Modulation of sarcoplasmic reticulum Ca2+ release in skeletal muscle expressing ryanodine receptor impaired in regulation by calmodulin and S100A1

Abstract

In vitro, calmodulin (CaM) and S100A1 activate the skeletal muscle ryanodine receptor ion channel (RyR1) at submicromolar Ca(2+) concentrations, whereas at micromolar Ca(2+) concentrations, CaM inhibits RyR1. One amino acid substitution (RyR1-L3625D) has previously been demonstrated to impair CaM binding and regulation of RyR1. Here we show that the RyR1-L3625D substitution also abolishes S100A1 binding. To determine the physiological relevance of these findings, mutant mice were generated with the RyR1-L3625D substitution in exon 74, which encodes the CaM and S100A1 binding domain of RyR1. Homozygous mutant mice (Ryr1(D/D)) were viable and appeared normal. However, single RyR1 channel recordings from Ryr1(D/D) mice exhibited impaired activation by CaM and S100A1 and impaired CaCaM inhibition. Isolated flexor digitorum brevis muscle fibers from Ryr1(D/D) mice had depressed Ca(2+) transients when stimulated by a single action potential. However, during repetitive stimulation, the mutant fibers demonstrated greater relative summation of the Ca(2+) transients. Consistently, in vivo stimulation of tibialis anterior muscles in Ryr1(D/D) mice demonstrated reduced twitch force in response to a single action potential, but greater summation of force during high-frequency stimulation. During repetitive stimulation, Ryr1(D/D) fibers exhibited slowed inactivation of sarcoplasmic reticulum Ca(2+) release flux, consistent with increased summation of the Ca(2+) transient and contractile force. Peak Ca(2+) release flux was suppressed at all voltages in voltage-clamped Ryr1(D/D) fibers. The results suggest that the RyR1-L3625D mutation removes both an early activating effect of S100A1 and CaM and delayed suppressing effect of CaCaM on RyR1 Ca(2+) release, providing new insights into CaM and S100A1 regulation of skeletal muscle excitation-contraction coupling.

Fig. 1.

Generation of mice with an L3625D mutation in the calmodulin (CaM)-binding site of ryanodine receptor 1 (RyR1). A: schematic representation of the mouse Ryr1 gene and targeting construct. S and N represent SpeI and NdeI restriction enzyme sites, respectively. neo, tACE-cre, and TK denote neomycin-resistant gene, Cre recombinase gene driven by testis-specific angiotensin-converting enzyme promoter, and thymidine kinase gene, respectively. Arrows and “x” indicate the position of primers for screening and a mutation site, respectively. B: southern blot analysis of genomic DNA. After restriction enzyme digestion of genomic DNA with SpeI and NdeI, 5′- and 3′-probes identify 12.2-kb and 5.1-kb fragment in targeted allele, respectively. Both probes hybridize with 16.7-kb fragments in wild-type (WT) allele. C: sequence analysis of RT-PCR. cDNA encoding CaM-binding site of RyR1 was amplified from total RNA from homozygous (hybrid genetic background) mouse skeletal muscle and sequenced. The L3625D mutation was confirmed. A HinfI site created by the L3625D mutation (GATTC) was used for screening the mutant allele.

Fig. 2.

Effects of CaM on single Ryr1^+/+ and Ryr1^D/D channel activities. Membranes isolated from skeletal muscle of Ryr1^+/+ and Ryr1^D/D mice were fused with lipid bilayer as described in experimental procedures. Representative single-channel currents (downward deflections from closed levels, c–) were recorded at −35 mV at 0.2 μM (A) and 7 μM (B) cis, cytosolic free Ca²⁺ in the presence of 1 mM ATP as described in experimental procedures in the absence of CaM (top traces) and following the addition of 20 nM (middle traces) and 200 nM (bottom traces) cis CaM. Channel open probabilities (P_o) obtained from 2-min recordings in the absence of CaM were 0.15 ± 0.02 (n = 21) and 0.14 ± 0.02 (n = 29) at 50% threshold setting, 0.30 ± 0.04 and 0.23 ± 0.02 at 25% threshold setting at 0.2 μM Ca²⁺ and 1 mM ATP, and 0.27 ± 0.05 (n = 14) and 0.26 ± 0.05 (n = 11) at 50% threshold setting and 0.38 ± 0.05 and 0.48 ± 0.05 at 25% threshold setting at 7 μM Ca²⁺ and 1 mM ATP for Ryr1^+/+ and Ryr1^D/D, respectively. Normalized P_o values (bottom) were obtained from 2-min recordings by setting the threshold level at 25% (open bars) and 50% (gray bars) of the current amplitude between the closed (c) and open (o) channel states. Data are means ± SE of 9 and 15 (A) and 14 and 11 (B) channel recordings for Ryr1^+/+ and Ryr1^D/D, respectively. *P < 0.05 compared with respective control (−CaM).

