Ryanodine receptors, calcium signaling and regulation of vascular tone in the cerebral parenchymal microcirculation

Abstract

The cerebral blood supply is delivered by a surface network of pial arteries and arterioles from which arise (parenchymal) arterioles that penetrate into the cortex and terminate in a rich capillary bed. The critical regulation of cerebral blood flow, locally and globally, requires precise vasomotor regulation of the intracerebral microvasculature. This vascular region is anatomically unique as illustrated by the presence of astrocytic processes that envelope almost the entire basolateral surface of parenchymal arterioles. There are, moreover, notable functional differences between pial arteries and parenchymal arterioles. For example, in pial vascular smooth muscle cells (VSMCs), local calcium release events (“calcium sparks”) through ryanodine receptor (RyR) channels in sarcoplasmic reticulum membrane activate large conductance, calcium-sensitive potassium (BK) channels to modulate vascular diameter. In contrast, VSMCs in parenchymal arterioles express functional RyR and BK channels, but under physiological conditions these channels do not oppose pressure-induced vasoconstriction. Here we summarize the roles of ryanodine receptors in the parenchymal microvasculature under physiologic and pathologic conditions, and discuss their importance in the control of cerebral blood flow.

Introduction

In vascular smooth muscle cells (VSMCs), a global increase in cytoplasmic calcium concentration ([Ca²⁺]_i) leads to vasoconstriction by activation of Ca²⁺/calmodulin-dependent myosin light-chain kinase. However Ca²⁺ is a versatile second messenger, and vasomotor pathways involving local Ca²⁺ signals, such as Ca²⁺ sparks encoded by ryanodine receptors (RyRs), or propagated Ca²⁺ signals in the form of Ca²⁺ waves, have been the subject of many recent studies. The topic of Ca²⁺ signaling in the microcirculation is one of growing interest, particularly in the cerebral parenchyma as brain integrity is critically dependent on constantly adapting blood perfusion. Arteriolar tone significantly depends on the Ca²⁺ entry through ion channels of the VSMC plasma membrane, and also on the Ca²⁺-dependent vasodilator influences of endothelial cells. However these aspects of microvascular Ca²⁺ signaling are beyond the scope of the current review. The goal of this review is to provide a synopsis of current knowledge regarding RyR-dependent Ca²⁺ signals in smooth muscle cells of the brain parenchymal microvasculature and how these signals influence cerebral blood flow (CBF).

Part 1: Regulation of cerebral blood flow by parenchymal arterioles

Within the brain microcirculation, the parenchymal arterioles (PAs) are fundamental regulators of CBF. PAs account for approximately 40% of the total cerebral vascular resistance (33). Thus regulation of their diameter is a central process for (i) maintaining constant perfusion of brain tissue across a range of systemic arterial pressures (cerebrovascular autoregulation), and (ii) during neurovascular coupling (NVC), i.e. the mechanisms by which functional hyperemia occurs in the brain (34, 58).

1.1 Myogenic tone and cerebrovascular autoregulation

A function of arteries and arterioles is to constrict and relax in response to changes in intravascular pressure. An increase in intraluminal pressure constricts cerebral arteries and arterioles. A decrease in intravascular pressure has the opposite effect, inducing vasodilation. This essential regulatory mechanism, known as the vascular myogenic response was first described more than one hundred years ago (9) and ensures that blood flow remains nearly constant during moment-to-moment fluctuations in arterial pressure. PAs are the last smooth muscle-containing vessels upstream of the capillary bed and play a critical role in maintaining appropriate capillary perfusion pressure under both normal and pathologic conditions (33, 55). Additional anatomical features of the brain microcirculation, particularly the limited collateral supply blood flow, which has a so-called “bottleneck” effect on perfusion of the neocortex (87), further emphasize the relative importance of this part of the intracerebral microcirculation and its control mechanisms including those involved in regulation of myogenic tone.

