Early evidence for astrocyte calcium signaling came from studies that demonstrated glutamate-triggered intracellular calcium wave oscillation and propagation in cultured astrocytes [1,2] and rat hippocampal slices [3–8]. Shortly after the initial discovery of glutamate-induced calcium waves in astrocytes [2], many other chemicals were found to produce the same effect including noradrenaline [9,10], histamine [11], acetylcholine [5,11,12], ATP [13–15], GABA [16], endocannabinoids [17], nitric oxide (NO) [18,19], and brain-derived neurotrophic factor (BDNF) [20]. Stimulation of mossy fibers originating in the dentate gyrus (DG) of rat organotypic hippocampal slice cultures led to delayed calcium waves in astrocytes within CA3 [21]. Long-lasting calcium oscillations were also seen after neuronal stimulation in both hippocampal and visual cortex transverse brain slices [22]. Propagating calcium elevations within astrocytes were induced by uncaging astrocytic IP₃ [3] or mechanical stimulation of a single cell [1,23] and these waves were spatially restricted to functionally independent microdomains of astrocytes [24]. Nearby neurons also responded with increases in cytosolic calcium levels [22,23] which was blocked with the gap junction inhibitor octanol [23]. This suggests that the astrocyte-neuron communication occurred through intracellular connections rather than synaptic connections.

Parri et al. [25] discovered that astrocytes in rat ventrobasal (VB) thalamus slices displayed intrinsic intracellular calcium oscillations independent of action potential-evoked transmitter release and neuronal firing. Di Castro et al. [26] studied endogenous calcium activity in adult mouse hippocampal slice preparations and observed both focal (confined to a specific subregion) and expanded (spread to 9 or more contiguous subregions) events in astrocyte processes. In contrast to results presented by Parri et al. [25] in the VB thalamus, Di Castro et al. [26] found that while the focal events were dependent on synaptic neurotransmitter release, expanded events depended on nearby, individual action potentials. Calcium activity was abolished with an IP₃ receptor blocker and overcome with subsequent bath application of thapsigargin (which empties intracellular calcium stores). Blocking astrocytic calcium activity decreased basal transmitter release at local synapses, suggesting that calcium elevations participate in local fine tuning of basal neurotransmission at excitatory synapses [26]. IP₃ receptor subtype 2 (IP₃R2) is the primary IP₃ receptor expressed by astrocytes. IP₃R2 knockout mice exhibited reduced calcium events [26], suggesting that activation of the IP₃-sensitive calcium channel may be a major contributor to intracellular calcium release in astrocytes.

Gq-linked protein coupled receptor (GPCR) activation of phospholipase C leads to the release of IP₃, activation of the IP₃ receptor, and subsequent release of calcium from internal stores. Calcium imaging techniques were used to demonstrate that astrocytes in vitro and in situ respond to neuronal activity with GPCR-mediated intracellular calcium increases [5,6,12,16,22,26,32,33,48,49]. Specifically, neuronal stimulation led to activation of astrocytic mGluRs [6,17,22,40,49–51] and involved mGluR subtype 5 (mGluR5) [4]. It is likely that the astrocytic mGluRs are activated by glutamate released from neurons [6]. Stimulation of the somatosensory cortex led to increased astrocytic calcium, a response that was mediated by synaptic glutamate release and activation of mGluRs [45,46]. In response, astrocytes are capable of increasing basal synaptic transmission [4]. Furthermore, NO may play a role in astrocytic calcium rises through the NO-G-kinase signaling pathway [52,53]. Conversely, the polyunsaturated fatty acids, arachidonic acid and docosahexaenoic acid, but not eicosapentaenoic acid, blocked intracellular calcium oscillations, inhibited store-operated calcium entry, and reduced the amplitudes of GPCR-calcium responses [54]. The metabolic gliotoxin fluorocitrate suppressed astrocytic intracellular calcium signaling and calcium-dependent glutamate release [50].

Astrocytic intracellular calcium increases can modulate inhibitory synaptic transmission. Photostimulation of astrocytes in the primary visual cortex increased the spontaneous firing of inhibitory neurons in an mGluR-dependent manner [55]. In hippocampal slices, uncaging astrocytic calcium increased the frequency of kainate receptor-depended spontaneous inhibitory postsynaptic potentials (sIPSCs) in nearby interneurons [56] but decreased the amplitude of evoked IPSCs and frequency of miniature IPSCs in an mGluR-dependent manner [57].

