Inhibition of Cerebral Vasoconstriction by Dantrolene and Nimodipine

Abstract

Introduction

Cerebral vasoconstriction is associated with increased cytosolic Ca²⁺ concentration in vascular smooth muscle, presumably due to Ca²⁺ influx and Ca²⁺ release from intracellular stores. We tested the hypothesis that dantrolene (a blocker of Ca²⁺-induced Ca²⁺ release from the ryanodine receptor channel on the sarco-endoplasmic reticulum) would potentiate the action of nimodipine (a voltage-dependent L-type Ca²⁺ channel blocker, considered standard therapy for SAH) in inhibiting the vasoconstriction of isolated cerebral arteries.

Method

Sprague-Dawley rat basilar and femoral arteries were analyzed for ryanodine receptor expression by immunofluorescence and PCR. Vasoconstriction of basilar artery ex vivo was measured in a wire myograph while exposed to serotonin (5-HT) or endothelin-1 (ET-1) in the presence or absence of dantrolene (10–100μM) and/or nimodipine (30nM). Femoral artery was examined for comparison.

Results

Basilar and femoral arteries express only the ryanodine receptor 3 (RyR3) isoform. In both basilar and femoral arteries dantrolene significantly inhibited the constriction to 5-HT, whereas it poorly affected the constriction to ET-1. The inhibitory effect of dantrolene on 5-HT was substantially increased by nimodipine, inducing a 10-fold increase in the 50% effective concentration of 5-HT and a 46% reduction in maximum basilar constriction. In femoral artery, dantrolene modestly affected constriction to phenylephrine and there was no interaction with nimodipine.

Conclusion

Dantrolene has synergistic effects with nimodipine against 5-HT-induced vasoconstriction in isolated cerebral arteries. Dantrolene-nimodipine interaction will require testing in a pathophysiological model but might provide treatment for reducing SAH-related vasospasm or other 5-HT-related vasospastic syndromes, such as Call-Fleming syndrome.

Introduction

Dantrolene is an FDA approved drug for the treatment of malignant hyperthermia. It inhibits the ryanodine receptor Ca²⁺ channel (RyR) located on the sarco-endoplasmic reticulum, thereby inhibiting Ca²⁺ release from the large Ca²⁺ stores of the sarco-endoplasmic reticulum into the cytosol. ¹ The ryanodine channel is expressed in 3 isoforms: RyR1, RyR2, and RyR3. ² RyR1 isoform is predominantly found in skeletal muscle, RyR2 in myocardium, and RyR3 in neurons, smooth muscle cells and inflammatory cells.² Dantrolene blocks RyR1 and RyR3 isoforms, but not RyR2,³ consistent with its lack of significant cardiac effect when used in vivo.¹

Understanding cerebral artery physiology is imperative to understanding pathophysiology such as cerebral vasospasm following subarachnoid hemorrhage (SAH) is a frequent cause of secondary brain ischemia.⁴ The pathogenesis of vasospasm is incompletely understood, but it is associated with maintenance of tone due to persistent elevation of cytosolic Ca²⁺ concentration in vascular smooth muscle⁵ and involves multiple mediators, including 5-hydroxytryptamine (serotonin, 5-HT) and endothelin-1 (ET-1).^6–8 Nimodipine, a 1,4-dihydropyridine blocker of voltage-dependent L-type Ca²⁺-channel (VDCC), is currently indicated after SAH and modestly improves outcomes, but the outcome remains generally poor⁹ and despite the ability to induce vasorelaxation it is not clear whether this or a neuroprotective effect is responsible for improved outcomes.

VDCC blockers (also referred to as calcium antagonists) relax vascular smooth muscle, by inhibiting Ca²⁺ influx, a pathway shared by different vasoconstrictor agonists.¹⁰ Because Ca²⁺-induced Ca²⁺ release (CICR) is a recognized mechanism of excitation-contraction coupling in vascular smooth muscle,^{11, 12} including cerebral arteries,^{13, 14} we hypothesized that dantrolene i) might inhibit cerebrovascular constriction to agonists and; ii) might potentiate the action of nimodipine, because the two drugs moderate sequentially linked physiological steps. Nimodipine inhibits Ca²⁺-influx resulting in lower cytosolic Ca²⁺ concentrations, thereby reducing sarco-endoplasmic reticulum Ca²⁺ release, which in turn is blocked by direct inhibition of RyR3 channel by dantrolene, eventually leading to reduce smooth muscle contraction.

