Astrocytic Ca2+ Signaling and its Role in Modulating Cerebral Blood Flow
- 1. Dalton Cardiovascular Research Center, USA
- 2. Department of Bioengineering, University of Missouri-Columbia, USA
Citation
Ding S (2014) Astrocytic Ca2+ Signaling and its Role in Modulating Cerebral Blood Flow. Ann Vasc Med Res 1(2): 1006
Abbreviations
GAFP: glial fibrillary acidic protein; CNS: central nervous system; 2-P: two-photon; L1: layer 1; L2/3: layer 2/3; GPCRs: G-protein coupled receptors; mGluR5: metabotropic glutamate receptors; PLC/IP3 : phospholipase-C/inositol 1,4,5-triphosphate; TRP: transient receptor potential; NMDARs: GECIs: genetically encoded Ca2+ indicators; CBF: cerebral blood flow; PLA2: phospholipase A2; BG cells: Bergmann glial cells; Kir channel: inward rectifier K+ channel; COX1: cyclooxygenase; fMRI: Functional magnetic resonance imaging
Introduction
Astrocytes are predominant glial cell type in the central nervous system (CNS) [1-3]. Protoplasmic astrocytes in grey matter and fibrous astrocytes in white matter are the major type of astrocytes which are morphologically different. Protoplasmic astrocytes are complex (sponge like) and highly branched with numerous fine processes and their endfeet wrap around blood vessels, while fibrous astrocytes are less complex and have thicker and less branched processes. Under normal conditions, protoplasmic astrocytes occupy distinct non-overlapping domains in vivo, and their processes completely wrap or ensheath synapses as well as blood vessels [4-6]. Studies using immunofluorescence labeling of neuronal somata in mouse brains revealed that a single astrocyte enwraps on average four neuronal somata with an upper limit of eight. Halassa et al. [5] determined from singleneuron dye-fills that one astrocyte contacts 300–600 neuronal dendrites. The processes from an individual astrocyte envelope approximately 140,000 synapses from multiple neurons [6]. Thus astrocytes can be stimulated by synaptic activities. On the other hand, it has been known for a long time that astrocytes and blood vessel have intimate anatomic relationship. Recent studies using different approaches including fluorescence imaging and electron microscopy revealed that the cerebral vascular surface is almost completely covered by astrocytic endfeet [7-10]. In vivo imaging using 2-P fluorescence microscopy also indicate that astrocyte endfeet wrap around the blood vessels in the mouse brain (Figure 1). Importantly, the endfeet of perivascular astrocytes express high levels of aquaporin and connexin 43 [7]. The spatial occupation and the intimate contact with both synapses and blood vessels render astrocytes as ideally situated to relay neuronal signals to blood vessels and regulate cerebral blood flow. This article reviews the astrocytic Ca2+ signaling pathway and recent advances regarding the role of Ca2+ signaling in the regulation of cerebral blood flow (CBF). The various aspects of neurovascular coupling and the signaling pathways have been reviewed previously [11,12]. Analysis of data from different studies (especially from those in vivo studies) suggests that the involvement of astrocytic Ca2+ in functional hyperemia can be affected by tissue metabolism, animal species, age, brain region and wakefulness of animals. Thus the precise mechanisms by which astrocytic Ca2+ regulates cerebral blood flow can only be elucidated in a defined preparation.
Astrocytes and astrocytic Ca2+ signaling in the CNS
It is well known that astrocytes act as a K+ sink to maintain extracellular K+ homeostasis [13] and remove glutamate from the synaptic cleft by their glutamate transporters to avoid glutamate toxicity [14-16]. Astrocytes also provide nutritional and structural support for neurons. Some astrocytes express the glial fibrillary acidic protein (GFAP), which is used as a specific marker to distinguish them from other cell types; however, its expression levels are different in astrocytes in different regions. For example, under normal conditions, cortical astrocytes express much lower levels of GFAP than do protoplasmic astrocytes in the hippocampus of a mouse brain, although the densities of astrocytes in these two regions are similar [17,18]. It is also clear now that not all astrocytes express GFAP and vice versa, not all cells that express GFAP are astrocytes [19]. It has been recently shown that spinal cord astrocytes express region specific genes that are functionally different, e.g., a recent study revealed that ventral astrocytes in mouse spinal cord express Semaphorin 3a that is important for sensorimotor circuit organization [20]. Electrophysiological recordings also showed that astrocytes even in the same region have differential patterns of current-voltage relationship known as outward rectifying astrocytes and variably rectifying astrocytes [21]. Similarly, in vivo intracellular recording of astrocytic membrane potential revealed a significant variation in fluctuations depending on the local field potential state and cell body location [22]. Astrocytes also exhibit regional heterogeneity in spontaneous Ca2+ signaling. Using two photon (2-P) laser-scanning fluorescence microscopy, Takata and Hirase [23] used an anesthetized adult rat to show how the astrocytes in the cortical layer 1 (L1) exhibited distinct Ca2+ dynamics in vivo when compared to astrocytes in the cortical layer 2/3 (L2/3). They found that astrocytes in L1 had nearly doubled the Ca2+ activity of astrocytes in L2/3 [23]. Furthermore, Ca2+ fluctuations in the processes within an astrocyte were independent in L1, while those in L2/3 were more synchronous [23]. In urethane, anesthetized young mice (P9-25), hippocampal astrocytes exhibited synchronized Ca2+ oscillations and intercellular waves [24,25]. Furthermore, subcellular Ca2+ analysis revealed a complex dynamics in different domains within an astrocyte [26,27]. The difference in astrocytic Ca2+ activities in different brain regions and different microdomains within an astrocyte reflects a functional heterogeneity, which may be the result of different neuronal activities or synaptic integrations, microenvironment and/or different properties of astrocytes per se. These findings demonstrate that astrocytes in the CNS are heterogeneous in morphology, molecular expression, and function.
