical activity), the associated pathology (with vascular disorders and ischemic lesions being the most prominent factors [O'Brien and Thomas 2015]), but also the ability of the nerve tissue to withstand the environmental stresses as well as the ability of the brain to compensate for age-dependent deterioration. These intrinsic properties of the brain are generally described as a cognitive reserve that applies to all neurological diseases and defines the extent of neurological deficit when similar brain damage produces very different cognitive outcomes in different subjects (Stern 2009; Zorec et al. 2018). The cognitive reserve reflects previous experiences as well as intrinsic defensive capacity of the nervous system and is determined by (1) neuronal reserve represented by neural networks predating pathology, and (2) neural compensation, defined by an overall defensive and regenerative capacity of the nervous tissue (Stern 2009). Both components rely on the performance of neuroglial cells that provide neuroprotection, limit the lesioned area, and which are indispensable for regeneration and remodeling of neuronal networks. Arguably, the neuroglia determine the brain senescence, and neuroglial failure to protect and sustain the neuronal networks, the nerve tissue, and the central nervous system (CNS) as an organ is an important factor in the progression from physiological to pathological brain aging (Verkhratsky et al. 2015a).

HOMEOSTATIC ASTROGLIA PROVIDES THE BRAIN WITH LIFELONG SUPPORT

The neuroglia of the CNS are represented by microglia (cells of mesodermal origin responsible for nervous tissue defense and immunity [Kettenmann et al. 2011]) and macroglia (cells of ectodermal neuroepithelial origin) that are further divided into astroglia, oligodendroglia, and oligodendroglial precursors also known as NG2 glia (Kettenmann and Ransom 2013; Verkhratsky and Butt 2013). Astrocytes are the main homeostatic cells of the CNS; they are quite heterogeneous in their form and function. Astrocytes are classified into subtypes, which include protoplasmic astrocytes of the gray matter, fibrous astrocytes of the white matter, radial astrocytes (cerebellar Bergmann glia, retinal Müller glia, radial glia-like neural stem cells of the neurogenic niches and tanycytes of the hypothalamus, hypophysis, and the raphe part of the spinal cord), velate astrocytes of cerebellum and olfactory bulb, pituicytes of the neurohypophysis, Gomori astrocytes, surface-associated astrocytes, perivascular and marginal astrocytes and ependymocytes, choroid plexus cells, and retinal pigment epithelial cells (Verkhratsky and Nedergaard 2018). In addition, the brains of higher primates contain interlaminar astrocytes and polarized or varicose projection astrocytes (Oberheim et al. 2012; Colombo 2018; Verkhratsky et al. 2018a).

Despite this prominent regional heterogeneity, the common feature of astroglia is their function, which lies in the maintenance of CNS homeostasis at all levels of organization from molecular to the whole organ; notably astroglial homeostatic functions are linked to neuronal activity (Verkhratsky and Nedergaard 2018). Astrocytes are key regulators of molecular homeostasis of the nervous tissue; they transport ions thus supporting neuronal excitability, they accumulate and catabolize neurotransmitters and supply neurones with neurotransmitter precursors thereby sustaining synaptic transmission, they support neuronal metabolism and detoxify ammonium and reactive oxygen species, and they secrete numerous trophic factors, neuroactive agents, and even hormones (see Parpura et al. 2012; Pellerin and Magistretti 2012; Hertz 2013; Rose et al. 2013; Hertz et al. 2015; Verkhratsky and Nedergaard 2016; Verkhratsky et al. 2015b, 2016a,b for literature and details). At the network level, astrocytes contribute to cell renewal through adult neurogenesis (the stem cells of neurogenic niches being essentially radial astroglia) and, through astroglial cradle, support synaptogenesis, synaptic maturation, and extinction (Verkhratsky and Nedergaard 2014; Kempermann 2015; Falk and Götz 2017). Astroglial cells support the cytoarchitecture of nerve tissue; through the process of tiling they parcellate the gray matter into neuro-glio-vascular units, while astroglial end-feet contribute to functional hyperemia (Bushong et al. 2002; MacVicar and Newman 2015). Astrocytes constitute the glia limitans that form the pial coverage of the CNS; astrocytes line up ventricles, produce and move cerebrospinal fluids, and provide a foundation for the glymphatic system (Iliff and Nedergaard 2013; Nedergaard 2013; Rangroo Thrane et al. 2013). Astrocytes as chemosensors participate in systemic homeostatic responses (Angelova et al. 2015; Turovsky et al. 2016) and contribute to the regulation of feeding behavior and energy balance (García-Cáceres et al. 2016). Through these multiple functions, astroglia support all aspects of CNS operation, including its development, experience-dependent adaptation, and aging.