To our knowledge, dantrolene has not been investigated as a treatment for any cerebral vasospasm syndrome, while nimodipine is now standard treatment for SAH.^{9, 15} Another potential advantage of dantrolene is its potential for neuroprotection (same can be noted for nimodipine), as demonstrated in in vitro and in vivo models of injury.¹⁶

In this physiology study we challenged ex vivo basilar arteries, isolated from rats, with serotonin (5-HT) or endothelin-1 (ET-1), in a wire myograph in the presence or absence of dantrolene and/or nimodipine. To determine whether the effects of dantrolene and nimodipine were specific for the preparation and/or for the agonist, we applied the same experimental protocol to rat femoral arteries for comparison. Our data support the hypothesis that the association dantrolene+nimodipine has a synergistic inhibitory effects on vasoconstriction induced by 5-HT in the cerebrovascular bed.

Methods

The subcommittee for research and animal care at Massachusetts General Hospital approved animal use. Rats (Sprague-Dawley, male, 200–300 g) were sedated with inhaled chloroform followed by decapitation. A total of 73 rats were used for this study.

Reverse transcription and Polymerase Chain Reaction (PCR)

Cerebral or femoral arteries were removed and cleaned to remove surrounding adventitial tissue (n=2). Cortex, striatal muscle and cardiac muscle were also removed as positive controls. All tissue was homogenized in RNA lysis buffer and processed as recommended by Ilustra RNAspin Mini kit (GE Healthcare; Piscataway, New Jersey, USA). RNA was eluted with 40 μl of water. cDNA was made with 18ul of RNA as directed using Superscript III kit (Invitrogen; Carlsbad, California, USA). Total cDNA was measured with a BioPhotometer (Eppendorf; Westbury, New York, USA). A total of 1 μg of cDNA in a 20μl PowerSybr reaction (Applied Biosystems; Foster City, California, USA) was used in real-time PCR analysis on a 7500 Real-Time PCR System (Applied Biosytems). Cycling parameters for all genes were 50°C for 10min, 95°C for 10min and 40 cycles of 95°C for 15s and 60°C for 1min. Sequence for the following genes were:

RyR1: forward 5′ GAAGGTTCTGGACAAACACGGG 3′, reverse 5′ TCGCTCTTGTTGTAGAATTTGCGG 3′,

RyR2: forward 5′ GAATCAGTGAGTTACTGGGCATGG 3′, reverse 5′ TTGATCTCTGAGTTCTCCAAAAGC 3′

RyR3: forward 5′ ACTGGGTATATGGCACCAACACT 3′, reverse 5′ CCACACAGACCAGAGAGATGACA 3′

18s rRNA (endogenous control for cDNA loading, Applied Biosystems, #4310893E). Specificity of amplicon verified by dissociation curves and 2% agarose gel electrophoresis. Agarose gel imaged with Typhoon 9410 (GE Healthcare) with Sybr Green, (image settings: 488nm laser, 520nm filter, 500V photomultiplier tube, normal sensitivity, 50μm resolution).

Immunoflourescence

Basilar artery was removed as described above (n=2). It was post-fixed in 4% paraformaldehyde for 24 hours at 4°C followed by treatment for cryopreservation in 30% sucrose at 4°C. Sections were cut on a cryostat at 20μm thick. Sections were blocked in 10% normal donkey serum (Vector Laboratories, Burlingame, California, USA) and 0.3% Triton X-100 for 30 minutes at room temperature (RT). Slides were then incubated in same solution as above with mouse anti-RyR3 at 1:100 dilution (Developmental Studies Hybridoma Bank at University of Iowa) overnight at 4°C. After washing, slides were incubated with donkey anti-mouse biotinylated antibody (Vector Laboratories) at 1:250 dilution in PBS and 2% normal donkey serum for 2 hours at RT. After washing, slides were incubated with Alexa546 streptavidin at a 1:500 dilution for 1 hour at RT. Nuclei were stained with Sytox Green (Invitrogen). Specificity of staining was confirmed by above protocol but only the secondary antibody was eliminated. Images were taken with a confocal Zeiss LSM5 system. Images were taken with 100x oil objective 2μm thick z-stack for 5 sections and projected into a single image.

