Triad Biogenesis, Repair, and Maintenance
The events that characterize TT maturation and triads formation have been morphologically defined by electron and confocal microscopy studies. These revealed that the juxtaposition of j-SR and TT begins in the embryonic (E) period of development and evolves through sequential steps in the postnatal life
[14][15][14,15]. In mammals, the first step in triads biogenesis is represented by the formation of a tubular SR adjacent to myofibrils, which will subsequently develop into reticular structures surrounding the myofibrils
[16]. In mouse embryos, a detectable SR is observed as early as day E14; by day E15, immunohistochemical analysis reveals the presence of RyR1 clusters positioned at the periphery of developing myotubes, and, by day E16, RyR1 clusters start aligning with the A-I borders of the sarcomeres. On day E17 and continuing after birth, the j-SR progressively acquires a transverse orientation, with RyR1 showing its typical double-row distribution
[14][15][14,15].
TT development appears to start immediately after that of the SR. At day E15, small tubular invaginations can be observed at the periphery of developing muscle fibers. By day E19, these invaginations expand deeper into the fiber, although remaining almost longitudinally oriented. After birth, the TT system starts acquiring its final transverse position over the borders between the A and I bands, achieving its definitive localization at around three weeks of age in mice
[2][14][2,14]. A recent study performed in zebrafish that combined live imaging with three-dimensional electron microscopy proposes that membrane-derived tubules start forming at transient sarcolemma endocytic nucleation sites. These structures further grow by membrane traffic-based mechanisms and interact with the sarcomere, likely via the SR, finally resulting in an ordered transverse organization
[17]. The association between SR and sarcolemma membranes is already evident as early as E14–E15, mainly at the periphery of the fiber. Starting on E16, internal junction sites appear as diads, formed by one developing TT flanked by a single terminal cisterna. On days E17–E18, most SR/TT junctions are structured as triads
[18]. During in vitro differentiation of skeletal muscle, myoblasts recapitulate the main steps of in vivo SR and TT development, as reported in
Figure 2.
Figure 2. In vitro differentiation of primary rat myocytes.Primary rat myocytes induced to differentiate for 1 to 18 days were decorated with antibodies against RyR1 to label the j-SR. At the beginning of differentiation, RyRs show a diffuse distribution in the SR. Starting from day 4 of differentiation, RyRs form clusters, and the SR progressively forms peripheral couplings and/or diads with the TT. Triads form during the following days, and at the end of differentiation, they acquire their transverse orientation and localize at the borders between the A and I bands of the sarcomeres.
While several details of the morphological events that characterize TT and SR maturation have been defined, the molecular mechanisms that drive this process remain more elusive. RyR1 and DHPR, although essential for ECC, have been shown to be dispensable for the formation of triads; indeed, triad formation was observed in knockout mice for RyR1, also called “dyspedic” mice since RyRs were defined as “junctional feet”. Triad formation was also observed in mice carrying muscular dysgenesis in DHPR, a natural mutation that causes lethal paralysis in mice. Finally, mice lacking both RyR1 and DHPR also show triad formation, further supporting the idea that these two proteins are not directly involved in triad formation or maintenance
[19][20][21][19,20,21]. Several lipid-binding proteins and enzymes regulating phospholipid metabolism have been proposed to participate in TT and j-SR biogenesis and maintenance, such as amphiphysin 2, Caveolin-3, Dysferlin, Mitsugumins, Myotubularin, and Junctophilins
[22] (
Figure 3).
