Introduction

In skeletal muscle and the heart, the transient movement of calcium ions from sarcoplasmic reticulum (SR) calcium stores to the cytosol results in myofiber contraction and, to relax these fibers, the active transport of calcium ions back into the SR1. This transport of calcium ions relies heavily on the sarco/endoplasmic reticulum Ca2+ ATPase pump (SERCA)1,2. SERCA activity is inhibited by the direct binding of two small proteins, sarcolipin and phospholamban—which are expressed in skeletal and cardiac muscle3. Both sarcolipin and phospholamban inhibit SERCA, which leads to calcium accumulation in the cytosol and the contraction of myofibers—underscoring the importance of these proteins for proper muscle mechanics, and, in the context of skeletal muscle, organismal movement4,5,6. Much work has linked disruptions in sarcolipin expression to the pathophysiology of Duchenne muscular dystrophy (DMD), obesity, and type 2 diabetes7,8,9,10,11,12,13. In the context of DMD, increased sarcolipin expression is associated with worsened disease progression; while in metabolic disease, increased sarcolipin levels correlate with improved whole-body metabolism7,8,9,10,11,12,13. Despite the importance of sarcolipin to maintain normal skeletal muscle function, movement, and its link to disease, little is understood about how sarcolipin is regulated14,15.

Site-1 protease (S1P; also known as subtilisin/kexin-isozyme 1 or PCSK8) is a Golgi-resident protease that regulates cellular homeostasis by cleaving and activating a variety of transcription factors16. S1P has been classically studied in liver and bone as a regulator of cellular lipid homeostasis and the unfolded protein response17,18,19,20. We recently demonstrated that S1P controls skeletal muscle function and size and that excessive S1P activity impairs exercise tolerance in a person with an S1P gain-of-function mutation, emphasizing the importance of S1P in mammalian muscle biology21,22. We also reported that skeletal muscle-specific deletion of S1P in mice does not impact muscle cholesterol or lipid homeostasis at baseline, suggesting that S1P controls muscle biology via alternative mechanisms21. Here, we show that skeletal muscle-specific S1P knockout mice (S1PsmKO) have increased sarcolipin expression in multiple muscle groups yet have normal skeletal muscle SR calcium flux and contractile function. Our data indicate that S1P inhibits sarcolipin expression by preventing activation of the sarcolipin promoter and that activation of the S1P-substrate ATF6 acts as an inhibitor of calcineurin-dependent sarcolipin transcription. Our study unveils a function for S1P in the control of sarcolipin and identifies a mechanism by which this occurs.

Results

Sarcolipin expression is increased in S1P depleted skeletal muscle

We recently reported that S1P is a key component of skeletal muscle biology and identified 75 significantly differentially expressed genes in the gastrocnemius of S1P skeletal muscle-specific knockout (S1PsmKO) mice using bulk RNA sequencing21,22. From these 75 genes, sarcolipin was one of the most highly expressed genes in S1PsmKO muscle compared to WT control muscle (Fig. 1A). This finding, coupled with the fact that S1P and sarcolipin are both important for mammalian muscle biology, led us to hypothesize that one mechanism by which S1P controls muscle function is by regulating sarcolipin expression7,8,9,10,11,12,13,21,22. To test this hypothesis, we first validated our RNASeq data by measuring sarcolipin mRNA and protein levels by qPCR and western blotting, respectively and confirmed that sarcolipin expression is increased in the gastrocnemius of S1PsmKO mice compared to WT mice (Fig. 1B, C). Muscle group-specific differences in sarcolipin expression have been reported, and we previously showed that S1P can alter expression of genes in a muscle group-specific manner, particularly in muscle that contains a mixture of glycolytic and oxidative fibers, like the gastrocnemius and quadraceps21,23. Thus, we measured the levels of sarcolipin protein in other muscle groups and observed increased sarcolipin levels in the quadriceps, but not in the soleus or tibialis anterior, muscles composed primarily of oxidative and glycolytic fibers, respectively (Fig. 1D and Supplemental Fig. 1A, B). Together, these data indicate that S1P negatively regulates sarcolipin expression in a subset of skeletal muscle groups that contain a mixture of both oxidative and glycolytic fibers, such as the gastrocnemius and quadricep muscles.