Measurement of contractile tone in isolated vessels

Basilar and femoral arteries were removed and immersed in physiological solution (composition, mM: NaCl, 118; KCl, 4.6; NaHCO₃, 25; MgSO₄, 1.2; KH₂PO₄, 1.2; CaCl₂, 1.25; glucose, 10; EDTA, 0.025; pH 7.4 at 37 °C). Arteries were cleaned of surrounding tissue, cut into segments (1.5 – 2 mm long), threaded onto 40 μm stainless steel wires and mounted in a isometric myograph (610M, Danish Myo Technology, Aarhus, Denmark). After mounting, each preparation was equilibrated, unstretched, for 30 min, in physiological solution, maintained at 37°C and aerated with a gas mixture of 95% O₂ - 5% CO₂. Then, the normalized passive resting force and the corresponding diameter were determined for each preparation from its own length-pressure curve, according to Mulvany and Halpern.¹⁷ Contractile responses were recorded into a computer, by using a data acquisition and recording software (Myodaq and Myodata, Danish Myo Technology). After tension normalization and 30-min equilibration in physiological solution, the preparations were stimulated with isotonic depolarizing 100 mM KCl solution, in which part of NaCl had been replaced by equimolar amount of KCl. After washout, the preparations were incubated with or without dantrolene, nimodipine or both, for 30-min; then the preparations were exposed to vasoconstrictor agonists (5-hydroxytryptamine, 5-HT, 1 nM – 10 μM; phenylephrine, PE, 10 nM – 100 μM; endothelin-1, ET-1 10 pM – 100 nM). The amount of vasoconstriction induced by each drug was normalized to the maximum contraction induced by high KCl for each individual preparation.

5-HT, PE, ET-1, dantrolene and nimodipine were from Sigma (St Louis, MO); millimolar stock solution were prepared in H₂O for 5-HT, PE, ET-1, and further diluted as required; dantrolene was dissolved in dimethylsulfoxyde as a 3 mM stock solution, nimodipine was dissolved in ethanol as a 0.1 mM stock solution; experiments with nimodipine were performed by protecting the organ chamber from light, to avoid drug photoinactivation; final concentrations of dimethylsulfoxyde and ethanol in the organ chamber did not exceed 0.33 and 0.03%, respectively. Nimodipine final concentration of 30 nM for all experiments was chosen based on a low-end drug concentration of oral dosed nimodipine in humans.¹⁸

Data was graphed as a percentage of high KCl-induced vasoconstriction against a log molar concentration of drug. Each set of data points was curve-fitted by a non-linear regression, best-fit, sigmoidal dose-response curve with no constraints, with the use of GraphPad Prism version 4.0c for Macintosh (GraphPad Software, San Diego, CA). Each curve is represented by 6 or greater individual preparations from 3 or more rats. Analysis is performed on the average of the number of preparations, as the preparation of the segments generated greater variability than individual normal rat basilar or femoral artery. Whole curves were compared by analysis of variance (ANOVA), with significance set at p<0.05. Those curves with differences were compared by ANOVA or t-test for logEC₅₀ (concentration producing 50% of maximum contraction) and maximum contraction (E_max), with significance set at p<0.05.

Results

Ryanodine Receptor Expression

Cerebral and femoral vessels were isolated and analyzed for expression of the ryanodine receptor channel. Evidence of ryanodine receptor expression throughout the smooth muscle wall was obtained in basilar artery (Figure 1A) and in femoral artery (not shown). However, because this antibody does not discriminate among the three ryanodine isoforms, RyR1, RyR2, or RyR3, we further investigated the expression of each isoform at the gene level. Neither cerebral nor femoral vessel smooth muscle cells expressed mRNA for RyR1 or RyR2 despite expression noted in our positive controls of skeletal muscle and cardiac muscle, respectively (data not shown). Both cerebral and femoral smooth muscles cells expressed RyR3 at relatively the same levels (4.66×10⁻⁶ and 3.91×10⁻⁶ per 18s rRNA), as measured by real-time PCR (Figure 1B).