The precise nature of the mechanisms and modulators of myogenic vasoconstriction is still a matter of some debate. However, myogenic constriction occurs in blood vessels studied in isolation, demonstrating that mechanisms intrinsic to the vascular wall are sufficient to induce this response (90, 119). Disruption of the endothelium does not impair pressure-induced constriction suggesting that both sensor and effector mechanisms responsible for the myogenic response reside within smooth muscle cells (75). In turn, it is broadly accepted that the pressure-induced constriction involves a depolarization of the arterial myocyte cell membrane (52) that activates voltage-dependent Ca²⁺ channels (VDCC), resulting in Ca²⁺ influx and subsequent vasoconstriction (62). Interestingly, compared with pial arteries, PAs depolarize and constrict to lower levels of intravascular pressure leading to a high amount of tone (between 30% and 40% at 40 mm Hg) (23, 29, 32, 51, 86, 88). Physiologically, this creates a high “vasodilator reserve”, i.e. a substantial capacity of these arterioles to dilate in response to locally generated vasodilator signals. This is particularly adapted to CBF regulation and the functional hyperemic response, as the vascular resistance is inversely proportional to the fourth power of the arterial radius according to the Hagen-Poiseuille equation.

1.2 The neurovascular unit regulates CBF locally

The ability of the brain to coordinate neural activation state with local CBF has been recognized since the end of the nineteenth century (102). This linkage is called neurovascular coupling. It underlies functional hyperemia in the brain, which ensures that local increases in metabolic demand are satisfied by increased substrate delivery by the blood. Impairment of this coupling between neuronal metabolism and cerebral perfusion, even if brief, generally trigger dramatic consequences. Anatomic and functional observations support a role of capillaries in functional hyperemia (59, 94). However, PAs are thought to mediate a large part of the local CBF increase in response to increased neuronal activity, due to their high level of vascular resistance and ability to dilate rapidly in response to a variety of endogenous substances (34, 37, 44, 110, 133). In contrast to pial arteries, PAs are not supplied with perivascular nerves (49) but are wrapped in astrocytic processes, called endfeet, which cover nearly their entire basolateral surface (58, 106). Studies over the last ten years have illuminated the importance of this anatomical configuration in mediating NVC. Synaptic release of neurotransmitters such as glutamate during brain neuronal activity stimulates metabotropic receptors on astrocytes and generates an increase in cytoplasmic Ca²⁺ which propagates through astrocytic processes by activation of inositol 1,4,5-trisphosphate receptors (InsP₃Rs), ultimately elevating [Ca²⁺]_i in the endfeet. There, this Ca²⁺ signal activates pathways leading to dilation of the PAs, involving various factors such as locally released K⁺ ions, adenosine, prostaglandins or epoxyeicosatrienoic acids (EETs) (37, 58, 96, 108). Neurons also release several substances able to regulate CBF such as nitric oxide (NO), prostanoids and vasoactive neurotransmitters (5, 57). Therefore neurons, astrocytes, myocytes and endothelial cells form a neurovascular unit that collectively controls local blood flow. These various pathways leading to vasodilation all act, to some extent, by hyperpolarizing the VSMC membrane which deactivates VDCC and then decreases [Ca²⁺]_i.

Thus, it is clear that the intracerebral microcirculation participates actively and substantially in the global and local regulation of CBF. The control of parenchymal arteriolar diameter by the VSMCs within the vessel wall is a process that is fundamental to this regulation. In turn, regulation of VSMC tone depends on variations in cytoplasmic [Ca²⁺]_i. Therefore, understanding the regulation of [Ca²⁺]_i in smooth muscle cells of PAs is of central importance.

Part 2: Roles of the sarcoplasmic reticulum, inositol 1,4,5-trisphosphate receptors and ryanodine receptors in regulation of parenchymal arterioles

2.1 The sarcoplasmic reticulum and regulation of intracellular Ca²⁺

The sarcoplasmic reticulum (SR) is constantly involved, through the release and re-uptake of Ca²⁺ into the cytoplasm, in vascular smooth muscle Ca²⁺ homeostasis (123). At low intraluminal pressure (e.g. 10 mm Hg), VSMC [Ca²⁺]_i has been estimated to be ~120 nM (62, 88). The sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA) transports Ca²⁺ from the cytoplasm into the SR. The presence in the SR lumen of the Ca²⁺ binding proteins, calreticulin and calsequestrin (65), also facilitates the active pumping of Ca²⁺ against its gradient by decreasing the luminal Ca²⁺ concentration (123). As a result, the available Ca²⁺ in the SR is significant, and therefore the opening of SR Ca²⁺ channels leads to the release of Ca²⁺ into the cytosol (54, 97).

