Does a lack of RyR3 make mammalian skeletal muscle EC coupling a ‘spark-less’ affair?

In mammalian skeletal muscle, cellular depolarization initiates calcium (Ca²⁺) release from the sarcoplasmic reticulum (SR), which activates contraction during the process of excitation–contraction (EC) coupling. The Ca²⁺ release from the SR occurs through the ryanodine receptor (RyR) Ca²⁺ release channels. RyR type 1 (RyR1), the predominant isoform found in adult mammalian muscle, is directly controlled by membrane potential through protein–protein interactions (i.e. voltage initiated Ca²⁺ release, VICR). RyR3 is found at high levels during development but is limited in adult mammalian skeletal muscle. In contrast to RyR1, it is less clear how RyR3 is activated during depolarization and this is the focus of the paper by Legrand et al. (2008) in the current issue of The Journal of Physiology: how do RyR1 and RyR3 contribute to the elementary local Ca²⁺ signals in muscle called ‘Ca²⁺ sparks’?

Ca²⁺ sparks were first discovered in heart (Cheng et al. 1993), and then in both smooth muscle (Nelson et al. 1995) and amphibian skeletal muscle (Tsugorka et al. 1995; Klein et al. 1996). Ca²⁺ sparks appear as brief (∼30 ms), spatially restricted (∼2 μm) Ca²⁺ signals arising from a cluster of RyRs. In heart, voltage-dependent Ca²⁺ influx across the sarcolemma through L-type Ca²⁺ channels (dihydropyridine receptor, DHPR) activates a cluster of RyR2 channels through local Ca²⁺-induced Ca²⁺ release (CICR) to trigger a Ca²⁺ spark. With increasing DHPR activation, the Ca²⁺ spark frequency rises to produce the cell-wide macroscopic [Ca²⁺]_i transient that activates contraction.

In contrast, amphibian skeletal muscle, which contains both RyR1 and RyR3 homologues (RyR α, β), exhibits ‘dual mode’ RyR activation: VICR and CICR are apparent. VICR, which is most clearly visualized at very low levels of depolarization, is devoid of ‘Ca²⁺ sparks’ and thought to result from direct DHPR-RyR1 activation (Shirokova & Rios, 1997). The second mode involves CICR and is composed of Ca²⁺ sparks activated from the Ca²⁺ arising from the ‘mode 1’ activated RyR1s (Klein et al. 1996; Shirokova & Rios, 1997). Detailed evaluation of voltage elicited Ca²⁺ sparks in amphibian skeletal muscle confirms a ‘dual mode’ of activation by identifying a voltage-dependent, low amplitude, spatially restricted Ca²⁺ signal underscoring the Ca²⁺ spark (Gonzalez et al. 2000).

Structural detail of RyR localization within the SR–transverse tubule junction (Felder & Franzini-Armstrong, 2002) reveals that in addition to the RyR1 located in the triad junction, skeletal muscles expressing both RyR1 and RyR3 (i.e. amphibian, avian and embryonic/neonatal mammalian and limited adult mammalian) express RyR3 in a ‘parajunctional’ arrangement: an anatomical specificity for spatially localized RyR signalling. A conservative model in which voltage activated RyR1 triggers parajunctional RyR3 during a macroscopic global Ca²⁺ release as well as in a local Ca²⁺ spark is generally accepted. In this model both RyR1 and RyR3 would be requisite for the genesis of a voltage elicited Ca²⁺ spark in skeletal muscle (see Ward & Lederer, 2005).

In adult mammalian myofibres solely expressing RyR1, voltage elicited Ca²⁺ sparks are not present. This finding is most consistent with the lack of the RyR3 protein expressed. Although RyR3 has been shown in some adult mammalian myofibres (e.g. diaphragm/soleus) with the same parajunctional specialization as in amphibian muscle, a direct assessment of the role of RyR3 in the occurrence of voltage elicited Ca²⁺ sparks in adult mammalian muscle is difficult due to low levels of total RyR3 expression and our inability to identify the limited number of myofibres within a muscle that express RyR3.

In this issue of The Journal of Physiology,Legrand et al. (2008) directly address the role of RyR3 in voltage elicited EC coupling. In their model, either RyR1 or RyR3 protein was overexpressed in adult mammalian skeletal myofibres in which RyR1 was the sole native isoform. These authors demonstrate that in contrast to RyR1 overexpression, which had no apparent effect, the overexpression of RyR3 endowed the myofibre with spontaneous macroscopic CICR behaviour and a secondary CICR amplification of the voltage elicited calcium transient, behaviour similar to that observed in amphibian muscle (Klein et al. 1996; Shirokova & Rios, 1997). Despite the gross alterations in EC coupling behaviour, voltage-dependent Ca²⁺ sparks were not seen during RyR3 overexpression. This is the opposite of results reported by Pouvreau et al. (2007) using a similar RyR3 overexpression strategy in the same muscle type where voltage elicited Ca²⁺ sparks were observed.

Resolution of the differences in the studies of Pouvreau et al. 2007) and Legrand et al. (2008) is not simple. Numerous methodological differences must be considered including the delivery of the RyR cDNA (i.e. electroporation methodologies, time of measurement after gene delivery), the cellular conditions under which the electrophysiology experiments took place (e.g. dye loading, Ca²⁺ buffering), and the relative expression level and localization of RyR3 protein secondary to overexpression and more. The apparently contradictory findings regarding Ca²⁺ sparks of two outstanding groups doing state-of-the-art experiments suggest to us that the 3-D molecular and physical architecture of the junctional SR and its very nearby regions are exquisitely sensitive to how molecular manipulations are carried out. The functional consequences of the types of RyRs present and their organization is great as suggested by Felder & Franzini-Armstrong (2002). The question of how RyR3 contributes to the details of EC coupling is certainly better understood now but not yet answered.