Figure 1.

Ryanodine Receptor Expression. A, cross-section of rat basilar artery immunostained for ryanodine receptor (red) and a nuclear counterstain (green). B, Agarose gel electrophoresis of PCR products from 2 samples of femoral (f) and cerebral (c) arteries. The left panel shows a 70 bp PCR fragment of the RyR3 sequence and the corresponding loading control of 18s rRNA in the right panel. Closest ladder bands to amplicon fragments are labeled 75 bp and 200bp, respectively.

Vasoconstriction Studies

Basilar Artery

Our goal was to test the effect of dantrolene and nimodipine on the normal physiological contractile response to the potent vasoconstrictors, serotonin (5-HT) and endothelin-1 (ET-1). These two agents have been chosen because they cause reproducible contraction of cerebral arteries ex vivo^{8, 19–21} and, most importantly, because they have both been implicated in the pathogenesis of cerebrovascular vasospasm. ^{6–8, 22, 23} The solvents for dantrolene and nimodipine are DMSO and ethanol (ETOH), respectively. We tested whether these solvents effect the normal contractile response, at final maximum concentration of 0.33% (DMSO) and 0.03% ETOH. Figure 2 shows that neither solvent nor the combination of the two had a significant effect for either 5-HT or ET-1, at the maximum solvent doses used for dantrolene and nimodipine.

Figure 2.

Vehicle effect on concentration-response curves. A, Cumulative serotonin (5-hydroxytryptamine, 5-HT) concentration-response curve of basilar artery constriction normalized to 100 mM KCl induced-contraction, control arterial segments are compared to segments (n=5) incubated with 0.33% DMSO, 0.03% ethanol (ETOH) or both (n=6, for all vehicle groups), there are no statistical differences among curves. B, Cumulative endothelin-1 (ET-1) concentration-response curve of basilar artery constriction normalized to 100 mM KCl induced-contraction, control arterial segments (n=5) are compared to segments incubated with 0.33% DMSO, 0.03% ethanol (ETOH) or both (n=6, for all vehicle groups), there are no statistical differences among curves.

Next, isolated basilar artery was challenged with cumulative concentrations of 5-HT (1nM – 10μM) or ET-1 (10pM – 30nM). The addition of either 5-HT (Figure 3) or ET-1 (Figure 4) caused a concentration-dependent increase in contractile tone, reaching approximately 100%–120% of the contraction previously induced in the same preparations by high K⁺ (100 mM). Preincubation with 10 or 30 μM dantrolene did not significantly affect the contractile response to 5-HT, whereas 100 μM dantrolene produced a significant inhibition (Figure 3A); both the 50% effective concentration (EC₅₀) and the maximum contraction to 5-HT (E_max) were significantly changed (p<0.01).

Figure 3.

Serotonin (5-hydroxytryptamine, 5-HT) concentration-response curves in basilar artery. A, Cumulative concentration-response curve of basilar artery constriction normalized to 100 mM KCl induced-contraction, control arterial segments are compared to segments incubated with increasing concentration of dantrolene (10–100μM) (*EC₅₀ and E_max compared to control). B, The effect of dantrolene (10μM) plus nimodipine (30nM) (*EC₅₀ and E_max of nimodipine compared to control). C, The effect of dantrolene (30μM) plus nimodipine (30nM) (*EC₅₀ of dantrolene+nimodipine compared to nimodipine or dantrolene alone). D, The effect of dantrolene (100μM) plus nimodipine (30nM) (*EC₅₀ and E_max of dantrolene+nimodipine compared to nimodipine or dantrolene alone). The effective concentration of 5-HT that gave 50% contraction (EC₅₀) for each treatment is listed below the respective graph.

Figure 4.