Various smooth muscle cellular functions, including contraction, relaxation, proliferation and differentiation, involve Ca²⁺ release from the SR (13, 54, 123). Two families of SR Ca²⁺ release channels have been identified so far: the InsP₃R and the RyR.

2.2 Inositol 1,4,5-trisphosphate receptors

The InsP₃R consists of 4 subunits with a molecular weight of ~310 kDa each (40). Three distinct genes code three different types of subunits leading to three receptor subtypes (InsP₃R1-3). Although all three isoforms appear to be expressed in vascular myocytes, InsP₃R1 seems to be the predominant subtype in adult VSMCs, and InsP₃R2 and InsP₃R3 appear to be expressed primarily during proliferative phases of cellular activity (69, 79, 83, 112, 129). The possible formation of heterotetrameric complexes has been demonstrated by co-immunoprecipitation of InsP₃R expressed in cell lines (77).

All InsP₃R subtypes are activated by inositol 1,4,5-trisphosphate (InsP₃) (38), which is produced by phospholipase C (PLC) activation in response to Gα_q/11 protein coupled receptor activation (11, 13, 40). Therefore, InsP₃Rs are involved in the responses to several endogenous vasoactive molecules such as endothelin 1, acetylcholine, noradrenaline, and 5-HT (12). InsP₃Rs also display a bell shaped Ca²⁺ sensitivity: InsP₃R-Ca²⁺ release is potentiated at low Ca²⁺ concentrations (< 500 nM) whereas it is inhibited at higher Ca²⁺ concentrations (> 1~10 μM), with differences between subtypes (15, 38, 40, 115). These properties appear to be fundamental in the generation of Ca²⁺ oscillations observed in VSMCs in response to vasoconstrictors (12). Recently, in pial arteries, InsP₃R-mediated Ca²⁺ release has been shown to activate transient receptor potential TRPM4 channels (46) and large conductance Ca²⁺-activated potassium (BK) channels (130). Activation of transient receptor potential TRPC3 channels by direct physical coupling with InsP₃R has also been described (124).

2.3 Ryanodine receptors

Similarly to InsP₃Rs, three different genes encode three different RyR subtypes (RyR1-3) and functional channels are formed by the association of four subunits (35, 50). Possible heteromeric RyR2/RyR3 and RyR2/RyR1 formations, but not RyR1/RyR3 formations, have been demonstrated by immunoprecipitation (127). To date, the RyR is the only ion channel discovered by electron microscopy (41), presumably due to the strikingly large size of RyR protein subunits (550–660 KDa) (89). RyR subunits have cytosolic C- and N-termini and between four and twelve transmembrane domains, depending on the four different models proposed (19, 111, 117, 134). The reason(s) for the well-conserved large size of the RyRs, and the associated substantial metabolic cost of production and maintenance, remain unclear but underscores the physiological importance of these proteins.

RyR1 and RyR2 have been extensively studied in skeletal muscle and cardiac muscle, respectively, (66) where they are highly expressed and play a fundamental role in excitation-contraction (E-C) coupling. RyR3 appears to be more broadly expressed, with predominance in neurons (43, 68). The three subtypes are expressed in large cerebral arteries (118) as well as in the cerebral microcirculation (Dabertrand & Brayden, unpublished observations). Their expression in VSMCs also appears to be modulated by pathologic and environmental conditions (25, 26, 30, 65, 78). The channel opening is regulated by several agents including SR Ca²⁺ load (35, 89). But RyRs seem regulated primarily by binding of Ca²⁺ on the cytosolic face of the channel. [Ca²⁺]_i ranging from 1 to 10 μM triggers the opening of the channel, whereas millimolar [Ca²⁺]_i inhibits it (35, 60). Consequently, RyRs can be activated by Ca²⁺ entry through plasma membrane Ca²⁺ channels or by SR Ca²⁺ release from adjacent InsP₃Rs or RyRs (13). This mechanism, called Ca²⁺-induced Ca²⁺ release (CICR), is fundamental in cardiac contraction (66, 98). CICR has been described in several types of smooth muscle including cerebral arteries (61), in which bolus application of caffeine, which simultaneously activates all the RyRs, causes transient increases in [Ca²⁺]_i and vasoconstriction in both pial and parenchymal arterioles (29, 63). However, in contrast to cardiac and skeletal myocytes, global RyR-Ca²⁺ release appears to be minimal during E-C coupling in VSMCs under physiological conditions (60, 120).

