The distal C terminus of the dihydropyridine receptor β_1a subunit is essential for tetrad formation in skeletal muscle

Significance

Vertebrate skeletal muscle excitation–contraction coupling (ECC) is based on Ca²⁺-influx–independent interchannel cross-talk between DHPR and RyR1. The skeletal muscle DHPR complex consists of the main, voltage-sensing, and pore-forming α_1S subunit, the auxiliary β_1a, α₂δ-1, γ₁ subunits, and Stac3. The DHPRβ_1a subunit plays an essential role in full triad targeting of DHPRα_1S, voltage sensing, and tetrad formation (grouping of four DHPRs)—the three prerequisites for skeletal muscle ECC. Hence, a lack of DHPRβ_1a results in a lethal phenotype in both β₁-null mice and zebrafish. Here, we identified the nonconserved, distal C terminus of DHPRβ_1a as playing a pivotal role in the formation of DHPR tetrads, and thus allosteric DHPR–RyR1 coupling, essential for proper skeletal muscle ECC.

Abstract

The skeletal muscle dihydropyridine receptor (DHPR) β_1a subunit is indispensable for full trafficking of DHPRs into triadic junctions (i.e., the close apposition of transverse tubules and sarcoplasmic reticulum [SR]), facilitation of DHPRα_1S voltage sensing, and arrangement of DHPRs into tetrads as a consequence of their interaction with ryanodine receptor (RyR1) homotetramers. These three features are obligatory for skeletal muscle excitation–contraction (EC) coupling. Previously, we showed that all four vertebrate β isoforms (β₁–β₄) facilitate α_1S triad targeting and, except for β₃, fully enable DHPRα_1S voltage sensing [Dayal et al., Proc. Natl. Acad. Sci. U.S.A. 110, 7488–7493 (2013)]. Consequently, β₃ failed to restore EC coupling despite the fact that both β₃ and β_1a restore tetrads. Thus, all β-subunits are able to restore triad targeting, but only β_1a restores both tetrads and proper DHPR–RyR1 coupling [Dayal et al., Proc. Natl. Acad. Sci. U.S.A. 110, 7488–7493 (2013)]. To investigate the molecular region(s) of β_1a responsible for the tetradic arrangement of DHPRs and thus DHPR–RyR1 coupling, we expressed loss- and gain-of-function chimeras between β_1a and β₄, with systematically swapped domains in zebrafish strain relaxed (β₁-null) for patch clamp, cytoplasmic Ca²⁺ transients, motility, and freeze-fracture electron microscopy. β_1a/β₄ chimeras with either N terminus, SH3, HOOK, or GK domain derived from β₄ showed complete restoration of SR Ca²⁺ release. However, chimera β_1a/β₄(C) with β₄ C terminus produced significantly reduced cytoplasmic Ca²⁺ transients. Conversely, gain-of-function chimera β₄/β_1a(C) with β_1a C terminus completely restored cytoplasmic Ca²⁺ transients, DHPR tetrads, and motility. Furthermore, we found that the nonconserved, distal C terminus of β_1a plays a pivotal role in reconstitution of DHPR tetrads and thus allosteric DHPR–RyR1 interaction, essential for skeletal muscle EC coupling.

Excitation–contraction (EC) coupling in skeletal muscle is initiated by depolarization of the muscle cell membrane induced by motor neuron input, which subsequently induces myofibril contractions. This transduction event depends on junctions between the surface membrane and its invaginations (transverse [T] tubules) and the sarcoplasmic reticulum (SR), in structures termed Ca²⁺ release units. The dihydropyridine receptor (DHPR) in the T-tubular membrane of the muscle cell functions as voltage sensor for this excitation signal. EC coupling in vertebrate skeletal muscle is based on Ca²⁺-influx–independent interchannel protein–protein interaction between the DHPR and ryanodine receptor (RyR1) in the SR membrane (1–3). Because of this physical interaction, the depolarization-induced conformational change of the DHPR is transmitted to the RyR1 channel, which opens to release large amounts of Ca²⁺ ions from the SR Ca²⁺ stores—a process that is the final trigger for myofibril contraction (4, 5).

The skeletal muscle DHPR complex consists of the central, pore-forming, and voltage-sensing α_1S subunit and the accessory subunits β_1a, α₂δ-1, and γ₁ (6–8). Among them, the α_1S and the β_1a subunits are indispensable for skeletal muscle EC coupling (9–11). Akin to DHPRα_1S-null (dysgenic) (9) and RyR1-null (dyspedic) (12) mice, β₁-null mice (10) and β₁-null zebrafish (strain relaxed) (11) show a lethal phenotype due to complete absence of skeletal muscle contractility that leads to asphyxia. Besides the two canonical DHPR subunits, the junctional proteins Stac3 and junctophilin-2 (JP2) are also crucial for proper DHPR–RyR1 interaction that enables concerted voltage-induced SR Ca²⁺ release in skeletal muscle (13).

In DHPRβ₁-null zebrafish strain relaxed, a lack of the β_1a subunit results in 1) reduced DHPRα_1S expression in the T-tubular membrane, 2) elimination of α_1S charge movement, and 3) a lack of the arrangement of DHPRs into groups of four (tetrads) opposite every other RyR1 (11). These three features are prerequisite for the tight protein–protein interaction between the DHPR and RyR1 and thus form the structural–functional basis for skeletal muscle EC coupling. Using zebrafish strain relaxed as a very convenient expression system, we previously showed that all four vertebrate β-isoforms (β₁–β₄), and also the ancestral β-subunit of Musca domestica (β_Μ) (14), are able to fully target α_1S into triads (15). Additionally, except for β₃, all other vertebrate β-isoforms are able to restore full charge movement (16) (SI Appendix, Fig. S1). Consequently, despite the surprising fact that β₃, akin to β_1a, is able to accurately cause the organization of DHPRs into tetrads, it is unable to restore EC coupling (16). Interestingly, only expression of β_1a fulfills all the three structural–functional prerequisites, i.e., proper DHPR triad and tetrad restoration, as well as proper charge movement facilitation and consequently, accurate DHPR–RyR1 interaction (15). As a result, native skeletal muscle β_1a is the only DHPRβ subunit that supports proper skeletal muscle EC coupling (SI Appendix, Fig. S1).

