Structural Arrangement of the Sarcoplasmic Reticulum

The sarcoplasmic reticulum (SR) is a membrane-delimited intracellular organelle present in striated muscle of almost all species.¹ Its name literally means “network of structures located in the cytoplasm of a muscle fiber,” referencing the interconnected network of tubules and vesicles that spans the sarcomere and wraps up the contractile myofilaments. The SR is not continuous with the external membrane, but recent data indicate that it is continuous with the nuclear envelope.² In cardiac and skeletal muscle, the main function of the SR is to provide the majority of Ca²⁺ ions that are needed to activate the contractile proteins of the myofilaments, and to resequester Ca²⁺ from the myoplasm to allow for relaxation. The Ca²⁺ concentration within the SR ([Ca²⁺]_SR ) is approximately 1 mM, with a larger pool of Ca²⁺ bound to calsequestrin and other Ca²⁺-binding proteins.¹ Therefore, when myofilaments are relaxed and bathed in cytoplasmic milieu containing approximately 0.1 µM Ca²⁺, there is an approximately 10,000-fold [Ca²⁺] gradient across the SR membrane. The compartmentalization of muscle fibers into small (~2 µm) structural-functional units (sarcomeres) ensures that Ca²⁺ diffusion from release sites is not a limiting step in muscle contraction. Similarly, the enveloping of myofilaments by the SR ensures rapid reuptake of Ca²⁺, leaving rate-limit steps of relaxation to contractile proteins.

It is useful to compare the structural arrangement of the SR in cardiac and skeletal muscle to facilitate discussion of the mechanisms that trigger Ca²⁺ release in these two muscles. The series of events by which depolarization of the sarcolemma generates a mechanical contraction is termed excitation-contraction coupling (E-C coupling). Although common subcellular structures participate in E-C coupling, different processes link membrane depolarization to Ca²⁺ release in cardiac and skeletal muscle. Some of these differences stem in part from the ultrastructural arrangement of the muscle fiber’s organelles. It is well accepted that E-C coupling in skeletal muscle is greatly dependent on Ca²⁺ release from the SR, with little or no Ca²⁺ entry across the sarcolemma during a single twitch. Accordingly, the SR of skeletal muscle is robust and contains large saccular enlargements, whereas T-tubules, which are invaginations of the external membrane that carry the depolarizing stimulus to the interior of the cell, are narrow (~20 to 40 nm) and elongated. By contrast, Ca²⁺ entry across the sarcolemma is an obligatory step for E-C coupling in cardiac muscle.³ T-tubules, which not only carry the electrical stimulus to the interior of the cell but must serve also as Ca²⁺ reservoirs, are therefore larger in diameter (~200 nm) in cardiac muscle. The SR appears thinner and less organized than in skeletal muscle, but exhibits multiple points of contact with the sarcolemma and T-tubules.⁴ Depending on the animal species and the type of cardiac cell, the SR volume can be as high as 12% of cell volume (mouse atrium), to as little as 0.8% (finch ventricle).¹ These sharp differences likely reflect the relative importance of Ca²⁺ entry versus SR Ca²⁺ release for contraction among different cells: whereas mouse cardiomyocytes rely heavily on Ca²⁺ release from the SR, finch (and all avian) ventricular cells lack T-tubules altogether and thus rely almost entirely on Ca²⁺ fluxes across the sarcolemma.

Molecular Players of Excitation-Contraction Coupling

In both cardiac and skeletal muscle, the L-type Ca²⁺ channels or dihydropyridine receptors (DHPRs) are the voltage sensors of sarcolemma and T-tubules that initiate E-C coupling, and the SR Ca²⁺ release channels that provide the majority of Ca²⁺ for contraction are also known as ryanodine receptors (RyRs).⁴ The manner in which DHPRs trigger RyRs to open is particular to each muscle. Because skeletal muscle fibers can twitch in the absence of external Ca²⁺, and Ca²⁺ release follows sarcolemmal depolarization with virtually no delay, a physical coupling between DHPRs and RyRs is thought to initiate E-C coupling in skeletal muscle.⁵ Indeed, identification of the structural domains of DHPRs and RyRs involved in their mechanical interaction continues in earnest,⁶ although the presence of proteins “sandwiched” between these channels has not been completely ruled out. By contrast, RyRs of cardiac muscle are closely apposed but do not appear mechanically joined to DHPRs⁴; instead, they are activated by a small influx of external Ca²⁺ (the inward Ca²⁺ current provided by DHPRs, or I_Ca), releasing greater amount of Ca²⁺ from the SR.⁷ This process, Ca²⁺-induced Ca²⁺ release (CICR),⁸ is a signature event of RyRs but is not exclusive to cardiac muscle. In fact, CICR was first described in skeletal muscle,⁹ where it apparently serves to activate RyRs that are not coupled to DHPRs.¹⁰