It was discovered more than two decades ago that cultured astrocytes could mediate Ca2+ signaling (i.e., transient Ca2+ increase) [28,29], suggesting that astrocytes can play more active roles in the CNS than previously found. More recent studies using 2-P microscopy have found that astrocytes can mediate Ca2+ signaling and intercellular waves in vivo by the activation of a variety of G-protein coupled receptors (GPCRs) including Group I metabotropic glutamate receptors (mGluRs) [30-32], P2Y receptors [32-36], GABAB receptors (GABABRs) [33;37], α1-adrenergic receptors [38-40], cholinergic receptors [41;42], and dopamine receptors [43;44]. GPCR-mediated Ca2+ signaling is now considered a primary form of Ca2+ signaling pathways in astrocytes as shown by Ca2+ imaging. GPCR stimulation activates phospholipase-C/inositol 1,4,5-triphosphate (PLC/IP3 ) pathway to release Ca2+ from the ER through the activation of IP3 Rs [1,3,45] for Ca2+ signaling pathways (for review see Verkhratsky et al. [46] and Ding [47]). Among the three subtypes of IP3 R (IP3 R1-3), IP3 R2 seems to be predominant in stimulating astrocytes in the rodent brain [48-50]. IP3 R2 knock-out (IP3 R2 KO) mice did not exhibit GPCR agonists-evoked Ca2+ release in astrocytes in brain slice preparation and in studies of live mice, demonstrating that IP3 R2 is a key mediator of intracellular Ca2+ release in astrocytes [36;45]. In addition to GPCR stimulations, astrocytic Ca2+ signals can also be evoked by sensory stimulations, mechanical stimulations, and photolysis of caged compounds [30,32,35,51,52]. Sensory stimulations, including whisker deflection [35], locomotion [53,54], limb stimulation [38;55;56], light illumination of visual cosrtex [42;57], and odor stimulation of olfactory glomeruli [8] can all induce Ca2+ elevation and intercellular waves in astrocytes in vivo in anesthetized animals. A number of plasma membrane proteins also control Ca2+ homeostasis through regulating Ca2+ influx from the extracellular side. Those proteins include the Na+ / Ca2+ exchanger [58-60], plasma membrane Ca2+ ATPase [58], store operated channels [61], P2X purinoceptors [62,63], transient receptor potential A (TRPA) [64] and C (TRPC) channels [61,65- 67], and N-methyl-D-aspartate (NMDA) receptors [63]. Astrocytic Ca2+ signaling has been extensively studied in cultured astrocytes as well as in live animals using fluorescence imaging. Both synthetic organic Ca2+ probes and genetically encoded Ca2+ indicators (GECIs) can be used to label astrocytes. The acetoxymethyl (AM) ester form of synthetic organic Ca2+ indicators such as fluo-4, Oregon green BAPTA-1 (OGB), Rhod2, or x-Rhod-1 have been largely used for in vitro and in vivo Ca2+ imaging due to their high sensitivity and speed [23,30,32- 35,41,51,68-72]. The selective labeling of astrocytes in vivo by synthetic organic Ca2+ indicators can be confirmed by an astrocyte selective dye sulforhodamine 101 (SR101) [70]. Recently developed GECIs including FRET-based GECIs such as YC3.6 and intensity-based single-fluorophore GECIs such as GCaMP provided an alternative way to image Ca2+ signaling in astrocytes in vivo. The major advantage for using GECIs is that they can be selectively expressed in astrocytes using an astrocyte-specific promoter. They can be expressed in astrocytes using viral transduction, in utero electroporation or transgenic mice [40,73-78]. Astrocyte-specific expression of GECIs using adeno-associated viral vectors can be achieved by astrocytespecific promoters [18,76,79]. The most recently developed GCaMP5 and GCaMP6 yielded a cytosolic Ca2+ increase in a neuron after triggering a single action potential and sensory stimulation [75,80]. Transgenic mice expressing floxed GCaMP3 and GCaMP5G have been crossed with Cre mice having neuronand astrocyte-specific promoters for neuronal and astrocytic Ca2+ imaging [40,74,77]. For astrocytic specific expression of GCaMP3 and GCaMP5G, GFAP-Cre mice are commercially available [81,82]. Although Ca2+ signals in the processes of astrocytes can be readily observed using bulky loading of organic synthetic Ca2+ indicators [30,33-35,38], GECIs especially GCaMP can reveal more detailed features of Ca2+ signaling in microdomains than organic synthetic Ca2+ indicators [40,76-78]. Since GECIs can