Malignant hyperthermia-associated mutations in the S2-S3 cytoplasmic loop of type 1 ryanodine receptor calcium channel impair calcium-dependent inactivation

Abstract

Channel activities of skeletal muscle ryanodine receptor (RyR1) are activated by micromolar Ca²⁺ and inactivated by higher (∼1 mM) Ca²⁺. To gain insight into a mechanism underlying Ca²⁺-dependent inactivation of RyR1 and its relationship with skeletal muscle diseases, we constructed nine recombinant RyR1 mutants carrying malignant hyperthermia or centronuclear myopathy-associated mutations and determined RyR1 channel activities by [³H]ryanodine binding assay. These mutations are localized in or near the RyR1 domains which are responsible for Ca²⁺-dependent inactivation of RyR1. Four RyR1 mutations (F4732D, G4733E, R4736W, and R4736Q) in the cytoplasmic loop between the S2 and S3 transmembrane segments (S2–S3 loop) greatly reduced Ca²⁺-dependent channel inactivation. Activities of these mutant channels were suppressed at 10–100 μM Ca²⁺, and the suppressions were relieved by 1 mM Mg²⁺. The Ca²⁺- and Mg²⁺-dependent regulation of S2–S3 loop RyR1 mutants are similar to those of the cardiac isoform of RyR (RyR2) rather than wild-type RyR1. Two mutations (T4825I and H4832Y) in the S4–S5 cytoplasmic loop increased Ca²⁺ affinities for channel activation and decreased Ca²⁺ affinities for inactivation, but impairment of Ca²⁺-dependent inactivation was not as prominent as those of S2–S3 loop mutants. Three mutations (T4082M, S4113L, and N4120Y) in the EF-hand domain showed essentially the same Ca²⁺-dependent channel regulation as that of wild-type RyR1. The results suggest that nine RyR1 mutants associated with skeletal muscle diseases were differently regulated by Ca²⁺ and Mg²⁺. Four malignant hyperthermia-associated RyR1 mutations in the S2–S3 loop conferred RyR2-type Ca²⁺- and Mg²⁺-dependent channel regulation.

ryanodine receptor calcium release channels (RyRs) play pivotal roles in releasing Ca²⁺ from the sarcoplasmic reticulum, an intracellular membrane compartment, during action potentials in skeletal and cardiac muscle. RyRs are homotetramers of 565-kDa polypeptides and their channel activities are regulated by a number of physiological molecules and proteins, which include Ca²⁺, Mg²⁺, ATP, protein kinases, and calmodulin (4, 16, 21). Intracellular Ca²⁺ activates both skeletal (RyR1) and cardiac (RyR2) isoforms of RyR at micromolar concentrations. Both isoforms are inhibited by millimolar Ca²⁺, but inhibition of RyR2 requires ∼10 times higher concentrations of Ca²⁺ than RyR1. Generation of recombinant RyR1/RyR2 chimera channels allowed for the identification of regions involved in isoform-specific Ca²⁺-dependent channel inactivation (7, 8, 24). We found that two distinct domains in the C-terminal ∼1,000 amino acids of RyRs are involved in isoform-specific Ca²⁺-dependent channel inactivation (14). One domain contained two EF-hand-type Ca²⁺ binding motifs [RyR1 amino acids (aa) 4065–4298], and the other domain included the second transmembrane region (RyR1 aa 4630–4688). Replacing these two regions in RyR1 with the corresponding RyR2 sequences resulted in RyR2-type Ca²⁺-dependent inactivation (14).

Missense mutations in the human RYR1 gene have been found to be associated with skeletal myopathies including malignant hyperthermia (MH), central core disease (CCD), and centronuclear myopathy (CNM) (17, 19, 20, 27, 28). MH is a life-threatening condition in which anesthetic reagents such as halothane cause abnormal sarcoplasmic reticulum Ca²⁺ release and increase muscle oxidative metabolism, resulting in an uncontrollable increase in body temperature. CCD is a congenital myopathy that causes gradually worsening muscle weakness after birth. Skeletal muscle cells of CCD patients show damaged areas, called cores, that are missing mitochondria and oxidative enzymes. A number of patients with CCD have a risk for MH susceptibility as well. CNM is another congenital myopathy in which there are a prominent number of myofibers with centrally located nuclei (19, 20). Phenotypes of CNM observed in muscle biopsies sometimes overlap with other congenital myopathies. In total, more than 200 point mutations in the RYR1 gene have been found to be linked to patients with MH or congenital myopathy (17, 27, 28).