To identify a structural domain(s) of β_1a essential for restoration of DHPR voltage sensing, and hence to probe how the DHPRα_1S–β_1a interaction affects this initial step of EC coupling, we previously performed reconstitution studies in the relaxed system using chimeras between β_1a and β₃ (16). Voltage-gated Ca²⁺ channel β-subunits are intracellular proteins with a five-domain organization and two conserved domains, the src homology 3 (SH3) and guanylate kinase (GK) domains that are connected by the variable HOOK region and flanked by variable N and C termini (17–20). The outcome of systematic domain swapping between β_1a and β₃ in the study of Dayal et al. (16) revealed a pivotal role of the β_1a SH3 domain and the C terminus in charge movement restoration. The results indicate that this domain–domain interaction is dependent on a SH3-binding polyproline (PXXP) motif in the proximal C terminus of the β_1a subunit. Consequently, it was concluded that the β_1a subunit, apparently via its SH3–C-terminal PXXP interaction, adopts a discrete conformation required for inducing a proper conformational change in the α_1S subunit crucial for “turning on” its voltage-sensing function (16).

Nevertheless, we are just beginning to understand the importance of distinct molecular domains of the β_1a subunit in skeletal muscle EC coupling. In the present study, we characterized the second crucial structural prerequisite, tetrad formation, which contrary to the promiscuous structural property of DHPR triad targeting by all β-subunits, is shared by only β_1a and β₃ (SI Appendix, Fig. S1). To identify β_1a domains responsible for proper DHPR tetrad formation and thus proper DHPR–RyR1 protein–protein interaction as a basis for induction of SR Ca²⁺ release and finally muscle contractility/motility, we expressed putative loss- and gain-of-function chimeras with systematically swapped domains between β_1a and β₄ in zebrafish strain relaxed for patch clamp, cytoplasmic Ca²⁺ transients, motility, and freeze-fracture electron microscopy (EM) analyses.

Here we report that our loss- and gain-of-function chimeras indicate the importance and exclusivity of the nonconserved distal C terminus of β_1a in DHPR tetrad formation and thus a DHPR–RyR1 interaction essential for proper skeletal muscle EC coupling. Within the distal C terminus, we found that a hydrophobic surface (L₄₉₆L₅₀₀W₅₀₃), previously postulated to be important for activation of RyR1 (21), does not appear to play a role in EC coupling Ca²⁺ release. Based on these results, we propose a model in which the distal β_1a C terminus enables a conformation of the β-subunit, which in turn causes the intracellular domains of α_1S to assume the positioning required for the interaction with RyR1 and thus the tetradic arrangement of DHPRs (22, 23).

Results and Discussion

The β₄ Subunit Is the Apt Isoform for Mapping the Domain(s) of β_1a Crucial for Tetrad Formation in Skeletal Muscle.

To elucidate the importance of distinct β_1a domain(s) for DHPR tetrad formation the first step was to identify the most apt β-isoform, which lacked this property and thus could serve as a molecular tool for chimerization with β_1a. Proper DHPR tetrad formation is a common attribute of β_1a and β₃, but is missing in β_2a (15, 16). Although not directly tested, we postulated that β₄ might also be poor at restoring DHPR tetrads because it only restored ∼50% of cytoplasmic Ca²⁺ transients upon expression in relaxed myotubes despite its ability to completely support DHPR triad targeting and charge movement (16). β₄ was given preference over β_2a as a chimerization partner with β_1a, because it is phylogenetically older than β_2a (16) and thus has overall lower amino acid homology to β_1a than β_2a (60.1% compared to 65.7%, respectively), which also holds true for the aligned C termini (24.2% versus 31.8%, respectively). Equally important, when exploring the role of the C terminus in charge movement restoration (16), we saw that β₄ does not have as long a C terminus as β_2a (117 versus 193 residues, respectively).

As we postulated, DHPR tetrads were not detectable in β₄-expressing relaxed myotubes (Fig. 1A), making β₄ useful for mapping the molecular domain(s) of β_1a essential for tetrad formation via chimeric constructs. In β₄-expressing relaxed myotubes, cytoplasmic Ca²⁺ transients [(ΔF/F₀)_max = 1.02 ± 0.15, n = 13] was significantly larger (P < 0.001) than in untransfected relaxed myotubes (below detection limit [bdl], n = 10) but significantly smaller when compared (P < 0.001) to β_1a [(ΔF/F₀)_max = 2.30 ± 0.15, n = 9] (Fig. 1B). Moreover, ΔF/F₀ in the β₄-expressing myotubes had a voltage dependence (Fig. 1B) that was >12 mV rightwardly shifted compared to β_1a (V_1/2: β₄, 6.05 ± 3.42 mV, n = 13; β_1a, −6.87 ± 2.65 mV, n = 9; P < 0.01). The transients also had a different time course in β_1a- and β₄-expressing myotubes. In the β_1a-expressing myotubes, the transients had a steep rise followed by a plateau during the 200-ms depolarization (Fig. 1 C, Upper Right), which presumably represents a rapid, transient release of Ca²⁺ into the cytoplasm followed by a lower, sustained release just sufficient to balance the Ca²⁺ removal mechanisms (24, 25). In the β₄-expressing myotubes both the transient and sustained release appear to be reduced so that the initial rise is smaller and that the transient decays during the pulse because the sustained release is outweighed by the removal processes (Fig. 1 C, Lower Right). To obtain a signal related roughly to total release (transient plus sustained), we integrated (intg.) the transients and plotted the area versus test potential, which also revealed a significant difference (P < 0.001) between β₄ and β_1a (intg.ΔF/F₀: β₄, 0.64 ± 0.10, n = 12; β_1a, 1.97 ± 0.18, n = 9; Fig. 1C).

Fig. 1.

