A large increase in [Ca²⁺]_cyto has been found in muscle fibers from MHS individuals (Lopez et al., 1992), myotubes derived from MHS patient biopsies (Figueroa et al., 2019) and in MHS animal models (Lopez et al., 1986). MHS animals also showed increased calpain activity (Michelucci et al., 2017). Given that junctophilins are targets of Ca²⁺-activated calpain proteases (Guo et al., 2018; Lahiri et al., 2020), we hypothesized that excess [Ca²⁺]_cyto, by promotion of calpain activity, enhances JPh1 cleavage in MHS individuals. As a test, JPh1 was quantified by Western blotting (WB) on the same 25-lane gel, in total protein extracts from biopsied muscle of 13 MHN (normal) and 12 MHS subjects. The immunoblot, stained using a polyclonal antibody raised against a human JPh1 immunogen (residues 387–512), referred to here as ‘A’, is shown in Figure 1A. A visible reduction in the MHS of a~72 kDa stained band revealed an almost twofold, statistically significant difference in content (Figure 1C). Because the 72–75 kDa migration weight is consistent with the full JPh1 sequence, we refer to this band as ‘JPh1’. That the band is in some columns a doublet, and that it migrates slightly faster than the corresponding band in mouse muscle gels (Figure 1—figure supplement 1) indicates that proteolytic degradation may take place in the human biopsies and this band may not correspond to the full sequence of JPh1.

Figure 1 with 3 supplements see all

Download asset Open asset

Figure 1—source data 1

JPh1 raw blot shown in Figure 1A.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig1-data1-v1.zip

Download elife-78874-fig1-data1-v1.zip

Figure 1—source data 2

Gel for JPh1 blot in Figure 1A.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig1-data2-v1.zip

Download elife-78874-fig1-data2-v1.zip

Figure 1—source data 3

JPh44 raw blot shown in Figure 1B.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig1-data3-v1.zip

Download elife-78874-fig1-data3-v1.zip

Figure 1—source data 4

Gel for JPh44 blot of Figure 1B.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig1-data4-v1.zip

Download elife-78874-fig1-data4-v1.zip

Figure 1—source data 5

Raw blot in Figure 1F.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig1-data5-v1.zip

Download elife-78874-fig1-data5-v1.zip

Figure 1—source data 6

Gel for blot of Figure 1F.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig1-data6-v1.zip

Download elife-78874-fig1-data6-v1.zip

A blot from a different gel, stained with anti-JPh1 antibody ‘B’ (raised against a more distal mouse JPh1 immunogen, 509–622) showed a greater content of a band with an effective migration size of ~44 kDa, referred to as JPh44 (Figure 1B and D, p=0.06). Blots in Figure 1—figure supplement 2 demonstrate that ab A reacts with both JPh1 and JPh44, while ab B fails to react with the 72 kDa protein, therefore constituting a useful tool for detection of JPh44 by immunofluorescence. The 72 kDa protein is overwhelmingly more abundant than its fragments, including JPh44 in nearly every condition (as shown in Figure 1, Figure 1—figure supplement 2 and multiple documents later). Accordingly, the fluorescence of ab A can be used as a monitor of JPh1, with minimal error. Herein and unless noted differently, the labels ‘JPh1’ and ‘JPh44’ tag images of fluorescence of ab A and B, respectively.

While these results suggest that a greater cleavage of JPh1 leads to a reduction in its content, with consequent increase in the 44 kDa fragment, there is a poor negative correlation in individual subjects between the two changes (Figure 1E). A tentative explanation is that the cleavage process, which may operate at multiple sites and sequentially on cleaved products, results in faster degradation of the full protein, blurring any detailed correspondence between its disappearance and the increase in content of some fragments.

Phosphorylase kinase (PhK) reciprocally activates glycogen phosphorylase and inhibits glycogen synthase (GS). In previous work we demonstrated that this enzyme is activated in the muscle of MHS patients (Tammineni et al., 2020). Its effects are consistent with an observed shift of the glycogen ←→glucose balance towards glycogenolysis (Tammineni et al., 2020), which is presumed responsible for the hyperglycemia and diabetes that develops in many of these patients (Tammineni et al., 2020; Altamirano et al., 2019). Here, we compared the contents of GSK3β, a serine/threonine protein kinase that directly inhibits GS by phosphorylation. The kinase activity of GSK3β is controlled by its phosphorylation at Ser 9 and also promoted by truncation of the original ~47 kDa molecule to a form of ~40 kDa (Jin et al., 2015; Goñi-Oliver et al., 2007; Ma et al., 2012).

WB of protein extracts (Figure 1F) from muscle biopsies of the same patients of Figure 1A, reveal both forms of GSK3β. The signal in the 47 kDa band of full-size GSK3β was reduced in MHS, while that of the activated 40 kDa fragment was increased. The changes in the two proteins were highly negatively correlated (Figure 1I); consequently, the ratio of signals (40 kDa/47 kDa) was more than 3-fold greater in the MHS group, with high statistical significance. Regardless of diagnosis (MHS or MHN), there was a significant positive correlation between content of the activated GSK3β form and serum glucose (Figure 1—figure supplement 3). These results are consistent with activation of the GS kinase by calpain proteolysis, which contributes to the shift in glycogen ←→glucose balance in favor of glycogenolysis, adding to the previously reported effect of inhibitory phosphorylation of GS (Tammineni et al., 2020).

In cardiac muscle, the full-length JPh2 protein is located at terminal cisternae of the SR, while cleaved JPh2 fragments can be found inside nuclei (Guo et al., 2018; Lahiri et al., 2020). Here we defined the location of JPh1 and JPh44 by immunostaining combined with 3D high-resolution imaging (a procedure used for every fluorescence image shown in this study, consisting in acquisition of vertical ‘z-stacks’ of confocal x-y images, followed by a correction algorithm described in Materials and methods). Moderately stretched thin myofiber bundles were fixed in PFA and stained with antibodies A or B. As shown in Figure 2 (panels A), the ab A signal was largely located at T-SR junctions, where it was highly colocalized with RyR1. In contrast, ab B (JPh44-specific) detected discrete particles at variable locations within the sarcomeric I band, as well as nuclei (panel 2Bb), and accordingly failed to colocalize with RyR1 (panels 2 Cc, Cd).

Figure 2 with 6 supplements see all

Download asset Open asset

Colocalization between the junctophilin forms and RyR1 was quantified by four measures, with average results listed in Table 1. First: R, Pearson’s correlation coefficient of immunofluorescence intensities, calculated pixel by pixel after subtraction of a small background signal. This index had high value for the ab A fluorescence, (indicating a much greater quantity of full-size JPh1 than JPh44) and low for ab B (JPh44), with high statistical significance of the difference. Second: the Intensity Correlation Quotient (ICQ), emergent from an approach by Li et al., 2004 illustrated in panel 2Cb, whereby the intensity of fluorescence of ab A (JPh1) is plotted, pixel by pixel, against the covariance of ab A and RyR1 intensities (Materials and methods). The comma-shaped cloud with negative curvature, shown in 2Cb, is a characteristic of high colocalization, contrasting with the funnel-shaped cloud found for ab B vs RyR1 (panel 2 Cd), which includes pixels with negative covariance, reflecting mutual exclusion rather than colocalization. The ICQ, which varies between –0.5 (reflecting exclusion) and 0.5 (perfect colocalization), was close to 0.5 for ab A, while that of ab B was smaller but still positive, a difference with high statistical significance (Table 1; here and elsewhere significance was established through biological replications).

A third approach, introduced by van Steensel et al., 1996 and illustrated in Figure 2 panels Ca, c, plots the correlation coefficient of the two signals, averaged over all pixels, vs. a variable shift of one of the images in one direction (in the example, the shift is in the x direction, parallel to the fiber axis). The curve generated is then fitted by a Gaussian. The x-axis location of the apex of the Gaussian (or that of the actual correlation, when the fit is poor), named ‘VS shift’, provides a rough measure of the average separation of the two fluorescent species in the x direction. In the example, the distance was 21 nm for JPh1 (antibody A, panel 2Ca) and 154 nm for JPh44 (2 Cc), a significant difference (Table 1). Fourth: the FWHM of the fitted Gaussian is a second measure of dispersion or de-localization, also greater for the JPh fragment. Further use of these techniques will show that the calculation of four different measures provides a multidimensional view of colocalization — or its absence — revealing differences between proteins and treatments not reflected in the usual correlation analysis.

Images of patient-derived primary myotubes, stained with antibodies A and B and co-stained for RyR1, are illustrated in Figure 2D and E. Here too, the differences were clear. Ab A staining in cytosol was highly colocalized with RyR1 (2D, 2 F, quantification in Table 1), in clusters that presumably correspond to developing junctions. In agreement with the observation in myofibers, ab B did not colocalize with RyR1 and was found mostly within nuclei, in fine granular form (panel 2Eb). Unlike myofibers, myotubes had some intra-nuclear staining with ab A, which adopted a fine granular pattern similar to that of ab B. Observations presented later assign this ab A signal to JPh44.