In contrast to the excitatory role of RyRs in skeletal and cardiac muscle, RyRs mediate local Ca²⁺-release events (Ca²⁺ sparks) that activate BK channels to oppose pressure-induced depolarization and contraction of cerebrovascular smooth muscle (60, 63, 85, 120). The serial linkage of RyR and BK in this negative feedback vasodilator mechanism is well established, and clearly demonstrated by the robust and non-cumulative constrictions induced by RyR and BK channel blockers in intact pressurized pial arteries (63, 64). Interestingly, small caliber pial arterioles and PAs do not constrict to BK channel blockers in vivo (91, 109) or in vitro (29, 44, 51, 56), despite functional BK channels in VSMCs of these arterioles (29, 51). This suggests that under basal conditions the functional role of RyRs in the cerebral microcirculation is distinct from pial arteries.

Part 3: Calcium signaling in parenchymal arteriolar myocytes

3.1 Waves rather than sparks in the microcirculation

Under normal conditions of pH and temperature, pressure-induced constriction of intracerebral parenchymal arterioles is poorly counterbalanced by Ca²⁺ spark-driven BK current (29, 51, 56). This absence of negative feedback likely contributes to the greater myogenic tone observed in PAs compared to pial arteries (23, 24, 88). Using the intracellular fluorescent Ca²⁺ dye Fluo-4, we found that VSMCs in pressurized PAs exhibited primarily Ca²⁺ waves (78% of cells) propagating through the cytoplasm over a duration of several seconds, and very few localized Ca²⁺ events (12% of cells) (29). Ca²⁺ waves and Ca²⁺ sparks were not observed in PAs when SR Ca²⁺ was depleted by cyclopiazonic acid or when the arterioles were exposed to RyR blockers (ryanodine or tetracaine). These results indicate that RyRs mediate both Ca²⁺ waves and Ca²⁺ sparks in PAs (29). However, it is premature to conclude that waves rely exclusively on RyRs with no InsP₃R involvement (47). Interestingly, a similar Ca²⁺ signal topography, i.e. Ca²⁺ sparks in large arteries and Ca²⁺ waves in the microcirculation, has also been described in cremaster muscle feed arteries and arterioles (121, 122). However, Westcott and Jackson (2011) attributed the waves in the arterioles not to RyR, but to InsP₃R activation. Differences in InsP₃R and RyR subtype expression between the cremaster and the cerebral microcirculation may account for this divergence (25, 27, 28, 73, 121). Moreover, the presence of perivascular nerves in cremaster arterioles (39) may facilitate the engagement of the InsP₃R in the generation of Ca²⁺ waves by releasing Gα_q/11 protein coupled receptor agonists, and thus maintaining a high level of cytoplasmic InsP₃. Receptor tyrosine kinase activation also generates InsP₃ (13) and possible contributions of this signaling pathway should be considered. In contrast to cremaster arterioles, PAs lack extrinsic innervation (49) and demonstrate elevated cytoplasmic [Ca²⁺]_i (88) which may promote RyR activation. The large variability observed in the mechanisms triggering and propagating Ca²⁺ waves in smooth muscle makes it difficult to develop a global model, but the variation in RyR function between large and small caliber arteries remains noteworthy and requires further detailed investigation.

In a recent study, Mufti et al. provided interesting insights into the role of the RyR-mediated Ca²⁺ waves in cerebral arteries. They demonstrated their importance in maintaining myogenic tone under low pressure by increasing phosphorylation of myosin light chain (82). Parenchymal arterioles depolarize and constrict to a greater extent than pial arteries at lower pressure (23, 51, 88). These observations are consistent with (i) the predominance of Ca²⁺ waves in PA VSMCs and (ii) the lack of effect of RyR and BK channel blockers on PA diameter. A likely explanation for the inability of Ca²⁺ waves to activate BK channels is that Ca²⁺ waves do not deliver sufficient Ca²⁺ to activate BK channels (95). In contrast, Ca²⁺ sparks deliver micromolar Ca²⁺ to closely apposed BK channels to cause substantial channel activation.