Figure 3. Schematic representation of the main proteins accommodated in TT, j-SR, and l-SR. Protein localization and reciprocal interactions are schematized as detailed in the text. Red arrows indicate Ca
2+ fluxes (red dots) through RyR1, Orai1, and SERCA pumps. RyR1 opens following interaction with DHPR; Orai1 opens following interaction with STIM1 aggregates, which in turn are induced by a reduction in Ca
2+ levels in the SR; SERCA pumps actively transport Ca
2+ from the cytosol to l-SR; PLN or SLN act as SERCA inhibitors. DNM2, Cavin-4, BIN1, CAV-3, and MTM1 are involved in the maintenance of TT architecture and stability. They also participate in TT formation (not shown) and, together with DYSF, contribute to vesicle trafficking during the repair of the damaged plasma membrane (see text for additional details). For simplicity, not all proteins and/or protein complexes depicted, including cytoskeleton components, are positioned on both sides of the triad, as it occurs physiologically. The following is a list of acronyms depicted in
Figure 3: BIN1 (Bridging integrator-1/Amphiphysin 2); CASQ1 (Calsequestrin 1); CAV-3 (Caveolin-3); CKAP4 (Cytoskeleton-associated protein 4/Climp63); DHPR (dihydropyridine receptor); DNM2 (Dynamin 2); DYSF (Dysferlin); HRC (Histidine-Rich Calcium binding protein); JNT (Junctin); JP45 (J-SR protein 1); JPH1 (Junctophilin 1); j-SR (junctional sarcoplasmic reticulum); l-SR (longitudinal sarcoplasmic reticulum); MG29 (Mitsugumin-29); MG53 (Mitsugumin-53); MTM1 (Myotubularin); PLN (Phospholamban); RyR1 (Type 1 Ryanodine Receptor); SAR (Sarcalumenin); SERCA (Sarco/Endoplasmic Reticulum Calcium ATPase); SLN (Sarcolipin); STIM1 (Stromal Interaction Molecule 1); TRDN (Triadin); TRPC3 (Transient Receptor Potential Cation Channel 3); T-tubule (transverse tubule). Adapted from
[23].
Amphiphysin 2, also known as Bridging integrator-1 (Bin1), is a member of a family of ubiquitous proteins able to shape lipid membranes; thanks to their BAR (Bin/Amphiphysin/Rvs) domain, they can interact with membrane lipids and detect and/or generate membrane curvatures
[24]. Bin1 can drive membrane infoldings, thus supporting the generation of TT in muscle cells
[25]. In skeletal muscle cell differentiation, Bin1 shuttles between the nucleus and cytoplasm
[26] and can recruit and partially inhibit the embryonic isoform of Dynamin2 (DNM2), a GTPase involved in membrane fission and endocytosis
[27][28][27,28]. Interestingly, Bin1 is not able to affect the activity of the adult isoform of DNM2, suggesting that it supports the first steps of membrane tubulation while, in the adult, it contributes to stabilizing the TT system
[27]. The concurrent role of Bin1 and DNM2 in TT biogenesis has been confirmed by studies performed in Bin1 knockout mice
[29] and by the identification of mutations in both genes that are causative of human centronuclear myopathy
[30][31][30,31]. The SH3 domain of Bin1 also forms a complex with N-WASP that regulates the assembly of the actin cytoskeleton required for structural stability and organization of triads. This complex appears to include gamma-actin and tropomyosin Tm5NM1 in the proximity of the triads
[32][33][32,33].
Caveolin-3 (Cav-3) is the muscle-specific isoform of caveolins, which mediates the formation of peripheral invaginations in cholesterol-rich regions of the plasma membrane
[27][34][35][27,34,35]. Cav-3 knockout mice develop morphological abnormalities in the TT system consisting of tubule dilation and loss of transversal orientation
[36]. In the adult, caveolae concentrate at sites of membrane injury where the resorption of damaged material is mediated through caveolae-dependent endocytosis
[37]. The process of caveolae-mediated repair is inhibited or impaired in caveolinopathies, including limb–girdle muscular dystrophy (LGMD1C)
[38], rippling muscle disease
[39], and distal myopathy
[40], all characterized by TT structural alterations. In this context, a key role is also played by another protein family, the cavins, which have been described as essential structural components of caveolae
[41]. Cavin-4 is expressed predominantly in muscle, and its distribution is perturbed in human muscle diseases associated with Cav-3 dysfunction
[42]. Cavin-1 (also known as Polymerase I and transcript release factor) is also expressed in skeletal muscle, where it works as a docking protein, together with Mitsugumin 53, during acute cell damage and membrane repair
[43]. In addition to cavins, Cav-3 also interacts with Dysferlin and Mitsugumin 53 during membrane repair
[44][45][46][44,45,46].
