A novel saline-soluble, rapidly-metabolized RyR1 inhibitor rescues volatile anesthesia-induced death and environmental heat stroke in a mouse model relevant to malignant hyperthermia

Mutations in the type 1 ryanodine receptor (RyR1), a Ca2+ release channel in skeletal muscle, hyperactivate the channel to cause malignant hyperthermia (MH) and are implicated in severe heat stroke. Dantrolene, the only approved drug for MH, has the disadvantages of having very poor water solubility and long plasma half-life. We show here that a novel RyR1-selective inhibitor, 6,7-(methylenedioxy)-1-octyl-4-quinolone-3-carboxylic acid (Compound 1), effectively rescues MH and heat stroke in several mouse models relevant to MH. Compound 1 reduced resting intracellular Ca2+, inhibited halothane-induced Ca2+ release, suppressed caffeine-induced contracture in skeletal muscle, reduced sarcolemmal cation influx, and prevented or reversed the fulminant MH crisis by isoflurane anesthesia and heat stroke by environmental heat stress. Notably, Compound 1 has great advantages of better water solubility and rapid clearance in vivo over dantrolene. Compound 1 has the potential to be a promising new candidate for effective treatment of patients carrying RyR1 mutations.


Introduction
The type 1 ryanodine receptor (RYR1; MIM# 180901) is a Ca 2+ -release channel in the sarcoplasmic reticulum (SR) of skeletal muscle that plays a central role in muscle contraction 1,2 .
Genetic mutations in the RYR1 gene are associated with malignant hyperthermia (MH; MIM# 145600) 7 . MH is a life-threatening disorder characterized by skeletal muscle rigidity and elevated body temperature in response to halogenated anesthetics such as halothane or isoflurane 8 . MH mutations are thought to hyperactivate CICR, which causes massive Ca 2+ release by halogenated anesthetics under resting conditions 6,9,10 . Heat stroke is a medical emergency with a high body temperature and altered mental status, which is triggered by exercise or environmental heat stress 11,12 . It has been reported that MH mutations in the RYR1 gene are also implicated in some heat stroke [13][14][15] .
The only drug approved for ameliorating the symptoms of MH is dantrolene 16,17 .
Dantrolene was first developed in 1960s as a muscle relaxant 18 and was later shown to prevent Ca 2+ release by direct interaction with RyR1 19,20 . However, dantrolene has several disadvantages for clinical use: the main disadvantage is its poor water solubility which makes rapid preparation difficult in emergency situations 21 . In addition, dantrolene has long plasma half-life (t 1/2 ~12 h), which causes long-lasting side effects such as muscle weakness 22 . To date, no alternative drugs improving these disadvantages have been developed over 50 years since the first discovery. 4 We have recently identified oxolinic acid as a novel RyR1-selective inhibitor with better water solubility using efficient high-throughput screening platform for RyR1 inhibitors 23,24 . We synthesized a series of oxolinic acid derivatives by structural development, and successfully identified 6,7-(methylenedioxy)-1-octyl-4-quinolone-3-carboxylic acid (Compound 1) with extremely high potency which is comparable to dantrolene in in vitro study 25 .
In this study, we tested whether Compound 1 has therapeutic effects in treating the MH crisis by creating a novel MH mouse model carrying RYR1-p. R2509C, a corresponding mutation (p.R2508C) in humans 26,27 . Evaluations were also done with two other MH mouse models, RYR1-p.R163C 28 and RYR1-p.G2435R 29 . We found that Compound 1 effectively prevents and rescues mice from fulminant MH crisis triggered by isoflurane anesthesia. The drug also treats mice with heat stroke caused by environmental heat stress. A notable finding is that Compound 1 is rapidly metabolized in mice. Our results provide crucial evidence for Compound 1 as a RyR1 inhibitor that may prove to be clinically-useful.

Generation of novel MH model mice carrying RYR1-p.R2509C mutation
A MH mouse model carrying p.R2509C mutation in the RYR1 gene (RYR1-p.R2509C) was created using CRISPR/Cas9 system (see Materials and Methods, Fig.   1a). An additional Xba I site was introduced to screen the genotype by PCR-RFLP. The homozygous (Hom) RYR1-p.R2509C mice died in late embryonic stage, but heterozygous (Het) RYR1-p.R2509C mice grew normally and were as fertile as wild type (Wt) (Fig. 1b).
It has previously been shown when MH susceptible mice, swine or humans with mutations in the RYR1 gene are exposed to volatile anesthetics, it causes a MH crisis and if untreated can cause death [28][29][30][31][32][33] . As with other animal models, when anesthetized by isoflurane,