3.2 Local Ca²⁺ release initiates Ca²⁺ waves

The ability of Ca²⁺ sparks to initiate Ca²⁺ waves which travel through the cytoplasm following a CICR mechanism was first described in cardiac myocytes (20, 21) and is still an area of active investigation (18). Similarly, Ca²⁺ spark-initiated Ca²⁺ waves have been observed in several vascular beds including the hepatic portal vein (4, 48), small mesenteric arteries (99), and retinal arterioles (116). The prevalent model explaining the propagation of the initial Ca²⁺ release to the entire cell is that a local increase in Ca²⁺ activates neighboring RyRs because resting open state probability (P_o) of RyRs is high (Figure 1) (10, 13, 16). When resting P_o is low, local RyR Ca²⁺ release (e.g. a spark) is not sufficient to activate nearby RyRs to cause a Ca²⁺ wave (Figure 1) (10, 13). It is also possible that ambient InsP₃ levels are elevated in PAs, and that a combination of InsP₃Rs and RyRs in non-junctional SR elements is causing Ca²⁺ waves (47).

Figure 1. Effect of pH on RyR-mediated Ca²⁺ signals in smooth muscle cells of pressurized pial arteries and parenchymal arterioles.

By modulating RyR excitability, protons can influence the shape of the Ca²⁺ signals in cerebrovascular smooth muscle cells. At the physiologic pH of cerebrospinal fluid (pH=7.3) myocytes in large (pial) arteries display spatially limited and brief Ca²⁺ release events (sparks). When pH increases (alkalosis), RyR open probability (P_o) increases as well and the Ca²⁺ ions released by a cluster of RyRs activate neighboring clusters, forming a Ca²⁺ wave by propagation of the signal. In contrast, parenchymal arterioles display Ca²⁺ waves under normal conditions. When pH decreases (acidosis), RyR P_o is low and this spatially restricts the Ca²⁺ release (Ca²⁺ spark). This reshaping of Ca²⁺ signals in the brain microcirculation plays a fundamental role during the vasodilation induced by acidosis.

The fundamental role played by the RyR P_o in shaping intracellular Ca²⁺ signals is particularly well illustrated in the observations reported by Heppner et al. in VSMCs of pial arteries (53). These investigators found that a small increase in pH (7.4 to 7.5) increases Ca²⁺ spark frequency. However, further increases in pH beyond 7.6, shifted Ca²⁺ signaling from stationary sparks to Ca²⁺ waves (Figure 1) (53). The simplest explanation for this phenomenon is that a decrease in intracellular proton (H⁺) levels increased RyR P_o, and therefore the likelihood of Ca²⁺ waves. However, the induced Ca²⁺ waves contributed little to the observed vasoconstriction (53).

3.3 Acidosis converts Ca²⁺ waves to sparks to activate BK channels

Because smooth muscle cells in PAs exhibit mostly Ca²⁺ waves at physiological pH, we hypothesized that acidic pH might dilate PAs by reshaping the predominant intracellular Ca²⁺ signal from Ca²⁺ waves to Ca²⁺ sparks, which would activate BK channels to cause membrane potential hyperpolarization and vasodilation (Figure 1). Using pharmacological and genetic approaches, we found that this mechanism accounts for 60% of the vasodilation observed by acidifying the external pH from 7.4 to 7.0 (29). The concept that acidosis, over a narrow and physiologically relevant pH range, can induce vasodilation principally by remodeling Ca²⁺ signals is entirely novel, and as described below, may not be limited to the intracerebral microcirculation.