Absence of DHPR tetrad restoration in β₄-expressing relaxed myotubes. (A) Freeze-fracture replicas of peripheral couplings in tail myotomes of 27- to 30-hpf zebrafish. Control myotomes (Top) show arrangement of DHPR particles in tetrads (center indicated by red dots), organized in orthogonal arrays. In β₄-expressing relaxed zebrafish (Bottom) DHPR tetrads show a lack of tetrad formation. (Scale bar, 50 nm.) (B) Quantification of voltage dependence of cytoplasmic Ca²⁺ transients yielded (ΔF/F₀)_max values that are significantly lower (P < 0.001) in β₄ (n = 13)- compared to β_1a (n = 9)-expressing relaxed myotubes. ΔF/F₀ values recorded from untransfected relaxed myotubes were below detection level (n = 10). (C) Similarly, plots of voltage dependence of the integral of the ΔF/F₀ transients in response to 200-ms test depolarizations indicate a highly significant difference (P < 0.001) in the total amount of Ca²⁺ released between relaxed myotubes expressing β_1a (n = 9) or β₄ (n = 12) subunit. (Right) Representative ΔF/F₀ recordings from relaxed myotubes expressing β_1a or β₄. (Scale bars, 50 ms [horizontal], ΔF/F₀ = 1 [vertical].) Error bars indicate SEM. P determined by unpaired Student’s t test.

The C Terminus of the β_1a Subunit Is Key for Proper DHPR–RyR1 Coupling.

To explore the role of the β_1a domain(s) in tetrad formation, we constructed a set of β_1a/β₄ chimeras in which the N terminus (N), SH3 domain (SH3), HOOK region (H), GK domain (GK), and C terminus (C) of β_1a were systematically swapped with corresponding β₄ sequences (Fig. 2A). To test whether all the β_1a/β₄ chimeras were functionally expressed in relaxed myotubes, we measured DHPRα_1S outward (on) charge movement (Q_on). The Q_max values displayed by all β_1a/β₄ chimeras (n = 13 to 24) were not significantly different (P > 0.05) from the basis constructs β_1a (10.28 ± 1.07 nC/μF; n = 16) and β₄ (10.63 ± 0.85 nC/μF; n = 12) (Fig. 2B).

Fig. 2.

Loss-of-function β_1a/β₄ chimeras revealed the importance of the β_1a C terminus in skeletal muscle DHPR–RyR1 coupling. (A) Block schemes of domain organization of putative loss-of-function β_1a/β₄ chimeras with systematic exchange of N terminus (N), SH3 domain (SH3), HOOK region (H), GK domain (GK), or C terminus (C) of β_1a (blue) by β₄ sequences (orange). Homologous SH3 and GK domains are represented by hatched boxes. (B, Left) Analyses of voltage dependence of integrated outward gating currents normalized to cell capacitance exhibited maximum charge movement (Q_max) values indistinguishable (P > 0.05) between relaxed myotubes expressing β_1a (n = 16), β₄ (n = 12), β_1a/β₄(N) (n = 21), β_1a/β₄(SH3) (n = 19), β_1a/β₄(H) (n = 24), β_1a/β₄(GK) (n = 15), or β_1a/β₄(C) (n = 13). Q_max values from untransfected relaxed myotubes were slightly above detection level (P < 0.001, n = 11). (Right) Representative Q recordings from relaxed myotubes expressing either β_1a or β₄. (Scale bars, 5 ms [horizontal], 3 pA/pF [vertical].) (C and D) Cytoplasmic Ca²⁺ transient restoration was comparable (P > 0.05) between relaxed myotubes expressing β_1a (n = 9), β_1a/β₄(SH3) (n = 17), β_1a/β₄(GK) (n = 11), β_1a/β₄(N) (n = 14), or β_1a/β₄(H) (n = 16). By contrast, ΔF/F₀ values were significantly lower (P < 0.001) for chimera β_1a/β₄(C) (n = 12) and similar (P > 0.05) to those of β₄ (n = 13). Exemplar Ca²⁺ transient recordings from relaxed myotubes expressing β_1a/β₄(SH3) (C, Right) or β_1a/β₄(C) (D, Right). (Scale bars, 50 ms [horizontal], ΔF/F₀ = 1 [vertical].) Error bars indicate SEM. P determined by unpaired Student’s t test, ***P < 0.001.

β_1a/β₄ chimeras with SH3 and GK domains derived from β₄ completely restored cytoplasmic Ca²⁺ transients in relaxed myotubes [(ΔF/F₀)_max: 2.28 ± 0.19, n = 17 and 2.36 ± 0.32, n = 11, respectively] to the level of β_1a (P > 0.05) (Fig. 2 C, Left). Profiles of these cytoplasmic Ca²⁺ transients exhibited kinetics typical for β_1a with a sustained plateau (Fig. 2 C, Right), indicating normal DHPR–RyR1 interaction. Chimeras in which either the nonconserved N terminus or HOOK region of β_1a was replaced by corresponding β₄ sequences also displayed restoration of ΔF/F₀ [(ΔF/F₀)_max: β_1a/β₄(N), 2.53 ± 0.24, n = 14; β_1a/β₄(H), 2.29 ± 0.25, n = 16], not significantly different (P > 0.05) from β_1a control myotubes (Fig. 2D). Notably, in contrast to the other constructs, chimera β_1a/β₄(C), carrying the nonconserved C terminus of β₄ restored Ca²⁺ transients [(ΔF/F₀)_max 0.88 ± 0.13, n = 12] that did not differ significantly from β₄ (P > 0.05) but were significantly (P < 0.001) smaller than those of β_1a (Fig. 2D). The results above emphasize the importance of the β_1a C terminus in proper DHPR–RyR1 coupling.

The Greater Length of the β₄ C Terminus Is Not Responsible for the Impairment of DHPR–RyR1 Coupling.