Parpura et al. and others applied bradykinin to cultured astrocytes to stimulate internal calcium elevations, which resulted in astrocytic glutamate and aspartate release into the media [58–60]. Bradykinin also led to NMDA receptor-mediated neuronal calcium elevations in neuron-astrocyte cocultures, but not in neurons cultured alone [60]. Photolysis of caged astrocytic calcium increased astrocyte membrane capacitance [61] and potentiated transmitter release at synapses [49]. Innocenti et al. [62] used an enzyme-linked assay system involving glutamate dehydrogenase, an enzyme that metabolizes glutamate and consequently reduces NAD⁺ to NADH, to measure extracellular glutamate levels in astrocyte cultures. Stimulation of intracellular calcium produced increased levels of NADH, corresponding to glutamate release. Similarly, mGluR agonist trans-aminocyclopentane-trans-1,3-dicarboxcylic acid (trans-ACPD)-induced release of glutamate evoked increases in extracellular NADH and, in most cells, an increase in whole cell capacitance [63]. Propagation of calcium waves occurred at a rate of 10–30 μm/s and glutamate release underlying NADH detection propagated at an average speed of 26 μm/s [64]. Depletion of internal calcium stores reduced the accumulation of NADH [62]. Other gliotransmitters may be released in a similar mGluR- and calcium-dependent manner including D-serine [65], ATP [66–68], and, in the rat olfactory bulb, GABA [69].

The release of gliotransmitters may play a role in plasticity. D-serine released from astrocytes acts as a coagonist for NMDA receptors; therefore, the degree of astrocytic coverage of neurons may govern the level of activation [70]. Clamping internal astrocytic calcium, depletion of D-serine, or disruption of D-serine release from astrocytes blocked long-term potentiation (LTP) in hippocampal slices from rats [71]. This could be reversed by endogenous application of D-serine or glycine and subsequent binding to the NMDA receptor coagonist sites [71]. An mGluR-mediated rise in intracellular calcium and subsequent ATP release multiplicatively scaled glutamate synapses [68]. This effect occurred quickly but was long lasting. Whisker stimulation induced delayed onset local field potentials (LFPs) in the barrel cortex whereas the combination of whisker stimulation with electrical stimulation of the nucleus basalis of Meynert (NBM) induced surges of intracellular calcium levels, elevated extracellular D-serine levels, and LTP in the cortex [72]. These effects were lost in the IP₃R2 knockout mice [72].

Astrocytes possess the machinery necessary for exocytosis of gliotransmitters, including the protein machinery necessary to form the soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complex [63,73–75]. ATPase, cellubrevin, ras-related protein rab3a, secretory carrier membrane protein, synapsin I, synaptic vesicle glycoprotein 2 A (SV2), synaptotagmin I, synaptobrevin II, synaptophysin, vesicular glutamate transporter 1 (VGLUT1), and VGLUT2 were all expressed in cultured astrocytes [58,63,76–82]. Astrocytic VGLUT3 expression was reported in slices from rats [83]. VGLUT and cellubrevin in astrocytic vesicles were found in close proximity to NMDA receptor-containing neuronal membranes [77]. Glutamate [76] and D-serine [65,84] colocalize with many of these exocytosis markers in cultured astrocytes. Pharmacological stimulation of intracellular calcium led to recruitment of synaptobrevin to the plasma membrane with concomitant disappearance of D-serine, suggesting calcium-dependent exocytosis [84].