The reshaping of the Ca²⁺ signaling in VSMCs of PAs by protons likely occurs because changes in extracellular pH (pH_o) lead to changes in intracellular pH (pH_i). The relationship between pH_o and pH_i has been established in VSMCs from many tissues including the brain (3, 6, 92, 113). This relationship appears to be linear with a ratio ΔpH_i/ΔpH_o of about 0.73 (7). Moreover, both normocapnic and hypercapnic acidosis induce a RyR-BK channel component of acidic pH-induced dilation. In a 1989 study, Toda et al. investigated the relaxation of dog cerebral artery strips in response to acidosis induced by hypercapnia or manipulation of the bicarbonate concentration (114). These investigators showed that the increase in CO₂ from 5% to 15%, as in our study (29), has a vasorelaxant effect mediated by decreasing pH rather than acting through direct effects of CO₂ in the bathing media. This finding is consistent with our observations concerning the similarity between normocapnic and hypercapnic acidosis effects on Ca²⁺ signaling in PAs. Additionally, a reduction in pH_o lowers cytoplasmic Ca²⁺ (8) which would decrease RyR activity. Interestingly, Heppner et al. showed that inhibition of Ca²⁺ influx through VDCC did not prevent the induction or the persistence of alkaline pH-induced Ca²⁺ waves in rat cerebral pial arteries (53). Thus the reshaping of RyR-dependent Ca²⁺ signaling by pH_o appears to be due primarily to the direct regulation of RyR P₀. The simplest interpretation of the observed remodeling of Ca²⁺ signals at pH 7.0 is that protons lower RyR P₀ and then prevent local Ca²⁺ release to activate neighboring RyRs (Figure 1) and Ca²⁺ release remains limited to single RyR (Ca²⁺ quark) or clusters of RyRs (Ca²⁺ sparks). However we have not directly assessed this mechanism and more complex models can be considered. Intracellular acidification in combination with the subsequent reduction of intracellular Ca²⁺ could also decrease the P_o of both RyRs and InsP₃Rs, and thereby terminate Ca²⁺ waves. The subsequent reduced leak of Ca²⁺ from the SR would then elevate SR luminal [Ca²⁺], and thereby may enhance local spark probability at the junctional SR elements which abut the plasma membrane with BK channels (131). Indeed it is well demonstrated and recognized that an increase in the SR Ca²⁺ load activates RyRs (35) and increases Ca²⁺ spark frequency along with spontaneous transient outward currents occurrence (131). Unfortunately the influence of SR Ca²⁺ load on InsP₃Rs is less clear and experimental evidence is scarce. A luminal Ca²⁺ binding site has been described (104) and Bezprozvanny & Ehrlich showed that InsP₃R activity is inhibited when luminal Ca²⁺ concentration is raised from 3 μM to 10 mM (14), whereas Bottman and coworkers reported a potentiation of the InsP₃-induced Ca²⁺ release by luminal Ca²⁺ (17). These issues remain to be resolved and are important topics for future investigation.

3.4 RyRs and neurovascular coupling

Parenchymal arterioles play a critical role during NVC (34). These arterioles dilate in response to neuronal activation, which results in a local increase in blood flow (58, 67, 96). Activation of BK channels by Ca²⁺ sparks through RyR is a powerful vasodilator mechanism and can be activated in PAs by acidification (29). However, the origin and nature of Ca²⁺ signals that occur in parenchymal VSMCs during NVC remain to be identified. Interestingly, in brain slices, VSMCs of PAs exhibit synchronous Ca²⁺ oscillations (36, 37) which are rapidly suppressed during NVC simulated through neuronal depolarization by electric field stimulation. This observation suggests that Ca²⁺ signals in these arterioles are influenced by neural stimulation via vasoactive substances released from astrocytes and neurons. In newborn piglets, glutamate increases Ca²⁺ spark frequency and decreases cytoplasmic Ca²⁺ concentration in the VSMCs of brain slice arterioles (126). This effect is thought to be mediated by carbon monoxide (CO) derived from astrocytic heme oxygenase metabolism (70, 125, 126). However, CO has been described as a tonic vasoconstrictor, preventing H₂S-induced vasodilation in cerebellar slices from neonatal mice (80) and CO failed to dilate isolated pial arteries from adult rats and mice (2). Therefore, the involvement of CO during NVC in the mature brain remains to be established. Others have suggested that epoxyeicosatrienoic acids (EETs) produced by cytochrome P450 epoxygenase may mediate NVC (72, 76, 93, 103). The mode of action of EETs in NVC has not been fully resolved, but these compounds induce BK channel-mediated hyperpolarization of VSMCs (42), and increase Ca²⁺ spark frequency in pial arteries (31). Consequently, EETs are potential modulators of RyR activity during NVC. As in the case of acidosis, Ca²⁺ spark-driven BK currents can be activated in PAs and can induce large, sustained and reversible vasodilation (29) (Figure 1). Then Ca²⁺ spark-driven activation of BK channels may also play a role in the vasodilator effects of substances such as CO and EETs, whereby VSMC relaxation occurs.