In the present study, we hypothesized that nine skeletal disease-associated mutations, located in RyR1 regions (or flanking regions), previously shown to be involved in Ca²⁺-dependent channel inactivation (14) alter Ca²⁺-dependent regulation of RyR1 channel activity, resulting in abnormal Ca²⁺ homeostasis. We constructed three mutations in the EF-hand domain (T4082M, S4113L, and N4120Y) (18, 20, 29), four mutations in the S2–S3 cytoplasmic loop (F4732D, G4733E, R4736W, and R4736Q) (12, 23, 27, 29), and two mutations in the S4–S5 loop (T4825I and H4832Y) (1, 3). The [³H]ryanodine binding method was used to determine RyR1 channel activities, as [³H]ryanodine prefers to bind to RyRs in the open channel state (22, 33). We found that RyR1 mutations in the three different domains affected Ca²⁺- and Mg²⁺-dependent regulation of RyR1 differently.

MATERIALS AND METHODS

Materials.

[³H]Ryanodine was obtained from PerkinElmer (Waltham, MA), unlabeled ryanodine from Calbiochem (La Jolla, CA), protease inhibitors from Thermo Scientific (Waltham, MA) and Sigma-Aldrich (St. Louis, MO), and human embryonic kidney (HEK) 293 cells from American Type Culture Collection (Manassas, VA). Full-length wild-type RyR1 and RyR2 cDNAs (13, 25) were kindly provided by Dr. Gerhard Meissner of University of North Carolina at Chapel Hill and Dr. Junichi Nakai of Saitama University, Japan, respectively.

Construction of RyR cDNAs.

Full-length rabbit RyR1 and RyR2 were cloned into the mammalian expression vectors pCMV5 and pCIneo, respectively. Single and multiple base changes were introduced by Pfu-turbo polymerase-based chain reaction, by using a small DNA fragment of RyR1 as a template, mutagenic oligonucleotides, and a QuikChange site-directed mutagenesis kit (Agilent, Santa Clara, CA). The complete mutated DNA fragments amplified by PCR were confirmed by DNA sequencing and were cloned back to the original positions. Full-length expression plasmids were constructed by ligating three RyR1 fragments [ClaI-XhoI (polylinker-6598), XhoI-EcoRI (6598–11767), EcoRI-XbaI (11767-polylinker) carrying mutation] and pCMV5 expression vector (ClaI/XbaI cloning sites) as described previously (13). Sequences and numbering are consistent with previous studies (25, 34). In this paper, disease-associated RyR1 mutations in human patients were described with rabbit RyR1 sequence and numbering.

Expression of full-length RyRs in HEK293 cells.

RyR1 cDNAs were transiently expressed in HEK293 cells with FuGENE6 (Promega, Madison, WI) according to the manufacturer's instructions. The expression level of endogenous RyR protein is background level in HEK293 cells, which enables us to measure the channel activities of the recombinant wild-type and mutant RyRs by heterologous expression. Cells were maintained at 37°C with 5% CO₂ in high-glucose Dulbecco's modified eagle medium containing 10% fetal bovine serum and plated the day before transfection. For each 10-cm tissue culture dish, 3.5 μg cDNA was used, and cells were harvested 48 h after transfection. To prepare crude membrane fractions, cells were homogenized in 0.3 M sucrose, 150 mM KCl, 20 mM imidazole (pH7.0), 0.1 mM EGTA, 1 mM oxidized glutathione, and protease inhibitors. Homogenates were centrifuged for 45 min at 100,000 × g, and pellets were resuspended in the aforementioned buffer solution without EGTA and glutathione.

[³H]Ryanodine binding assay.

Ryanodine binds specifically to RyRs and is widely used as a probe for RyR channel activity because of its preferential binding to the open state of the RyR channels (22, 33). The bound amount of [³H]ryanodine at equilibrium represents the RyR channel open probability. [³H]Ryanodine binding experiments were performed with crude membrane fractions at a relatively low [³H]ryanodine concentration to determine regulation of wild-type and mutant RyRs by Ca²⁺ and Mg²⁺ (14, 40). Unless otherwise indicated, membranes were incubated with 2.5–3 nM [³H]ryanodine in 20 mM HEPES (pH 7.4), 0.15 M sucrose, 200 mM KCl, 0.3 mM oxidized glutathione (GSSG), protease inhibitors, and various Ca²⁺ and Mg²⁺ concentrations. Nonspecific binding was determined with an 800- to 1,000-fold excess of unlabeled ryanodine. After 20 h, samples were diluted with six volumes of ice-cold water and placed on Whatman GF/B filters preincubated with 2% polyethyleneimine in water. Filters were washed three times with 5-ml ice-cold 100 mM KCl, 1 mM K-PIPES (pH 7.0) solution. The radioactivity remaining within the filters was determined by liquid scintillation counting to obtain bound [³H]ryanodine.