RYR1-p.R2509C
Het mice rapidly increased their rectal temperature to over 42°C (Fig. 1c), exhibited muscle rigidity (Fig. 1d) and died after 65±31 min (mean±SD, n=14) of exposure ( Fig. 1e). In contrast, Wt mice showed no elevation in rectal temperature in response to isoflurane. Time from start of anesthesia to death varied largely among individual mice and took significantly longer in females (87±33 min, n=5) than males (53±24 min, n=9) (p<0.05) ( Fig. 1f). Interestingly, time from rectal temperature of 39°C to death was similar among all individuals, and no sex difference was observed (p=0.19) (Fig. 1g). These findings suggest that RYR1-p.R2509C mice can reproduce human MH symptoms in response to exposure to volatile anesthetics and thus are a useful model for test of Compound 1.
Dantrolene is known to suppress twitch and tetanic contractions, which causes muscle weakness as a major side effect of the drug 36,37 . We therefore tested whether Compound 1 affects twitch and tetanic contractions. Untreated soleus muscles from Het RYR1-p.R2509C mice showed similar twitch and tetanic contraction force to that recorded from Wt soleus muscles (Extended Data Fig. 2 and Fig. 2e). 3 µM Compound 1 reduced both twitch and tetanic contractions. The effect was stronger with twitch (72% inhibition) than with tetanic (35% inhibition at 100 Hz) contractions, which is similar to dantrolene 36 .

In vivo isoflurane challenge
We next tested whether Compound 1 can prevent or treat MH episodes in RYR1-p.R2509C Het mice in vivo. Since Compound 1 is a free acid having carboxyl group ( Fig. 2a), and is virtually water insoluble, we prepared its sodium salt, which neutralizes the carboxyl group and significantly increases water solubility. Compound 1 sodium salt also exhibited good solubility (849±58 μ g/mL, n=4) in normal saline (0.9% NaCl) ( Table 1).
Dantrolene sodium salt, in contrast, was hardly soluble in normal saline (25.7±2.1 μ g/mL, n=4) under the same conditions, which is in accord with the previous study 38 (Extended Data Table 1).
We first tested the preventive effect of Compound 1 on MH episodes in response to isoflurane exposure. Doses of either 3 mg/kg or 10 mg/kg of Compound 1 solubilized in normal saline were administered i.p. 10 min before isoflurane challenge. The 3 mg/kg dose did not prevent rise in rectal temperature and only 1 out of 6 mice survived ( Fig. 3a-3c). Time to death for mice treated at 3 mg/kg (47±15 min, n=5) was not significantly different than that for control mice (53±24 min, n=9). However, with the 10 mg/kg dose, Compound 1 successfully prevented any rise in rectal temperature and all the mice survived 90 min after inducing anesthesia ( Fig. 3a-3c).
We next tested whether Compound 1 could rescue mice from MH episodes. Het mice were anesthetized with isoflurane and the drug was administered i.p. when rectal temperature reached 39°C. The 3 mg/kg dose was able to stop the rise in rectal temperature (-0.03±0.81°C at 10 min after administration, compared to 1.79±0.82°C in controls), but rectal temperature started to rise again ~30 min after administration of the drug (Fig. 3d, 3e).
The 10 mg/kg dose decreased body temperature (-0.64±0.29°C at 10 min after administration) and maintained it at that level until the experiment was terminated at 60 min. At that time 50% of the mice in the 3 mg/kg group and all the mice in the 10 mg/kg survived at the 60 min time point (Fig. 3f). 8

Environmental heat stress
It has been shown that environmental heat stress causes an increase in body temperature and death of existing MH model mice 28,29,32,33 . We therefore tested whether  Fig. 3a, 3b). The time from start of heat stress to death largely varied among individual mice (74±21 min, n=13), but the variation became smaller for time to death after reaching body temperatures of 38°C (36±13 min, n=13) or 39°C (22±8 min, n=13) (Extended Data Fig. 3c).
Using this model, we initially tested preventive effect of Compound 1 on the heat stroke. Pretreatment of mice with Compound 1 (10 mg/kg) slowed the rate of temperature rise ( Fig. 4a), but none of the treated mice (0 out of 4) survived 120 min of heat stress (Fig. 4b).
We next tested whether Compound 1 could rescue mice from environmental heat stroke. The mice were incubated at the test chamber and the drug was administered i.p. when rectal temperature reached 38°C. All the untreated mice died within 60 min (time to death, 36±13 min, n=13). Compound 1 at the 3 mg/kg dose suppressed rise in the body temperature ( Fig. 4d, 4e) and 5 out of 6 mice survived 60 min after administration (Fig. 4f). Compound 1 at the 10 mg/kg dose showed more pronounced effect: it transiently dropped the body temperature and all the mice (6 out of 6) survived for 60 min after administration. The rescue effect of the drug was also observed when the drug was administered when body temperature reached 39°C, after which most untreated mice died within 30 min (time to death, 22±8 min, 9 n=13). When given at this time, Compound 1 (10 mg/kg) strongly suppressed the continued rise in the body temperature and 5 out of 6 mice survived 60 min after its administration (Extended Data Fig. 4).

