Silencing genes of sarcoplasmic reticulum proteins clarifies their roles in excitation–contraction coupling

In skeletal muscle, the release of Ca²⁺ ions from the sarcoplasmic reticulum (SR) through ryanodine receptor ion channels (RyR1s) into the cytoplasm leads to muscle contraction. Skeletal muscle SR has two structurally and functionally distinct regions, the longitudinal and terminal cisternae SR. The longitudinal SR is a sarcotubular network that surrounds the myofibrils and contains a SR Ca²⁺ pump (SERCA) responsible for the re-uptake of Ca²⁺ by the SR. The terminal SR is mainly involved in Ca²⁺ release and contains RyR1 as a large multi-protein complex composed of four 560 kDa RyR and four 12 kDa FK506 binding protein (FKBP) subunits (Franzini-Armstrong & Protasi, 1997; Meissner, 2002). In addition to FKBP, RyR1 is associated with at least three other SR proteins, calsequestrin, triadin and junctin. This perspective discusses the work of our laboratory on calsequestrin, triadin and junctin, in view of the symposium articles by Knollmann (2009), Protasi et al. (2009), Royer & Rios (2009), Marty et al. (2009) and Pritchard & Kranias, 2009 in this issue of The Journal of Physiology.

As reviewed by Marty et al. (2009) and Pritchard & Kranias (2009), triadin and junctin are two related integral membrane proteins that directly interact with RyR1 (Lee et al. 2004; Goonasekera et al. 2007). Mice lacking triadin showed changes in Ca²⁺ handling but survived to adulthood (Shen et al. 2007). CSQ is a SR lumenal low-affinity Ca²⁺ binding protein that is thought to increase Ca²⁺ storing capacity (MacLennan & Wong, 1971; Meissner et al. 1973; reviewed by Royer & Rios 2009 in this issue). Consistent with its high Ca²⁺ binding capacity, overexpression of CSQ1 in skeletal muscle-derived C₂C₁₂ myotubes increased Ca²⁺ store size (Shin et al. 2003). CSQ1-null mice are viable but show major SR structural changes and reduced SR Ca²⁺ release (Paolini et al. 2007). In addition, CSQ1 has been reported to modulate RyR1 channel activity directly and through RyR associated proteins such as triadin and junctin (Beard et al. 2004). To better understand the roles of triadin, junctin and calsequestrin in stored Ca²⁺ release, we knocked down each or a combination of these proteins using siRNA methodology in mouse skeletal muscle derived C₂C₁₂ myoblasts. Because the C₂C₁₂ myoblasts also express the cardiac isoform of CSQ (CSQ2), we also studied the role of CSQ2. Thus, C₂C₁₂ myoblasts were infected with double-stranded AAV plasmids encoding siRNAs specific for triadin, junctin, CSQ1 and/or CSQ2 and allowed to differentiate into myotubes over 8 days (Wang et al. 2006, 2009).

Immunoblot and immunofluorescence analysis in 8-day-old myotubes either singly or doubly infected with siRNAs specific for triadin or junctin showed a reduction in the protein levels by 80% and 100%, respectively (Wang et al. 2009). Junctin knockdown (but not triadin knockdown) reduced RyR1 and CSQ1 protein levels but not those of SERCA, Cav1.2 α1 subunit or CSQ2. Knocking down triadin or junctin reduced Ca²⁺ release induced by KCl depolarization (in Ca²⁺-free buffer) by 20–25%. Knocking down both proteins reduced release by 35%, thus suggesting that these two proteins have non-overlapping roles in Ca²⁺ release of these cells. Pharmacological agents (caffeine, 4-chloro-m-cresol, thapsigargin) indicated that junctin knockdown reduced SR Ca²⁺ store size, whereas store size was maintained in triadin knockdown myotubes. This study therefore strengthens the consensus described by Marty et al. (2009) of a triadin role in facilitating depolarization-induced Ca²⁺ release, and suggests that junctin is required to maintain Ca²⁺ store size.

Delivery of siRNAs specific for CSQ1 or CSQ2 reduced the protein levels of the two CSQs close to background levels, while affecting the morphology and function of 8-day C₂C₁₂ myotubes differently (Wang et al. 2006). C₂C₁₂ myotubes were shorter and wider upon CSQ2 but not CSQ1 knockdown. RyR and SERCA protein levels were reduced in CSQ2 and dual knockdown (CSQ1 and CSQ2) cells. No differences were observed in the protein levels in CSQ1 knockdown cells compared to controls. The expression levels of other proteins including triadin and junctin were maintained in both cases. Ca²⁺ release induced by caffeine, 4-chloro-m-cresol or KCl in Ca²⁺-free buffer was significantly decreased in CSQ2 and CSQ1/CSQ2 knockdown C₂C₁₂ myotubes, whereas there were no significant changes in CSQ1 knockdown C₂C₁₂ cells.

To determine whether the two CSQ isoforms differently regulate RyR1, single channel measurements were performed using the lipid bilayer method. RyR channels from membrane vesicles isolated from double knockdown myotubes were activated by CSQ1 and CSQ2 when added to the lumenal side of the bilayer. Activation was dependent on the presence of lumenal Ca²⁺. In contrast, the purified RyR1, which was separated from junctin and triadin, was not activated by lumenal CSQ1. This suggests that the functional interaction of RyR with CQS depends on other SR junctional proteins such as triadin and junctin.

In summary, the results from our analysis of silencing SR proteins in C₂C₁₂ myotubes demonstrate several functional differences in two sets of SR proteins. Our observations predict that the two related RyR1-associated proteins, triadin and junctin, have non-redundant roles in forming a functional SR Ca²⁺ release complex. Knockdown of skeletal and cardiac CSQ isoforms suggests that the cardiac CSQ2 is required to maintain normal C₂C₁₂ morphology and Ca²⁺ store size. While a reduction in store size is not observed in cardiomyocytes devoid of CSQ2, deletion resulted in increased SR volume (Knollmann, 2009). Surprisingly, skeletal-type calsequestrin makes only a slight contribution to SR Ca²⁺ store size and Ca²⁺ release in C₂C₁₂ myotubes. Functional changes in triadin, junctin and CSQ knockdown myotubes correlated with alterations in RyR1 and SERCA protein expression levels. Considering the balance of evidence discussed in this symposium, it is clear that the details of Ca²⁺ buffering in the SR as well as the nature of the mechanisms that remodel Ca²⁺ handling in myotubes and adult cells with reduced amounts of these three proteins remain to be determined.