A nearly saturating [³H]ryanodine concentration of 20 nM was used to determine wild-type and mutant RyR expression levels. Maximal binding (B_max) values of [³H]ryanodine were determined by incubating homogenates for 4 h with 20 nM [³H]ryanodine in 20 mM imidazole, pH 7.0, 0.6 M KCl, protease inhibitors, and 100 μM Ca²⁺. Nonspecific binding of [³H]ryanodine was determined in the presence of 2.5 mM EGTA and 2.5 μM unlabeled ryanodine. All experiments were performed at room temperature (22–24°C). The gain of activity for each RyR preparation was calculated by normalizing the peak values of Ca²⁺-dependent activity to B_max values.

Free Ca²⁺ concentrations and data analysis.

Free Ca²⁺ concentrations in the solutions were obtained by including 5 mM EGTA and the appropriate amounts of CaCl₂ in the absence of Mg²⁺ and ATP (pH 7.4), as determined by using the stability constants and the computer program published by Shoenmakers et al. (32) and Maxchelator (http://maxchelator.stanford.edu). Free Ca²⁺ concentrations were verified in the absence of Mg²⁺ and ATP with a Ca²⁺ selective electrode (Model 9720BNWP, Thermo Scientific). A standard curve for Ca²⁺ concentration was obtained from buffer solutions (pH 7.2–7.4) with known Ca²⁺ concentrations by using a Calcium Calibration Buffer kit (Biotium, Hayward, CA) (for <100 μM) and 1 M CaCl₂ solution stock (Sigma-Aldrich) (for ≥100 μM). To determine and compare EC₅₀ and IC₅₀ of Ca²⁺ for each mutant, we prepared the Ca²⁺/EGTA buffer solutions in bulk for each free Ca²⁺ concentration and used them for [³H]ryanodine binding experiments of each mutant RyR. We assumed the same free Ca²⁺ concentrations in the presence and absence of Mg²⁺. Free Ca²⁺ concentrations below 1 μM may be slightly affected by 1 mM Mg²⁺. However, we found that these slight changes, which were calculated by the computer program (Maxchelator), did not affect the statistical analysis of EC₅₀.

Ca²⁺-dependent activities of RyRs were normalized to the peak activity of each Ca²⁺-dependent curve, and half maximal concentrations of Ca²⁺ for activation and inactivation were derived as EC₅₀ and IC₅₀, respectively. Caffeine-dependent activities of RyRs were normalized to the activities at 10 mM caffeine in each experiment by (B − B_caf0)/(B_caf10 − B_caf0) × 100, where B is the bound amount of [³H]ryanodine, B_caf0 is bound amount at 0 mM caffeine, and B_caf10 is bound amount at 10 mM caffeine. The data were fit by the Hill equation as follows

$$B=B_{\text{peak}}\times \left\{{\left[\text{caffeine}\right]}^n/\left({\left[\text{caffeine}\right]}^n+{K_A}^n\right)\right\}$$ (1)

where B_peak is the peak activity, K_A is the activation Hill constant of caffeine, and n is a Hill coefficient.

Results are given as means ± SE. Significant differences in the data (P < 0.05) were determined by one-way ANOVA followed by Tukey's test.

RESULTS

We previously reported that replacing the EF-hand and the second transmembrane (S2) regions in RyR1 with the corresponding RyR2 sequence impaired Ca²⁺-dependent inactivation of the RyR channel activity (14). In the present study, we tested the hypothesis that nine human disease-associated RyR1 mutations located in these regions or flanking regions, namely the EF-hand domain, S2–S3 loop, and S4–S5 loop of RyR1, alter the Ca²⁺-dependent regulation of RyR1 (Fig. 1). We used [³H]ryanodine binding assay, which has been widely used to determine the regulation of RyRs by Ca²⁺ and Mg²⁺ (5, 8, 22, 33). All mutant RyR1s retained the ability to bind [³H]ryanodine and yielded a peak-to-B_max ratio of [³H]ryanodine binding similar to WT-RyR1 (Table 1).

Fig. 1.

Skeletal disease-associated mutations in RyR1. Amino acid sequences of three domains are shown. Underlines indicate EF-hand motifs. Bold letters denote MH, CCD, or CNM-linked mutation sites (https://cardiodb.org/Paralogue_Annotation/). Mutation sites characterized in this study are highlighted in gray, and mutations are shown in italic below the sequence.

Table 1.