Since the C terminus of β₄ is markedly longer (117 residues) than that of β_1a (66 residues) (Fig. 3A), the question arose whether the difference in length between the two isoforms is responsible for the significant difference in cytoplasmic Ca²⁺ transient restoration (Fig. 1 B and C). Consequently, we removed the distal 51 residues from the β₄ C terminus to yield construct β₄(Δ51). Full restoration of charge movement (Q_max: 9.94 ± 0.91, n = 17) upon expression of β₄(Δ51) in relaxed myotubes demonstrated that the expression of the deletion mutant did not significantly (P > 0.05) differ from β_1a and β₄ (Q_max: 10.28 ± 1.07, n = 16 and 10.63 ± 0.85, n = 12, respectively) (Fig. 3B). Nonetheless, peak Ca²⁺ transients for the mutant β₄(Δ51) were significantly (P < 0.001) smaller than for β_1a [(ΔF/F₀)_max of 1.30 ± 0.14, n = 10 compared to 2.30 ± 0.15, n = 9 for β_1a] and not significantly different (P > 0.05) from β₄ (1.02 ± 0.15, n = 13) (Fig. 3C). The same is true after comparing the integral of the Ca²⁺ transients during the 200-ms test pulses. Maximal intg.ΔF/F₀ for truncation mutant β₄(Δ51) was comparable (P > 0.05) to β₄ (0.92 ± 0.12, n = 10 and 0.64 ± 0.10, n = 12, respectively), and significantly smaller (P < 0.001) than for β_1a (intg.ΔF/F₀ of 1.97 ± 0.18, n = 9) (Fig. 3D). Thus, the greater length of the C terminus of β₄ does not appear to be responsible for impairing DHPR–RyR1 coupling.

Fig. 3.

Length of the β₄ C terminus is not crucial for skeletal muscle DHPR–RyR1coupling. (A) Amino acid sequence alignment depicting variable lengths of the C termini of β_1a and β₄ subunits (GenBank accession nos.: rabbit β_1a, M25514; rat β₄, L02315). To determine whether the length of the C terminus was functionally critical, the last 51 amino acids from the β₄ C terminus were deleted, yielding mutant β₄(Δ51). (B) Q_max values were indistinguishable (P > 0.05) between relaxed myotubes expressing the deletion mutant β₄(Δ51) (n = 17), β_1a (n = 16), or β₄ (n = 12). (C) Maximal Ca²⁺ transients (ΔF/F₀)_max for β₄(Δ51) expressing relaxed myotubes (n = 10) was similar (P > 0.05) to that of β₄ (n = 13). (D) Similarly, total maximal Ca²⁺ transients (intg.(ΔF/F₀)_max) for β₄(Δ51) expressing relaxed myotubes was statistically indistinguishable (P > 0.05) from that of β₄. Error bars indicate SEM. P determined by unpaired Student’s t test.

The Distal C Terminus of β_1a Is Crucial for the Functional and Structural Interactions between DHPRs and RyR1.

We next constructed and tested a mirror chimera to the loss-of-function chimera β_1a/β₄(C), namely chimera β₄/β_1a(C), where the β₄ C terminus was exchanged with the corresponding β_1a sequence (Fig. 4A). Relaxed myotubes expressing chimera β₄/β_1a(C) showed functional DHPR membrane expression as indicated by full restoration of charge movement comparable (P > 0.05) to β_1a (Q_max: 10.07 ± 0.83, n = 18 and 10.28 ± 1.07, n = 16, respectively) (Fig. 4B). Moreover, relaxed myotubes expressing chimera β₄/β_1a(C) exhibited Ca²⁺ transient levels ((ΔF/F₀)_max: 2.34 ± 0.32, n = 12) that were significantly larger (P < 0.001) than those of β₄ (1.02 ± 0.15, n = 13) and comparable (P > 0.05) to β_1a [(ΔF/F₀)_max: 2.30 ± 0.15, n = 9] (Fig. 4C). As a guide for identifying the regions of the β_1a C terminus most important for interaction with RyR1, we aligned the C termini of β_1a and β₄, which reveals 45% overall homology in the proximal C terminus and only 6% in the overlapping region of the distal C terminus (Fig. 4D). Although divergent from β₄, the distal C terminus of β_1a shows an overall homology of 34% among various phylogenetically diverse vertebrates (SI Appendix, Fig. S2), including complete identity of the initial 10 residues (indicated by the red bracket in Fig. 4D). Thus, we hypothesized that the distal β_1a C terminus (dist.C) would have a stronger impact on EC coupling than the proximal C terminus (prox.C). To test this hypothesis, we constructed chimera β₄/β_1a(prox.C), containing the first 31 C-terminal amino acid residues of β_1a (459 to 489), and chimera β₄/β_1a(dist.C), carrying the subsequent 35 C-terminal residues of β_1a (490 to 524) in an otherwise β₄ sequence background (Fig. 4E).

Fig. 4.