Fig. 1: Deletion of S1P in skeletal muscle increases sarcolipin expression and does not impact baseline SR calcium flux.
figure 1

A Volcano plot of genes identified from RNA-Seq as significantly differentially increased (red dots) and decreased (blue dots) in gastrocnemius of S1PsmKO mice relative to WT mice. Sarcolipin is indicated. n = 4 per genotype. B qPCR of sarcolipin mRNA expression in gastrocnemius of S1PsmKO and WT mice, n = 5 per group. Western blots with corresponding densitometry of sarcolipin protein levels in (C) gastrocnemius and (D) quadriceps of S1PsmKO and WT mice. n = 3–5 per group. E Peak tension of twitch and tetanic contractions and time to fatigue of S1PsmKO and WT gastrocnemius. F Rates of ATP-stimulated free calcium SR uptake with area under the curve and uptake half-life at baseline. Graphs are normalized to cyclopiazonic acid (CPA)- treated samples. n = 4, 5 per group. Data are reported as ± SEM. Statistical significance was determined by an unpaired t-test with Welch’s correction for unequal variances and a Wilcoxon two-sample test.

Full size image

S1P deletion in skeletal muscle does not impact contractile function or SR calcium flux

Disruptions in skeletal muscle calcium regulators, such as sarcolipin, compromise contractile function in muscles like the soleus and plantaris24,25,26,27. Because the gastrocnemius and quadricep muscles of S1PsmKO mice have high levels of sarcolipin compared to WT mice, we examined whether this increase in sarcolipin correlated with changes in muscle contractile function. To do this, we measured in vivo plantarflexor torque of S1PsmKO and WT mice. The mouse plantar flexor muscles consist of gastrocnemius, soleus, and plantaris muscles, of which the gastrocnemius accounts for 70.6% of total muscle mass28. Twitch kinetics (peak twitch torque, time to peak torque and half-relaxation time) were unchanged between S1PsmKO and WT muscles. Additionally, we observed no differences in isometric tetanic contraction or fatigability between S1PsmKO and WT mice. Together, these data indicate that the measured change in sarcolipin expression in the gastrocnemius of S1PsmKO mice did not manifest in differences in peak torque or fatiguability (Fig.1E).

Sarcolipin inhibits SERCA activity in skeletal and cardiac muscle, and overexpressing sarcolipin in these tissues has been shown to decrease rates of SR calcium uptake4,5,29. Based on these data and the increased sarcolipin levels in S1PsmKO gastrocnemius and quadricep muscles, we examined whether SR calcium uptake was altered in S1PsmKO gastrocnemius. To test this, we quantified SR ATP-stimulated free calcium uptake in gastrocnemius homogenates of S1PsmKO and WT mice. We observed no difference in the rates of SR calcium uptake or half-life between S1PsmKO and WT tissues, even when we examined SERCA-dependent ATP-stimulated free calcium uptake by treating homogenates with the SERCA-specific inhibitor cyclopiazonic acid (CPA) (Fig. 1F). Together, these data indicate that in the gastrocnemius of S1PsmKO mice, an increase in sarcolipin protein is not associated with changes in SR calcium flux or contractile function.

S1P inhibits sarcolipin promoter activity

S1P controls an array of transcription factors that drive the expression of genes required for cellular homeostasis16. However, it has been reported that some of these transcription factors also inhibit gene expression by suppressing promoter activity30,31. Because sarcolipin expression is increased in S1PsmKO muscles, we examined whether S1P was indirectly inhibiting sarcolipin promoter activity. To measure sarcolipin transcription, we created a sarcolipin luciferase reporter construct that contains 1339 base pairs of the mouse sarcolipin promoter upstream of a cDNA that encodes for luciferase (Fig. 2A). We validated this construct in murine C2C12 cells by co-expressing it with constitutively active calcineurin (CnA), a known activator of the sarcolipin promoter14 and observed a 51-fold increase in sarcolipin promoter activity compared to empty vector control cells (Fig. 2B, C). To further validate the sarcolipin reporter construct, we tested the ability of the calcineurin inhibitor RCAN1 to suppress sarcolipin promoter activity. Indeed, overexpressing RCAN1 completely inhibited calcineurin-dependent sarcolipin promoter activity, consistent with previous studies (Fig. 2B, C)14.