Activities of wild-type and mutant RyRs

	Bpeak/Bmax, %
WT-RyR1	47 ± 6 (5)
T4082M	39 ± 3 (5)
S4113L	39 ± 3 (5)
N4120Y	49 ± 9 (5)
F4732D	63 ± 7 (4)
G4733E	62 ± 7 (4)
R4736W	55 ± 8 (7)
R4736Q	74 ± 7 (4)
T4825I	61 ± 9 (7)
H4832Y	57 ± 10 (4)
WT-RyR2	66 ± 4 (5)

Values are means ± SE of the number of experiments indicated in parentheses. Activities of wild-type (WT) and mutant ryanodine receptors (RyRs) were obtained by normalizing the peak values of Ca²⁺-dependent activity in Fig. 2 to the B_max values. None of the data of the mutants were significantly different from WT-RyR1 (one-way ANOVA).

Mutations in EF-hand domain.

RyRs have two EF-hand motifs that bind Ca²⁺ with an affinity of 60 μM to 3.8 mM (38, 39). Two mutations in EF1 (T4082M) and EF2 (N4120Y), and another one between EF1 and EF2 (S4113L), are associated with MH or CNM (Fig. 1, Table 2) (18, 20, 29). We constructed three recombinant RyR1s carrying each individual mutation, expressed them in HEK293 cells, and determined Ca²⁺-dependent activity changes of the mutants by using the [³H]ryanodine binding method (Fig. 2A). We observed essentially no change in their apparent Ca²⁺ affinities for channel activation (EC₅₀) compared with that of WT-RyR1 (Table 2). Affinities of Ca²⁺ for channel inactivation (IC₅₀) were slightly but not significantly decreased (Fig. 2, A and D).

Table 2.

Activation of wild-type and mutant RyRs by Ca²⁺

			EC50, μM
	Mutation	Myopathy	No Mg2+	1 mM Mg2+
WT-RyR1			1.0 ± 0.1 (5)	2.0 ± 0.1 (4)
EF-hand	T4082M	MH	0.74 ± 0.08 (5)	ND
	S4113L	CNM	1.2 ± 0.1 (5)	ND
	N4120Y	MH	0.91 ± 0.11 (5)	ND
S2–S3 loop	F4732D	MH	0.66 ± 0.13 (4)	1.4 ± 0.2 (7)
	G4733E	MH	0.73 ± 0.05 (4)	1.4 ± 0.2 (5)
	R4736W	MH	0.96 ± 0.09 (7)	1.4 ± 0.1 (5)
	R4736Q	MH	1.2 ± 0.1 (5)	1.4 ± 0.2 (6)
S4–S5 loop	T4825I	MH	0.22 ± 0.02*(6)	0.94 ± 0.05* (4)
	H4832Y	MH	0.41 ± 0.09*(5)	0.51 ± 0.03* (4)
WT-RyR2			4.2 ± 2.5 (5)	1.9 ± 0.04 (8)

Values are means ± SE of the number of experiments indicated in parentheses. EC₅₀ values for Ca²⁺-dependent activation of wild-type and mutant RyRs were determined in the absence and presence of 1 mM Mg²⁺. RyR1 mutations are associated with either malignant hyperthermia (MH) or centronuclear myopathy (CNM).

^*P < 0.05 compared with WT-RyR1 by one-way ANOVA followed by Tukey's test among RyR1 groups (wild-type and mutant RyR1s). ND, not determined.

Fig. 2.

Ca²⁺-dependent activities of MH/CNM-associated RyR1 mutants. Ca²⁺-dependent changes in the activities of wild-type (WT) and mutant RyR1 were determined by the [³H]ryanodine binding method. Activities of mutant RyR1s carrying disease-associated mutations in EF-hand domain (A), S2–S3 loop (B), and S4–S5 loop (C) are shown together with that of WT-RyR1 (dotted line, mean values). A dashed line in B represents a mean value of WT-RyR2. D: IC₅₀ values of wild-type and mutant RyRs. E: activities at 40 nM free Ca²⁺ were normalized to the peak activities for wild-type and each mutant RyR1. *P < 0.05 compared with WT-RyR1 by one-way ANOVA followed by Tukey's test among 10 sample groups (wild-type and mutant RyR1s). Data are shown as means ± SE (n = 4–7).

Mutations in the S2–S3 loop.