IONIC SIGNALING AS A SUBSTRATE OF ASTROGLIAL EXCITABILITY

Astrocytes, similarly to all other neuroglial cells, are unable to generate action potentials (and in their physiological context, astrocytes probably cannot even be substantially depolarized [Huang et al. 2018]), and hence are defined as electrically nonexcitable cells. Nonetheless, astroglial cells utilize an alternative mechanism for their excitability, which allows them to respond to neuronal activity. This mechanism relies upon spatially and temporally controlled fluctuations of intracellular concentration of ions, most notably of Ca²⁺ and Na⁺ (Kirischuk et al. 2012; Volterra et al. 2014; Rusakov 2015; Rose and Verkhratsky 2016; Verkhratsky et al. 2018b).

Dynamic changes in astroglial cytosolic ionized Ca²⁺ ([Ca²⁺]_i) in response to chemical or mechanical stimulation were discovered at the end of the 1980s (Enkvist et al. 1989; Charles et al. 1991; Dave et al. 1991; Verkhratsky and Kettenmann 1996) very soon after functional neurotransmitter receptors had been identified in cultured neuroglial cells (Bowman and Kimelberg 1984; Kettenmann et al. 1984). Almost simultaneously, the ability of Ca²⁺ signals to travel through astroglial syncytium in vitro was described (Cornell-Bell et al. 1990; Cornell-Bell and Finkbeiner 1991), and the concept of astroglial Ca²⁺ excitability was soon formalized (Deitmer et al. 1998; Verkhratsky et al. 1998). Astrocytes in vitro, in situ, and in vivo have been found to mount Ca²⁺ signals in response to physiological stimulation (Bekar et al. 2008; Ding et al. 2013; Semyanov 2019); these signals have complex spatiotemporal organization. At the cellular level, astroglial Ca²⁺ signals are represented by local microdomains, intracellular propagating waves, Ca²⁺ oscillations, and global Ca²⁺ signals (Grosche et al. 1999; Nimmerjahn et al. 2009; Shigetomi et al. 2010; Srinivasan et al. 2015). The molecular mechanisms responsible for such hierarchical organizations are distinct. The global signals and propagating waves rely almost solely on Ca²⁺ release from the endoplasmic reticulum (ER) Ca²⁺ store (mostly, but not exclusively, mediated by InsP₃ receptor type 2 [InsP₃R2]). The local Ca²⁺ microdomains, however, often result from Ca²⁺ entry through ionotropic receptors, transient receptor potential channels, store-operated Ca²⁺ entry (SOCE) or reversed Na⁺/Ca²⁺ exchanger (see Verkhratsky and Nedergaard 2018 for a detailed discussion). The targets of Ca²⁺ signals (the “Ca²⁺ sensors”) include regulators of metabolism, gene expression, possibly exocytosis, and astroglial reactivity.