In earlier studies of JPh2, both N-terminal (Guo et al., 2018) and C-terminal fragments (Lahiri et al., 2020) were shown to translocate to nuclei, in conditions of heart stress. To identify the JPh1 segment cleaved as JPh44 and follow its movements, a fusion of JPh1 with GFP (at the N terminus) and the FLAG tag (at the C terminus) was expressed in myotubes derived from patients’ muscle or a C2C12 line, and in mouse adult myofibers. Panels A and B in Figure 3, of myotubes expressing GFP-JPh1-FLAG, show GFP fluorescence (green) largely in the cytosol, in the form of small clusters or puncta. Instead, the C-terminal FLAG (red) appeared in both nuclear and extra-nuclear regions — corresponding to the cytosol and other organelles. The intensity of the intranuclear FLAG signal was highly variable in individual nuclei (Figure 3—figure supplement 1). Within the cytosol, the FLAG tag was distributed in two forms: punctate, colocalized with (N-terminal) GFP fluorescence, as well as a finely particulate disperse form, away from GFP, obviously corresponding to a fragment cleaved at a position distal to the N terminus.

Figure 3 with 4 supplements see all

Download asset Open asset

Figure 3—source data 1

JPh1 raw blot in Figure 3D.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig3-data1-v1.zip

Download elife-78874-fig3-data1-v1.zip

Figure 3—source data 2

Originating gel for blot in Figure 3D.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig3-data2-v1.zip

Download elife-78874-fig3-data2-v1.zip

Figure 3—source data 3

Raw blot in Figure 3E (boxed region).

The bands of upper molecular weight (particularly at 70 kDa) result from earlier incubation of same membrane with JPh1 abA and its incomplete stripping. Last 3 lanes in the blot are unrelated to the experiment.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig3-data3-v1.zip

Download elife-78874-fig3-data3-v1.zip

Figure 3—source data 4

Gel for Figure 3E.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig3-data4-v1.zip

Download elife-78874-fig3-data4-v1.zip

Figure 3—source data 5

Raw blot for Figure 3G.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig3-data5-v1.zip

Download elife-78874-fig3-data5-v1.zip

Figure 3—source data 6

Originating gel for blot in Figure 3G.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig3-data6-v1.zip

Download elife-78874-fig3-data6-v1.zip

The asymmetry was quantified by the ratio of signal densities, extranuclear (loosely called ‘cytosolic’)/nuclear, which in C2C12 myotubes was approximately threefold greater for GFP, a highly significant difference (Table 2). When the plasmid was expressed in myotubes derived from human muscle, a difference greater than twofold, again in favor of GFP, was found (panels 3B and Table 2). A clear nuclear internalization of FLAG, with exclusion of GFP, was also found in adult mouse muscle expressing the plasmid (Figure 3C). In this case, the ratio (cytosolic/nuclear) of signal densities was nearly 1000-fold greater for the N-terminal GFP than for the C-terminal FLAG (Table 2).

An additional observation (illustrated with panels 3Ca, d, e, and repeated in replications) was that in myofibers of adult mice the C-terminal fragment accumulated in perinuclear regions, in addition to entering nuclei. This was not the case for the native JPh44 in myofibers (Figure 2), which suggests that the perinuclear buildup be a consequence of the steep gradient generated by the transient increase of concentration of the fragment, combined with a diffusion barrier at the nuclear membrane. The accumulation might also reflect a sequence difference between the exogenous piece and JPh44. This explanation is refuted however, by the absence of perinuclear FLAG accumulation in myotubes expressing the construct (Figure 3Ad, Bd), which instead suggests a slower production of the protein or a lower barrier to nuclear entry in developing cells.

In the transfected mouse myofibers we also found the N-terminal GFP strictly colocalized with endogenous RyR1 at triads, reproducing the high colocalization of endogenous JPh1 and RyR1 found in human myofibers (illustrated with Figure 2 and Figure 2—figure supplement 1). These observations allow three conclusions: (1) in both developing and adult tissue, the full-size protein is attached to junctions between plasma or T membrane and SR. (2) The C terminus is part of the cleaved fragment that migrates into the I band and enters nuclei. (3) The N terminus of JPh1 stays with the plasma or T membrane, whether as part of the full protein or after cleavage.

Resting [Ca²⁺]_cyto in myofibers of MHN and MHS human subjects was reported as 112 nM and 485 nM, respectively (Lopez et al., 1992). (Murphy et al., 2013) reported some cleavage of JPh1 at [Ca²⁺]_cyto of 500 nm. Hypothesizing that the increased cleavage of JPh1 in MHS subjects is associated with increased [Ca²⁺]_cyto, we quantified the effect of this level of cytosolic [Ca²⁺] on the fragmentation of JPh1. Biopsied myofiber bundles from three subjects, pinned in chambers and exposed to saponin for permeabilization, were superfused with either 100 or 500 nM [Ca²⁺]. After 10 min of exposure, the bundles were processed separately to extract protein (hereon ‘whole-protein extracts’). Quantitative WB of the extract found increased JPh1 cleavage upon exposure to higher [Ca²⁺] (Figure 3D–F).

Together with an excess JPh44 in whole-protein extracts from MHS patients’ muscle, we found a significantly higher JPh44 content in nuclear fractions from these biopsies (Figure 3G and H). The altered phenotype of MHS patients, including elevated [Ca²⁺]_cyto, is reproduced to a large extent in myotubes derived from their muscle biopsies (Figueroa et al., 2019; Figueroa et al., 2021). Two remarkable observations affirm the relevance of the elevated cytosolic calcium in defining the altered phenotype. First, there was a positive correlation between the nuclear JPh44 content in the muscle of patients and [Ca²⁺]_cyto in their derived myotubes (Figure 3I). There was also a significant positive correlation between nuclear content of JPh in the myotubes (as quantified by the nuclear/cytosolic ratio of JPh densities) and their [Ca²⁺]_cyto (Figure 3J). Taken together, the observations indicate that cleavage of JPh1 and nuclear internalization of the 44 kDa fragment are higher in MHS muscle, driven by the higher [Ca²⁺]_cyto found in these patients.

Murphy et al., 2013 showed that activation of calpain 1, a heterodimer that incudes a main (~80 kDa) subunit, is associated with autocatalytic proteolysis and results in cleavage of junctophilins 1 and 2. Later studies confirmed that JPh2 in mammalian heart is cleaved by calpains (Guo et al., 2018; Lahiri et al., 2020), most effectively by calpain1 (Yoshimura et al., 1983), activated upon autolysis to a~76 kDa isoform (Goll et al., 2003; Suzuki et al., 1981; Moldoveanu et al., 2002). Because this activation is reported to occur in the 50–300 nM range of [Ca²⁺]_cyto, we tested whether calpain1 autolysis occurs and is responsible for the increased cleavage of JPh1 in MHS.

The prediction tool GPS-CCD 1.0 (Liu et al., 2011) applied to JPh1 — the human sequence of which (Yang et al., 2022) is diagrammed in Figure 4A — revealed several conserved calpain1 cleavage sites in multiple JPh1 orthologs (Figure 4B). The one with highest score is at R240-S241, within the cytosolic domain, in-between MORN motifs 6 and 7 (Figure 4A–C). In all these orthologs, cleavage at the R-S site will generate a C-terminal fragment of between 45.92 and 46.28 kDa, consistent with the ~44 kDa migration size of JPh44 and its C-terminal location in the sequence of JPh1.

Figure 4 with 2 supplements see all

Download asset Open asset

Figure 4—source data 1

Raw blot for Figure 4D (boxed region).

Lower part of the blot, with calpain fragments of small molecular weight is shown in Figure 4—figure supplement 2.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig4-data1-v1.zip

Download elife-78874-fig4-data1-v1.zip

Figure 4—source data 2

Originating gel for blot in Figure 4D.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig4-data2-v1.zip

Download elife-78874-fig4-data2-v1.zip

To test whether the greater cleavage of JPh1 in MHS is due to activation of calpain1, we analyzed by Western blotting the calpain1 content of muscle from the subjects with JPh content represented in Figure 1. The immunoblot revealed a double band, with a component at ~80 kDa and another at ~76 kDa, the sizes of the protease and its activated fragment (Figure 4E). We adapted a custom method to quantify immunoblots (Tammineni et al., 2020) to automatically compute the signal in closely placed double bands. Details of the procedure are in Figure 4—figure supplement 1. Comparisons of the signal in the two bands between MHS and MHN patients are represented in Figure 4F. The signal differences between MHS and MHN were not statistically significant for either calpain band. However, the paired ratios of band signals (76 kDa / 80 kDa, Figure 4G) were on average 74% higher in the MHS (p=0.029). This ratio, which can be taken as a measure of calpain activation, was negatively correlated with both the (72 kDa) JPh1 and GSK3β contents, and positively correlated with content of the 40 kDa GSK3β fragment, quantified in the same muscle samples (Figure 4H–J). The relative increase in the 76 kDa fragment was also accompanied by a higher content of smaller polypeptides detected by the calpain1 antibody (Figure 4—figure supplement 2).