Again, the structural arrangement of DHPRs and RyRs is congruent with the type of E-C coupling they exert in skeletal and cardiac muscle. In the former, a great proportion of DHPRs appears organized as tetrads that rest on top of a single RyR channel in the junctional SR, the specialized region of the SR that is proximal to T-tubules.⁴ Because RyRs are homotetrameric protein complexes (discussed later), each DHPR in a tetrad is presumed to interact with one RyR monomer. RyRs, on the other hand, are clustered in a paracrystalline lattice where they touch each other at the corners.¹¹ In this orderly array, every other RyR interacts with a tetrad of DHPRs and is directly activated by the physical coupling mechanism; the DHPR-free RyR is presumably activated by CICR and cooperative interaction between RyRs.^4,10 Multiple (i.e., 10 to 20) such DHPR-RyR interactions occur in a couplon, the structural domain of the junctional SR where E-C coupling actually takes place.¹² In cardiac muscle, DHPRs are irregularly and sparsely distributed (not forming tetrads) in a couplon, but RyRs keep their lattice array, thus changing the stoichiometry and the mode of interaction of these two channels. A typical couplon in cardiac myocytes can contain approximately 100 RyRs and only approximately 20 DHPRs.¹ Thus, in principle, one DHPR injects Ca²⁺ into a couplon to activate approximately 5 RyRs, but current estimates indicate that E-C coupling is normally effected with a high safety factor—that is, only a portion of the existing DHPRs is needed to fully activate RyRs¹³ (discussed later ).

In addition to the junctional SR, DHPRs and RyRs interact at the peripheral couplings, the points of contact between SR and sarcolemmal surface.⁴ Despite the fact that T-tubules contain the highest density of DHPRs, couplons of the peripheral couplings apparently have DHPR/RyR stoichiometry that is similar to those of junctional SR couplons. Thus, RyRs of peripheral couplings are expected to release Ca²⁺ by a model similar to that described earlier for RyRs of junctional SR, although this has not been tested rigorously. A different mechanism, however, must operate to activate RyRs in the corbular SR, a basketlike form of SR that is connected to the SR network at one point only, and completely lacking T-tubular or sarcolemmal contacts. Activation of corbular RyR channels probably requires a combination of Ca²⁺ diffusion from release sites and a propagating wave of CICR. Corbular SR can contain up to approximately 35% of the total RyRs and is particularly prominent in atrial cells and Purkinje fibers.¹⁴

Reconstitution of Sarcoplasmic Reticulum Channels in Lipid Bilayers

Advancements in confocal imaging techniques have started to shed new light on SR ion channels in their cellular environment, but at present the majority of their functional and biochemical properties have been defined using isolated SR vesicles. Homogenization of cardiac and skeletal muscle produces fragmented SR that can be partially segregated by sucrose density centrifugation into RyR-enriched “heavy” SR (corresponding to junctional SR and peripheral couplings), and Ca²⁺ pump–enriched “light” SR (corresponding to longitudinal SR). Because RyRs are confined to regions of the cell that are largely inaccessible to patch electrodes, reconstitution of heavy SR into artificial lipid bilayers remains the most versatile and effective method to characterize the biophysical and pharmacologic properties of RyRs at the single channel level.15,16 A vast body of knowledge has already been gained with this technique,^16,17 which remains the method of choice in situations that require precise control of modulators on both faces of the channel (e.g., in cytosolic or luminal Ca²⁺ titrations), and in measurements of unitary channel conductance under quasi-physiologic conditions. In addition, defining the potency and mechanism (open- or closed-channel block), which is indispensable for the rational design of drugs for various RyR-linked diseases, would be extremely difficult without single-channel recordings in lipid bilayers. Nevertheless, it is important to recognize the limitations of this technique, most of which stem from the extremely difficult recreation of the conditions under which RyRs operate in situ. In their intracellular environment, RyRs are activated by fast and transient Ca²⁺ stimuli (as opposed to stationary Ca²⁺ levels as is routinely used in lipid bilayer experiments), which modify RyR activity importantly (discussed further later). Furthermore, RyRs seldom work in isolation, as in single-channel experiments, but are clustered in arrays that exhibit cooperativity and synergism.^11,12 Lastly, various accessory proteins of the RyR control its response to cytosolic and luminal Ca²⁺, and these proteins might not be present in reconstituted channels.