The distal C terminus of β_1a is crucial for skeletal muscle EC coupling. (A) Block scheme of domain organization of gain-of-function chimera β₄/β_1a(C), where the the C terminus of β₄ (orange) was replaced by a corresponding β_1a sequence (blue). (B) Q_max values in relaxed myotubes expressing either chimera β₄/β_1a(C) (n = 18) or β_1a (n = 16) were comparable (P > 0.05). (C) Quantification of voltage dependence of cytoplasmic Ca²⁺ transients yielded significantly higher (P < 0.001) (ΔF/F₀)_max values for chimera β₄/β_1a(C) (n = 12) compared to β₄ (n = 13) but indistinguishable (P > 0.05) from that of β_1a (n = 9) expressing relaxed myotubes. (Right) Exemplar cytoplasmic Ca²⁺ transient recordings from relaxed myotubes expressing chimera β₄/β_1a(C). (Scale bars, 50 ms [horizontal], ΔF/F₀ = 1 [vertical].) (D) Amino acid sequence alignment of C termini of β_1a and β₄ depicting the homologous proximal C terminus (green box) and heterologous distal C terminus (blue box). Red bracket indicates the highly homologous sequence in the distal C terminus of β_1a revealed from sequence alignments of β_1a from several vertebrate species (fish to mammals) (SI Appendix, Fig. S2B). (E) Block scheme of domain organization of chimeras β₄/β_1a(prox.C) and β₄/β_1a(dist.C), where the proximal and distal C terminus of β₄ (orange) were exchanged by corresponding β_1a sequences (blue). (F) Q_max values were indistinguishable (P > 0.05) between relaxed myotubes expressing chimera β₄/β_1a(prox.C) (n = 11), β₄/β_1a(dist.C) (n = 19), or β_1a (n = 16). (G) Quantification of voltage dependence of cytoplasmic Ca²⁺ transients yielded (ΔF/F₀)_max values that were significantly lower (P < 0.001) for chimera β₄/β_1a(prox.C) (n = 15)- compared to β_1a (n = 9)-expressing relaxed myotubes. However, relaxed myotubes expressing chimera β₄/β_1a(dist.C) (n = 14) exhibited pronounced Ca²⁺ transients, equivalent (P > 0.05) to β_1a transfected myotubes (n = 13). (Right) Exemplar Ca²⁺ transient recordings from relaxed myotubes expressing chimera β₄/β_1a(dist.C) or β₄/β_1a(prox.C). (Scale bars, 50 ms [horizontal], ΔF/F₀ = 1 [vertical].) (H) Quantification of spontaneous or touch-evoked coiling of 27- to 30-hpf relaxed zebrafish injected with β_1a (n = 35), β₄ (n = 202), β₄/β_1a(C) (n = 79), and β₄/β_1a(dist.C) (n = 58) mRNA. Degree of motility was indistinguishable (P > 0.05) between relaxed zebrafish expressing β₄/β_1a(C) or β_1a. Relaxed zebrafish expressing β₄/β_1a(dist.C) displayed robust spontaneous coiling only slightly lower (P = 0.02) than β_1a. Conversely, β₄-injected relaxed zebrafish showed either no (n = 151) or very weak (n = 51) coiling following tactile stimulation and thus, highly significantly lower motility compared to (P < 0.001) β_1a-expressing relaxed zebrafish. Uninjected relaxed zebrafish displayed neither spontaneous nor tactile-induced motility (P < 0.001, n = 28). Error bars indicate SEM. P determined by unpaired Student’s t test, *P < 0.05; ***P < 0.001.

Upon transfection in relaxed myotubes, chimeras β₄/β_1a(prox.C) and β₄/β_1a(dist.C) were equivalent in their ability to support full membrane expression of functional DHPRs as indicated by full charge movement restoration (Q_max: 9.25 ± 0.81, n = 12 and 10.21 ± 0.92, n = 19, respectively) comparable (P > 0.05) to β_1a (Q_max: 10.28 ± 1.07, n = 16) (Fig. 4F). However, chimera β₄/β_1a(prox.C) did not restore Ca²⁺ transients above the β₄ level [(ΔF/F₀)_max: 1.06 ± 0.10, n = 15 and 1.02 ± 0.15, n = 13, respectively; P > 0.05] (Fig. 4G). In contrast to the proximal C-terminal construct, chimera β₄/β_1a(dist.C) led to complete restoration of cytoplasmic Ca²⁺ transients comparable (P > 0.05) to β_1a [(ΔF/F₀)_max: 2.17 ± 0.25, n = 14 and 2.30 ± 0.15, n = 9, respectively] (Fig. 4G). Furthermore, we performed motility tests on 27- to 30-h postfertilization (hpf) whole zebrafish. In zebrafish expressing β_1a, the degree of motility was high, whereas in those expressing β₄ it was only marginally greater than in the relaxed zebrafish (Fig. 4H). Also congruent to the Ca²⁺ transient data (Fig. 4C), the degree of motility restored was indistinguishable between relaxed zebrafish expressing β_1a and chimera β₄/β_1a(C) (both 4.00, n = 35 and n = 79, respectively) (Fig. 4H). Moreover, chimera β₄/β_1a(dist.C) resulted in a high extent of zebrafish motility (3.13 ± 0.24, n = 79), nearly (P = 0.02) reaching β_1a and β₄/β_1a(C) levels, but highly significantly (P < 0.001) above the very marginal β₄-induced motility (0.26 ± 0.06, n = 202) (Fig. 4H).

After determining that both β₄/β_1a(C) and β₄/β_1a(dist.C) restored EC coupling Ca²⁺ transients that differed little from that in muscles of wild-type (WT) animals, we next assessed their ability to cause the tetradic organization of DHPRs. We found that tetrads were present in relaxed myotubes expressing either β₄/β_1a(C) (Fig. 5A) and β₄/β_1a(dist.C) (Fig. 5B), in contrast to the absence of tetrads in relaxed myotubes expressing β₄ (Fig. 1 A, Lower). For a more quantitative comparison, unidentified images were provided to two investigators who counted the number of tetrads that were complete (four particles) or nearly complete (three particles). They were able to identify almost no tetrads in myotubes expressing β₄, but found that tetrads were present in myotubes expressing β₄/β_1a(C) at levels only slightly lower than in myotubes from normal animals (Fig. 5C). They detected tetrads in myotubes expressing β₄/β_1a(dist.C) at levels about half those of normal myotubes but still substantially above those of myotubes expressing β₄ (Fig. 5C).

Fig. 5.

The distal C terminus of β_1a is crucial for DHPR tetrad formation. (A and B) Representative freeze-fracture replicas from tail muscle tissue of 27- to 30-hpf relaxed zebrafish expressing β₄/β_1a(C) (A) or β₄/β_1a(dist.C) (B) reveal accurate arrangement of DHPR particles in tetrads. The red dots (Bottom) indicate the centers of three- or four-particle tetrads and additional particles that are in the expected position for an orthogonal array. (Scale bar, 50 nm.) (C) Numbers of tetrads (three or four particles) determined by two independent investigators from 95 anonymized freeze-fracture images acquired from zebrafish tails, either normal controls (normal), uninjected (relaxed), or injected with β₄, β₄/β_1a(C), or β₄/β_1a(dist.C) mRNA. Each bar represents mean of the counts normalized to normal zebrafish (where the mean of the two investigators’ counts was defined as 100%) and the two arrows (red and green) depict the counts of the two individual investigators (SI Appendix, Table S2). (D) Counts of DHPR particles per junction from zebrafish tails, either uninjected (relaxed), injected with β₄, β₄/β_1a(C), or β₄/β_1a(dist.C) mRNA, or normal controls (normal). Error bars indicate SEM. P determined by unpaired Student’s t test, *P < 0.05; **P < 0.01.