To test the ability of calpain1 to cleave human JPh1, total protein extracts from patients’ muscle were incubated with 0.5 units of calpain1 at different time intervals (Figure 5A and B). Incubation resulted in cleavage, as reflected in reduction in content of full-length JPh1 and increase of the JPh1 44 kDa fragment (still referred to as ‘JPh44’, even though it was here obtained by a different procedure). The changes were essentially completed in 5 min.

Figure 5 with 1 supplement see all

Download asset Open asset

Figure 5—source data 1

JPh raw blot incubated with JPh abA shown in Figure 5A.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig5-data1-v1.zip

Download elife-78874-fig5-data1-v1.zip

Figure 5—source data 2

originating gel for blot shown in Figure 5A.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig5-data2-v1.zip

Download elife-78874-fig5-data2-v1.zip

Figure 5—source data 3

JPh1 raw blot in Figure 5C.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig5-data3-v1.zip

Download elife-78874-fig5-data3-v1.zip

Figure 5—source data 4

JPh44 raw blot in Figure 5C.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig5-data4-v1.zip

Download elife-78874-fig5-data4-v1.zip

Figure 5—source data 5

Originating gel for blots in Figure 5C.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig5-data5-v1.zip

Download elife-78874-fig5-data5-v1.zip

Figure 5—source data 6

JPh1 blot in Figure 5E.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig5-data6-v1.zip

Download elife-78874-fig5-data6-v1.zip

Figure 5—source data 7

JPh44 blot in Figure 5E.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig5-data7-v1.zip

Download elife-78874-fig5-data7-v1.zip

Figure 5—source data 8

Originating gel for Figure 5E.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig5-data8-v1.zip

Download elife-78874-fig5-data8-v1.zip

The effects of calpain 1 on the two forms were large and reciprocal, and increased with protease concentration (5 C, D).

Finally, calpain-dependent JPh1 fragmentation was effectively inhibited in the presence of the calpain inhibitor MDL 28170 (Figure 5E and F). While the observation is consistent with calpain1-specific cleavage of human JPh1, it does not exclude effects of calpain 2 or cathepsin B, as these are also targets of the inhibitor.

While calpain1 is freely diffusible in the presence of normal cytosolic concentrations of Ca²⁺, it binds to cellular structures immediately upon elevation of [Ca²⁺]_cyto, which implies that its proteolytic activity can only be exerted on substrates that are close-by at the time of activation (Murphy et al., 2006b; Murphy, 2009). Therefore, to target JPh1 effectively, calpain1 must be present near T-SR junctions. To define this location precisely, human myofibers and human-derived primary myotubes were stained for calpain1, junctophilin (using antibodies A or B), and RyR1 as junction marker, and 3D-imaged at high resolution. Images of endogenous calpain1 in myofibers show that it colocalized with JPh1 (Figure 5G, Figure 5—figure supplement 1) — confirming the presence of the protease near junctions — while JPh44 was found in the I band, not colocalized with calpain1 (5 H). The quantitative measures described for colocalization of JPh and RyR1 were also indicative of colocalization of calpain1 with JPh1, but not with JPh44 (Figure 5I–K and Table 3).

The effects were studied by heterologous expression of tagged calpain1 in patient-derived myotubes. The main effect of the overexpressed calpain1 was an increase in intranuclear localization of JPh (Table 4), illustrated in Figure 6 with images of a human-derived culture transfected with FLAG-calpain1 and additional comparisons with control cultures in Figure 6—figure supplement 1. Panels 6A-D are 3D renderings of the z-stacks of confocal images. A and B show FLAG-calpain1 and junctophilin stained with ab A. Panel C renders RyR1 in the same cells. The images allow direct comparison, within the same field, of a myotube (Cell 1) that expressed strongly the FLAG-tagged calpain1, and one (Cell 2) that had no trace of the protein. Calpain1 (panel A) was widely distributed in the cytosol; nuclei can be recognized in A and C by the absence of calpain1 and RyR1. While the ab A signal (panel B) was widely distributed in Cell 2, in Cell 1 it resided mostly inside nuclei, presumably because it tagged largely JPh44, cleaved by the excess calpain. The intranuclear junctophilin in Cell 1 adopted a diffuse, fine-grained form, while the cytosolic junctophilin in Cell 2 often formed puncta or clusters. The increase in nuclear junctophilin in cells expressing calpain was large and highly significant (Table 5).

Figure 6 with 1 supplement see all

Download asset Open asset

RyR1 is clustered in both cells (Figure 6C and G). The overlay (D) shows that the ab A signal colocalizes with RyR1 (red) in Cell 2, corresponding to the more abundant JPh1 full-size form, but not in calpain-expressing Cell 1, where the signal is largely inside nuclei, presumably marking the JPh44 fragment. As illustrated with Figure 6—figure supplement 1, colocalization of JPh and RyR1 was stronger in control cells (not transfected with FLAG-calpain1), with highly significant differences in all 4 measures (Table 5).

Panels 6E-G show fluorescence in one slice of the z-stack, with pairwise overlays in H-J. The calpain1-rich cytosol of Cell 1 shows JPh1 in fine granular form, without clusters (Panel H). The same region shows abundant clusters of RyR1 (6 G, I). Cell 2, which lacks exogenous calpain1, shows abundant JPh1 in clusters (F), in almost perfect colocalization with RyR1(panel J), which identifies these puncta as JPh1 in developing T-SR junctions.

In Cell 1, where calpain1 was expressed, the protease was also clustered (panel E), highly colocalized with RyR1, that is, located at T-SR junctions (panel 6 I, colocalization quantified in Table 6). As argued by Murphy, 2009, the location of calpain1 at calcium release sites is consistent with the idea that Ca²⁺ activation of calpain1 causes its binding to structures, an effect that places the protease at an ideal location for cleaving JPh1.

Figure 6 illustrates an additional, frequent observation that we did not pursue further: the cell that expressed the calpain construct (Cell 1) had much greater density of RyR1, consistent with other evidence of a role of the protease in control and promotion of muscle development (e.g. Goll et al., 1992).

Together, the results reflect calpain1 activation and autolysis in MHS patients by virtue of their elevated [Ca²⁺]_cyto, accompanied by localization of the protease at T-SR junctions, where it cleaves JPh1 to produce JPh44. The activated calpain1 also produces the 40 kDa, activated form of GSK3β.

The presence of the JPh44 fragment in the nucleus suggested that it plays a role in gene transcription. The calpain algorithm locates the likely cleavage site at between R240 and S241. To test the prediction and produce a stand-in for the JPh44 fragment, we generated the plasmid GFP-∆(1-240) JPh1, coding for the N-terminal fusion of GFP with the human JPh1 deletion variant that starts at S241, and studied its expression in FDB muscles of adult mice transfected by electroporation. Panels A and B in Figure 7 compare expression of GFP-JPh1, the full-size construct, with the GFP-tagged deletion variant in myofibers co-stained for RyR1. While GFP-JPh1 remained near RyRs, in triad junctions, the variant moved away, into the I band, in a manner reminiscent of the movement of JPh44 (compare with Figure 2A and B). The movement away from triads (marked by RyR1) was quantified by colocalization metrics (Table 7). Panels 7 C and D demonstrate intranuclear localization of the deletion variant, reaching concentrations that saturate the light detector at the excitation intensities required to reveal the sarcomeric expression of the protein (e.g. 7 C).

Figure 7

Download asset Open asset

The similarity of movements, cellular location and antibody reactivity between the native JPh44 and the exogenous deletion mutant are consistent with a tentative identification of JPh44 as the C-terminal piece of JPh1 with N terminus at S241. The sole alternative sequence consistent with the predictive algorithm, starting at S234, would only differ for having 7 additional residues at the N terminus. Based on this likely identification, we used GFP-JPh1 ∆(1-240) as surrogate to uncover the effects of the native JPh44 fragment on gene transcription and translation.

To this end, GFP ∆(1-240) JPh1 was transfected into C2C12 cells, where its expression was compared with that of GFP-JPh1, using antibodies A and B (Figure 8—figure supplement 1). The deletion construct expressed well and could be found inside nuclei (Figure 8A). Gene expression analysis revealed that, relative to control cells, 121 and 39 genes were respectively repressed or induced (p<0.01) in the transfected myoblasts (Figure 8B). KEGG pathway-enrichment analysis identified these genes as intimately related to multiple processes, including regulation of the PI3K-Akt-glucose signaling pathway, energy metabolism, muscle growth and lipid metabolism (8 C). Specifically, transcription of GSK3β and proteins inhibitory of phosphorylation of Akt/protein kinase B (Pck1 (phosphoenolpyruvate carboxykinase 1, PEPCK-C)) (Gómez-Valadés et al., 2008), RBP4 (retinol-binding protein 4) (Graham et al., 2006), Repin1 (replication initiator 1) (Kern et al., 2014), APOC3 (guanidinylated apolipoprotein C3) (Botteri et al., 2017), TRAF3 (TNF receptor-associated factor 3) (Chen et al., 2015), Ces3a/TGH (carboxylesterase 3 A or triacylglycerol hydrolase) (Lian et al., 2012), and NGFR (nerve growth factor receptor) (Baeza-Raja et al., 2012; Baeza-Raja and Akassoglou, 2012) were inhibited in response to expression of the construct.