Confocal Imaging of RyR Ca²⁺ Fluxes

The optical detection of Ca²⁺ fluxes has provided new insight into the working principles of the supramolecular Ca²⁺ signaling system of many cells. The discrete, transient, and presumably elemental Ca²⁺ signaling events of cardiac myocytes, also known as Ca²⁺ sparks, are indisputable signs of RyR gating in situ.¹⁸ The introduction of fluorescein- and rhodamine-based Ca²⁺ indicators of high dynamic range and the arrival of low-cost versatile confocal microscopes greatly facilitated the discovery of Ca²⁺ sparks, first detected in ventricular myocytes¹⁹ and later in smooth²⁰ and skeletal²¹ muscle cells. The low myoplasmic [Ca²⁺] of resting cells keeps the open probability (P_o) of RyRs extremely low; still, RyR channels open with a finite rate that depends on several factors (most notably, [Ca²⁺] on the cytoplasmic and luminal side of the RyR), giving rise to spontaneous Ca²⁺ sparks. An estimate of spontaneous spark rates in ventricular myocytes was initially 100/cell/sec,¹⁸ which suggested an opening rate for RyRs of 10⁻⁴ s⁻¹, assuming that a typical ventricular myocyte contains approximately 1 million RyRs.¹⁹ However, spark rates vary widely among investigators apparently using the same conditions and can even be influenced by experimental artifacts such as cell damage during isolation, making this estimate less reliable. Initially, Ca²⁺ sparks were thought to originate from the opening of a single RyR,¹⁸ but the unitary RyR channel conductance in near physiologic conditions (~0.5 pA)²² appears to be too low to deliver the fluorescence signal mass of a typical Ca²⁺ spark.²³ Additional studies detected smaller RyR-originated Ca²⁺ signals (e.g., “Ca²⁺ quarks,”²⁴ “Ca²⁺ embers,”²⁵ “Ca²⁺ syntilla,”²⁶), somewhat demoting the elemental adjective bestowed on Ca²⁺ sparks as indivisible events of E-C coupling in cardiac cells. Because Ca²⁺ sparks typically give rise to a twofold increase in fluorescence intensity in an area of approximately 2 µm, it is likely that they result from the coordinated opening of a portion of (or all) RyRs that are clustered in a couplon (~100). Regardless of the number of RyRs that intervene to form a Ca²⁺ spark, these Ca²⁺ signaling events have brought fresh insights into the mechanisms that modulate the activity of RyRs in their intracellular environment.

[³H]Ryanodine Binding As an Index of RyR Activity

Measurements of RyR density and activity can be readily obtained by performing [³H]ryanodine binding assays in purified SR vesicles or even in whole-tissue homogenates. This is possible because of the high affinity and specificity of [³H]ryanodine for its receptor, which yield an excellent signal-to-noise ratio, and because the alkaloid binds to open RyR channels only.²⁷ The binding of [³H]ryanodine is enhanced by activators of Ca²⁺ release (Ca²⁺, adenosine triphosphate [ATP], caffeine) and decreased by inhibitors of Ca²⁺ release (Mg²⁺, H⁺, calmodulin), suggesting that the alkaloid binds to a conformationally sensitive domain of the RyR protein.²⁸ Therefore, [³H]ryanodine can be used as a probe of the functional state of the RyR channel. This approach has contributed to the purification of the RyR itself and to the identification of novel ligands, endogenous modulators, and accessory proteins of RyRs^17,27,29; however, ryanodine also displays some disadvantages as ligand of RyRs. Mainly, ryanodine has slow association and dissociation rates, which require long incubation times to reach equilibrium (>15 h at room temperature) and predisposes the RyR to protein degradation. Furthermore, its sluggishness to bind to its receptor renders ryanodine incapable of tracking dynamic changes in RyR activity, which are usually transient and swift (discussed later). Despite these drawbacks, [³H]ryanodine binding is the assay of choice to measure the averaged activity of literally billions of RyRs at the same time and under multiple conditions.