The S2–S3 cytoplasmic loop of RyR1 is composed of 118 amino acids and contains nine reported MH/CCD-linked point mutations (28). We determined the Ca²⁺ affinities for channel activation/inactivation of four RyR1 mutants (F4732D, G4733E, R4736W, and R4736Q) which are localized in a five amino acid stretch (Fig. 1) (12, 23, 27, 29). Note that F4732D of rabbit RyR1 corresponds to human MH mutation Y4733D. All four mutant RyR1s showed similar Ca²⁺-dependent activation and inactivation profiles of [³H]ryanodine binding (Fig. 2B). The activation phase was initially the same as WT-RyR1; however, it was followed by a plateau phase at 10–100 μM Ca²⁺ before reaching a peak value at ∼1 mM Ca²⁺. Ca²⁺ affinities for channel inactivation were greatly reduced compared with WT-RyR1. Average IC₅₀ values were 5.2 ± 0.2, 6.9 ± 0.4, 6.7 ± 0.6, and 5.0 ± 0.3 mM (means ± SE, n = 4–7) for F4732D, G4733E, R4736W, and R4736Q mutants, respectively, that were five- to eightfold larger than that of WT-RyR1 (0.86 ± 0.03 mM, n = 5). Overall, the four mutants had a Ca²⁺ activation/inactivation profile more resembling WT-RyR2 than WT-RyR1, which suggested that each of the four RyR1 mutations conferred RyR2-type Ca²⁺-dependent regulation on RyR1.

Mutations in the S4–S5 loop.

Biochemical analyses and cryo-electron microscopy has shown that RyR1 has a short cytoplasmic loop linking the S4 and S5 transmembrane segments (S4–S5, Fig. 1) (9, 10, 41, 44). We constructed and characterized two MH-associated S4–S5 loop mutants, T4825I and H4832Y, using the [³H]ryanodine binding method (Fig. 2C) (1, 3) to compare their Ca²⁺-dependent activities with WT-RyR1 and the other RyR1 mutants (EF-hand and S2–S3 loop). In agreement with previous studies (2, 30, 31, 42, 43) the two mutants showed changes in both their Ca²⁺-dependent activation and inactivation compared with WT-RyR1 (Fig. 2C). However, changes in Ca²⁺-dependent inactivation were not as prominent as those of G4733E and R4736W mutants (Fig. 2D). Apparent EC₅₀ values for Ca²⁺ activation of T4825I and H4832Y RyR1 channels were significantly different from WT-RyR1 (Table 2), as activities of these two mutants at 40 nM Ca²⁺ were significantly higher than WT-RyR1 (Fig. 2E). The results indicated that Ca²⁺-dependent activation of T4825I and H4832Y mutant channels were enhanced.

Effect of Mg²⁺ on RyR1 mutants.

Chugun et al. (5) have reported that RyR2 activities in the rat ventricular microsomal fractions and recombinant RyR2 were suppressed at 10–100 μM Ca²⁺ and that the suppression was relieved by 1 mM Mg²⁺. As observed for WT-RyR2, we found that the suppression for S2–S3 loop mutants (F4732D, G4733E, R4736W, and R4736Q) at 10–100 μM Ca²⁺ was relieved by the addition of 1 mM Mg²⁺ to [³H]ryanodine binding media, which resulted in more bell-shaped Ca²⁺ activation/inactivation curves (Fig. 3). Average IC₅₀ values of Ca²⁺ for channel activities were 3.7 ± 0.3, 5.7 ± 0.3, 4.8 ± 0.3, and 4.3 ± 0.5 mM (n = 5–7) for F4732D, G4733E, R4736W, and R4736Q mutants, respectively; they were significantly larger than that of WT-RyR1 (0.70 ± 0.09 mM, n = 4) (Fig. 3I). The affinities of Ca²⁺ for channel activation of these for S2–S3 loop mutants were essentially the same as WT-RyR1 (Table 2).

Fig. 3.

Ca²⁺-dependent activities of MH-associated RyR1 mutants in the presence of 1 mM Mg²⁺. A–H: Ca²⁺-dependent activities with and without 1 mM Mg²⁺ were determined for WT-RyR1, WT-RyR2, and six MH-associated RyR1 mutants. Data of Ca²⁺-dependent activities without 1 mM Mg²⁺ are from Fig. 2. Data are shown as means ± SE (n = 4–8). I: IC₅₀ values of wild-type and mutant RyRs in the presence of 1 mM Mg²⁺ are shown. *P < 0.05 compared with WT-RyR1 by one-way ANOVA followed by Tukey's test among seven sample groups (wild-type and mutant RyR1s). Data are shown as means ± SE (n = 4–8).

We also determined the Ca²⁺-dependent channel activities of S4–S5 mutants (T4825I and H4832Y) in the presence of 1 mM Mg²⁺. The IC₅₀ values of the two mutants were substantially higher than that of WT-RyR1 in the presence of Mg²⁺ (Fig. 3I), but the H4832Y mutant was not significantly different. Apparent affinities for Ca²⁺ activation (EC₅₀) of the two mutants were significantly higher than WT-RyR1 (Table 2).