Fig. 3.

CaM regulation of RyR1 from skeletal muscle of Ryr1^+/+ and Ryr1^D/D mice. Specific [³H]ryanodine binding to crude membrane fractions from skeletal muscle of Ryr1^+/+ (•) and Ryr1^D/D (○) mice was determined at the indicated CaM concentrations in the presence of 0.15 μM Ca²⁺ and 1 mM AMPPCP (a nonhydrolyzable ATP analog) (A) and 25 μM (B) free Ca²⁺ as described in experimental procedures. Data are means ± SE of 4 experiments. *P < 0.05 compared with control (−CaM).

Fig. 4.

CaCaM and CaS100A1 bound to the CaM/S100A1-binding domain of RyR1. A: space-filling model of dimeric CaS100A1 with each subunit bound to a peptide (RyRP12; shown in red) from the CaM/S100A1-binding domain of RyR1. In the foreground, one subunit is shown illustrating the location of RyR1 residues W3621 and L3625 on the hydrophobic side of an amphipathic helix, which in turn interact with hydrophobic residues on CaS100A1 (shown in blue). B: space-filling diagram of CaCaM bound to a peptide (red) from the CaM/S100A1-binding domain of RyR1 illustrating that these two residues (W3621, L3625) also interact directly with hydrophobic residues on CaCaM (shown in blue). C: ribbon diagram illustrating a close-up view of the CaS100A1-RyRP12 interaction. D: ribbon diagram illustrating a close-up view of the CaCaM-RyR1 peptide interaction. In A–D, calcium ions are colored orange. E: binding of CaS100A1 to WT and mutant RyRP12 peptides derived from RyR1 as determined using isothermal titration calorimetry (ITC). Shown is a representative trace together with a theoretical fitted curve for the binding of the WT-RyRP12 peptide to CaS100A1 (▴). These data were duplicated to provide the thermodynamic binding parameters (n = 0.75 ± 0.12; K_D = 75 ± 12 μM; ΔH = −5.8 ± 0.6 kcal/mol). Also shown are representative data for the binding of the RyR1-L3625D mutant peptide together with a theoretical curve for binding. In this case, the binding stoichiometry could not be quantitatively determined since the binding was weak and was assumed to be one (n = 1) in all the titrations measured; therefore, these data provided only a limiting value for binding (□ K_D > 0.9 ± 0.1 mM). No detectable binding was observed for the RyR1-W3621D (•) or the RyR1-W3621D + L3625D [double mutant (DM); ◊] mutant peptides.

Fig. 5.

Effects of S100A1 on single Ryr1^+/+ and Ryr1^D/D channel activities at 0.2 μM Ca²⁺. Membranes isolated from skeletal muscle of Ryr1^+/+ (left) and Ryr1^D/D (right) mice were fused with lipid bilayer. Representative single-channel currents (downward deflections from closed levels, c–) were recorded at −35 mV at 0.2 μM cis, cytosolic free Ca²⁺ in the presence of 1 mM ATP as described in experimental procedures in the absence of S100A1 (top traces) and following the addition of 0.3 μM (middle traces) and 1 μM (bottom traces) S100A1 to the cis chamber. Normalized P_o (bottom) was obtained from 2-min recordings by setting the threshold level at 25% (open bars) and 50% (gray bars) of the current amplitude between the closed (c) and open (o) channel states. Data are means ± SE of 3–14 channel recordings. *P < 0.05 compared with control (−S100A1).

Fig. 6.

Biochemical properties of membrane fractions isolated from Ryr1^+/+ and Ryr1^D/D mice. A: maximal specific binding capacity (B_max) values of [³H]ryanodine binding to crude skeletal muscle membrane fractions of 1.5- to 5.5-mo-old Ryr1^+/+ and Ryr1^D/D mice were determined as described in experimental procedures. Normalized data are means ± SE of 4–6 experiments. *P < 0.05 compared with Ryr1^+/+ mice. B: ⁴⁵Ca²⁺ uptake rates by crude membrane fractions of 1.5- to 5.5-mo-old Ryr1^+/+ and Ryr1^D/D mice were determined as described in experimental procedures. Normalized data are means ± SE of 4–5 experiments.

Fig. 7.