Electron Probe X-Ray Microanalysis of Ca²⁺ and Other Ions Within the Sarcoplasmic Reticulum

Ca²⁺ movements across the SR can be measured in situ by electron probe analysis of ultrathin cryosections of muscle rapidly frozen at different times in the contraction and relaxation cycle.³⁰ These pioneering experiments have provided critical data on the spatial and temporal movement of Ca²⁺ and its countercharge ions (mainly K⁺ and Mg²⁺) across the distinct regions of SR during the E-C coupling cycle. Foundational notions, such as the SR undergoing no charge deficit during Ca²⁺ release (and thus maintaining no transmembrane potential) and the terminal cisternae being the main Ca²⁺ release site of the SR, were derived largely from electron probe analysis experiments.³⁰

Regulation of Isolated RyR2 channels

Ca²⁺ is the most potent, versatile, and multifaceted regulator of RyR2 channels. Ca²⁺ ions turn on, turn off, and permeate through the RyR2. Ca²⁺ binds to cytosolic and luminal domains of RyR2s and regulates the activity of the channel in a concentration- and time-dependent manner. Because RyRs work cooperatively in their native environment and are activated by I_Ca, which is an extremely fast and transient Ca²⁺ stimulus, this mode of regulation is extremely complex.

It is necessary to discuss the effect of Ca²⁺ on isolated RyRs before considering the intricacies of Ca²⁺ regulation of RyRs in situ. Under stationary (nonfluctuating) [Ca²⁺], the activity of single RyR2 channels is a bell-shaped function of cytosolic [Ca²⁺] (Figure 6-3) because of the presence of Ca²⁺-activating and inactivating sites.^17,29,51 Ca²⁺ in the range of 0.1 to 10 µM binds to at least one Ca²⁺-binding domain that activates RyR2s. Higher [Ca²⁺] (100 µM to 3 mM) then inactivates the channel.⁵¹ The affinity and cooperativity of the activating and inactivating Ca²⁺-binding sites vary greatly under the presence of other relevant modulators (e.g., ATP, Mg²⁺, H⁺) and constitute powerful mechanisms by which posttranslational modifications of the channel protein (e.g., oxidation, phosphorylation, nitrosylation) modulate Ca²⁺ release. Among the RyR isoforms, RyR2 is particularly recalcitrant to Ca²⁺-dependent inactivation, requiring supraphysiologic [Ca²⁺] (~10 mM) for complete inactivation.^17,29,51 For this reason, the functional role of this process is questionable, although its participation in CICR termination has not been ruled out (discussed later). This overall picture represents the response of RyR2 channels to stationary levels of Ca²⁺, but is different from that obtained under dynamic (fast and transient) Ca²⁺ stimuli. In what are now landmark studies delineating CICR, Fabiato^8,52 observed that the magnitude of Ca²⁺ release was dependent on the velocity of the Ca²⁺ pulse (d[Ca²⁺]/dt) applied to skinned cardiac fibers. Similarly, single-channel studies have revealed that a fast Ca²⁺ stimulus elicits a transient increase in the P_o of RyR2s. If Ca²⁺ is applied rapidly but remains sustained for long periods, RyR2 P_o will relax, or adapt, to a new steady state that follows the activation curve obtained under stationary [Ca²⁺]. Thus, P_o–[Ca²⁺] curves of greater magnitude emerge from titration of RyR2 activity with fast, calibrated Ca²⁺ pulses compared with those obtained under similar but steady applications of Ca²⁺ (Figure 6-3).^53–55 The implication of these findings is that there are components of RyR2 activity that remain undetected in most of the assays presently used, and these aspects of RyR regulation might explain seemingly discrepant results, as in the case of RyR2 phosphorylation (discussed later).