Figure 8 with 1 supplement see all

Download asset Open asset

Figure 8—source data 1

GSK3b raw blot for Figure 8E.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig8-data1-v1.zip

Download elife-78874-fig8-data1-v1.zip

Figure 8—source data 2

GFP-∆(1-240) JPh1 blot for Figure 8E.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig8-data2-v1.zip

Download elife-78874-fig8-data2-v1.zip

Figure 8—source data 3

Originating gel for blots in Figure 8.

GSK3B blot is derived from the left part and JPh from right part of the membrane.

https://cdn.elifesciences.org/articles/78874/elife-78874-fig8-data3-v1.zip

Download elife-78874-fig8-data3-v1.zip

These results were validated via qRT-PCR. Consistent with the digital gene expression sequencing analysis, quantification of mRNA content by RT-PCR indicated that the contents of mRNAs for GSK3β, TRAF3, NGFR, and Repin1 were significantly reduced by 50% in cells transfected with GFP-∆(1-240) JPh1 relative to that in control cells transfected with empty vector (Figure 8D). (Transcription of the other antagonists of Akt phosphorylation listed above could not be evaluated due to failure of all primers tested).

As expected from the effects on transcription, we found by Western blot analysis a reduction in GSK3β content in cells transfected with GFP-∆(1-240) JPh1, by 40% of the content in cells expressing GFP alone (Figure 8E). The difference was significant (Figure 8F). By high-resolution imaging we found that cells expressing the JPh deletion construct had reduced levels of GSK3β compared to the neighboring non transfected cells, whereas cells expressing the vector coding for GFP and otherwise empty had similar GSK3β levels to its neighboring cells (Figure 8Hc and Table 8).

Junctophilin (Takeshima et al., 2000) is one of the five proteins deemed essential for functional, skeletal-muscle-style calcium signaling for EC coupling. We now show that a 72 kDa form of JPh1 (the top size detected in our samples) was reduced by ~50% in total protein extracts of muscles of MHS patients (Figure 1A and C). A 44 kDa fragment increased in the same patients, albeit in a lesser proportion, a discrepancy that we attribute to proteolysis at other sites in the full-length molecule (Figure 1B, D and E).

These changes join a wide-ranging pathogenic alteration of cellular metabolism in these patients, which comprises changes in content, distribution and posttranslational modification of multiple proteins. The observed modification of multiple molecular players of glucose utilization by muscle is the likely cause of the hyperglycemia and diabetes that disproportionally affects MHS patients (Tammineni et al., 2020; Altamirano et al., 2019). Here we demonstrate a conversion of the original ~47 kDa GSK3β molecule to a~40 kDa form, a truncation that activates the enzyme (Jin et al., 2015; Goñi-Oliver et al., 2007; Ma et al., 2012). Muscle-specific ablation of this protein activates glycogen synthase in mice, improving glucose tolerance, correlated with enhanced insulin-stimulated glycogen synthase activation and glycogen deposition (Patel et al., 2008). Therefore, we surmise that the activation of GSK3β found in these patients constitutes an additional link in the pathogenic chain that leads from increased [Ca²⁺]_cyto to alteration in muscle utilization of glucose, then to hyperglycemia and diabetes (Tammineni et al., 2020; Altamirano et al., 2019; DeFronzo and Tripathy, 2009).

The present study used techniques for imaging protein localization that, as documented in our earlier work (Tammineni et al., 2020), reached a spatial resolution beyond the theoretical optical limit, unprecedented in studies of live muscle. From this vantage point, we demonstrated consequences of cleavage of junctophilin1. While the full-size molecule is located at T-SR junctions (as demonstrated by quantitative measures of colocalization with RyR1; Figure 2A and C, and Table 1), the fragment JPh44 leaves the junctions and migrates into the I band (Figure 2B and C, and Table 1).

The fate of JPh44 was further followed on primary myotubes derived from patients’ muscle, where it was seen to migrate inside nuclei (Figure 2E and F). By contrast, JPh1 remained outside, largely in clusters corresponding to developing T-SR junctions, as revealed by high colocalization with RyR1 (Figure 2Dd, Fa, b and Table 1). To define the primary sequence of JPh44, the construct (N)GFP-JPh1-FLAG(C) was expressed in myotubes and adult mouse muscle. The consistent presence of the FLAG tag inside nuclei (Figure 3A–C and Table 2) plus the persistence of the N-terminal GFP in the cytosol, forming clusters analogous to those of JPh1 and RyR1 (Figure 2) in myotubes, and staying aligned with junctional triads in myofibers (Figure 3C), unambiguously indicate that the doubly-tagged exogenous protein undergoes cleavage similar to the endogenous JPh1, and that the cleaved fraction that enters nuclei -- corresponding to JPh44 -- includes the C terminus of JPh1. The persistence of GFP outside nuclei indicates that the N-terminal fragment cleaved from JPh1 conserves at least some of the features, presumably the MORN motifs, by which JPh1 attaches to T tubule and plasma membranes (Rossi et al., 2019). However, the stable location of the N-terminal cleavage fragment of JPh1 at triad junctions is inconsistent with the proposal that assigns the specific triadic location of JPh1 to its ability to dimerize (Rossi et al., 2019), as this ability is likely lost in the N-terminal fragment.

Direct perfusion of permeabilized human myofiber bundles with elevated Ca²⁺ concentrations resulted in reciprocal changes in JPh1 and JPh44 content (Figure 3D–F). While the application of Ca²⁺ was only transient, the major, statistically significant changes that resulted are consistent with a causative involvement of the chronically elevated [Ca²⁺]_cyto in the pathogenic process that takes place in MHS muscle. This process included a large increase in JPh44 content in blots of the nuclear fraction extracted from MHS patients muscle (Figure 3G and H), which positively correlated with the [Ca²⁺]_cyto measured in myotubes derived from the same patients (Figure 3I). This correlation linked a feature of adult muscle with a measure in the derived culture, and notably also applied to JPh localization in patients’ myotubes relative to their [Ca²⁺]_cyto values (Figure 3J and Figure 3—videos 1–3).

The identification of elevated [Ca²⁺]_cyto as mediator of the MHS cellular phenotype pointed at the Ca²⁺-activated calpains as enzymatic agents of JPh proteolysis. An algorithm predictive of calpain cleavage sites within protein sequences (Liu et al., 2011) contributed evidence of involvement of calpain1. Among the predicted cleavage sites in JPh1 (Figure 4B), the one with the highest score, at R240-S241, is highly conserved in mammals (Figure 4C), producing in every ortholog a C-terminal fragment of approximately 46 kDa. This exercise proposes JPh44 as the C-terminal fragment of JPh1 with N terminus at S241. The cleavage site S233-S234, also of high probability score, suggests an alternative sequence, identical except for 7 additional aminoacid residues at the N terminus. The extra residues would neither cause a difference in molecular weight detectable by electrophoresis nor be likely to change other physicochemical properties, including putative gene regulatory functions suggested by the nuclear localization. Both cleavage sites predict an N-terminal fragment that contains 6 MORN motifs, a prediction consistent with the observation that the N-terminal fragment of JPh1 stays at T-SR junctions (Figure 3). Recent cryo-EM evidence of the tertiary structure of JPh isoforms 1 and 2 (Yang et al., 2022) contradicts the conventional view of the helical backbone as spanning the inter-membrane gap (Lehnart and Wehrens, 2022), suggesting instead a detailed, presumably tight association of this region with the β sheet formed by the MORNs. If this were the case, a single cut at either site might not be sufficient to allow separation of the distal fragment, suggesting instead that both cuts might be necessary to free JPh44.

A key aspect of this process is the link between elevated [Ca²⁺]_cyto and activation of calpain to directly cleave JPh1. The link was first described in skeletal muscle, starting from the notion that acutely elevated [Ca²⁺]_cyto impairs EC coupling (Blazev and Lamb, 1999; Lamb et al., 1995), followed by the demonstration of calpain activation and junctophilin proteolysis at Ca²⁺ concentrations in a range ≥500 nM (Murphy et al., 2013; Verburg et al., 2009). Activation of calpain1 requires only moderate elevation of [Ca²⁺] (50–300 nM) and includes autolysis of its main subunit, which removes an N-terminal piece to bring the molecular weight from 80 to 76 kDa (Goll et al., 2003; Suzuki et al., 1981; Moldoveanu et al., 2002). We found that in MHS muscle, the fraction of calpain1 in 76 kDa form is about 80% higher than in controls (Figure 4D and F). That this piece represents activated protease was affirmed by the high negative correlation between the drop in JPh1 content and the fraction of truncated protease in MHS muscle (Figure 4H). A negative correlation of similar significance between calpain truncation and content of the full-length GSK3β is evidence of an additional role of calpain1, namely proteolysis and activation of the GS kinase (Figure 4I).