A count by one of the two investigators of the average number of DHPR-like particles per putative junction in unidentified images (Fig. 5D) indicated that the accumulation of DHPRs in the junctions of muscles expressing β₄ is comparable (P > 0.05) to what was observed in uninjected relaxed zebrafish (9.05 ± 1.04 and 12.37 ± 1.35 particles/junction, n = 18, respectively). At the other end of the spectrum, β₄/β_1a(C) expressing zebrafish show a similar number (P > 0.05) of DHPR-like particles per junction to that found in normal zebrafish muscles (19.32 ± 2.43 and 20.12 ± 2.03 particles/junction, n = 18, respectively). In the case of β₄/β_1a(dist.C), the clustering of DHPR-like particles in putative junctions (17.37 ± 1.36 particles/junction, n = 18) was substantially higher (P < 0.01) than that of uninjected relaxed zebrafish and β₄-injected zebrafish, and comparable (P > 0.05) to what was measured in normal zebrafish tail muscles. These data suggest that the C-terminal domain of β_1a substantially contributes to increase the efficiency of DHPR junctional targeting, a contribution that is dependent on its distal part. Since junctional particle density was earlier shown to be independent of the fact of whether DHPR particles are organized in tetrads or not (11), the observed differences in particle counts per junction mirrors the differences in sizes of the junctions.

The Hydrophobic Surface Motif (L₄₉₆L₅₀₀W₅₀₃) in the Distal β_1a C Terminus Is Not Essential for EC Coupling.

The results described so far demonstrate that the distal C terminus of β_1a plays a critical role in the physical interactions between the DHPR and RyR1, which are responsible for cytoplasmic Ca²⁺ transients and tetrad formation. As to why this might be, one possibility is that the β_1a C terminus adopts a structure specifically suited for this role. Unfortunately, the structure of the β_1a C terminus has not been resolved in the cryo-EM studies (26). However, the predicted secondary structures of the distal C termini of β_1a and β₄ are very similar (SI Appendix, Fig. S3) despite low sequence homology. Even with an overall similar structure, a more limited motif within the distal C terminus β_1a could be of importance. One candidate for such a role is a hydrophobic surface identified in previous work from other laboratories. In particular, using NMR spectroscopy, affinity chromatography, and RyR1 single-channel recordings in lipid bilayers, Karunasekara et al. (21) showed that a peptide corresponding to the distal 35 residues of the β_1a C terminus adopted a nascent α-helix, in which three hydrophobic residues (L₄₉₆L₅₀₀W₅₀₃) (Fig. 6A) align to form a hydrophobic surface that binds to isolated RyR1 with high affinity and increases its channel activity. This effect declined significantly upon substitution of the hydrophobic residues by alanines, a swap that did not destroy the α-helical structure (21). In a follow-up study of Hernández-Ochoa et al. (27), application of a peptide corresponding to the truncated β_1a C terminus (V₄₉₀–A₅₀₈), which contained the hydrophobic LLW motif, caused a similar increase of RyR1 channel activity in lipid bilayers. Perfusion of this 19-residue peptide into murine adult skeletal muscle fibers significantly increased cytoplasmic Ca²⁺ transients, which was not observed with a scrambled control peptide. Consequently, the authors of both the studies concluded that the hydrophobic motif L₄₉₆L₅₀₀W₅₀₃ is critical for EC coupling.

Fig. 6.

Hydrophobic residues (L₄₉₆L₅₀₀W₅₀₃) in the β_1a distal C terminus are not important for skeletal muscle EC coupling. (A) Amino acid sequence of rabbit β_1a C terminus depicting the position of the three hydrophobic residues LLW (red box with yellow filling), which were exchanged with alanines (AAA). (B) Relaxed myotubes expressing triple mutant β_1a(LLW-AAA) (n = 16) displayed Q_max values similar (P > 0.05) to β_1a (n = 16). (Right) Exemplar charge movement recording from relaxed myotubes expressing β_1a(LLW-AAA). (Scale bars, 5 ms [horizontal], 3 pA/pF [vertical].) (C) Plots of voltage dependence of maximal Ca²⁺ transients were indistinguishable (P > 0.05) between β_1a(LLW-AAA) (n = 13) and β_1a (n = 9)-expressing relaxed myotubes. (Right) Exemplar Ca²⁺ transient recordings from relaxed myotubes expressing mutant β_1a(LLW-AAA). (Scale bars, 50 ms [horizontal], ΔF/F₀ = 1 [vertical].) Error bars indicate SEM. P determined by unpaired Student’s t test.

To test the importance of the LLW motif, we generated the mutant construct β_1a (LLW-AAA), in which the LLW motif was ablated by substitution with alanines (Fig. 6A) and expressed it in zebrafish relaxed myotubes. Whole-cell patch-clamp recordings revealed that charge movement restored by the mutant construct β_1a(LLW-AAA) (Q_max: 9.93 ± 0.89 nC/μF, n = 16) was not distinguishable (P > 0.05) from that restored by β_1a (10.28 ± 1.07 nC/μF, n = 16) (Fig. 6B), indicating that the triple alanine substitution did not affect the membrane expression of functional DHPRs. Moreover, there were no significant differences (P > 0.05) in cytoplasmic Ca²⁺ transients (Fig. 6C) between relaxed myotubes expressing β_1a(LLW-AAA) or β_1a with respect to either magnitude [(ΔF/F₀)_max of 2.31 ± 0.27, n = 13 and 2.30 ± 0.15, n = 9, respectively] or voltage dependence (V_1/2 of −3.46 ± 1.82 mV, n = 13, and −6.87 ± 2.65 mV, n = 9, respectively). Thus, in contrast to the isolated, freely floating peptides (21, 27), the substitution of alanines for the LLW motif had no detectable effect on cytoplasmic Ca²⁺ transients when introduced into full-length β_1a expressed as part of the DHPR complex in intact muscle cells. Therefore, our data provide strong evidence that the L₄₉₆L₅₀₀W₅₀₃ motif in the distal C terminus of the DHPRβ_1a subunit is not important for DHPR–RyR1 interaction that underlies skeletal muscle EC coupling.