Regulation of RyR2 channels by luminal (intra-SR) Ca²⁺ has gained preeminence as a control mechanism of Ca²⁺ release, but this process remains as complex as cytosolic Ca²⁺ regulation. The crucial observation in ventricular myocytes is that, beyond a certain threshold, small changes in SR Ca²⁺ load result in far greater increases in Ca²⁺ release, with the relationship described as an inverse hyperbole with a high degree of cooperativity⁵⁶ (see Figure 6-3). This is clearly an adaptive mechanism that promotes greater Ca²⁺ release (and stronger contractions) under conditions of increased Ca²⁺ uptake (such as occurs under β-adrenergic stimulation) and suggests that RyR2 channels increase their P_o in response to elevated luminal [Ca²⁺]. The latter has been detected in single channel experiments,^57,58 but the mechanism underlying luminal Ca²⁺ regulation at the molecular level remains uncertain and is possibly due to a combination of several of the following processes: (1) direct Ca²⁺ binding to RyR2 domains available only from inside the SR^58,59; (2) activation of cytosolic sites by the Ca²⁺ ions being permeated by the channel (feed-through mechanism)⁶⁰; and (3) binding of Ca²⁺ to calsequestrin and subsequent regulation of RyR2 through junctin, triadin, or both.^47,61,62

Regulation of RyR2 Channels In Situ

In each of the approximately 10,000 individual couplons that occur in a typical ventricular myocyte, RyR2 channels are packed forming a paracrystalline array that promotes synchronized Ca²⁺ release.4,11 Each of these couplons is independently activated by Ca²⁺ entering the cell through the juxtaposed DHPRs of the T-tubules (junctional SR) or sarcolemma (peripheral couplings).^1,4 The rapid spread of sarcolemmal depolarization into the interior of the cell by the T-tubules ensures that all individual couplons are simultaneously activated by I_Ca, generating a synchronized wave of Ca²⁺ release that spreads quickly to virtually all corners of the cardiomyocyte. In adult ventricular cells, I_Ca is insufficient to induce full contractions but is amplified threefold to eightfold (depending on the species) by Ca²⁺ release from the SR.¹ Because Ca²⁺ is simultaneously the input and output signal of RyR2s, this amplification process (CICR) should intuitively be self-regenerating and all-or-none.⁶³ In other words, the Ca²⁺ released by an RyR2 channel should activate further the same channel or its neighbors in an apparently interminable cycle. However, this is not observed experimentally; instead, CICR is finely graded by I_Ca and quickly terminates in intact cells. What counters the inherently positive feedback of CICR? At least three mechanisms have been proposed and, given the importance of CICR termination, it is likely that several mechanisms intervene (even redundantly) to avoid overflowing the cytosol with Ca²⁺.

Ca²⁺-dependent inactivation of Ca²⁺ release was envisioned by Fabiato in his studies of CICR.3,8,52 The simplicity of this scheme makes it elegant. It involves two types of Ca²⁺ sites controlling the RyR2 channel: a fast-action, low-affinity Ca²⁺ activation site and a high-affinity but slow-action Ca²⁺ inactivation site.⁸ Accordingly, a fast Ca²⁺ stimulus evokes a transient burst of RyR2 activity because the channel is activated by Ca²⁺ acting on the activation site and then shut down as Ca²⁺ slowly binds to the inactivation site. On the other hand, a slow Ca²⁺ stimulus binds to the higher-affinity inactivation site and prevents channel openings, at least until the stimulus is removed. At sustained and high Ca²⁺ levels, estimated by Fabiato to be approximately 100 µM, the inactivation sites are saturated, leaving the channel in an absorbing inactivated state.⁵² Thus, this straightforward scheme allows for activation of Ca²⁺ release by the fast I_Ca and termination of CICR by the [Ca²⁺] lingering in the couplon after Ca²⁺ release. Recovery (repriming) of RyR2 channels from this inactivated (refractory) state requires removal of Ca²⁺ from the couplon, as expected to occur on a beat-to-beat basis. As intelligible as it sounds, this mechanism has encountered experimental and theoretical challenges. First, SR Ca²⁺ release was shown to be independent of the interval between two consecutive Ca²⁺ stimuli,⁶⁴ which was not expected if RyR2 channels were refractory for a certain time after the first stimulus. In addition, when a sustained Ca²⁺ stimulus was used to inactivate the RyR2 channel, a subsequent Ca²⁺ stimulus reactivated Ca²⁺ release,⁶⁵ which again was unexpected of a refractory process. Perhaps the most irreconcilable data come from single-channel experiments in which RyR2 channels inactivate at [Ca²⁺] >1 to 3 mM, which are levels unlikely to be reached during a normal contraction. Still, it is possible that at least some level of Ca²⁺-dependent inactivation of Ca²⁺ release occurs in vivo, as mathematical models estimate the [Ca²⁺] in the dyadic cleft at levels that are compatible with those required for partial inactivation of single RyR2 channels.⁶⁶ In addition, RyR2 channels in situ may be more sensible to Ca²⁺ than those recorded under artificial environments.