be expressed in astrocytes for prolonged times, another advantage for using GECIs is that it is feasible to conduct long-term and repeated in vivo Ca2+ imaging in astrocytes. Usually, for in vivo imaging, an open skull or thin skull cranial window in the cortex must be prepared for optical access and/or loading organic Ca2+ indicators [68,83-86]. It is also worth mentioning that GECIs are also advantageous over synthetic organic Ca2+ indicators when they are used for mitochondrial Ca2+ uptake as GECIs can be selectively targeted in mitochondrial matrix, overcoming the partial localization of synthetic organic Ca2+ indicators in the cytosol [87,88].
The role of astrocytic Ca2+ signaling in the regulation of CBF CBF
CBF is regulated by cerebrovascular autoregulation and functional hyperemia. The latter refers to matched delivery of blood flow to the brain regions with different activity levels. Given that Ca2+ signaling is the primary form of astrocytic excitability, the role of astrocytic Ca2+ signaling in regulating CBF has been studied using in vitro and in vivo preparations.
Using brain slices, Stobart et al. [89] and Zonta et al.[90] found that Ca2+ elevation in astrocytes induced by neuronal afferent stimulation and photolysis of caged Ca2+ induced vasodilation. Studies from brain and retina slices suggest that the polarity of astrocytic Ca2+-dependent regulation of blood flow is dictated by tissue metabolism [91-93]. Brain slice study has also shown that the levels of astrocytic Ca2+ increase determine vessel dilation or constriction regardless of the mechanism by which astrocytic endfeet Ca2+ is elevated [94]. The modest increases in Ca2+ induced dilation, whereas larger increases lead to constriction. These studies suggest complex mechanisms of blood flow regulation by astrocytic Ca2+ signals [11]. Different pathways and vasoactive mediators derived from astrocytes are involved in vessel dilation and constriction. Synaptically released glutamate activates metabotropic glutamate receptors (mGluRs) and stimulates astrocytic Ca2+ elevation that activates downstream pathways through phospholipase A2 (PLA2). PLA2 mediates arachidonic acid generation and subsequently three types of metabolite:1) Prostaglandins (PG) by cyclooxygenase (COX1) [51,90,91] and 2) exoxyeicosatrienoic acids (EETs) by cytochrome P450 (CYP-450) epoxygenase [93] in astrocytes dilate vessels, and 3) 20-hydroxyeicosatetraeonic acid (20-HETE) by CYP-450 ω-hydroxylase in smooth muscle, which constricts vessels [91,93]. Nitric oxide (NO) can regulate vascular tone by inhibiting the synthesis of the vasoconstricting 20-HETE as well as the vasodilating EETs [95]. Oxygen levels affect the vasoactive mediators [91,92]. When oxygen levels are lowered and astrocyte Ca2+ concentration is increased, glycolysis is dominated in astrocytic energy metabolism and lactate is accumulated in extracellular space. Extracellular lactate attenuates transportermediated uptake of extracellular prostaglandin E2 , leading to accumulation and subsequent vasodilation [91]. These studies indicate that cellular energy metabolism regulate astrocytic Ca2+ mediated vascular changes. Astrocytic K+ also plays a role in vascular tone. Using rat brain slice preparations, it was reported that neuronal activity-induced astrocytic Ca2+ signals opened large-conductance Ca2+-sensitive K+ (BK) channels in astrocytic endfeet [96]. BK channels in turn activated inward rectifier K+ (Kir) channels in smooth muscle cells and cause vasodilation. However, further study showed that a high concentration of astrocytic Ca2+ (induced by either uncaging or electric field stimulations) and perivascular K+ caused vessel constrictions, while low astrocytic Ca2+ perivascular K+ caused vasodilation [94]. A study from retinal slice preparation provided contradictive evidence for the role of astrocytic K+ in functional hyperemia. Depolarization-induced release of astrocytic K+ from endfeet did not cause vasodilation in arterioles [97]. Furthermore, the magnitude of light-evoked vasodilations was identical in Kir4.1 knock-out and wild-type animals, indicating that astrocytic K+ in the retina does not contribute to neurovascular coupling. Thus the role of astrocytic K+ in regulating functional hyperemia is not defined.
