Triadin, not essential, but useful

As presented in the review article by Marty et al. (2009), triadins are a family of proteins that are derived from tissue specific alternatively spliced products of a single gene. All triadins share a common amino terminus and differ primarily in the C-terminal regions. What are triadins supposed to do?

The predominant isoforms of triadin are co-localized with RyR and calsequestrin, using immunohistochemistry and can be immuno-precipitated from heart and skeletal muscle with Anti-RyR and Anti-Casq antibodies (Jones et al. 1995; Zhang et al. 1997; Groh et al. 1999). Because of its ability to be immunoprecipitated with calsequestrin, triadin has been suggested by many studies to function as a protein that binds to calsequestrin to bring it to the mouth of the calcium release channel and that binds to the RyRs themselves to modulate their gating. In skeletal muscle key structural aspects that are involved in both associations and triadin's RyR1 binding site have been defined and the motifs in triadin were shown to be the same in the predominant isoforms expressed both in skeletal muscle (Trisk95) and in heart (CT1) (Kobayashi & Jones, 1999). A second calsequestrin binding protein, junctin, was found to be expressed in both skeletal muscle and heart in addition to triadin and together with triadin, calsequestrin and the RyR makes up a quartenerary complex that was thought to be essential for proper EC coupling. Additional isoforms of triadin (Trisk 51, Trisk49 and Trisk32) have been identified in skeletal muscle by one group (Marty et al. 2000; Vassilopoulos et al. 2005). Despite evidence of the expression of three mRNA isoforms in the heart, the predominant isoform in heart is CT1, and there are also traces of CT3 (Kobayashi & Jones, 1999).

It is important to note that there is some controversy associated with the expression of the additional isoforms in skeletal muscle related to the antibodies used for their detection. Trisk 95, 51 and 32 have been identified in rat and rabbit skeletal muscle with an N-terminal antibody raised to a sequence common to all triadin isoforms. Trisk 49, which shares this epitope with the other three isoforms, could not be detected with the N-terminal antibody but has only been detected with an isoform specific antibody raised to a sequence in a non-common region (Marty et al. 2000; Vassilopoulos et al. 2005). In earlier publications using antibodies that are common to all triadins in the N-terminal and luminal regions, only Trisk 95 and a band at ∼66 kDa were detected (Jones et al. 1995; Zhang et al. 1997; Shen et al. 2007). Based on comparison of figures, the smaller isoform appears to have the same mobility as Trisk 51. However, the 32/44 kDa pair was only detected in heart with these antibodies and no smaller isoform was found in skeletal muscle. Even though the differences in isoforms could be species and/or developmentally specific, the fact that both groups used adults and significant species overlap doesn't resolve the controversy.

Early studies of triadin function were based on the effects of acute overexpression of triadin in myotubes and heart (Zhang et al. 2001; Kirchhefer et al. 2003; Tijskens et al. 2003; Kirchhefer et al. 2004; Rezgui et al. 2005). In this model EC coupling was almost abolished in response to depolarization in the absence of extracellular Ca²⁺ while the direct response of RyRs to caffeine and DHPR currents were maintained. This suggested that triadin was essential for skeletal-type EC coupling. The consequences of mutating the putative RyR1 binding site(s) (D4878A, D4907A or E4908A) that were identified earlier as binding to the KEKE motif of triadin are associated with discordant results (Lee et al. 2006; Goonasekera et al. 2007). In the study of Lee et al., mutation of all three amino acids partially disrupted co-immunoprecipitation of triadin by RyR1 in a RyR1-null cell line and variably (25–100%) inhibited caffeine induced Ca²⁺ release in myotubes, but not in HEK cells, where no triadin is expressed. When single mutations were studied individually D4907A had the largest effect. The results of the study of Goonasekera et al. expressing the mutated RyRs in RyR1−null primary myotubes agreed that D4907A had the largest effect and that caffeine induced Ca²⁺ release was depressed in myotubes but not HEK cells. However they found that either the triple mutant or double mutants containing the D4907A mutation abolished any co-immunoprecipitation of triadin with RyR1. They also showed that expression of selected mutants abolished EC coupling in response to electrical stimulation and amplified the reduction of Ca²⁺ transients caused by Cd²⁺/La³⁺ or ryanodine pretreatment in voltage clamp experiments. Despite inconsistencies, both concluded that triadin binding to the second intracellular loop of RyR1 was important for EC coupling. However, the fact that EC coupling was suppressed while direct activation of RyR1 was not could be due to several factors including (1) the need for triadin/RyR1 interactions to support EC coupling in the presence of triadin, (2) that the mutations cause conformational changes in RyR1 that directly block passage of the orthograde signal from the DHPR preventing channel opening irrespective of triadin, and (3) perhaps that without binding to RyR1 there is a relative overexpression of triadin.