ASTROGLIA IN AGING

It is now universally acknowledged that physiological aging is not associated with any significant change in the number of neurones (West 1993; Pakkenberg and Gundersen 1997; Verkhratsky and Toescu 1998; Toescu et al. 2004; Dickstein et al. 2006; von Bartheld et al. 2016). At the same time, oligodendrocytes do suffer age-dependent decline, with ∼11% reduction in the volume of white matter (Haug and Eggers 1991) and ∼30% decrease in the number of oligodendroglial cells (Pelvig et al. 2008; Fabricius et al. 2013). Similarly, human aging is associated with degeneration of microglia (Tremblay et al. 2012; Streit and Xue 2013). The number of astrocytes seems not to be affected in physiological aging (Pelvig et al. 2008; Fabricius et al. 2013). Morphological changes of astroglia associated with aging are multifaceted and most likely brain-region dependent. In old animals, both astroglial hypertrophy (manifested by an increase in the sickness of astroglial processes and indicative of astrogliosis) and atrophy have been observed (Diniz et al. 2010; Cerbai et al. 2012; Grosche et al. 2013). The morphometric analysis of profiles of astrocytes labeled with antibodies against either glial fibrillary acidic protein (GFAP), protein S100B, or glutamine synthetase (GS) revealed complex and region-specific changes in the astroglial appearance (Rodríguez et al. 2014) with signs of astroglial reactivity in the hippocampus and astroglial atrophy in entorhinal and prefrontal cortices. The GFAP-positive astrocytes in the entorhinal cortex, for example, had fewer main processes with a very low number of secondary branches. The GS-positive and GFAP-negative astrocytes in the hippocampus also showed signs of atrophy. GS is a central enzyme of the glutamate/GABA-glutamine shuttle and is fundamental for detoxification of ammonium (Rose et al. 2013). An age-dependent decrease in GS expression may hint at some abnormalities of both glutamatergic and GABAergic transmission.

Little is known about the physiology of aging astrocytes. The general electrophysiological properties (hyperpolarized resting membrane potential, input resistance, and overall predominance of K⁺ permeability) do not change with age (Lalo et al. 2011). Similarly, astrocytes retain their neurotransmitter receptors (Lalo et al. 2011; Gómez-Gonzalo et al. 2017). The functional expression of major receptors, however, changes with age: the density of ionotropic glutamate and purinergic receptors (as well as the density of glutamate-transporter currents) increased in mice between 1 month and 3–6 months of age and then rapidly declined (Lalo et al. 2011). Likewise, the amplitude of neurotransmitter-induced Ca²⁺ signals decreased with age (Palygin et al. 2010). This decrease of Ca²⁺ signals, in turn, led to an age-dependent decline in astroglial secretion of ATP and aberrant astroglial regulation of metaplasticity (Lalo et al. 2014, 2018). Aging was found to decrease the density of astroglial aquaporin 4 and hence impair glymphatic clearance (Kress et al. 2014), which may exacerbate retention of toxic misfolded proteins (such as, for example, amyloid-β) in the senescent brain.

ASTROGLIOPATHOLOGY IN ALZHEIMER'S DISEASE

Pathological changes in astroglia (astrogliopathies) are complex and context dependent. Conceptually, astrogliopathology can be classified into (1) astroglial reactive response, and (2) astrocytopathies, which can be further subdivided into (2a) astrodegeneration with astroglial atrophy and loss of function and (2b) pathological remodeling of astroglial cells (Verkhratsky et al. 2013, 2016b, 2017b; Pekny et al. 2016; Verkhratsky and Parpura 2016; Ferrer 2017, 2018). Reactive astrogliosis is characterized by biochemical, functional, and morphological reshaping of astrocytes essentially aimed at neuroprotection and restriction of brain damage. Astroglial reactivity is generally associated with modification of gene expression profile, increased expression of intermediate filaments (GFAP and vimentin) and morphological hypertrophy. Astroglial reactivity produces multiple pathological phenotypes, which are context- and disease-specific, and may exert either neuroprotective or neurotoxic effects (Zamanian et al. 2012; Burda and Sofroniew 2014; Pekny and Pekna 2014; Sofroniew 2014a,b, 2015; Pekny et al. 2016; Liddelow and Barres 2017). Astrodegeneration with loss of function has been discovered rather recently (Olabarria et al. 2010), and is detected in a wide spectrum of diseases including neurotoxic damage, neuropsychiatric disorders, and neurodegenerative pathologies (Rajkowska and Stockmeier 2013; Verkhratsky et al. 2014; Scofield et al. 2016; Plata et al. 2018; Rajkowska et al. 2018). Finally, in pathological remodeling, astrocytes acquire new qualities that drive pathological progression of the disease, with Alexander disease, megalencephalic leukoencephalopathy with subcortical cysts, and vanishing white matter syndrome being examples (Messing et al. 2012; Lanciotti et al. 2013).