Endothelin-1 (ET-1) concentration-response curves in basilar artery. A, Cumulative concentration-response curve of basilar artery constriction normalized to 100 mM KCl induced-contraction, control arterial segments are compared to segments incubated with increasing concentration of dantrolene (10–100μM). B, The effect of dantrolene (10μM) plus nimodipine (30nM) (*EC₅₀ of nimodipine compared to control). C, The effect of dantrolene (30μM) plus nimodipine (30nM) shows no interaction. D, The effect of dantrolene (100μM) plus nimodipine (30nM) shows no interaction. The effective concentration of ET-1 that gave 50% contraction (EC₅₀) for each treatment is listed below the respective graph.

In contrast to dantrolene, preincubation with nimodipine (30nM) alone did not affect the EC₅₀ of 5-HT, but significantly reduced E_max by 27% (p<0.01, Figure 3B).

The effect of adding low concentrations of dantrolene (10 or 30 μM) with nimodipine (30 nM) was studied on the effect of 5-HT-induced constriction. The maximal inhibition was not greater than with nimodipine (30 nM) alone (Figure 3B, C); however, dantrolene (30μM) did reduce the EC₅₀, p=0.01 (Figure 3C). The highest dantrolene dose (100μM) with nimodipine (30nM) caused a further 10-fold shift to the right in the EC₅₀ of 5-HT and a further reduction of the maximal contraction to 46% (p<0.01; Figure 3D).

When tested on ET-1, the effects of dantrolene and nimodipine, either individually or in combination, were barely detectable (Figure 4).

Femoral Artery

To determine whether the effects of dantrolene and nimodipine on 5-HT constriction were specific to the basilar artery, we investigated the response to phenylephrine (PE), 5-HT and ET-1 in femoral artery. Rat basilar artery is known to not constrict in response to alpha adrenergic stimulation.^{24, 25}

5-HT (1nM – 10μM), PE (10nM – 100μM) or ET-1 (100pM – 100nM) induced concentration-dependent increase in contractile tone of isolated femoral artery. The potency (EC₅₀) of 5-HT was similar to that estimated in basilar artery, but the efficacy (E_max) was higher, reaching 120–140% of the contraction previously induced in the same preparations by high K⁺ (100 mM) (Figure 5). In contrast, ET-1 was about 10-fold less potent in femoral than in basilar artery (P<0.01) (data not shown).

Figure 5.

Serotonin (5-hydroxytryptamine, 5-HT) concentration-response curves in femoral artery. A, Cumulative concentration-response curve of basilar artery constriction normalized to 100 mM KCl induced-contraction, control arterial segments are compared to segments incubated with increasing concentration of dantrolene from 10–100μM (*EC₅₀ of 30 and 100 μM compared to control). B, The effect of dantrolene (10μM) plus nimodipine (30nM) (*EC₅₀ of nimodipine compared to control). C, The effect of dantrolene (30μM) plus nimodipine (30nM) (*EC₅₀ of dantrolene+nimodipine compared to nimodipine or dantrolene alone). D, The effect of dantrolene (100μM) plus nimodipine (30nM) (*EC₅₀ and E_max of dantrolene+nimodipine compared to nimodipine or dantrolene alone). The effective concentration of 5-HT that gave 50% contraction (EC₅₀) for each treatment is listed below the respective graph.

Preincubation with dantrolene, 30 and 100 μM but not 10 μM, produced a significant (p<0.001) inhibition of the contractile response to 5-HT, with a parallel rightward shift of the concentration-constriction curve, without a significant change of E_max (Figure 5).

Preincubation with nimodipine alone (30nM) produced a parallel rightward shift in the response to 5-HT (P< 0.01) similar in amplitude to that observed with dantrolene 100 μM (Figure 5).

Dantrolene 30 and 100 μM + nimodipine (30 nM) together produced an additive inhibitory effect on the contractile response to 5-HT (Figure 5).

When tested individually on PE-induced constriction, 100μM dantrolene or 30 nM nimodipine produced a modest inhibition (Figure 6); dantrolene (at all concentrations tested) + nimodipine (30nM) did not produce an inhibitory effect stronger than nimodipine (30nM) alone, except for a further slight depression of E_max (Figure 6).

Figure 6.