In agreement with the calpain1 predictive algorithm, the protease cleaved the JPh1 protein extracted from human muscle in vitro, in a time- and concentration-dependent manner, to a form with the same apparent molecular weight of the native fragment we call JPh44, an action suppressed by a calpain inhibitor (Figure 5).

Additional evidence linking calpain activation to JPh1 cleavage was obtained by its expression in human myotubes. As illustrated in Figure 6, heterologous expressed calpain caused a visible increase of JPh44 — distinct from the full-size protein for adopting a fine-grained appearance— which moved massively into nuclei (Cell 1 of panels 6B and D, Table 4; see also Figure 6—figure supplement 1). Concomitantly, the colocalization of JPh and RyR1 was diminished (Table 5). Also consistent with cleavage of JPh1 by the heterologous calpain1 is the high colocalization of the two molecules, contrasting with the low colocalization of JPh44 and the protease (Table 3).

Previous studies reported that calpains cleave couplon proteins, including RyR1 (Place et al., 2015) and STAC3 (Ashida et al., 2021). The present findings show that the protease is at the right location for those actions. Both endogenous calpain1, imaged by immunofluorescence in human muscle (Figure 5G and Figure 5—figure supplement 1), and FLAG-tagged calpain1 (Figure 6I) were found precisely located at triad junctions; the location was confirmed by its high colocalization scores with JPh1 and with RyR1 (Tables 3 and 6). The observations support the proposal that the protease is freely diffusible in apo form but binds to cellular structures immediately upon exposure to high [Ca²⁺] (van Steensel et al., 1996). Indeed, the quantitative measures of colocalization with RyR1, especially the spatial shift and Gaussian spread obtained with the Van Steensel method, show that the junctional protein JPh1 has a tighter overlap with RyR1 than calpain1 (Table 6), which suggests the coexistence of bound and diffusible forms of the protease.

The results reviewed above establish with a high degree of confidence that the reduction in JPh1 content observed in MHS patients is caused by calpain1, activated by excess cytosolic calcium, and that the activated calpain also cleaves the specific kinase of glycogen synthase GSK3β, with inhibitory consequences on glycogen synthesis. This effect, together with the putative cleavage of the regulated glucose transporter GLUT4 (Otani et al., 2004), impair glucose utilization by muscle.

In 2019, Altamirano et al., 2019 first called attention to the high incidence of hyperglycemia and diabetes developing in patients years after they were diagnosed with MHS. Aware that the main proximate cause of insulin resistance and hyperglycemia is failure of glucose processing by muscle (e.g. DeFronzo and Tripathy, 2009), we set out to understand the pathogenic pathway linking MHS and diabetes. The question is relevant beyond the MHS syndrome, as a similar dysregulation of Ca²⁺ homeostasis is found in other conditions, including related inheritable diseases with mutations in couplon proteins (Ríos et al., 2015; Dowling et al., 2014; Lawal et al., 2020), DMD (Edwards et al., 2010; Boittin et al., 2006; Bellinger et al., 2009), Exertional and Non-Exertional Heat Stroke (Leon and Bouchama, 2015; Hopkins et al., 1991) and Statin-related Myotoxicity (Turner and Pirmohamed, 2019). We found a wide-ranging alteration of location and phosphorylation of the enzymes that manage storage of glucose as glycogen in muscle, namely glycogen synthase, glycogen phosphorylase and their controlling kinase PhK (Brushia and Walsh, 1999), leading to a shift of the glucose ←→ glycogen balance towards glycogenolysis (Tammineni et al., 2020). Additionally, we found a decrease in the deployment (translocation) of GLUT4 (Tammineni et al., 2020).

The present study adds two nested mechanisms (schematically represented in Figure 9 together with those previously revealed by Tammineni et al., 2020), which operate in the same direction: first, the activation by Ca²⁺ and autolysis of calpain1 (Figure 4D–F). Because calpain1 is known to lyse GLUT4 and increase its turnover (Otani et al., 2004), its activation is a likely explanation for the observed decrease in translocation of the transporter (Tammineni et al., 2020). The second link to diabetes is the observed activation, again by proteolysis mediated by the activated calpain, of GSK3β (Figure 1F-I, Figure 4H, I). The activated kinase would add its inhibitory effect on GS to that of phosphorylation by the activated PhK (Tammineni et al., 2020).

Figure 9

Download asset Open asset

The link between the high activity of GSK3β and diabetes in human skeletal muscle is well established (Henriksen and Dokken, 2006), as is the therapeutic potential of GSK3β inhibitors (Dokken and Henriksen, 2006; Maqbool and Hoda, 2017). It is manifested in our patients by the correlation observed between content of activated kinase and FBS (Figure 1—figure supplement 3). Regulation of GSk3β activity by proteolysis was first reported in the brain, where it was found associated with Tau hyperphosphorylation in Alzheimer’s disease (Jin et al., 2015); the present study provides the first demonstration of its occurrence in skeletal muscle, where it leads to phosphorylation of glycogen synthase, as indicated by the correlation found between GSK3β cleavage and blood sugar.

Taken together, the present observations and those in Tammineni et al., 2020 identify a multi-lane pathway, where Ca²⁺ activation, proteolysis and phosphorylation, involving at a minimum calpain 1, GSK3β, GS, GP, PhK, and GLUT4, lead from the primary Ca²⁺ dysregulation to hyperthermia and diabetes (Figure 9).

Calpains 1 and 2 activate in heart muscle damaged by ischemia (e.g. Singh et al., 2004; Singh et al., 2012) or in conditions of heart failure, to proteolyze JPh2 (Murphy et al., 2013; Lahiri et al., 2020; Guo et al., 2013). The lysis of JPh2 is associated with structural remodeling, loss of dyadic junctions and deficit of EC coupling function, all expected from the breakage of a structural brace of the junction (Wu et al., 2014). Surprisingly, however, L-S Song’s group went on to demonstrate that a large N-terminal calpain1-cleaved fragment of JPh2, named JPH2-NTP, enters nuclei, where it regulates transcription of multiple genes, with consequences that oppose the damaging effects of activated proteases (Guo et al., 2018). In contrast, Lahiri et al., 2020 reported the presence of a ~25 kDa C-terminal product of cleavage by calpain2, which was shown to be causative of cellular hypertrophy, rather than compensatory.

The present observations establish similarities and differences between the roles of JPh44 and those of the fragments described in cardiac muscle. The distribution of exogenous doubly tagged JPh1 and its fragments (Figure 3A and C and Table 2) establishes that JPh44 is a C-terminal fragment. The calpain predictive algorithm, together with the observed distribution of the GFP-tagged Δ(1-240) JPh1, also establish with high likelihood the site of JPh1 cleavage, at either R240-S241 or S233-S234 (Figure 4A). In addition to matching the apparent molecular weight of the endogenous JPh44, the C-terminal fragment resulting from cleavage at either point contains two nuclear location sequences, allowing nuclear import of the fragment (Figure 4A). Like JPH2-NTP, the fragment will have an alanine-rich region (ARR), with helix-turn-helix (HTH) structure, characteristic of DNA-binding proteins. These properties establish with reasonable likelihood the primary structure of JPh44 as the C-terminal fragment of JPh1 that starts at S241 or S234. We used this identification to probe directly the possible roles of JPh44 in adult mice muscle and C2C12 myoblasts. When expressed in muscle of adult mice, the GFP-tagged Δ(1-240) JPh1 appeared located similarly as JPh44 in human muscle, largely within the I band and inside nuclei, eschewing triadic junctions and colocalization with its marker, RyR1 (Figure 7B–D and Table 1).

In C2C12 myoblasts, the construct likewise distributed inside nuclei, as well as in a fine-grained cytosolic form (Figure 8A and Figure 8—figure supplement 1). Consistent with a gene regulatory role of JPh44, the transcription profile (obtained by LC Sciences) of myoblasts expressing GFP-tagged Δ(1-240) JPh1 showed significant alterations in multiple pathways (8B, C). Notably, the profile included significantly reduced mRNA levels of GSK3β and at least 9 proteins that act as brakes on the PI3k-AKT pathway, among which TRAF3, Ces3a (Lian et al., 2012), Repin1 (Kern et al., 2014) and Pck1 (Gómez-Valadés et al., 2008) stand out, given their role in muscle metabolism. By quantitative PCR we confirmed the substantial reduction in transcription of four of those genes, as represented schematically in Figure 9.