SR Ca²⁺ depletion has attracted attention as a mechanism to terminate CICR, but nagging issues remain. Although it is clear that RyRs terminate Ca²⁺ release once the Ca²⁺ inside the SR drops to a certain threshold (~50% of the total Ca²⁺ inside the store),67,68 the mechanisms underlying luminal Ca²⁺ regulation of RyRs remains unclear. In his classical studies that characterized calsequestrin (the major Ca²⁺-binding protein of the SR), Ikemoto et al.⁶⁹ observed that the amount and the speed of Ca²⁺ release from SR vesicles depended directly on vesicular calsequestrin content rather than Ca²⁺ content, thus implying that calsequestrin “senses” the SR Ca²⁺ load and regulates the activity of RyRs. Boding well with this notion, Ca²⁺ sparks lasted longer or terminated prematurely after overexpression or partial depletion of calsequestrin levels in ventricular myocytes, respectively.⁷⁰ Furthermore, purified RyR2 channels reconstituted in lipid bilayers exhibited little activation by luminal [Ca²⁺], but the addition of calsequestrin and its “anchors” junctin and triadin restored luminal Ca²⁺ sensitivity.⁵⁷ Thus, the above data portrays calsequestrin as an indispensable component of the signaling mechanism that allows RyRs to terminate Ca²⁺ release upon [Ca²⁺]_SR depletion; however, other data bestow little role on this protein. For example, RyR2 channels expressed in heterologous systems (and bearing no calsequestrin) exhibit robust luminal Ca²⁺ regulation and are activated by Ca²⁺ overload (store-overload induced Ca²⁺ release).^58,59 In addition, in direct contradiction with some of these data, recombinant RyR2 channels (calsequestrin-free) reconstituted in lipid bilayers were activated by luminal [Ca²⁺],⁵⁹ implying that luminal Ca²⁺ regulation is an intrinsic property of the channel protein. As a result, it is clear that RyRs are capable of sensing dropping levels of [Ca²⁺]_SR to terminate Ca²⁺ release, but it is unclear whether they do it directly or through the calsequestrin-junctin-triadin ternary complex. As derangement of RyR refractoriness leads to Ca²⁺-dependent arrhythmias and may be involved in heart failure progression,^71,72 luminal Ca²⁺ regulation of RyRs appears to be critical in the overall scheme to terminate CICR and likely depends on multiple factors.

Stochastic attrition, the random closing of individual RyR2 channels, has been deduced from mathematical modeling as a potential mechanism to terminate CICR.⁶³ High local [Ca²⁺] gradients build up in the couplon immediately right after CICR, but they quickly dissipate.⁶⁶ If the number of RyR2 channels in a couplon was only one, it should be obvious that a random closure of that one RyR2 channel could be responsible for Ca²⁺ release termination (and dissipation of the [Ca²⁺] gradient). Stern⁶³ inferred that stochastic attrition could terminate local Ca²⁺ release if the number of RyR2 channels in a couplon was fewer than 10, but higher number of channels would decrease its probability precipitously. Because ultrastructural and functional data estimate this number to be close to 100,^1,18

Stay updated, free articles. Join our Telegram channel

Join