Phenylephrine (PE) concentration-response curves in femoral artery. A, Cumulative concentration-response curve of basilar artery constriction normalized to 100 mM KCl induced-contraction, control arterial segments are compared to segments incubated with increasing concentration of dantrolene from 10–100μM (*EC₅₀ of 100μM compared to control). B, The effect of dantrolene (10μM) plus nimodipine (30nM) (*EC₅₀ and E_max of nimodipine compared to control; *E_max of dantrolene+nimodipine compared to nimodipine or dantrolene alone). C, The effect of dantrolene (30μM) plus nimodipine (30nM) (*E_max of dantrolene+nimodipine compared to nimodipine or dantrolene alone). D, The effect of dantrolene (100μM) plus nimodipine (30nM) (*E_max of dantrolene+nimodipine compared to nimodipine or dantrolene alone). The effective concentration of PE that gave 50% contraction (EC₅₀) for each treatment is listed below the respective graph.

As for basilar artery, the effects of dantrolene and nimodipine, either individually or in combination, on ET-1 induced constriction of the femoral artery, were barely detectable (data not shown).

Discussion

The main finding of this study is that dantrolene + nimodipine exerts a synergistic inhibition on 5-HT-induced vasoconstriction. This synergy was observed in either basilar or femoral artery, though with pharmacological differences (non-parallel shift of the concentration-contraction curve in basilar with a decrease in E_max versus a parallel shift in femoral without change in E_max) that remain to be elucidated. In vascular smooth muscle cells, 5-HT potentially stimulates 5-HT_1B/D receptors coupled to G_i (adenylate cyclase inhibition and cAMP decrease) and/or 5-HT_2A coupled to G_q (phospholipase C (PLC) stimulation with IP₃ and DAG generation).^{21, 26, 27} The vasoconstrictive effect of 5-HT in the basilar is mediated only by 5-HT_1B/D, whereas in the femoral artery it is mediated by both 5-HT_1B/D and 5-HT_2A;^{26, 28, 29} this difference might explain why E_max is variably altered in the two vascular beds. If so, the mechanism of this difference is unclear.

5-HT exerts multiple effects on membrane cation channels, leading to membrane depolarization and subsequent activation of VDCC.^{30, 31} Following 5-HT stimulation of rat basilar artery we therefore would expect an increase in cytosolic Ca²⁺ concentration due to opening of VDCC, IP₃-induced Ca ²⁺ release from intracellular stores and Ca²⁺-induced Ca²⁺ release (CICR) through RyR. Nimodipine blocks Ca²⁺ influx through VDCC.¹⁰ Dantrolene blocks CICR by binding to RyR3, although it does not directly block IP₃-induced-Ca²⁺ release in smooth muscle, it may block CICR due to cytosolic increases in Ca²⁺ from IP₃.³² Worthy of note, we found that smooth muscle cells in rat basilar and femoral arteries express only the RyR3 isoform, which is an isoform inhibited by dantrolene.³ Albeit excitation-contraction coupling following 5-HT has not yet been studied in detail in rat basilar artery, we speculate that simultaneous inhibition of Ca²⁺ influx via nimodipine and of CICR via dantrolene may affect sequential steps leading to synergistic inhibition of vasoconstriction. Our data show, not only that such an inhibitory synergism occurs, but also that it is most pronounced in the cerebral vasculature.