Sarcolemmal cation influx in skeletal muscle
It has been shown that sarcolemmal cation influx is accelerated in skeletal muscle of MH  (Fig. 5c, 5d).
Compound 1 reduced Mn 2+ quench in Hom RYR1-p.G2435R fibers in a dose-dependent manner and the Mn 2+ quench rate at the 10 µM dose was the same as untreated Wt fibers.

Pharmacokinetics
In in vivo experiments, we observed transient effects of Compound 1 in our isoflurane challenge and heat challenge experiments (see Fig. 3, 4). This implies fast metabolism or excretion of the drug in mice. We therefore measured plasma concentration of Compound 1 after i.p. administration. Compound 1 rapidly declined from mouse blood following first order kinetics with t 1/2 of ~10 min at both the 3 mg/kg and 10 mg/kg dose (Fig. 6a). Since Compound 1 has hydrophobic moiety of long alkyl chain of 8 carbons (see Fig. 2), it can be metabolized in the liver. In vitro drug metabolism assay using liver microsomes demonstrated that Compound 1 was reduced to 20% and 5% of the original amount at 10 min and 60 min, respectively, after incubation of mouse liver microsomes (Fig. 6b). Compound 1 was metabolized more slowly by human liver microsomes; it was reduced to 20% at 60 min.
Duration of the effect of Compound 1 in mice was also evaluated measuring in vivo muscle strength, since the drug inhibited twitch and tetanic contractions of the isolated skeletal muscle (see Fig. 2). Grip strength tests demonstrated that muscle strength of Wt mice was reduced by 15% 10 min after administration of 10 mg/kg Compound 1, and that this deficit was almost completely recovered at 60 min, suggesting short duration of the medicinal effect in vivo (Fig. 6c).
The above results indicate that transient effects of Compound 1 in the isoflurane challenge and heat challenge experiments is due to rapid metabolism of the drug in mice. We therefore tested whether additional administrations of Compound 1 could prevent or treat mice from the crisis. Repeated administrations of the 3 mg/kg dose successfully reduced the rectal temperature and prevented death during isoflurane-induced MH crisis (Extended Data   Fig. 6).

Discussion
MH is a life-threatening disorder in response to halogenated anesthetics such as halothane or isoflurane 8  mice grew normally and were as fertile as Wt (Fig. 1). Het mice showed a halothane-induced increase in [Ca 2+ ] i and caffeine-induced contracture in skeletal muscle (Fig. 2). In addition, RYR1-p.R2509C mice exhibited a MH crisis when exposed to isoflurane anesthesia (Fig. 3) and heat stroke when exposed to environmental heat stress (Fig. 4). These properties are common to those seen in the other model mice carrying severe MH mutations in the RYR1 gene, e.g., p.R163C 28 and p.Y524S 32 and p.G2435R 29 . Thus, the RYR1-p.R2509C mice are a useful model for MH research.
A major problem of dantrolene in clinical use is its poor aqueous solubility 21 . First, dantrolene must be solubilized with sterile water and injected or infused by the intravenous route, since it is virtually insoluble in normal saline. Second, dantrolene is prepared at a diluted concentration (0.33 mg/mL) due to its low water solubility, which makes rapid preparation difficult in emergency situations. The second problem has recently been solved by nanocrystalline suspension of dantrolene (Ryanodex ® ) which can be solubilized at 50 mg/mL 41 . We found that Compound 1 exhibits good solubility (~1 mg/mL) in normal saline, which is threefold higher compared to dantrolene in water (Extended Data Table 1). This enables continuous administration of the drug by infusion using the intravenous route and expands usage of the drug before and after surgery for prevention of MH and postoperative MH 42 .
Another drawback of dantrolene in clinical use is its long plasma half-life (10-12 h in humans after intravenous injection) 43 . Therefore, it is difficult to control the appropriate plasma concentration and reduce side effects such as muscle weakness 21 . Pharmacokinetic analysis revealed that Compound 1 is rapidly metabolized in mice with plasma half-life of ~10 min, probably by metabolism by the liver (Fig. 6). Indeed, because of this a single dose of Compound 1 showed only a transient effect during MH episodes (Fig. 3), and 13 environmental heat stress (Fig. 4) and the muscle weakness was transient (Fig. 6). Thus, Compound 1 may be beneficial for a temporary treatment of MH episodes. The slower metabolism of the drug with human liver microsomes (Fig. 6) suggests longer plasma half-life in humans compared to mice. Pharmacokinetic analysis in humans awaits completion.
Heat stroke is a life-threatening condition clinically diagnosed as a severe elevation in body temperature with central nervous system dysfunction 11,12 . It has been reported that MH mutations in the RYR1 gene is implicated in some heat stroke [13][14][15] . Indeed, dantrolene has therapeutic effects on some patients with severe exertional heat stroke 44 . Compound 1 effectively prevents and treats fulminant heat stroke caused by environmental heat stress in RYR1-p.R2509C model mice (Fig. 4). Since Compound 1 is soluble in normal saline, it has the potential to be administered by continuous infusion which could be necessary for treatment of severe heat stroke.
Dantrolene is also used for the treatment of neuroleptic malignant syndrome, a life-threatening neurologic emergency associated with the use of antipsychotic agents 45 as well as overdose of 2,4-dinitrophenol (a prohibited weight loss agent that interrupts ATP synthesis and causes hyperthermia) 46 . Since Compound 1 has similar effects in preventing and treating MH and heat stroke, it might also be a potential candidate for treatment of these emergency situations.