Subsuming all our previous (15, 16) and current observations of the role of β-subunits in functional skeletal muscle DHPR expression, we postulate a molecular model of conformational modifications of DHPRα_1S by β-subunits (Fig. 7). In normal muscle cells at rest, DHPRα_1S appears to be anchored strongly to RyR1, which results in the arrangement of DHPRα_1S in tetrads aligned with RyR1 homotetramers. Functionally, this anchoring is a necessary precondition for coupling depolarization-driven conformational changes of DHPRα_1S to the activation of RyR1. In the relaxed (β₁-null) muscle cell, both the membrane-embedded hydrophobic core of α_1S and its cytoplasmic domains have nonfunctional conformations so that there is neither charge movement nor tetrad formation, respectively (Fig. 7A) and a complete lack of EC coupling. Expression of the β₄ isoform facilitates a conformation of DHPRα_1S, which is distinct from the completely nonfunctional conformation in the relaxed system (Fig. 7 A and B). In particular, upon expression of the β₄ isoform (Fig. 7B), domain cooperativity between the SH3 and the PXXP motif in the proximal C terminus induces steric rectification of the hydrophobic core region, enabling the voltage sensing/charge movement function (16). Nevertheless, β₄ expression is not sufficient to promote accurate conformational restoration of the intracellular regions (loops and C and N terminus) of the α_1S subunit (Fig. 7B) and consequently is unable to restore full interaction of the DHPR complex with RyR1 resulting in greatly reduced Ca²⁺ transients and impaired tetrad formation. We observed similar behavior for the construct β₄/β_1a(prox.C) in which the proximal C terminus of β₄ is replaced with β_1a sequence (Fig. 7C). However, some anchoring of α_1S to RyR1 must occur for both β₄ and β₄/β_1a(prox.C) because both these β-constructs supported depolarization-induced calcium transients, although attaining peak levels that were only ∼40% of those for β_1a. A reasonable explanation for the reduced size of these transients is that the anchoring of α_1S to RyR1 is weaker for β₄ and β₄/β_1a(prox.C), which would also explain why these constructs did not result in tetradic arrays of α_1S. In particular, if the probability of α_1S binding to one subunit of RyR1 were 40% relative to β_1a, then the probability of three- and four-particle tetrads would be only 6.4% and 2.6%, respectively.

Fig. 7.

Model of conformational modification of α_1S by the β_1a distal C terminus—prerequisite for proper skeletal muscle EC coupling. (A) In zebrafish mutant relaxed due to the absence of the DHPRβ_1a subunit, the α_1S subunit is in a distorted conformation. This causes impediment of charge movement (Q) and of arrangement of DHPR into tetrads (tetrads) that accounts for the lack of skeletal muscle EC coupling (ECC). The distorted conformation of the membrane spanning hydrophobic core regions of the four homologous α_1S repeats (I–IV) is depicted by rectangular boxes. The primary and unspecified numbers of secondary α_1S-specific RyR1 interaction sites (32) are indicated with bold and normal black arrows, respectively. (B) β₄ is unable to reinstate full EC coupling [(+)/−] due to impaired DHPR tetrad formation. According to our model, β₄ (symbolized in orange) induces proper conformation of the hydrophobic α_1S core regions (depicted with cylinders) required for charge movement function, but is unable to reconstitute accurate conformation of the intracellular α_1S loops facilitating RyR1 anchoring (tetrad formation). Improper DHPR–RyR1 interaction (tilted arrows) leads to weak EC coupling and impaired tetrad formation. (C) Likewise, chimera β₄/β_1a(prox.C) in which the proximal C terminus of β₄ is swapped with corresponding β_1a sequence (blue), was unable to reinstate intact tetrad formation and thus full ECC. Yellow dots on the proximal C terminus of the β-subunit depict the intramolecular SH3–PXXP interaction sites critical for charge movement function (16). (D) However, the distal C terminus of β_1a (blue) enables proper conformation of the intracellular α_1S loops crucial for RyR1 anchoring (tetrad formation). Consequently, EC coupling is highly restored upon expression of chimera β₄/β_1a(dist.C). The direct DHPR–RyR1 interaction depicted in the model is still obscure. However, it is irrelevant for our conclusions whether the two channels interact directly or via an intermediate protein.

In stark contrast to the proximal β_1a C terminus, the distal C terminus of β_1a in chimera β₄/β_1a(dist.C) enables appropriate tertiary conformation of the β-subunit (Fig. 7D), apt for induction of an accurate conformation of the intracellular molecular regions (loops and termini) of the α_1S subunit. This conformational correction finally enables accurate anchoring of the DHPR to RyR1, allowing proper DHPR tetrad formation in orthogonal arrays strictly adjacent to every other RyR1 homotetramer—a key structural basis for full skeletal muscle EC coupling.

Since our results (Fig. 6) indicate that the previously proposed hydrophobic surface motif (L₄₉₆L₅₀₀W₅₀₃) in the distal β_1a C terminus (21) is not important for skeletal muscle EC coupling, the question remains as to which of the 35 residues of the distal β_1a C terminus are most directly involved for restoring interactions with RyR1. An alternative interaction motif to LLW might be formed by the first 10 residues of the distal C terminus, highly homologous in different β_1a distal C termini from fish to mammals (SI Appendix, Fig. S2). However, several motif search routines on various sequence databases did not yield promising motif predictions that would justify a targeted alanine replacement strategy on the β_1a distal C terminus. Particularly, we could not identify encouraging sequence homologies or motif identities in the C-terminal regions of β_1a and β₃, the only β-subunit beside β_1a that also promotes tetrad formation (SI Appendix, Fig. S1).