Dystrophy-associated fer-1-like protein (DYSF) is an integral protein belonging to the ferlin family, characterized by multiple C2 domains able to bind membrane phospholipids in the presence of Ca
2+ [47]. It is mainly involved in membrane repair. Indeed, DYSF-deficient myofibers show a slower sarcolemma resealing and contain more dilated, irregularly shaped, and longitudinally oriented TT, with abnormal accumulation of vesicles at damaged sarcolemma sites
[48][49][48,49]. Mutations in the human
DYSF gene cause myopathies, including limb–girdle muscular dystrophy type 2B (LGMD2B), Miyoshi myopathy (MM), and distal anterior compartment myopathy
[50][51][52][50,51,52]. DYSF, together with the membrane-associated actin-binding proteins annexin A6, annexin A2a, and annexin A1a, is concentrated in the vicinity of membrane injury, where it regulates Ca
2+-dependent vesicle fusion with the plasma membrane
[53]. Annexins are also recruited at the sarcolemma by Anoctamin 5 (ANO5/TMEM16E), a Ca
2+-activated chloride channel, which was also suggested to display phospholipid scramblase activity
[54]. Mutations in ANO5/TMEM16E are associated with autosomal recessive limb–girdle muscular dystrophy-12
[55][56][55,56]. More recently, Chandra and co-workers proposed that ANO5/TMEM16E may play a dual role in membrane repair since, in addition to recruiting annexins to the injury sites, it also mediates counter anion influx into the SR, supporting the activity of SERCA pumps in clearing the cytoplasm from Ca
2+ overload due to membrane damage
[57].
Mitsugumin-53 (MG53), also known as tripartite motif 72 (TRIM72), is a skeletal and cardiac muscle-specific protein that contributes to vesicle trafficking involved in the membrane repair machinery
[46][58][46,58]. MG53 expression increases during myogenesis, and, during membrane damage repair, it is recruited at damaged membrane sites by cavin-1 where, by binding to phosphatidylserine, it acts as a scaffold to recruit additional proteins, such as Cav-3, DYSF, and annexin V, to start vesicle trafficking
[49][59][60][49,59,60]. Indeed, MG53 knockout mice develop progressive muscle pathologies, likely correlated with an impediment in the membrane repair mechanism
[49]. Surprisingly, no MG53 mutation associated with the onset of skeletal muscle pathologies has been described so far.
Mitsugumin-29 (MG29), which belongs to the synaptophysin family and is characterized by early expression during myogenesis, associates with newly formed SR vesicles
[61][62][61,62]. During triad maturation, MG29 can be detected on both the SR and TT, while, in adult skeletal muscle, it localizes exclusively at triads
[63]. MG29 knockout mice are characterized by decreased muscle mass and overall muscle weakness, likely due to the presence of swollen TT and less organized triads
[64][65][64,65]. In vitro studies on mouse primary skeletal myotubes also showed that MG29 interacts with the canonical-type transient receptor potential cation channel 3 (TRPC3), localized at the TT, suggesting the existence of an additional mechanism able to regulate Ca
2+ transients during skeletal muscle contraction
[66].
Myotubularin (MTM1) is a lipid phosphatase localized at triads. Mutations in human
MTM1 are associated with X-linked myotubular myopathy, a skeletal muscle disorder that results in severe muscle weakness with abnormal nuclei positioning
[67]. Mice knockout for MTM1 present with progressive centronuclear myopathy characterized by SR and TT disorganization, centrally positioned nuclei, and disorganized mitochondria distribution. Interestingly, these structural alterations are more evident with aging, suggesting that MTM1 might play a role in later stages of TT formation and/or in maintaining the stability of TT, rather than in the early biogenesis of triads
[68]. MTM1 dephosphorylates phosphatidyl-inositol-3-phosphate (PtdIns3P) and phosphatidyl-inositol-3,5-bisphosphate (PtdIns3,5P2), two second messengers that regulate docking and fusion of vesicles with the plasma membrane
[69]. This suggests that muscle damage observed in MTM1 knockout mice might be due to abnormal endosomal vesicle trafficking and autophagic fluxes
[70]. On the other hand, overexpression of MTM1 also led to the formation of abnormal membrane structures that stained positive for Cav-3, dystrophin, and DHPR
[71], indicating that balanced expression levels of MTM1 are required for the proper assembly of SR and TT into triads. Interestingly, MTM1 co-localizes and interacts with BIN1 and can enhance BIN1-mediated membrane tubulation, indicating that MTM1 may also be involved in regulating membrane curvature by acting on PtdIns3P-rich membrane subdomains
[72][73][72,73]. Finally, MTM1 was found to interact with the striated muscle enriched protein kinase SPEG
[74]; mutations in
SPEG have been linked to centronuclear myopathies with disruptions of triadic structures and impairment of ECC
[75].