The many signaling pathways regulated by Akt importantly include glucose uptake by muscle and adipocytes. In muscle, Akt isoform 2 is activated via phosphorylation by a chain of events at the plasma membrane, started in response to insulin Boucher et al., 2014; this activation promotes translocation of GLUT4 to the membrane, enabling its transport role. Phosphorylation by Akt also inactivates GSK3β, which in turn releases glycogen synthase from inactivation, promoting storage of glucose. Hepatic and skeletal muscle silencing of the four genes named above improved the insulin resistance by inducing phosphorylation of Akt (Gómez-Valadés et al., 2008; Kern et al., 2014; Chen et al., 2015; Lian et al., 2012; Hesselbarth et al., 2017). Other genes repressed by the exogenous JPh1 fragment JPh44, including Rbp4 (retinol-binding protein 4), Apoc3 (apolipoprotein C-III), MEG3 (maternally expressed gene 3 — a long noncoding RNA) and NGFR contribute to insulin resistance by various actions in other tissues (Graham et al., 2006; Botteri et al., 2017; Baeza-Raja et al., 2012; Zhu et al., 2019). Together, the effects on transcription of the exogenous JPh1 fragment are consistent with a beneficial, ‘stress-adaptive’ role of the endogenous JPh44, similar to that ascribed to JPH2-NTP in stressed myocardium (Guo et al., 2018).

In spite of the similarity, there are many differences between the putatively regulatory fragments of the two junctophilin isoforms. Unlike JPH2-NTP, JPh44 loses an N-terminal portion and with it the main stretch of MORN motifs. If the helical backbone were tightly bound to the MORN β sheet, as argued in Yang et al., 2022, separation from the N-terminal fragment that includes most MORNs would free the helix for its putative interactions with DNA inside nuclei. Unlike JPH2-NTP, JPh44 retains a second NLS (K588-L614, Figure 4A). Therefore, the skeletal muscle fragment appears to have advantages over JPH2-NTP, to leave the junctional location (by losing most T-membrane anchors), and to reach inside nuclei, by having an additional NLS.

The degree of autolysis of calpain in the samples that we analyzed is substantially greater than that reported for muscle tissue processed immediately after excision (Murphy et al., 2013; Murphy et al., 2006a; Murphy et al., 2007). This fact indicates protein degradation in the human samples after excision, and is consistent with the difference in gel migration between our mouse and human samples (Figure 1—figure supplement 1). While these indications introduce some uncertainty about the molecular sequence of the JPh fragments of interest, they do not preclude the comparisons between samples from individuals with different diagnosis, as they are all treated in the same manner.

An additional difference between the MHS condition and the stress of heart failure reported to unleash calpain cleavage of JPh2 is that the cardiac condition is accompanied by structural remodeling with loss of couplons and reduction in JPh2 content (Wu et al., 2014; Zhang et al., 2013), while no structural remodeling has been reported in muscle of MHS patients. Moreover, studies of Ca²⁺ release function revealed an increase in muscle sensitivity to stimuli (membrane voltage or calcium, e.g., Figueroa et al., 2019; Figueroa et al., 2021) rather than the major deficits expected to arise from a loss of functional junctions (Perni et al., 2017) upon JPh1 loss in the proportion observed here.

A speculative explanation recalls the expression of both JPh1 and 2 in skeletal muscle, both capable of sustaining calcium release function (Perni and Beam, 2022). The presence of two molecules that share a function — a redundancy of paralogs — would combine with a different level of redundancy, the presence of multiple copies of the same protein, perhaps polymerized (Rossi et al., 2019), to provide a ‘sentinel’ role. When proteolysis is activated in pathophysiological situations, limited JPh1 cleavage would cause minimal or no deterioration of structure or function, while producing an adaptive gene-regulatory messenger.

We report that the elevated [Ca²⁺]_cyto in skeletal muscle of individuals with MHS activates calpain1, which associates to T-SR junctions. The protease then cleaves a kinase that inhibits glycogen synthase, an effect that correlates with and probably contributes to an increase in blood glucose concentration. Cleavage of the kinase and that of GLUT4 promote the transition to hyperglycemia and diabetes found in many of these patients. It also cleaves junctophilin1, to produce a C-terminal fragment of ~44 kDa that abandons junctions and enters nuclei. There, the fragment modulates transcription of multiple genes, resulting in effects opposing the deleterious alteration of the glucose utilization pathway. Borrowing a term applied to a fragment of junctophilin2 found to enter nuclei in myocytes of failing hearts (Guo et al., 2018), the JPh44 fragment is therefore ‘stress-adaptive’.

All patients were recruited and studied in the Malignant Hyperthermia Investigation Unit (MHIU) of the Canadian University Health Network, at Toronto General Hospital, in Toronto, Canada, under protocol 12–5474, last reviewed on 11/08/21. All patients consented to all aspects of the study, including publication. The use of tissue and the language of the informed consent were also approved by the Institutional Review Board of Rush University under protocol 16050502-IRB01. Patients were diagnosed with the CHCT and studied over the last 5 years. Criteria for recruitment of subjects included one or more of the following: a previous adverse anesthetic reaction, family history of MH without a diagnostic MH mutation (https://www.emhg.org), a variant of unknown significance (VUS) in RYR1 or CACNA1S, recurrent exercise- or heat-induced rhabdomyolysis and idiopathic elevation of serum creatine kinase.

The anonymized demographic data of the 44 individuals whose tissues were used for analyses, imaging or development of cultures, is provided in Supplementary file 1.

Biopsied segments of human Gracilis muscle were shipped to Rush University at 4 °C overnight in an optimized solution of composition described in Figueroa et al., 2019. Primary cultures, derived from the biopsies as described by Censier et al., 1998, were maintained at 37 °C in a humid atmosphere containing 5% CO₂. After 8–15 days, cells derived from the explants were transferred to a culture dish for proliferation in a growth medium. Myoblasts were expanded through up to four passages. For calcium concentration measurements in live cells, myoblasts were reseeded on collagen-coated dishes (MatTek, Ashland, MA, USA). For immunostaining experiments, myoblasts were seeded on 13 mm coverslips in 24-well plates. At 70% confluence, the cells were switched to a differentiation medium (DMEM-F12 with 2.5%horse serum). Studies were carried out 5–10 days thereafter, in myotubes showing a similar degree of maturation.w.

The mouse myogenic C2C12 cell line was obtained from ATCC (http://www.atcc.org/) and cells were used until passage number 20. Myoblasts were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 15% fetal bovine serum and 1% penicillin-streptomycin in a humidified incubator kept at 37 °C and 5% CO₂. C2C12 myotubes were obtained by culturing 70% confluent myoblasts in differentiation medium (DMEM, 2% horse serum, 1% penicillin-streptomycin) for at least 4days. Authentication of the cells can rely on that done by the supplier, as most were used within a year of their purchase. We also authenticated C2C12 cell lines based on the morphology observed during experiments. Indeed, these cells turned to multinucleated myotubes upon differentiation, the formation of which is a unique feature of C2C12 cell lines. The cells were tested for Mycoplasma contamination with a detection kit (MycoStrip, Invivogen) that uses isothermal polymerase chain reaction (PCR) and can detect over 95% of commonly occurring mycoplasma species contaminating cell line. Results were negative.

Transient transfections of C2C12 myoblasts were performed upon reaching 70% confluence using the K2 Transfection System (Biontex Laboratories GmbH, Munich, Germany) as described by the manufacturer. C2C12 myotubes and biopsy-derived human primary myotubes were transfected with plasmids using lipofectamine 3000 (Thermofisher, Waltham, MA, USA) following the supplier’s instructions.

Plasmids used in this study such as empty vector for GFP-JPh1, dual tagged JPh1 plasmid (GFP-JPh1-FLAG), and GFP tagged plasmid with JPh44 translating region (GFP-∆(1-240) JPh1) were created by OriGene Technologies (Rockville, MD, USA). p3XFlag-CAPN1 was a gift from Yi Zhang (Addgene plasmid # 60941; http://n2t.net/addgene:60941; RRID:Addgene_60941).

This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All the animals were handled according to approved institutional animal care and use committee (IACUC) of Rush University under protocols (#17–035, 18–065 and 21–068.) 6- to 10-week-old mice, Mus musculus, of the Black Swiss strain, sourced at Charles River Laboratories (Boston, MA, USA), were used to define the localization of JPh1 and JPh44 in living cells. Hind paws were transfected with plasmid vector for GFP-JPh1, dual tagged JPh1 plasmid (GFP-JPh1-FLAG), and GFP-tagged plasmid with JPh44 translating region (GFP-∆(1-240) JPh1) as described in Pouvreau et al., 2007. Transfection required electroporation via intramuscular electrodes, which was performed under anesthesia. Animals were euthanized and muscles collected and processed for imaging as in Manno et al., 2017. Every effort was made to minimize animal suffering and discomfort.