In Alzheimer's disease (AD), both astroglial atrophy and astroglial reactivity are featured, being present either alone or in combination with regional and disease-stage-dependent differences (Verkhratsky et al. 2010, 2015a, 2016c; Rodríguez-Arellano et al. 2016). Astroglial reactivity (imaged by PET or analyzed in postmortem tissues) is observed in AD patients during the initial phases of the disease from mild cognitive impairment to early dementia (i.e., Braak stages I to IV). Astroglial reactivity declines at the more advanced AD stages associated with profound dementia (i.e., Braak V and VI [Verkhratsky et al. 2015a; Rodríguez-Vieitez et al. 2016]). In animal models of AD, widespread atrophy of astrocytes has been detected in several brain regions, with entorhinal and prefrontal cortices showing the earliest signs of decrease in astroglial morphological profiles (Rodríguez et al. 2009; Yeh et al. 2011; Kulijewicz-Nawrot et al. 2012). Astroglial degeneration was also confirmed in cells differentiated from pluripotent stem cells obtained from patients with familial and sporadic AD. These astrocytes exhibited atrophic morphology and aberrant distribution of major homoeostatic molecules such as glutamate transporters, S100Β protein, and GS (Jones et al. 2017), and had a reduced ability to support neurones in culture (Oksanen et al. 2017; see also Mohamet et al. 2018 for further details). In addition, atrophic astrocytes were unable to mount an astrogliotic response to β-amyloid, which correlated with higher vulnerability of entorhinal and prefrontal cortical regions to AD-like pathology (Yeh et al. 2011; Kulijewicz-Nawrot et al. 2012; Verkhratsky et al. 2015a).

ASTROGLIAL CALCIUM SIGNALING IN AD

The “calcium hypothesis of aging” was proposed by Khachaturian (1987) who based his theory on experimental studies of Philip Landfield (Landfield 1987; Landfield and Pitler 1984). The theory postulated that aging neurones experience increased Ca²⁺ influx during depolarization, which elevates resting Ca²⁺ concentration ([Ca²⁺]_i), thus triggering excitotoxicity. If anything, physiological aging of neurones is associated with more subtle alterations of neuronal Ca²⁺ extrusion, which, although capable of handling normal Ca²⁺ loads, fail to effectively curtail excessive Ca²⁺ influx. This in turn results in higher vulnerability of old nerve cells to the periods of high activity (Verkhratsky and Toescu 1998; Toescu et al. 2004; Toescu and Verkhratsky 2007). In contrast, neuronal Ca²⁺ homeostasis and Ca²⁺ signaling are compromised in the AD, which is often regarded as a “chronic calciumopathy” (Stutzmann 2007; Stutzmann and Mattson 2011). Almost nothing is known about changes in Ca²⁺ homeostatic machinery, resting Ca²⁺ handling, and Ca²⁺ signaling in aged astrocytes. There are several sporadic reports demonstrating a decrease in evoked astrocytic Ca²⁺ signals in mice aged 16–21 months, when compared to adult animals (Lalo et al. 2011, 2014, 2018). Analysis of Ca²⁺ signaling in astrocytes from older animals is yet to be performed.

EXPOSURE TO β-AMYLOID DISTURBS ASTROGLIAL Ca²⁺ DYNAMICS

Accumulation of β-amyloid in both soluble and oligomerized forms is generally considered to be one of the main pathogenetic mechanisms of AD (Korczyn 2008; Selkoe and Hardy 2016), although the β-amyloid hypothesis is often criticized (Hardy 2009; Castellani et al. 2010; Castellani and Smith 2011). Numerous in vitro experiments (albeit often employing high β-amyloid concentrations) have documented acute disturbance of Ca²⁺ homeostasis and Ca²⁺ signaling in the presence of β-amyloid. Exposure to β-amyloid increased resting [Ca²⁺]_i in cultured astrocytes: treatment with β-amyloid (in concentrations between 100 nm and 5 µm) for several hours increased resting [Ca²⁺]_i two- to threefold over initial levels (Haughey and Mattson 2003; Lim et al. 2013). These data are not universally confirmed; several studies found that incubations of astroglial cultures with 100–200 nm of β-amyloid (or its toxic fragment β-amyloid_25-35) for 48–72 h did not alter resting [Ca²⁺]_i (Casley et al. 2009; Toivari et al. 2011).