We compared the Mg²⁺ dependence of [³H]ryanodine binding to recombinant RyR1s and RyR2 at a Ca²⁺ concentration of 100 μM (Fig. 4), which maximally suppressed activities of WT-RyR2 and four S2–S3 loop RyR1 mutants (Fig. 3). As previously recognized (5, 42), Mg²⁺ inhibited WT-RyR1 in a concentration-dependent manner, whereas rabbit recombinant WT-RyR2 was activated by 1–3 mM Mg²⁺ but inhibited by 10 mM Mg²⁺ (Fig. 4A).

Fig. 4.

Effect of Mg²⁺ on MH/CNM-associated RyR1 mutants. Mg²⁺-dependent changes in the activities of wild-type and mutant RyR were determined at 100 μM Ca²⁺ by the [³H]ryanodine binding method. Activities of wild-type RyRs (A), S2–S3 loop mutants (B), S4–S5 loop mutants (C), and EF-hand mutants (D) are shown. Dotted and dashed lines in B, C, and D are mean values of WT-RyR1 and WT-RyR2, respectively. The smaller dotted line indicates the level at 100%. *P < 0.05 compared with control (-Mg²⁺) by one-way ANOVA. In B, *P < 0.05 for F4732D, G4733E, R4736W, and R4736Q mutants; **P < 0.05 for G4733E mutant. Data are shown as means ± SE (n = 4–6). E: IC₅₀ values of WT-RyR1, EF-hand mutants, and S4–S5 loop mutants are shown. *P < 0.05 compared with WT-RyR1 by one-way ANOVA followed by Tukey’s test. Data are shown as means ± SE (n = 4–5).

The four S2–S3 mutants showed a Mg²⁺ activation/inhibition profile similar to WT-RyR2 (Fig. 4B). G4733E-RyR1 was significantly activated by 1–3 mM Mg²⁺, and the other three mutants were activated by 1 mM Mg²⁺. Both S4–S5 loop mutations (T4825I and H4832Y) showed significantly reduced affinities of Mg²⁺ for channel inhibition compared with WT-RyR1, but these mutant channels were not significantly activated by Mg²⁺ (Fig. 4C). The IC₅₀ values for Mg²⁺ inhibition were 6.1 ± 0.8 and 5.9 ± 0.4 mM for T4825I and H4832Y mutants, respectively, which were significantly higher than 0.82 ± 0.03 mM for WT-RyR1 (means ± SE, n = 4–5). The three EF-hand mutants were inhibited by Mg²⁺ in a concentration-dependent manner similar to WT-RyR1 (Fig. 4D). N4120Y mutations slightly reduced Mg²⁺ inhibitory effects, but the IC₅₀ of Mg²⁺ was not significantly changed compared with that of WT-RyR1 (Fig. 4E). Taken together, the results suggested that MH-associated point mutations in the S2–S3 loop of RyR1 conferred RyR2-type Ca²⁺ and Mg²⁺ regulation.

Effect of caffeine on G4733E and R4736W RyR1 mutants.

MH/CCD RyR1 mutations increased sensitivity for RyR agonists such as caffeine and halothane (35, 36, 42). We measured the effect of caffeine on G4733E and R4736W RyR1s, two mutants that showed the most pronounced changes in Ca²⁺ and Mg²⁺ regulations (Figs. 2–4).

Caffeine increased WT-RyR1 activity in a concentration-dependent manner with near saturating concentrations of 10–20 mM caffeine (Fig. 5), which is in agreement with previous observations (35, 36, 42). We found that both G4733E and R4736W mutations slightly increased caffeine affinities by ∼1.6-fold compared with WT-RyR1 (Fig. 5).

Fig. 5.

Effect of caffeine on activities of G4733E and R4736W RyR1 mutants. Caffeine-dependent changes in the activities of wild-type, G4733E, and R4736W RyR1s were determined at 0.4 μM Ca²⁺ by [³H]ryanodine binding assays. Changes in the amount of bound [³H]ryanodine were normalized and fit by Eq. 1 as described in materials and methods. Data are shown as means ± SE (n = 4–15). There are no significant differences among three types of RyR1 at any tested caffeine concentrations by one-way ANOVA. Maximal amounts of bound [³H]ryanodine are 121 ± 9, 109 ± 9, and 111 ± 9% (10 mM caffeine as 100%); activation Hill constants of caffeine are 2.2 ± 0.5, 1.4 ± 0.4, and 1.4 ± 0.3 mM, and Hill coefficients are 1.1 ± 0.2, 1.0 ± 0.2, and 1.0 ± 0.2 for WT-RyR1, G4733E, and R4736W, respectively (fitted value ± regression SE).