Materials
Compound 1 was synthesized as described previously 25

Determination of thermodynamic solubility
Test compound (2 mg

Generation of mutant mouse by gene editing using CRISPR-Cas9 system
Mouse genomic sequence within upstream-and downstream-50 bps each to that corresponding to human mutation site was searched by CRISPR design tool (http://crispr.mit/edu) to select PAM and its consequent guide sequence with high specificity.
CAATGCCATACACTCGGTCCAGG (bold: PAM sequence) was selected for crRNA synthesis. A single-stranded donor oligonucleotide carrying Xba I site (underlined), CCGAAGATGTCTGCATCCTTCGTGCCTGACCAGGCATCCATGGTGCTTTTTCTAGA CTGCGTGTATGGCATTGAGAACCAGGACTTCTTGCTGCATGTGCTGGATGTGGG was chemically synthesized for homologous recombination (Fig. 1). A RYR1-genomic fragment containing PAM and the consequent guide sequence were amplified by PCR from wild-type mouse genomic DNA. RNP complex formed by mixing of tracrRNA and crRNA with recombinant Cas9 protein, was added to the PCR fragment and incubated for 1 hour.
Cleavage of PCR fragment into two fragments was confirmed by agarose electrophoresis.
RNP complex and single-stranded donor oligonucleotide were introduced to mouse one-cell-stage zygotes by electroporation. The electroporated zygotes were transferred to the oviducts of pseudopregnant females. Genomic DNA was prepared from newborns' tail and knock-in mice were screened by PCR-RFLP method using a pair of primers (R2509C_Forward: GCGTCAGGAATGGTTAGAATAGAT; R2509C_Reverse: TTGTCTGATGCTAGGTAGAAGGTG) and Xba I digestion (Fig. 1). Crossing obtained knock-in founders with wild-type mice to get F1 generations were done by standard protocol.

Preparation of single flexor digitorum brevis (FDB) fibers
Mice were deeply anesthetized with intraperitoneal (i.

Measurement of muscle force in vitro
Muscle contraction measurement was assessed using intact soleus muscle as described either 10 min before the start of isoflurane anesthesia for test of the preventive effect or when the rectal temperature was achieved 39°C for treatment of MH episodes.

In vivo heat stress challenge
To investigate heat stress responses, mice were anesthetized (i.p.) with an anesthetic mixture (0.75 mg/kg medetomidine, 4 mg/kg midazolam and 5 mg/kg butorphanol) and transferred to a test chamber prewarmed at 35°C. Rectal temperature was measured continuously during the heat stress challenge (up to 90 min) or until the time of fulminant heat stroke. Compound 1 was administered (i.p.) either 10 min before the start of heat stress challenge for test of the preventive effect or when the rectal temperature was achieved 38 or 39°C for treatment of heat stroke.

Ca 2+ determinations in intact muscle fibers using ion-specific microelectrodes
Ca 2+ ion specific microelectrode recordings were performed as described previously 39,48 . For in vivo measurements, Wt and RYR1-p.R163C Het mice were anesthetized with 100 mg/kg ketamine and 5 mg/kg xylazine (i.p.) and placed on a 37 o C heating pad and their temperature was monitored. The vastus lateralis was exposed surgically after which its muscle fibers were

Mn 2+ quench assay
Assays were carried out as described previously 39

21
Data are presented as the mean ± SD. Statistical analysis was performed using Prism 8 (GraphPad Software, Inc., La Jolla, USA). Student's t-test was used for comparisons between two groups. One-way analysis of variance (ANOVA), followed by Dunnett's test, was performed to compare multiple groups. Statistical significance was defined as p < 0.05.