Ryr1^D/D muscle fibers exhibit normal Indo-1 resting ratio but decreased peak amplitude of the Indo-1 Ca²⁺ transient following a single action potential (AP). A: average Indo-1 ratio Ca²⁺ transients from Ryr1^+/+ (black trace; n = 20) and Ryr1^D/D (gray trace; n = 22) fibers. Isolated fibers were stimulated with field electrodes at 400 ms, and emission ratio was examined. B: bar plot summarizing resting Indo-1 ratio averages (left) and peak ΔIndo-1 ratio (right) of control and mutant fibers. No significant (NS) differences in resting Ca²⁺ concentration were detected, but a significant difference in peak ΔIndo-1 ratio was found (*P = 0.017). C: averaged Indo-1 ratio responses from control (black traces; n = 12) and mutant (gray traces; n = 12) fibers elicited with a two-AP protocol with recovery periods of 40, 80, and 160 ms between stimuli. Traces were normalized to amplitude of the first stimulus to appreciate relative summation in mutant fibers.

Fig. 8.

Ryr1^D/D muscle fibers exhibit pronounced summation of the Ca²⁺ transient during repetitive stimulation. A: Fluo-4 transient responses to single AP stimuli followed by a 100-Hz train of APs show reduced amplitude of Ca²⁺ twitch and tetanic transients in mutant fibers. B: traces from A normalized to initial transient amplitude to demonstrate greater summation of the Ca²⁺ transient during repetitive stimulation of mutant fibers when compared with controls. C: quantification of Ca²⁺ transient summation from B, calculated as the ratio of the amplitude of the last Ca²⁺ transient in the train compared with the first. Summation was increased approximately twofold in the mutant fibers. *P < 0.01.

Fig. 9.

Ryr1^D/D muscle fibers demonstrate summation of AP-evoked Ca²⁺ transients and slower inactivation of sarcoplasmic reticulum (SR) Ca²⁺ release flux during tetanic stimulation. A: average Fluo-4 transient responses of 7 control and 10 mutant fibers to a single AP stimuli followed by 5 repeats of 200 ms trains of 100 Hz stimuli, with 200 ms recovery between trains. Traces are normalized to the amplitude of the first transient to demonstrate differences in transient summation during repetitive stimulation. B, left: boxed traces from A, time expanded and normalized to amplitude of the initial peak to examine transient summation of last train. Right: quantification of transient facilitation between Ryr1^+/+ and Ryr1^D/D fibers (*P < 0.01, Ryr1^+/+ n = 7, Ryr1^D/D n = 10). C: average SR Ca²⁺ release flux of control and mutant fibers, quantified using a Ca²⁺ removal model to calculate release from fluorescence responses of fibers subjected to stimulation paradigm seen in A. Release flux is suppressed in mutant fibers compared with controls. D: representative SR Ca²⁺ release flux time course of first tetanic train in C from a control and mutant fiber, normalized to amplitude of initial flux. Peaks of release flux are fit to a single exponential function to quantify inactivation of release during the train. Tau of inactivation is significantly increased in Ryr1^D/D fibers, suggesting a slower rate of inactivation. Fractional inactivation was not significantly different between the groups.

Fig. 10.

Ryr1^D/D muscle fibers exhibit reduced Ca²⁺ release flux with slowed inactivation of Ca²⁺ release in response to voltage-clamp depolarizations. A: average Ca²⁺ release flux from control (left; n = 6) and Ryr1^D/D (middle; n = 8) fibers calculated from Fluo-4 fluorescent transients elicited by steplike 80-ms depolarizations from a holding potential of −80 mV. At right is the release vs. voltage (R-V) relationships of control and Ryr1^D/D fibers. SR Ca²⁺ release flux was significantly suppressed in mutant muscle fibers. *P < 0.05. B: Ca²⁺ release flux from A elicited by depolarizing pulses to −20, 0, and +20 mV from a holding potential of −80 mV, normalized to peak flux to investigate the relative time course of Ca²⁺ release. While the time course of release was similar between mutant and control fibers at lesser depolarizations (−20 mV, left), upon larger depolarizations (0 mV, +20 mV), mutant fibers demonstrate less rapid inactivation of release compared with controls, as seen by the pronounced shoulder during the fast inactivation of release flux in Ryr1^D/D fibers.

Fig. 11.

Ryr1^D/D mice exhibit decreased twitch force generation and increased force summation in vivo. A: representative twitch (single AP) and tetanic (100 Hz, 200 ms) force responses from the tibialis anterior muscles of anesthetized Ryr1^+/+ and Ryr1^D/D mice. B: force vs. frequency of stimulation relationships for Ryr1^+/+ (n = 5) and Ryr1^D/D (n = 6) mice. C: specific twitch force of Ryr1^+/+ and Ryr1^D/D mice. D: specific tetanic force of Ryr1^+/+ and Ryr1^D/D mice. E: tetanus-to-twitch ratio, an indicator of summation, of Ryr1^+/+ and Ryr1^D/D mice. *P < 0.05