Acute exposure to β-amyloid was reported to trigger changes in astroglial [Ca²⁺]_i. There are indications that β-amyloid evoked astroglial Ca²⁺ oscillations or [Ca²⁺]_i transients in cultured astrocytes and in astrocytes in organotypic slices (Jalonen et al. 1997; Abramov et al. 2003, 2004; Chow et al. 2010; Alberdi et al. 2013; Lim et al. 2014), although, again, several studies have not noticed such acute effects (Casley et al. 2009; Toivari et al. 2011; Lim et al. 2013). Addition of 1 µm of β-amyloid_1-40 to cultured rat cortical astrocytes evoked [Ca²⁺]_i elevations in 17% of the cells; application of β-amyloid_25-35 triggered Ca²⁺ signals in 36% of astrocytes (Jalonen et al. 1997). In primary cultured rat newborn astrocytes, acute application of 1 µm of β-amyloid_25-35 induced [Ca²⁺]_i transients in 27% of cells, whereas at higher concentrations (2–5 µm) ∼60% of astrocytes mounted [Ca²⁺]_i responses (Stix and Reiser 1998). At very low concentrations (200–300 pm), β-amyloid was also reported to activate astroglial α7 nicotinic cholinoceptors, which produced Ca²⁺ influx and [Ca²⁺]_i transients (Pirttimaki et al. 2013; Lee et al. 2014).

ABERRANT Ca²⁺ SIGNALING IN ASTROCYTES IN VITRO

Studies on astrocytes isolated from mouse models of familial AD similarly revealed some aberrant forms of Ca²⁺ signaling (Lim et al. 2014, 2016). Abnormally large Ca²⁺ signals have been detected in astrocytes isolated from newborn triple transgenic AD mice. Aberrant Ca²⁺ signaling in these 3xTG mice (harboring mutant genes for presenelin 1 [PS1], amyloid precursor protein [APP], and Tau, with astrocytes expressing solely the PS1 mutant gene), indicating intrinsic shortcomings in Ca²⁺ handling in cells expressing AD-associated mutant genes (Ronco et al. 2014). In particular, these astrocytes had increased SOCE. Astrocytes cultured from 3xTg-AD animals also demonstrated aberrant kinetics of ATP-induced Ca²⁺ signals (Stenovec et al. 2016). It seems that mutant PS1 expressed in the astroglial ER can be responsible for these alterations in Ca²⁺ handling and metabotropic Ca²⁺ signals.

In another mouse model of the AD, which overexpressed APP (the Tg5469-AD mouse), SOCE was not affected. However, deletion of APP resulted in SOCE inhibition (Linde et al. 2011), and the authors suggested that this inhibition may reflect down-regulated expression of TRPC1 or Orai 1 channels. Down syndrome shares some common features with AD, and in particular the abnormal amyloid processing associated with an extra allele of the APP gene. In astrocytes from trisomy 16 mice, a widely used model for Down syndrome, the resting [Ca²⁺]_i was found to be substantially elevated, and higher ER Ca²⁺ content was reported (Bambrick et al. 1997).

ABERRANT Ca²⁺ SIGNALING IN ASTROCYTES IN VIVO

Astrocytes in the brains of AD animal models generally demonstrate aberrant, hyperactive [Ca²⁺]_i dynamics, which is fundamentally similar to neuronal hyperexcitability routinely observed in AD-like experimental pathology (Zott et al. 2018). In particular, such aberrant hyperactivity is observed in reactive astrocytes associated with senile plaques. Plaque-associated astrocytes in the brain of APP/PS1 mice had increased (almost doubled compared to healthy cells) levels of resting [Ca²⁺]_i, and generated pathological [Ca²⁺]_i oscillations and long-projecting propagating Ca²⁺ waves (Kuchibhotla et al. 2009). The appearance of this [Ca²⁺]_i hyperactivity is claimed to be associated with abnormal purinergic signaling in reactive astroglial cells. It seems that reactive astrocytes release excessive amounts of ATP through connexin hemichannels. This ATP, in turn, acts in an autocrine fashion on astroglial P2Y purinoceptors, which mediate abnormal [Ca²⁺]_i dynamics (Delekate et al. 2014). An increased frequency of astroglial Ca²⁺ oscillations was also noted in AD animals in the pre-plaque stage, and these abnormal [Ca²⁺]_i dynamics coincided with the instability of vascular tone probably indicating that astrocytes in their ability to regulate local blood flow (Takano et al. 2007).