Susceptibility to MH was diagnosed at the MHIU following the North American CHCT protocol (Larach, 1989). Increases in baseline force in response to caffeine and halothane (F_C and F_H) were measured on freshly excised biopsies of Gracilis muscle with initial twitch responses that met viability criteria. Three muscle bundles were exposed successively to 0.5-, 1-, 2-, 4-, 8-, and 32 mM caffeine; three separate bundles were exposed to 3% halothane. The threshold response for a positive diagnosis was either F_H ≥0.7 g or F_C (in 2 mM caffeine)≥0.3 g. Patients were diagnosed as ‘MH-negative’ (MHN) if the increase in force was below threshold for both agonists, and ‘MH-susceptible’ (MHS) if at least one exposure exceeded the threshold. While prior work distinguishes the muscles that respond excessively to halothane but not to caffeine (named ‘HH’) from those that respond to both (the “HS”), in the present work the distinction is not made, and both groups together are classified as MHS.

Cytosolic Ca²⁺ concentration, [Ca²⁺]cyto, was monitored in myotubes by shifted excitation and emission ratioing (Launikonis et al., 2005) of indo-1 fluorescence as described in Zhou et al., 2006. Imaging was by confocal microscopy (scanner TCS SP2; Leica Microsystems; Buffalo Grove, IL, USA), using a 63X, 1.2 numerical-aperture water-immersion objective. [Ca²⁺]_cyto was derived from fluorescence signals (Figueroa et al., 2012). Indo-1 was procured from Invitrogen, Waltham, MA, USA.

For in vitro JPh1 cleavage experiments, the whole biopsied muscles from human subjects were homogenized in RIPA lysis buffer (Santa Cruz Biotechnology, Dallas, TX, USA) with protease and phosphatase inhibitors, and the supernatant of centrifugation at 13,000 g for 15 min was used for further process. The cleavage of JPh1 was induced by incubating 100 micrograms of protein supernatants with 0.3–1.0 µg of purified human erythrocyte calpain1 of activity 1 unit/µg (MilliporeSigma, Burlington, MA, USA) at 30 °C for 15 min. Supernatants are treated with or without DMSO dissolved 10 µg calpain inhibitor (MDL28170, Cayman Chemical Co., Ann Arbor, MI, USA). SDS sample buffer was added to stop the reaction.

Human biopsied muscle segments received from the MHIU were quick-frozen for biochemical studies and storage. For measuring total content of proteins in muscle, the tissue was chopped into small pieces in RIPA lysis buffer (Santa Cruz Biotechnology, Dallas, TX, USA) containing protease and phosphatase inhibitors, and homogenized using a Polytron disrupter. The homogenate was centrifuged at 13,000 g for 10 min and supernatant aliquots were stored in liquid nitrogen. Nuclear and cytosolic protein fractions from muscle biopsies were prepared using NE-PER Nuclear and Cytoplasmic Extraction Kit (Catalog. No 78835; ThermoFisher) according to the manufacturer instructions, using a Dounce homogenizer.

Protein content was quantified by the BCA assay (ThermoFisher). Proteins were separated by SDS–polyacrylamide gel electrophoresis, using 10% mini gels and 26-well pre-cast gels (Criterion TGX, Bio-Rad, Hercules, CA, USA), which enable separation in a broad range of molecular weights, and transferred to a nitrocellulose membrane (Bio-Rad). Membranes were blocked at 4.5% blotting grade blocker (Bio-Rad) in PBS and incubated with the primary antibody overnight at 4 °C. Thereafter they were washed in PBS containing 0.1% Tween 20 and incubated in horseradish peroxidase–conjugated anti-mouse or anti-rabbit secondary (Invitrogen, Carlsbad, CA, USA) for 1 hr at room temperature. The blot signals were developed with chemiluminescent substrate (MilliporeSigma) and detected using the Syngene PXi system (Syngene USA Inc, Frederick, MD, USA).

Immunoblot analysis was conducted using the following antibodies (antigen, commercial ID number, concentration, supplier): JPh1 ‘antibody A’, #PA5-52639, 1:1000, ThermoFisher; JPh1 ‘antibody B’, #40–5100, 1:1000, ThermoFisher; GSK3β, #9315, 1:1000, Cell Signaling Technology (Danvers, MA, USA); calpain1, #MA1-12454, 1:1000, ThermoFisher.

Quantitative analysis of Western blots was done with a custom application (written in the IDL platform, Harris Geospatiale, Paris, France; fully described and demonstrated with video in Tammineni et al., 2020) that combined information in the blot and the source gel. This method had less variance than the commercial (Syngene) software tools. The content of interest was measured in the blot by the signal mass within a rectangle that enclosed the protein band above a background. A normalization factor, quantifying the sample deposited in the lane, was computed on the gel as the average signal in a large area of the corresponding lane above background. A more advanced version was developed to automatically quantify closely spaced dual bands (as in Figure 4E); its use is illustrated in Figure 4—figure supplement 1.

Two anti-JPh1 antibodies, A and B, both supplied by ThermoFisher, serendipitously showed different reactivity. When used in Western blots, antibody A (#PA5-52639) reacted approximately equally with the full-size, ~72 kDa molecule and its 44 kDa fragment, while antibody B (#40–5100) marked exclusively JPh44 (Figure 1—figure supplement 1).

Immunofluorescence imaging was done on thin myofiber bundles dissected from human muscle biopsies, FDB muscles from mice, primary human myotubes, C2C12-line myotubes and C2C12 myoblasts. Human or mouse muscles were mounted moderately stretched in relaxing solution, on Sylgard-coated dishes. Relaxing solution was replaced by fixative containing 4% PFA for 20 min. Myotubes or myoblasts on coverslips were washed in 1 X PBS and fixed with 2% PFA for 20 min. Tissues or cell cultures on coverslips were transferred to a 24-well plate and washed three times for 10 min in PBS, then permeabilized with 0.1% Triton X-100 (Sigma) for 30 min at room temperature and blocked in 5% goat serum (Sigma) with slow agitation for 1 hr. The primary antibody was applied overnight at 4 °C with agitation, followed by three PBS washes for 10 min. Fluorescent secondary antibody was applied for 2 hr at room temperature. Dehydrated tissues or cell culture coverslips were mounted with anti-fade medium (Prolong Diamond, ThermoFisher). Immunofluorescence imaging used the following antibodies (antigen, commercial ID number, concentration, supplier): JPh1 ‘antibody A’, #PA5-52639, 1:100, ThermoFisher; JPh1 ‘antibody B’, #40–5100, 1:100, ThermoFisher; calpain1, #MA1-12454, 1:1000, ThermoFisher; Ryr1, #34 C, 1:200, ThermoFisher.

Immunostained myofibers and cell cultures, as well as live tissues expressing fluorescently tagged proteins were imaged confocally using a Falcon SP8 laser scanning system (Leica Microsystems) with a 1.2-numerical aperture, water-immersion, 63 x objective. Resolution was enhanced by high sensitivity hybrid GaAsP detectors (HyD, Leica), which allowed low intensity illumination for image averaging with minimum bleaching, optimal confocal pinhole size (below 1 Airy disk), collection of light in extended ranges (e.g. 470–580 nm), and acquisition of z-stacks (vertical sets of x-y images) at oversampled x–y–z intervals. The stacks usually included 40 x-y images at 120 nm z separation and 60 nm x-y pixel size or, for highest resolution imaging, 20 x-y images at 120 nm z and 36 nm x-y pixel size. Dual images were interleaved by line. Most cells were triply stained and correspondingly monitored at (excitation/emission) (405/430–470 nm), (488/500–550 nm) and (555/570–620 nm). The stacks were acquired starting nearest the objective, at or closely outside the lower surface of the myofiber.

Availability of stacks allowed for offline deblurring by a constrained iterative deconvolution algorithm that used all images in the stack (Voort and Strasters, 1995; Agard et al., 1989) and the point spread function (PSF) of the system, which was determined using 170 nm beads. PSF FWHM was 350 nm in x–y and 480 nm in z. After deblurring, the separation effectively resolved in the x-y plane was approximately 0.1 μm (Supplement 11 in Tammineni et al., 2020). The deblurred set was represented or ‘rendered’ in three dimensions using the ‘Simulated Fluorescence Process’ (Messerli et al., 1993) applied to the full deblurred stack. Determination of the point spread function (PSF) of the imaging system, deblurring, and rendering were done in the HuPro (SVI, Amsterdam, The Netherlands) programming environment.

Location analysis defined densities of protein content within nuclei or in extranuclear areas (named ‘cytosol’), by the ratio of integrated fluorescence signal over area in the respective regions. For comparisons among replications the ratio (nuclear/cytosolic) was used, as it was insensitive to inter-preparation variance in expression or staining intensity.