While brain slice preparations provide ready access for pharmacological manipulations, Ca2+ imaging and patch clamp recording, a major drawback in using brain slice preparations is that arterioles are usually preconstricted and lack spontaneous tone. In vivo approach using 2-P microscopy to image cellular Ca2+ signals and blood flow through a cranial window provides a unique tool to study the mechanism of neurovascular coupling in the intact brain. Different studies indicate that Ca2+ evaluation in astrocytes is involved in functional hyperemia. Takano et al. reported that photolysis of caged Ca2+ in the endfeet of astrocytes induced rapid dilation of the arteries, but not of the veins and capillaries in the mouse cortex [51]. Similar phenomenon was observed following neuronal stimulation. Electrical stimulation evoked an increase in local neuronal activity and a widespread increase in astrocytic Ca2+ and was associated with vasodilation. Combined with pharmacological manipulation, they further showed that astrocyte-mediated vasodilation is partially dependent on COX-1 activity. In cerebellum, Ca2+ signals of Bergmann glial (BG) cells are associated with motor behavior and neuronal activity [53]. In awake and behaving mice, BG cells have exhibited three different forms of Ca2+ excitation: flares, bursts, and sparkles. Bursts and sparkles were ongoing in awake mice at rest, whereas flares were initiated during locomotor behavior. Locomotor performance initiated synchronized Ca2+ signals in the network of BG cells. Motor behavior induced astrocytic Ca2+ flares were correlated with blood flow increase, which can be attenuated by tetrodotoxin through the inhibition of neuronal activity. Thus, the specific animal behavior noted in the experiments dealing with awake and behaving mice dictated Ca2+ excitability in BG cells and vascular response. Odor stimulation also induced Ca2+ transients in astrocyte endfeet and an associated dilation of upstream arterioles; furthermore, Ca2+ elevations in astrocytes and functional hyperemia depended on mGluR5 in astrocytes and cyclooxygenase activation [8]. Electrical stimulation of a forepaw of isoflurane-anesthetized mice induced vasodilation and an astrocytic Ca2+ increase in the forepaw region of their primary somatosensory cortex [36]. However, a recent study showed that IP3 R2 KO mice exhibited normal functional hyperemia in the same tests [36], indicating that the stimulus-induced vasodilation is independent of IP3 R-mediated Ca2+ increase since IP3 R2 KO mice lack cytosolic Ca2+ elevation in astrocytes. In addition, it was observed that the onset of vasodilation preceded astrocytic Ca2+ increase. Consistent with this study, Takata reported that IP3 R2 knockout mice showed similar changes in CBF with WT mice after a brief electrical stimulation of the nucleus basalis of Meynert (NBM), the primary source of cholinergic projection to the cerebral cortex [98]. Moreover, whisker stimulation resulted in similar degrees of CBF increase in IP3 R2 KO mice and WT mice. In vivo electrophysiological recording on dentate gyrus and fMRI showed that WT and IP3 R2 KO mice exhibited no difference in electrical responses and BOLD signals after electrical stimulation of perforant pathway [99]. Using in vivo 2-P imaging on awake mice and astrocyte-specific expression of GCaMP6s, a very recent study from Bonder and McCarthy further confirmed that IP3 R2-mediated Ca2+ signaling is not critically involved in meditating functional hyperemia [100]. These results from in vivo indicate that neural activity-driven CBF modulation could occur without large cytosolic increases of Ca2+ in astrocytes. Similarly, inhibition of group I mGluRs did not affect transient hemodynamic responses following a brief whisker stimulation in rats under isoflurane anesthesia [101]. This study suggests that group I mGluRs, whether they are expressed in neurons and/or astrocytes, do not play a role in early hemodynamic responses following sensory stimulation in rats, which is contradictive to other in vitro and in vivo studies [8,51,90] and raises the question whether neurovascular coupling involves signaling from neurons to glial cells to blood vessels.