AD PATHOLOGY RECONFIGURES THE ASTROGLIAL Ca²⁺ SIGNALING TOOLKIT

Besides acute effects on astroglial [Ca²⁺]_i dynamics, the AD pathology (in model animals) as well as exposure to β-amyloid remodels the astrocytic Ca²⁺ signaling toolkit by modifying expression of molecules responsible for Ca²⁺ homeostasis and Ca²⁺ signaling. For example, changes in expression level have been reported for ionotropic and metabotropic receptors, intracellular Ca²⁺ channels, store-operated Ca²⁺ channels, and Ca²⁺-sensors (Lim et al. 2014, 2016; Verkhratsky et al. 2017a).

Numerous experiments in vitro have demonstrated that chronic incubation with β-amyloid changes expression of neurotransmitter receptors. For example, 48–72 h exposure of astroglial cultures to 10–30 μm β-amyloid_1-40 selectively increased the amplitude of [Ca²⁺]_i transients in response to activation of mGluR5. This increase reflected up-regulation of mGluR5 expression, which was detected at both messenger RNA (mRNA) and protein levels (Casley et al. 2009). Incubation of cultured astroglial cells with 100 nm of oligomeric β-amyloid for 1 h instigated clustering and diffusional trapping of mGluR5, which stimulated the receptor and triggered secretion of ATP (Wevers and Schröder 1999). Chronic (24–72 h) exposure to oligomeric β-amyloid (100 nm–20 µm) up-regulated expression of astroglial mGluR5 (Grolla et al. 2013b; Lim et al. 2013). This up-regulation of mGluR5 expression was suppressed by the inhibitors of calcineurin and NF-κB (Lim et al. 2013). The same up-regulation of mGluR5 was found in astrocytes in the animal AD models and in postmortem human tissues. Increased presence of mGluR5 was detected in cortical reactive astrocytes associated with senile plaques in the brains of 6- to 16-month-old APPswe/PS1dE9 mice (Wevers and Schröder 1999). High levels of mGluR5 protein was found in the postmortem hippocampal samples of AD patients at advanced (Braak V–VI) stages of the disease (Casley et al. 2009; Lim et al. 2013). Chronic (24–72 h) treatment with nanomolar (0.1–100 nm) concentrations of β-amyloid_1-42 up-regulated the expression of several subunits of nicotinic cholinoceptors, including α7nAChR, α4nAChR, and β2nAChR (Xiu et al. 2005). Incidentally, increased presence of α7nAChR was detected in the postmortem brain tissue of patients with sporadic AD and familial AD associated with the Swedish APP (KM670/671NL) mutation (Yu et al. 2005).

The cellular remodeling caused by AD pathology also affects expression of astroglial Ca²⁺ release channels. The up-regulation of InsP₃R1 was detected in rat hippocampal astrocytes after 72 h incubation with 125 nm of Tat-ProADAM10_709-729 peptide, which promotes production of β-amyloid_1-40 and β-amyloid_1-42 by severing interactions between ADAM10 with SAP97 (Grolla et al. 2013a). Exposure of rat hippocampal cultured astrocytes to 100 nm oligomeric β-amyloid_1-42 led to an increase of expression of InsP₃R1 and InsP₃R2 mRNA (Lim et al. 2013). Changes in the astroglial [Ca²⁺]_i signaling toolkit differs between brain regions. For example, incubation with β-amyloid increased expression of InsP₃R1 in cells isolated from the hippocampus, but not from the entorhinal cortex, of healthy mice (Grolla et al. 2013b). However, in hippocampal astrocytes isolated from 3xTG-AD animals, incubation with β-amyloid did not affect expression of InsP₃R1, suggesting that exogenous β-amyloid and overexpression of mutated AD-related genes share common molecular pathways associated with deregulation of Ca²⁺ homeostasis. These in vitro data were not corroborated by the postmortem studies, which demonstrated overall decrease in the expression of InsP₃Rs throughout the brain, including the frontal and parietal cortex, hippocampus, and entorhinal cortex (Young et al. 1988; Haug et al. 1996; Kurumatani et al. 1998). In these experiments, however, the total brain tissue was analyzed without discriminating between neurones and nonneuronal cells. The AD pathology affects many other components of signaling systems associated with Ca²⁺ signaling, including calpain-10 (Garwood et al. 2013), NFAT (Abdul et al. 2009), NF-κB (Grolla et al. 2013a), calcineurin (Norris et al. 2005; Grolla et al. 2013a), L-type calcium channels (Daschil et al. 2014), and store-operated Ca²⁺ channels (Ronco et al. 2014). Analyses of the transcriptome of astrocytes microdissected from the brain samples of patients with different Braak stage of AD have found changes in expression of 32 genes associated with Ca²⁺ signaling. It appeared that expression of several isoforms of CaMKII, two isoforms of calmodulin, plasma membrane Ca²⁺-ATPases, ryanodine receptors and InsP₃Rs, was decreased at an advanced (Braak V–VI) stage when compared with an early (Braak I–II) stage of the disease (Simpson et al. 2011).