Junctophilins mediate the apposition between ER/SR and plasma membrane in different cell types. In mammals, Junctophilin 1 (JPH1) and Junctophilin 2 (JPH2) are expressed in striated muscles
[76][77][78][76,77,78]. All isoforms share a similar protein structure consisting of two groups of membrane occupation and recognition nexus (MORN) motifs, separated by a joining region followed by an α-helix, a divergent region, and a transmembrane domain at their C-terminus
[79]. MORN motifs can recognize and interact with membrane phospholipids
[76][80][76,80]. The α-helical region is highly conserved and provides mechanical elasticity to the protein, while the divergent region, which displays the lowest similarity among the four isoforms, still lacks a clear functional role
[80]. The short transmembrane domain anchors JPHs to the ER/SR membrane and, together with MORN motifs—which interact with plasma membrane phospholipids—allows JPHs to establish a stable contact site keeping the two cell membranes 12–15 nm apart
[77][80][77,80].
In skeletal muscle, both JPH1 and JPH2 are present at triads, where they are essential for the formation and maintenance of these contact sites. In addition to their structural role, JPHs can also bind and regulate various proteins at triads, acting as a scaffold for assembly of the Ca
2+ release complex
[81][82][83][81,82,83]. JPH1 and JPH2 form homo- and heterodimers that can interact with several proteins on the SR and the sarcolemma, including RyRs, DHPR, Cav-3, and the TRPC3 channel
[84][85][86][87][88][89][90][91][84,85,86,87,88,89,90,91]. JPHs are also required to support the organization of the Store Operated Calcium Entry (SOCE) pathway
[92]. Downregulation of JPHs results in the defective formation of triads and diads and altered Ca
2+ homeostasis due to a partial loss in RyR1 and DHPR co-localization
[84][85][86][84,85,86]. JPH1 knockout mice die shortly after birth, presenting triad alterations and impaired ECC
[93][94][93,94], while JPH2 knockout mice show embryonic mortality due to significant alterations in diads and cardiac ECC
[77]. In the heart, remodeling and loss of TT in chronic heart failure, in aging or in cardiomyopathies correlate with downregulation of JPH2 protein expression and/or loss of JPH2 localization at TT; in addition, mutations in JPH2 are linked to the development of cardiomyopathies
[95][96][97][98][99][100][95,96,97,98,99,100]. Two mechanisms have been suggested to explain JPH2 downregulation in cardiac diseases. The first is associated with the up-regulation of microRNA-24 (miR-24) observed in failing hearts; this microRNA can bind the 3′ untranslated region of JPH2 mRNA, resulting in the downregulation of JPH2 translation
[95]. The second mechanism is due to calpain-mediated protein cleavage
[95]; several studies showed that both JPH1 and JPH2 are targets of the calcium-depended protease μ-calpain (or calpain-1) and m-calpain (or calpain-2), whose activity increases in response to stress and/or unbalanced Ca
2+ homeostasis
[101][102][103][104][105][101,102,103,104,105]. In cardiac muscle, calpain-dependent proteolysis may explain the downregulation of JPH2 observed in heart failure
[103][104][103,104]. Nevertheless, while JPH degradation results in the modification of the geometry of triads and diads, leading to ECC alterations, the N-terminal fragment of JPH2 generated by calpain digestion in stressed cardiac muscle was found to translocate to the cell nucleus, where it acts as a transcriptional regulator mitigating the hypertrophic response to cardiac disease
[101][102][103][104][105][101,102,103,104,105].