The expression of neuronal syntaxin and SNAP-25 protein of the fusion complex have also been reported in cultured astrocytes [85] but their expression is controversial [58,80]. SNAP-23 is an analog of SNAP-25 and has been reported in both astrocyte cell cultures [79,86] and in the rat cerebellum [86]. Astrocytes analyzed immediately after isolation from 11-day-old animals were immunopositive for SNAP-23 and synaptobrevin II, almost entirely devoid of SNAP-25 and synaptophysin, and completely devoid of SV2 (Figs. 6.1 and 6.2) [82]. When cultured for various lengths of time, however, the expression of all proteins were abundant, suggesting that culturing astrocytes can induce the expression of certain proteins [82]. Jeftinija et al. [58] found that astrocyte cultures lacked SNAP-25 immunoreactivity, but pretreatment with BoTx-A (cleaves SNAP-25) and BoTx-C (cleaves syntaxins) decreased baseline and bradykinin-evoked glutamate release. After bradykinin-induced glutamate release in astrocyte cultures, α-latrotoxin stimulated calcium-independentglutamate release from astrocytes [59]. Since α-latrotoxin induces vesicle fusion and calcium-independent release of neurotransmitters from neurons, it suggests that astrocytes may release gliotransmitters in a similar manner. Furthermore, botulinum toxin B (cleaves SNAP-25) and tetanus toxin (cleaves synaptobrevin) decreased synaptobrevin II immunoreactivity and abolished glutamate release in cultured astrocytes [58,87]. The vacuolar type H⁺-ATPase inhibitor bafilomycin also reduced glutamate release from astrocytes [27,78,87]. The inhibition of glutamate transporters, on the other hand, had no effect on calcium-dependent glutamate release in rat hippocampal cultured astrocytes [87].

Both ATP- and glutamate-storing vesicles may be present in astrocytes, but they have distinct properties. For example, ATP release was only partially sensitive to tetanus neurotoxin in cultured astrocytes whereas glutamate release was almost completely impaired [66,81]. In addition, lysosomes found in astrocytes contain ATP and execute calcium-dependent exocytosis of gliotransmitters [88–90]. ATP mediated by P2YR1 is thought to be the major determinant of astrocyte calcium wave propagation (Fig. 6.3) [91]. Blocking lysosome release of ATP prevented the spread of calcium waves [92].

Secretogranin II, a marker for dense-core granules, was detected in the Golgi complex [93] and in a population of dense-core vesicles [93,94] in cultured astrocytes. These vesicles also contained neuropeptide Y (NPY) and underwent botulinum toxin B-sensitive calcium-dependent exocytosis [94]. The activation of mGluRs may lead to calcium-dependent fusion of NPY-containing dense-core granules with cell membrane followed by peptide secretion in astrocytes [95].

Immunogold analysis of rat tissue demonstrated glutamate and D-serine accumulation in synaptic-like microvesicles in the perisynaptic processes of astrocytes [96]. In the rat cerebral cortex, D-serine colocalized with astrocytic processes [79]. The amino acid content of immunoisolated glial vesicles consisted of both glutamate and D-serine but lacked GABA [79]. Furthermore, vesicular D-serine and glutamate uptake is driven by an electrochemical potential generated by V-ATPase, coupled to chloride transport, and is stereoselective [79].

Various chemical mediators have been found to modulate calcium-dependent exocytosis. Endocannabinoids activate CB1 receptors on astrocytes, which leads to phospholipase C-dependent release of intracellular calcium stores and release of glutamate [48,100,101]. This may lead to LTP if accompanied by postsynaptic NO production [100]. Prostaglandins mediate the activation of both AMPA/kainate receptors and mGluRs on astrocytes in acute slice preparations [102]. The application of prostaglandin E2 promoted calcium-dependent glutamate release whereas the inhibition of prostaglandin synthesis prevented the release [102].

Tumor necrosis factor α (TNFα) and its receptor TNF receptor 1 (TNFR1) play a role in the modulation of glutamate secretion from astrocytes. Using both wild-type and TNFα-deficient neurons and glia, Stellwagen and Malenka [103] demonstrated that synaptic scaling in response to activity blockade is mediated by TNFα released from astrocytes. In hemibrain horizontal slices from wild-type mice, TNFα activated astrocyte P2YR1 as well as induced astrocytic calcium elevations and glutamate release [104]. In mice lacking TNFα, however, intracellular calcium elevations could still be obtained, but glutamate release and neuromodulation could not. Cultured astrocytes from TNFα-deficient mice exhibited a defect in functional docking of VGLUT-expressing synaptic-like microvesicles [104]. Activation of the G-protein linked CXCR4 receptor by the chemokine stroma cell-derived factor 1 in hippocampal slices resulted in the rapid release of TNFα and subsequent stimulation of calcium-dependent glutamate exocytosis, a process that was absent in TNFα-deficient mice [105]. Interestingly, reactive microglia amplified this glutamate release.