DISCUSSION

Over 200 point mutations in RyR1 associated with MH or congenital myopathies including CCD and CNM have been determined by genetic screening of human patients. Of these, 35 mutations were validated to cause MH (i.e., causative mutations) by genetic, cellular, and biochemical characterizations (https://emhg.org/genetics/mutations-in-ryr1/). In the present study, we constructed and expressed seven mutant RyR1s carrying MH/CNM-associated mutations in EF-hand domain or S2–S3 loop as well as two MH causative mutants in S4–S5 loop (T4825I and H4832Y) and determined their Ca²⁺- and Mg²⁺-dependent regulation by [³H]ryanodine binding assay. Six mutations (T4082M, N4120Y, F4732D, G4733E, R4736W, and R4736Q) were identified in MH susceptible individuals/families in Europe, North America, or Japan (12, 18, 23, 27, 29). One mutation (S4113L) was found in an Asian female who was diagnosed with CNM (20). Patients were positive using the caffeine/halothane contracture test, Ca²⁺-induced Ca²⁺ release test, or histological analysis. The present study aimed to obtain a better understanding of how the mutations impair RyR1 channel activity. Two causative RyR1 mutations (T4825I and H4832Y) in the S4–S5 loop were included in the present study to compare their properties with those in the other two domains. As previously reported (2, 30, 31, 42, 43), the two mutants exhibited an altered Ca²⁺/Mg²⁺ regulation.

We found that MH-associated mutations in the S2–S3 loop (F4732D, G4733E, R4736W, and R4736Q) greatly reduced affinities of Ca²⁺ for channel inactivation (Figs. 2 and 3), whereas EC₅₀ of Ca²⁺, representing an initial phase of Ca²⁺ activation, was only slightly changed (Table 2). These four mutations suppressed RyR1 activities at 10–100 μM Ca²⁺, and the suppressions were relieved by 1 mM Mg²⁺. Together, all four S2–S3 loop mutants showed Ca²⁺/Mg²⁺-dependent regulation similar to recombinant WT-RyR2. It is possible that activities of S2–S3 loop mutants are enhanced particularly at ∼1 mM Ca²⁺, which results in apparent suppressions of activities at 10–100 μM Ca²⁺, though the gains of activity of RyR1 mutants are comparable to that of WT-RyR1 (Table 1). Although it is widely accepted that the bell-shaped Ca²⁺-dependent [³H]ryanodine binding to RyRs represents Ca²⁺-dependent activation of RyRs by cytosolic micromolar Ca²⁺ and inhibition by cytosolic millimolar Ca²⁺ (see review in Ref. 6), we cannot rule out a possible millimolar luminal Ca²⁺ effect in this equilibrium binding assay. Single amino acid substitution in the cytoplasmic S2–S3 loop of RyR1 might have allosterically altered luminal Ca²⁺ regulation of RyR1. While further clinical and functional tests remain to be done to validate the four S2–S3 loop mutations as causative MH mutations, our studies point out that the S2–S3 loop may participate in the regulation of RyR1 by Ca²⁺ and Mg²⁺.

In a previous study, we found that two isoform-specific regions involved in Ca²⁺-dependent inactivation were located in the EF-hand domain and the second transmembrane region (S2), but not the S2–S3 loop. While an RyR1/RyR2 chimera carrying the RyR2 sequence of the S2–S3 loop showed essentially the same affinity of Ca²⁺ for channel inactivation as WT-RyR1 (14), our current results indicate that the S2–S3 loop is involved in Ca²⁺-dependent inactivation of RyR1. One possible mechanism is that the S2–S3 loop mediates functional signal transduction between two domains, the EF-hand and the S2 transmembrane segment in RyR1, and mutations in the S2–S3 loop in RyR1 possibly alter the functional coupling between these two domains (Fig. 6). The S2–S3 loop of RyR1 was shown to contain four to five α-helixes by studies with secondary structure prediction and 3D image analysis of cryo-electron microscopy (10, 41, 44). One of these is formed by RyR1 aa 4734–4741. In the secondary structure prediction programs (SABLE, http://sable.cchmc.org/, and Jpred 4, http://www.compbio.dundee.ac.uk/jpred4/), S2–S3 loop mutations (F4732D, G4733E, R4736W, and R4736Q) do not affect formation of this α-helix and minimally modify the neighboring structure (i.e., only one amino acid longer or shorter in formation of a neighboring α-helix), suggesting that these mutations are unlikely to cause large structural change.

Fig. 6.

A possible model of Ca²⁺-dependent inactivation of RyR1. Disease-associated RyR1 mutations in the EF-hand region, S2–S3 loop, and S4–S5 loop are shown. The regions involved in Ca²⁺-dependent channel inactivation are highlighted in gray.