Based on these studies one might expect that the triadin KO mouse would very likely produce an embryonic lethal phenotype in homozygous mice due to an early cardiac death and a birth lethal phenotype if a skeletal muscle specific KO could be accomplished. We found neither phenotype. Moreover homozygous triadin KO mice lived, grew and reproduced normally (Shen et al. 2007). Western blot analysis of sarcoplasmic reticulum (SR) proteins in skeletal muscle showed that the absence of triadin expression was associated with a small down-regulation of Junctophilin-1, junctin, and calsequestrin in fast twitch muscles but resulted in very modest contractile dysfunction. Ca²⁺ imaging studies in null lumbricalis muscles and myotubes showed that the absence of triadin did not prevent skeletal type excitation–contraction coupling and only slightly reduced the Ca²⁺ transient amplitude in response to electrical stimulation and reduced the amplitude but not the characteristics of the caffeine response. This decrease was attributed to a modest decrease in Ca²⁺ stores. [³H]Ryanodine binding studies of skeletal muscle SR vesicles detected no differences in Ca²⁺ activation or Ca²⁺ and Mg²⁺ inhibition of binding between wild-type and triadin- null animals but B_max was significantly higher in null vesicles. The latter could be the result of the fact that null myotubes and adult fibres had significantly increased myoplasmic resting free Ca²⁺ as measured directly with microelectrodes. Subtle ultrastructural changes, evidenced by the appearance of rare longitudinally oriented triads and the occasional presence of calsequestrin in the sacs of the longitudinal SR, were present in fast but not slow twitch-null muscles. These data support an indirect role for triadin in regulating myoplasmic Ca²⁺ homeostasis and a minor role in organizing the molecular complex of the triad but not in regulating skeletal-type EC coupling. Two studies using shRNA knockdown of triadin in C₂C₁₂ cells confirm that suppression of triadin expression has small effects on the amplitude of the Ca²⁺ release in response to KCl depolarization or caffeine (Fodor et al. 2008; Wang et al. 2009). One shRNA study found that the mechanism for this change was a reduced Ca²⁺ store, the other did not. Both demonstrated that skeletal EC coupling was intact and one confirmed the increase in resting free Ca²⁺ previously observed in triadin-null myotubes and muscles (Wang et al. 2009). The frequency of localized Ca²⁺ release events was increased with suppression of triadin expression and the amplitude of both sparks and embers was larger.

Triadin overexpression clearly causes a detriment in Ca²⁺ homeostasis with a significant decrease in EC coupling efficiency, especially in the presence of reduced extracellular Ca²⁺. It has been recently shown by the Marty lab that overexpression of triadin in Cos7 cells causes a change in the architecture of the ER and a modification of the microtubules that is very similar to overexpression of an anchoring protein (see accompanying paper). This might suggest that one of the roles of triadin may be that of an anchoring protein. While modification in SR architecture may be the explanation for the deficits seen when triadin is overexpressed in skeletal myotubes and may also serve as a possible mechanism for the absence of EC coupling when RyR1's triadin binding sites are mutated, this remains to be demonstrated. It is unlikely, however, that this is triadin's primary role because although there was remodelling of the architecture of the SR in triadin-null animals, this was a rare event, and was completely absent in slow fibres.

What remains to be determined is what triadin really does. Our recent unpublished data in null animals in vivo and null-EDL muscles in vitro confirm that triadin is not essential for EC coupling and suggest that SR stores in triadin−null animals may be smaller and are depleted more rapidly than in WT muscles, which may be related to the fact that the rate of Ca²⁺ entry after store depletion is slower (Vassilopoulos et al. 2007). Together these studies demonstrate that triadin is required to maintain normal intracellular Ca²⁺ homeostasis when the muscle is stressed and suggest that the presence of triadin may be needed for proper mobilization of the SR Ca²⁺ sensor Stim1. The mechanism for how over-expression disrupts EC coupling remains to be determined but may be the result of a disruption of normal SR and microtubule architecture (see Marty et al. 2009).