The above data collectively constitute convincing evidence for the vesicular release of gliotransmitters in response to Gq-linked GPCR-dependent calcium elevations within astrocytes. This release can be sensed by neighboring neurons and is thought to play a role in the modulation of neuronal activity. The calcium-dependent mechanism of gliotransmission, however, has been called into question for several reasons [106]. First, transgenic mice that express a Gq-coupled metabotropic receptor (MrgA1) only in astrocytes were created [107]. Activation of these receptors produced robust, widespread astrocyte calcium increases in acute mouse hippocampal slices but failed to increase neuronal calcium, produce SICs in CA1 pyramidal neurons, or have any effect on excitatory synapse activity [107]. Second, IP₃ receptor 2 (IP₃R2) knockout mice lack spontaneous Gq-linked GPCR-mediated calcium increases in astrocytes but neuronal Gq GPCR calcium increases remain intact. In these knockout mice, baseline excitatory neuronal synaptic activity and glutamate levels were not altered [108]. Developmental adaptations in these mice cannot be ruled out until conditional knockouts are created [109]. Third, the majority of studies have been conducted in vitro and may not be representative of the true brain state [74,109,110]. One study reported the lack of vesicular glutamate transporter and synaptic protein (VGLUT1, VGLUT2, synapsin 1, synaptotagmin) mRNA in the astrocyte transcriptome in vivo [111]. These proteins were found in neurons. Fourth, experimental design and pharmacological agents used in preparations often have unintended consequences. For example, slice preparations that remove magnesium from the bath relieve the voltage-dependent block on NMDA receptors, but also may initiate spontaneous calcium signaling [112]. “Specific” agonists and antagonists often have inadvertent targets [23,110]. These considerations should be taken into account when interpreting astrocyte calcium signaling studies.

One major problem with studying astrocytic calcium signaling in vivo has been the lack of available tools. Recent developments of new tools that enable real-time monitoring of calcium dynamics, however, may alleviate some of the difficulties. First, genetically encoded calcium indicator (GECI) technology will help visualize thin astrocytic processes with an exceptional signal-to-noise ratio [113]. Second, genetically altered mice, when fully characterized, can prove to be powerful tools [107]. For example, the reporter mouse PC:G5-tdT when crossed with GFAP-Cre drivers can reliably report astrocytic calcium transients [114]. Third, in utero electroporation techniques have allowed for the stable transfection of cells with plasmids featuring the GECI GCaMP [114]. Fourth, excellent calcium imaging has been accomplished using two-photon microscopy after loading of fluorescent indicators in vivo [91]. The creation of better pharmacological, astrocyte-specific tools and the creation of conditional knockout mice combined with powerful imaging strategies will be required to help elucidate the role of astrocyte calcium signaling in healthy and diseased brain.

Astrocytes cultured from human epileptic foci exhibited spatially and temporally coherent calcium waves [120]. In coculture with neurons, astrocyte cytosolic calcium initially increased in response to high K⁺, glutamate, or bicuculline and slowly returned to baseline within 10 min (Fig. 6.4) [121]. Calcium imaging combined with electrophysiology revealed spontaneous calcium elevations in cortical and hippocampal slices from biopsies of epileptic patients [122]. These spikes were insensitive to the neuronal activity blocker tetrodotoxin (TTX), abolished by depleting internal calcium stores with thapsigargin, and propagated with the addition of ATP or glutamate. Electrical stimulation led to astrocytic calcium increases and glutamate release, which caused NMDA receptor-dependent SICs in neighboring neurons [122]. These SICs, however, were recorded in the absence of magnesium to maximize NMDA receptor activation, but it is important to remember that the lack of magnesium can have other unintended effects, such as the initiation of spontaneous calcium signaling [112].

Related posts:

Gliotransmitters Blood–Brain Barrier Disruption Inflammation Adenosine Metabolism Neuropathology of Human Epilepsy Types of Epilepsy

Stay updated, free articles. Join our Telegram channel

Join