Colocalization (of JPh1, its fragment JPh44, its expressed constructs GFP-JPh1-FLAG, (N)Myc-JPh1 and GFP-Δ(1-240)JPh1, calpain1, its expressed construct (N)FLAG-calpain1 and RyR1) was evaluated by 4 techniques: (Stern et al., 1997) the Pearson correlation coefficient, calculated as:

where $A_i$ and $B_i$ are intensities of two signals in the same pixel i, ${\overline{A}}_i$ and ${\overline{B}}_i$ the averages over all pixels, and the summation is extended to all pixels, or a ROI when the entire image is not usable. Subtler, but statistically significant differences in colocalization were detected using the Intensity Correlation Analysis (ICA) introduced by Li et al., 2004. This analysis produces a quantitative measure, the Intensity Correlation Quotient, ICQ, a correlation measure with a definition that reduces the influence of heterogeneous staining of protein expression.

The numerator contributes a 1 for every pixel where both signals are above or below average and a –1 for the situation where the differences are opposite. The ratio is normalized to 1 and the subtraction of 0.5 moves the range to [–0.5, 0.5], with the extremes corresponding to perfect exclusion and perfect colocalization. The ICA approach includes a graph, (with examples in panels Figure 2Cb, d), that plots for every pixel the product $\left(A_i-{\overline{A}}_i\right)\left(B_i-{\overline{B}}_i\right)$ , or ‘pixel covariance’ of A and B, vs. the intensity of either A or B. The result is strikingly different for cases of colocalization, where the points draw a noisy parabola (e.g. 2Cb) or lack of it (e.g. 2 Cd), where points to the left of the abscissa 0 correspond to mutual exclusion of the two signals.

The conventional correlation analyses have no provision for identifying the anisotropic de-localization of particles, which, if systematic, may reflect an actual association, with systematic displacement. This sensitivity to vectorial displacement is achieved here using an approach of van Steensel et al., 1996, which again produces a plot and, in this case, two numerical outputs of interest. The approach (illustrated in Figure 2 Ca, c) plots the correlation coefficient R between image A and image B shifted in one direction (say, x, defined as the longitudinal direction in myofibers) by variable amounts dx. The ‘Van Steensel plot’ thus plots R (dx) vs. dx. represented by individual symbols in Figure 2Ca. When the two signals are colocalized, the plot is narrow, and centered at dx = 0. If not, the plot displaces its peak (when the delocalization occurs in a preferred direction) and its width increases. Two numerical quantifiers can be derived on a Gaussian fitted to the points: the abscissa of the maximum (here the ‘VS shift’) and the FWHM. These quantifiers reveal vectorial aspects of the relationship between the two markers. The VS shift gives a rough measure of distance when the separation of the molecules has some vectorial regularity and sidedness. The FWHM is a measure of dispersion of one marker or both. Combined with the high spatial resolution achieved in our images, it allows to detect effects of interventions that fail to cause significant changes by conventional colocalization measurements.

Colocalization measures and plots were implemented with custom programs written in the IDL platform or with the ImageJ ‘plugin’ JACoP (Bolte and Cordelières, 2006).

RNA was isolated from GFP-empty-vector-transfected and GFP-∆(1-240) JPh1-transfected C2C12 myoblasts using the RNeasy plus mini kit (Qiagen, Hilden, Germany), based on manufacturer instructions. Samples were processed by LC Sciences (Houston, TX, USA). Processing included the generation of a library and sequencing of transcripts and their identification in the mouse genome (UCSC mm10). A cDNA library constructed from the pooled RNA from C2C12 cell samples of mouse species was sequenced run with Illumina Novaseq TM 6000 sequence platform. Raw paired-end RNA-seq data were firstly subjected to Cutadapt v1.9 to remove reads with adaptor contamination, low quality bases and undetermined bases. Filtered reads were aligned using HISAT2 to generate alignments tailored for transcript assembler. The mapped reads of each sample were assembled using StringTie with default parameters. Then, all transcriptomes from all samples were merged to reconstruct a comprehensive transcriptome using Gffcompare. After the final transcriptome was generated, StringTie and Ballgown were used to estimate the expression levels of all transcripts and perform expression abundance for mRNAs, by calculating FPKM (fragment per kilobase of transcript per million mapped reads). Differential expression analysis was performed by DESeq2 software between two groups. The genes with the parameter of false discovery rate (FDR) below 0.05 and absolute fold change ≥2 were considered differentially expressed. Differentially expressed genes were then subjected to enrichment analysis of GO functions and KEGG pathways.

RNA extraction and real-time quantitative PCR (qPCR): For qPCR, total RNA was isolated from C2C12 myoblasts transfected with GFP-empty vector or GFP-∆(1-240)JPh1 using the RNeasy-plus mini kit (QIAGEN) according to the manufacturer’s instructions. cDNA was synthesized using the iScript gDNA clear cDNA synthesis kit (BIORAD) according to the manufacturer’s instructions. The cDNA was then analyzed via real-time qPCR using the SYBR Green qPCR Master (BIORAD) with the ViiA 7 Real-Time PCR System, equipped with a FAST 96-well heated block (Applied Biosystems, Life Technologies). Expression levels of HPRT were used as an internal control to normalize the expression of different genes. Quantification was performed by the 2^−ΔΔCt method. The qPCR primers used in this study are provided in the ‘Resources’ table.

With few exceptions, due to limitations in the availability of human samples or their derived cultures, imaging and quantitative analyses were replicated in multiple cells from multiple individuals (biological replicates). In figures and tables, the numbers of individuals (patients, mice) is represented by N and the total numbers of cells by n. When multiple cells derive from multiple individuals, statistical measures are derived by hierarchical (nested) analysis implemented in the R environment (Sikkel et al., 2017). Significance of differences of averages or paired differences is established using the two-tailed Student’s t test, or, when the distributions of compared measures do not satisfy tests of normality and equal variance, the Mann-Whitney Rank Sum test. Correlation between variables is quantified by the first-order correlation coefficient R; this number is always accompanied by an estimate p of the probability of no correlation, a function of R and the sample size, calculated as described in Tammineni et al., 2020. In tables, sample medians are always provided after sample averages, to afford a first idea of the distribution.

The human samples for protein quantification include 12 ‘normal’ and 13 ‘HS’ individuals selected at random from available samples that satisfy size and preservation criteria. The numbers were chosen for efficiency (as they complete a 25 lane electrophoresis chamber), for consistency with prior work (Tammineni et al., 2020), and after analysis that evinced power consistent with study goals (a power of 0.8 for detecting differences of averages of 30%, with standard deviations of 27% at α=0.05, is achieved with a sample size of 24, 12 each).

There was no blinding and no exclusion criteria other than adequate size and preservation of the sample — no discoloration or gross damage — evaluated visually upon receipt of the biopsy.

All quantitative raw data, as well as statistical processing are contained in ‘JNB’ (Sigmaplot) and ‘XLSX’ (Excel) annotated worksheets, publicly available in Harvard Dataverse as ‘Study on junctophilin 1 and calcium stress’ at https://dataverse.harvard.edu/dataverse/Junctophilin1_and_calcium included in dataset ‘Raw and processed numerical data’. All raw image sets are in the same Dataverse, included in dataset ‘Raw image sets for Junctophilin 1 and calcium stress’. For ease of access, a ‘data trace’ line is added to every figure and table legend, identifying the corresponding raw data in the Dataverse and storage in the Rush University lab. Individual files in the Dataverse contain annotations for traceback to figures and tables in the article.

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

We are grateful to Prof. Manuel Díaz, for discussion and comments — akin to mentoring — on many aspects of this study, to Prof. Filip Van Petegem for helpful suggestions (notably, that multiple cuts in JPh1 may be needed to release the regulatory fragment), to Prof. Vincenzo Sorrentino, for his generous gift of an N-terminal Myc-tagged junctophilin1 vector, to Profs. Isabelle Marty and Graham Lamb, for detailed comments on the manuscript and its prepublication in bioRxiv. Funding: National Institute of Arthritis and Musculoskeletal and Skin Diseases (R01 AR 071381) to SR, ER, and M Fill (Rush University). National Institute of Arthritis and Musculoskeletal and Skin Diseases (R01 AR 072602) to ER, S L Hamilton, S Jung and F Horrigan (Baylor College of Medicine). National Center for Research Resources (S1055024707) to ER and others. Also supported by matching funds provided by philanthropy contributors to Rush University.

All patients were recruited and studied in the Malignant Hyperthermia Investigation Unit (MHIU) of the Canadian University Health Network, at Toronto General Hospital, in Toronto, Canada, under protocol 12-5474, last reviewed on 11/08/21. All patients consented to all aspects of the study, including publication. The use of tissue and the language of the informed consent were also approved by the Institutional Review Board of Rush University under protocol 16050502-IRB01.

This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All of the animals were handled according to approved institutional animal care and use committee (IACUC) protocols (#17-035, #18-065 and #21-068) of Rush University. The simple surgery (electroporation via intramuscular electrodes) was performed under anesthesia, and every effort was made to minimize suffering and discomfort.

© 2023, Tammineni et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.