ENDOPLASMIC RETICULUM Ca²⁺ RELEASE AND ASTROGLIOTIC RESPONSE

The astrogliotic component is a prominent part of AD histopathology and pathogenesis. Hypertrophic, highly GFAP-expressing astrocytes associate with, and surround, senile plaques in human tissue and in the brains of AD animal models, arguably forming a defensive barrier protecting neural networks (Heneka et al. 2010; Verkhratsky et al. 2010). Inhibition of astroglial reactivity (for example, following genetic deletion of GFAP and vimentin) exacerbates β-amyloid load and facilitates plaques dissemination (Kraft et al. 2013). The capacity of astroglia to mount the reactive response in the context of AD differs between brain regions.

Formation of senile plaques (and probably accumulation of soluble β-amyloid species in the extracellular space) triggers prominent reactive astrogliosis in the hippocampus. In contrast, in the entorhinal and prefrontal cortices, accumulation of β-amyloid in the form of either plaques or perivascular depositions does not induce astrogliotic remodeling; astrocytes do not associate themselves with β-amyloid depositions, do not demonstrate hypertrophic morphology and do not up-regulate GFAP (Yeh et al. 2011; Kulijewicz-Nawrot et al. 2012; Rodríguez et al. 2016). Molecular mechanisms of such diversity may be associated with deficient Ca²⁺ signaling or pathological remodeling of Ca²⁺ signaling toolkits.

It is generally accepted that exposure of astroglia to β-amyloid in vitro or in situ instigates reactive astrogliosis (Verkhratsky et al. 2010; Alberdi et al. 2013). At the same time, β-amyloid often evokes astroglial Ca²⁺ transients or oscillations as discussed above. These β-amyloid-induced Ca²⁺ signals are mainly associated with Ca²⁺ release from the ER Ca²⁺ store and these Ca²⁺ signals are directly linked to the astrogliotic response. Pharmacological inhibition of Ca²⁺ release from the ER effectively blocked astrogliosis induced by β-amyloid in both cultured astrocytes and in astroglia in organotypic slices (Alberdi et al. 2013). These data agree with the previous findings of the critical role of InsP₃Rs and ER calcium release in the initiation of astrogliosis. Genetic deletion of InsP₃R2 effectively suppressed astroglial reactivity (Kanemaru et al. 2013). Exposure of astrocytes from different regions of the brain to β-amyloid had a distinct effect on tissue-specific Ca²⁺ signaling toolkits. In hippocampal astroglia, β-amyloid up-regulated expression of molecules associated with ER Ca²⁺ signaling. Conversely, in the entorhinal cortex, no such up-regulation has been observed (Grolla et al. 2013b). This deficiency may explain the absence of astrogliotic defensive responses in astrocytes from cortical regions, which renders these parts of the brain vulnerable to the AD pathology (Verkhratsky et al. 2015a).

CONCLUDING REMARKS

Astroglial Ca²⁺ signaling underlies astroglial excitability (together with other forms of ionic signaling). The astroglial Ca²⁺ signaling toolkits undergo modification in neurodegenerative pathologies including AD. Pathological remodeling of Ca²⁺ signaling contributes to astroglial pathology associated with deficient neuroprotection (suppressed astrogliosis) or failure in glial homeostatic support (astroglial asthenia and paralysis). Predominance of either process may affect cognitive reserve and thus balance the evolution of the disease and the degree of neurological deficit.