Discussion and Conclusions
Several in vitro and in vivo studies demonstrated that astrocytic Ca2+ signals were involved in the regulation of CBF. However, data from different studies were not consistent. The fact that from multiple in vivo studies, IP3 R2 KO mice do not show a difference from the wild type mice in CBF after sensory and electrical stimulation suggests that functional hyperemia induced by these stimulations is mainly neuronal dependent, and IP3 R2-mediated Ca2+ increase may not be sufficient enough to exert an additional effect on the onset of vasodilation [36,98- 100]. One must realize that anesthetized or awake animals might have different responses to functional hyperemia as astrocytes exhibit different Ca2+ properties in anesthetized versus awake mice. In awake mice, cortical astrocytes exhibited much higher frequency of spontaneous Ca2+ signaling in the cell body and processes than in mice anesthetized by isoflurane, ketamine and urethane [34]. In addition, spontaneous somatic Ca2+ signals in astrocytes were expressed as intercellular wave characterized as synchronized Ca2+ elevation in the astrocyte network in awake mice. Thus, synchronization of cortical astrocytic Ca2+ activity is a hallmark of wakefulness of an animal and the concentration of vasoactive mediators released from astrocytes might also depend on the wakefulness of an animal. The role of astrocytic Ca2+ in CBF regulation might also be regional difference while most in vivo studies were conducted on the somatosensory cortex. Age of animals could also be a factor. A recent study showed that spontaneous Ca2+ wave in BG increased with age [102], thus, the threshold for the glial Ca2+ in aged mice to exert an effect on neurovascular couple might be different from young mice. The role of mGluR5 in CBF regulation is also controversial. Inhibition of group I mGluR reduced sensory stimulation-induced astrocytic Ca2+ signals in cortex and olfactory glomeruli [8,51,90], suggesting that signaling from neuron to glial cells to blood vessel is involved in neurovascular coupling. However, another study showed that the antagonists of mGluR5 did not affect functional hyperemia, suggesting that neuron-derived glutamate is not involved in hemodynamic response to sensory stimulation [101]. Since young mice express much higher mGluR5 than adult mice [32], the discrepancy could be due to the age difference of animals used. Thus it is unclear to what extent functional hyperemia is dependent on glutamate-stimulated astrocytic Ca2+. On the other hand, functional hyperemia is multifactorial, involving neurons, glial cells, smooth muscle cells and endothelial cells; furthermore, their coordinated actions are required. Thus a unifying mechanism can only be determined using a defined preparation. Tissue metabolism, animal species, age, brain region for study, and wakefulness of animals may all affect the involvement of astrocytic Ca2+ signaling in the regulation of CBF. It is also worth pointing out that although KO mouse models are particularly useful for studies astrocytic Ca2+ signaling and neuron-glial vasculature coupling, functional and developmental compensation may eliminate the phenotypic effect [100]. As multiple studies with negative results were from the same IP3 R2 KO mice, it is important to develop and use other animal models to further clarify the role and involvement of astrocytic IP3 R2-mediated Ca2+ in the regulation of CBF. Inducible IP3 R2 KO mice might be a direction as the target gene of interest can be inactivated in specific cell types at specific time points, thus reducing or eliminating the phenotypic effect. However, inducible IP3 R2 KO mice are not available for the time being. Another approach is to use viral transduction to disrupt IP3 R-mediated Ca2+ signaling in astrocytes by overexpressing IP3 sponge and IP3 phosphatase specifically in astrocytes [18,88,103], however, it is expectable that the effect can only be achieved when enough number of astrocytes around the blood vessels is transduced. While we have discussed the involvement of astrocytic Ca2+ in functional hyperemia in healthy animals, several studies showed that astrocytes have increased Ca2+ signals in the mouse models of neural diseases (for review see Ding [17]). Using in vivo 2-P imaging, Ding et al. [30,33] reported that astrocytes exhibited enhanced Ca2+ signaling and intercellular waves in vivo after status epilepticus and photothrombosis-induced ischemic stroke. Traumatic brain injury [104] and Alzheimer diseases [105] also dramatically increase astrocytic Ca2+ signaling. Therefore, it is important to determine whether and how astrocytic Ca2+ plays a role in CBF regulation in neural diseases. Understanding the mechanisms by which enhanced astrocytic Ca2+ isinduced and the role of astrocytic Ca2+ is played in the regulation of CBF should provide therapeutic implications for these and other neural diseases.
Acknowledgements
This work was supported by the National Institutes of Health [R01NS069726] and the American Heart Association Midwest Affiliate Grant in Aid award [13GRANT17020004] to SD.