Interestingly, dantrolene had no effect on ET-1 vasoconstriction and a modest inhibition on PE-induced vasoconstriction; and there was no potentiation of either vasoconstrictor with nimodipine. The reason for this peculiar sensitivity of 5-HT-induced vasoconstriction, relative to ET-1 and PE is at present unclear. Ca²⁺-independent modulation of the contractile apparatus may differ among various vasoactive agents.³³ ET-1 induces vasoconstriction through its ET_A receptor. ET_AR is a G-protein-coupled receptor, which modulates G_q and G₁₂.^{34, 35} ET-1 vasoconstriction is dependent on Ca²⁺; however it does not activate VDCC. It mobilizes Ca²⁺ through non-voltage dependent channels such as store-operated Ca²⁺ channels (SOCCs) and nonselective cation channels (NSCCs) each regulated by G_q and G₁₂, respectively.³⁴ Furthermore, ET-1 can increase myosin phosphorylation and the force of contraction through activation of the rho-rho kinase pathway in vascular smooth muscle and this mechanism is independent of cytosolic Ca²⁺ concentration;^{36, 37} the activation of this pathway by ET-1 may be particularly effective.^{38, 39} PE activates vasoconstriction through alpha-1 adrenergic receptors, which are G-protein-coupled receptors that modulate G_q, which in turns activates PLC and IP₃ mediated Ca²⁺ release. This mechanism is similar to 5-HT_2A receptors, which play a role in mediating the vasoconstrictive properties of 5-HT in the femoral artery, however there was no potentiation of dantrolene with nimodipine in response to PE in contrast to the EC₅₀ for 5-HT in the femoral artery. Possible explanations for this difference, although unexplored, are potential differential regulation of cyclic ADP ribose (cADPR) and/or nicotinic acid adenine dinucleotide phosphate (NAADP) by 5-HT receptors. Both cADPR and NAADP are secondary signaling molecules that can stimulate ryanodine receptor Ca²⁺ release and potentiate CICR.^40–42

The therapeutic significance of VDCC blockade has been well established for decades, because selective drugs, i.e. calcium antagonists, have been available.⁴³ In contrast, the therapeutic significance of RyR blockade is largely unknown, because, except for dantrolene, available RyR ligands are therapeutically unusable due to toxicity, non-specific blockade of RyR1, 2 or 3, or pharmacologically unsuitable (irreversible binding to RyR, opening/blocking mixed activity of RyR).⁴⁴ Furthermore, mutant mice lacking RyR3 do not show a dramatic vascular phenotype,⁴⁵ except for a modest increase in myogenic tone,⁴⁶ suggesting that pharmacological blockade of these channels might not produce detectable physiologic changes in vascular tone under most circumstances, despite ex vivo evidence that RYR regulates arterial diameter.⁴⁷ Activation of RyR located in proximity to the plasma membrane by extracellular Ca²⁺ influx through VDCC produces CICR (Ca²⁺sparks), which in turn modulates plasma membrane excitability (e.g., hyperpolarization) through activation of Ca²⁺-dependent K+ channels; hyperpolarization then produces vasodilatation, by decreasing the probability of opening of VDCC.⁴⁸ It therefore appears that regulation of K+ channels expression and function may profoundly affect the impact of CICR on vascular tone. Consistent with this view is evidence that vascular Ca²⁺-dependent K+ channels with reduced function are associated with greater depolarization and increased vascular tone.⁴⁹ At variance with CICR occurring close to plasma membrane, CICR from RyR at SR locations further away from the plasma membrane could propagate Ca²⁺ waves that induce vasoconstriction.⁵⁰

Although efficacy in a pathophysiological model was not tested in this paper, it is important to note that for SAH in animal models and humans, 5-HT, a potent vasoconstrictor, is released from activated platelets. After SAH evidence supports an upregulation of 5-HT1B receptors and enhanced vasoconstriction in response to 5-HT with significant reductions in cerebral blood flow.^51–53 Other drug treatments, particularly calcium channel blockers, have been proposed to prevent and/or treat the vasospasm of reversible cerebral vasoconstrictive syndromes (e.g., Call-Fleming)⁵⁴ and SAH⁵⁵ (despite lack of evidence of vasospastic relaxation in humans). However, one potential limitation these drugs are systemic vasodilatation that may lead to hypotension and further reduce cerebral perfusion pressure and cerebral blood flow.⁵⁶ Dantrolene at IV doses of 2.5mg/kg (10μM), lower than the most effective dose tested here, does not appear to cause hypotension, even in the presence of nimodipine and may have a vasorelaxation effect in humans.⁵⁷

In conclusion, dantrolene has synergistic effects with nimodipine against 5-HT-induced vasoconstriction in isolated cerebral arteries. The association dantrolene-nimodipine might be a promising treatment for reducing SAH-related vasospasm or other serotonin-related vasospastic syndromes, such as Call-Fleming syndrome, but requires testing in a pathophysiological model.