As mentioned above, one could also postulate that a specific secondary structure, adopted only by the distal β_1a C terminus might induce the conformation of the intracellular molecular regions (loops and N and C terminus) of the DHPRα_1S subunit required for accurate anchoring of the DHPR onto RyR1 (Fig. 7D). However, an algorithm that predicts the structure of isolated peptides (28) yielded structures that were roughly similar for the distal C terminus of β_1a and the corresponding region of β₄ despite the low amino acid sequence homology (SI Appendix, Fig. S3). Within the full-length proteins, AlphaFold2 predicts that β_1a residues V₄₉₀ to L₅₀₀ are alpha helical, whereas β_1a residues S₅₀₁ to M₅₂₄ are unstructured, as are all the corresponding residues (S₄₃₄ to K₄₆₈) of β₄ (β_1a: https://alphafold.ebi.ac.uk/entry/P19517; β₄: https://alphafold.ebi.ac.uk/entry/D4A055), but the confidence of the predictions for both β_1a and β₄ ranges from low to very low.

In summary, we found in this study that the heterologous distal C terminus of β_1a (amino acid residues V₄₉₀ to M₅₂₄) is critical both for arrangement of DHPRs into tetradic arrays and for full restoration of EC coupling Ca²⁺ release. We could exclude a proposed motif, consisting of the three amino acids L₄₉₆L₅₀₀W₅₀₃ (21) as relevant for accurate DHPR–RyR1 interaction and thus, tetrad formation. Because the currently available alignment and predictive methods did not identify a specific motif or structure, future studies with an allover alanine scan of the distal C terminus of β_1a may be necessary for identifying the motif(s)/structure(s) responsible for the key structural prerequisites for EC coupling—DHPR tetrad formation.

Materials and Methods

Zebrafish Care.

Breeding and maintenance of adult zebrafish, WT, and heterozygous for the DHPRβ₁-null mutation relaxed (red^ts25) (11) were performed according to established protocols (29, 30). One-day-old postfertilization homozygous relaxed zebrafish were recognized by their inability to move in response to tactile stimulation. Motile, heterozygous, and WT siblings, termed “normal” were used as controls. All experimental procedures were approved by the Tierethik-Beirat of the Medical University of Innsbruck and Bundesministerium für Bildung, Wissenschaft und Forschung (BMBWF-66.011/0140-V/3b/2019).

Expression Plasmids.

Detailed cloning strategies for generation of GFP-tagged cDNAs of β-subunits, chimeras, and mutants are described in SI Appendix, SI Materials and Methods.

Primary Culture of Myotubes.

Myoblasts from 1-dpf relaxed zebrafish were isolated, transfected with 2 µg of plasmid cDNA using the Rat Cardiomyocyte Neonatal Nucleofector Kit (Lonza) and cultured in L-15 medium supplemented with 3% fetal calf serum, 3% horse serum, 4 mM L-glutamine, and 4 U/mL penicillin/streptomycin for 4 to 6 d in a humidified incubator at 28.5 °C (30).

Whole-Cell Patch-Clamp Electrophysiology.

Recordings of intramembrane charge movement as a measure of functional DHPRα_1S membrane expression simultaneously with cytoplasmic Ca²⁺ transients were performed on transfected GFP-positive myotubes as previously described (30). Borosilicate glass patch pipettes had a resistance of 3.5 to 5 MΩ when filled with internal solution containing (in millimolar): 100 Cs-aspartate, 10 Hepes, 0.5 Cs-ethylene glycol-bis(-aminoethyl ether)-N,N,N',N'-tetracetic acid, 3 Mg-ATP, and 0.2 Fluo-4 (pH 7.4 with CsOH). N-benzyl-p-toluene sulphonamide, Myosin-II blocker (100 μM) was continuously present in the bath (external) solution containing (in millimoles): 10 Ca(OH)₂, 100 L-aspartate, and 10 Hepes (pH 7.4 with tetraethylammonium hydroxide). All recordings were performed at room temperature (RT).

mRNA Injection and Freeze-Fracture Electron Microscopy.

Freshly spawned zebrafish embryos were microinjected with in vitro synthesized RNA of GFP-tagged β-subunits, chimeras, or mutants and raised at 28 °C. At 27- to 30-hpf, tails of GFP-positive homozygous relaxed zebrafish were fixed in 9% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.2) for 30 min at RT and preserved in 4.5% glutaraldehyde at 4 °C. Tails were mechanically skinned, infiltrated in 30% glycerol in water, fractured in double replica holders, and shadowed with platinum at an angle of 45°, followed by replication with carbon, in a freeze-fracture unit (BFA 400, Balzers S.P.A.) (15). The replicas were examined at the electron microscopy facility of the University of Colorado, Anschutz Medical Campus, using a Tecnai FEI TF20 electron microscope.

Zebrafish Motility Analysis.

At 17 hpf, normal zebrafish exhibit slow, spontaneous coiling movements and by 21 hpf, multiple coils of the body in response to tactile stimulation can be observed (31). GFP-positive relaxed zebrafish, 27- to 30-hpf, injected with β-subunits, chimeras, or mutants were dechorionated using pronase and spontaneous or touch-evoked motility was visually evaluated and degrees of motility were judged according to an assigned scheme (SI Appendix, Table S1). Identification and confirmation of the rescued homozygous relaxed zebrafish were performed via restriction fragment length polymorphism (RFLP) test (30).

Statistical Analysis.

Data were analyzed using ClampFit (v10.7, Axon Instruments) and SigmaPlot (v11.0, Systat software, Inc.). Results are expressed as mean ± SEM and n = number of myotubes or individual zebrafish. Statistical significance was calculated using unpaired Student’s t test and P values were set as follows: *P < 0.05, **P < 0.01, and ***P < 0.001.

Data Availability

All study data are included in the article and/or SI Appendix.