Recently, the near-atomic level structure of RyR1 was shown by high-resolution cryo-electron microscopy and 3D image analysis (10, 41, 44). Zalk et al. (44) showed proximity between the EF-hand domain and the S2–S3 loop. While the structural data predicted a possible mechanism of two domains for Ca²⁺-dependent activation of RyR1, our current results from [³H]ryanodine binding assay suggested a potential role of S2–S3 loop for Ca²⁺-dependent inactivation. This cryo-electron microscopy study was done in the presence of 5 mM EGTA, yielding the structure of the closed channel (44). Thus we need to await the structure of the open RyR1 and the conformational changes between the EF-hand and S2–S3 loop upon Ca²⁺ binding to address the significance of two domains in regulating Ca²⁺-dependent gating. Recombinant studies with the EF-hand domains showed that the affinity of Ca²⁺ was 60 μM to 3.8 mM, which exceeds the concentration required for activation (38, 39). Scrambling two EF-hand motifs (EF1 and EF2) of RyR1 did not affect caffeine or depolarization-induced Ca²⁺ release in RyR1-expressing myotubes, whereas EF1-disrupted RyR1 showed slightly higher affinity of Ca²⁺ for channel activation (11). In addition, a mutant RyR2 lacking the entire EF-hand domain showed essentially the same Ca²⁺ activation as WT-RyR2, but the mutant was impaired in Ca²⁺-store overloaded-induced Ca²⁺ release (15). We observed no significant changes in the Ca²⁺-dependent activation and inactivation of ryanodine binding to EF-hand mutants T4802M, S4113L, and N4120Y. Channel activities of T4082M and S4113L-RyR1 were previously studied by using biopsies from patients harboring these mutations (18, 20). Skinned fiber preparation from a patient with T4082M mutation exhibited an enhanced Ca²⁺-induced Ca²⁺ release (18). In another study, KCl and RyR1 agonist (4-chloro-m-cresol)-induced Ca²⁺ release was increased in MyoD transfected skin fibroblasts that show a muscle-like phenotype from the patient with S4113L mutation compared with control cells (20). The apparent discrepancies between our results and the previous data suggest that other yet to be identified cellular factors have a role in the aberrant regulation of T4802M and S4113L-RyR1 mutants.

Two mutations in the S4–S5 loop were previously characterized in recombinant (T4825I and H4832Y) and in vivo (T4825I) studies (2, 30, 31, 42, 43). Our results with the two mutants were in agreement with these data and showed that the effects of two S4–S5 loop mutations differed from those of four S2–S3 loop mutations. All six mutations altered Ca²⁺-dependent inactivation, but impairments of Ca²⁺-dependent inactivation were more prominent in S2–S3 loop mutants. Changes in Ca²⁺ activation were observed only in S4–S5 loop mutants. Although allosteric effects of mutations have to be taken into account, taken together, the S4–S5 loop of RyR1 seems to be involved in Ca²⁺-dependent regulation, especially Ca²⁺ activation. In addition, 1 mM Mg²⁺ activated S2–S3 loop mutants but not S4–S5 loop mutants (Fig. 4).

In vitro experiments showed that RyR1 inhibition by Ca²⁺ required 0.5–1.0 mM, which was somewhat beyond the physiological concentration of intracellular Ca²⁺. However, single-channel studies showed that Ca²⁺ flux through the RyR1 channel from lumen to cytoplasm inhibited RyR1, which suggested that local Ca²⁺ concentration around the Ca²⁺ inactivation site of RyR1 reached millimolar levels (37). Termination of Ca²⁺ sparks, elementary Ca²⁺ release events, in frog skeletal muscle depended on the amount of released Ca²⁺ (26). Therefore, an altered Ca²⁺-dependent inactivation of RyR1, possibly caused by anesthetic reagents such as halothane, may result in the release of abnormal amounts of Ca²⁺ from the sarcoplasmic reticulum.

In conclusion, MH-associated RyR1 mutations in the S2–S3 cytoplasmic loop confer RyR2 type Ca²⁺ and Mg²⁺ regulation on RyR1. The results suggest that these mutant RyR1s do not readily close during excitation-contraction coupling, leading to intracellular Ca²⁺ overload, which is often observed in MH skeletal pathology.

GRANTS

Support for this study was provided by National Institutes of Health Grants R03-AR-061030 and UL1-TR-001450, American Heart Association Grant 10SDG3500001, and National Science Foundation Grant EPS0903795 (to N. Yamaguchi).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

A.C.G., T.W.H., and N.Y. performed experiments; A.C.G., T.W.H., and N.Y. analyzed data; A.C.G., T.W.H., and N.Y. interpreted results of experiments; A.C.G., T.W.H., and N.Y. edited and revised manuscript; A.C.G., T.W.H., and N.Y. approved final version of manuscript; N.Y. conception and design of research; N.Y. prepared figures; N.Y. drafted manuscript.