In their classic study, Roy and Sherrington suggested a possible role for H⁺ during NVC (102). Protons produced by neuronal metabolism could cause local acidification and then vasodilation. While this hypothesis has been challenged, (58, 74) local release of protons from the astrocytic endfoot onto the smooth muscle could produce dilation. Indeed, alkalinization of astrocytic cell body during cortical stimulation (22) or metabotropic glutamate receptor activation (1) has been reported, which would be consistent with the release of H⁺. However, using the fluorescent pH indicator SNARFR-5, we did not observe changes in endfoot pH in brain slices following electrical field stimulation (Dabertrand and Nelson, unpublished observations). It is more likely that vasodilation induced by a decrease in pH within the brain may occur during pathologic conditions.

In normal conditions, appropriate pH levels are maintained by the lungs and kidneys. The CO₂ resulting from metabolism is expired by the lungs, but if it is retained or rapidly eliminated, it causes respiratory acidosis or respiratory alkalosis, respectively, The kidney, as a regulatory organ, adjusts acid and base excretion and dysregulation in this balance leads to metabolic acidosis or metabolic alkalosis. Nevertheless, acidification of nervous tissue can be triggered rapidly with reduced respiratory rate, during CO₂ inhalation (132) or during cerebral lactic acidosis related to ischemia (101) and hypoxia (71). Indeed, without oxygen, the brain switches to anaerobic glycolysis, leading to accumulation of lactic acid. Therefore, during ischemia, tissue pH falls by as much as 1 unit (84, 105) and this acidification is thought to play a role in the injury caused, for example, by ischemic stroke (81, 107, 128). Consequently moderate hyperventilation of patients with cerebral ischemia has been considered in the past to counteract the acidification. However this also increases vascular tone and therefore globally decreases CBF, which worsens the ischemia. (100). Consequently, blockade of downstream targets activated by lactic acidosis appears currently more promising clinically (81). Nevertheless, in the case of cerebral hypoxia or ischemia, acidosis-related vasodilation of PAs might represent a mechanismto maintain perfusion by promoting collateral blood flow.

Conclusion

Since the discovery of Ca²⁺ sparks in smooth muscle in 1995 (85), many roles of RyRs in vascular function have come into focus. Ca²⁺ release through RyRs contributes to Ca²⁺ waves in VSMCs of pressurized arterioles, but the role of Ca²⁺ waves and their contribution to parenchymal arteriolar tone remains obscure. It is also possible that Ca²⁺ waves modulate the activity of Ca²⁺-dependent transcription factors such as Nuclear factor of activated T-cells (NFAT) (45), and thereby exert transcriptional control. It is, however, clear that at physiological external pH, Ca²⁺ waves do not activate BK channels in VSMCs of pressurized parenchymal arterioles. Acidification leads to a loss of Ca²⁺ waves and the appearance of Ca²⁺ sparks causing profound, rapid and reversible vasodilation. Over a narrow and physiologically relevant pH range, the majority of the proton-induced vasodilation depends on Ca²⁺ spark-driven activation of BK channels (29). These results suggest that the role of the RyRs in vascular smooth muscle is context-dependent. The lack of Ca²⁺ sparks and absence of BK channel-mediated hyperpolarization to oppose pressure-induced constriction at physiological pH can explain why PAs are more depolarized and constricted at lower pressures, an appropriate physiological response for arterioles that experience lower pressures. Acidification-induced Ca²⁺ spark engagement of BK channels provides a novel mechanism to enhance cerebral blood flow, and may play an important role in pathological responses in the brain. Therefore, elucidating the regulatory mechanisms of RyRs is crucial for understanding the control of cerebral blood flow.

Abbreviations used

[Ca²⁺]_i

intracellular calcium concentration

BK

large conductance Ca²⁺-activated potassium channel

CBF

cerebral blood flow

CICR

Ca²⁺-induced Ca²⁺ release

CO

carbon monoxide

CO₂

carbon dioxide

E-C

excitation-contraction

EET

epoxyeicosatrienoic acid

H⁺

proton

InsP₃

inositol 1,4,5-trisphosphate

InsP₃R

inositol 1,4,5-trisphosphate receptor

K⁺

potassium

NFAT

nuclear factor of activated T-cells

PA

parenchymal arteriole

pH_i

intracellular pH

pH_o

extracellular pH

P_o

open state probability

RyR

ryanodine receptor

SERCA

sarco/endoplasmic reticulum Ca²⁺-ATPase

SR

sarcoplasmic reticulum

VDCC

voltage-dependent Ca²⁺ channel

VSMC

vascular smooth muscle cell