Fig. 2: S1P inhibits basal sarcolipin transcription.
figure 2

A Schematic of sarcolipin luciferase reporter construct containing 1.339 kb of the mouse sarcolipin promoter. Created in BioRender. Brookheart, R. (2025) https://BioRender.com/ B Western blot of calcineurin and RCAN1 protein levels in C2C12 cells transiently transfected with EV, CnA, or RCAN1 plasmids. Representative blots of n = 3. C Quantification of sarcolipin promoter activity in C2C12 cells transfected with the indicated plasmids, n = 3 per group. D S1P (gene name: Mbtps1) knockdown in C2C12 cells as examined by qPCR. E Sarcolipin promoter activity in scrambled and S1P knockdown C2C12 cells. n = 6 per group. EV empty vector, CnA constitutively active calcineurin, SLN sarcolipin. Data are reported as ± SEM. Statistical significance was determined by unpaired t-test and unpaired t-test with Welch’s correction for unequal variances where necessary.

Full size image

With this validated sarcolipin promoter reporter system, we next examined sarcolipin promoter activity in the absence or presence of S1P. We transiently knocked down S1P in C2C12 cells using small interfering RNA (siRNA) oligos, as we previously reported (Fig. 2D)21. Cells transfected with a scrambled siRNA served as a negative control. Scrambled and S1P knockdown cells were transfected with the sarcolipin promoter construct, and promoter activity was measured. Relative to negative control cells, depletion of S1P in C2C12 cells increased sarcolipin promoter activity over 2-fold (Fig. 2E). These data indicate that loss of S1P de-represses sarcolipin promoter activity.

ATF6 restores repression of the sarcolipin promoter in S1P depleted cells

S1P is known to activate the transcription factor ATF6, which can selectively activate or repress promoter activity30,31,32,33. Unlike the other transcription factors that S1P activates, ATF6 inhibits calcineurin, a protein that drives sarcolipin promoter activity (Fig. 2C)14,34,35. To determine if the S1P substrate ATF6 is an inhibitor of the sarcolipin promoter, C2C12 cells were co-transfected with the sarcolipin promoter construct and an expression vector to express a truncated form of ATF6 (nATF6) or an empty vector control. Since nATF6 lacks the transmembrane domain that tethers ATF6 to the ER and Golgi, nATF6 constitutively localizes to the nucleus and controls gene expression in an unregulated manner36. In the presence of nATF6, sarcolipin promoter activity was inhibited compared to cells transfected with an empty vector control (Fig. 3B). Additionally, when cells were depleted of S1P, nATF6 still inhibited sarcolipin promoter activity (Fig. 3C). These data suggest that the S1P substrate ATF6 is an inhibitor of sarcolipin promoter activation.

Fig. 3: S1P inhibits sarcolipin transcription by activating ATF6.
figure 3

A Western blot of cells transfected with EV or constitutively active ATF6 (nATF6-HA). Representative blot of n = 3. B Sarcolipin promoter activity in cells expressing EV or nATF6. n = 5 per group. C Sarcolipin promoter activity in scrambled + EV, S1P knockdown + EV, and S1P knockdown + nATF6 cells. n = 5 per group. D ATF6-dependent transcription reporter activity in cells expressing EV, nATF6, or ATF6S1Pmut. n = 5 per group. Sarcolipin promoter activity in E cells expressing EV, nATF6, or ATF6S1Pmut and F cells expressing EV, CnA, or nATF6 as indicated. n = 5 per group. EV empty vector, CnA constitutively active calcineurin, ATF6S1Pmut ATF6 with mutated S1P cleavage site. Data are reported as ± SEM. Statistical significance was determined by unpaired t-test and unpaired t-test with Welch’s correction for unequal variances where necessary.

Full size image

S1P cleaves ATF6, triggering the subsequent translocation of ATF6 to the nucleus, where it can control promoter activity34,36. However, reports have shown that ATF6 can also be activated independently of S1P cleavage, and that this leads to selective but still significant changes in gene transcription34. To determine if ATF6’s inhibition of the sarcolipin promoter was dependent on S1P cleavage, we measured sarcolipin promoter activity in C2C12 cells expressing either empty vector, nATF6 (constitutively active), or an ATF6 mutant lacking the S1P cleavage site (ATF6S1Pmut)36. While nATF6 completely suppressed sarcolipin promoter activity, ATF6S1Pmut caused a slight, but significant reduction in sarcolipin promoter activity compared to control cells (Fig. 3E). As a control, we tested the ability of ATF6S1Pmut to activate an ATF6-responsive luciferase reporter construct, which, unlike nATF6, it did not (Fig. 3D). These data indicate that the S1P-dependent cleavage of ATF6 is a necessary precursor for ATF6’s inhibition of the sarcolipin promoter.

ATF6 suppresses calcineurin-dependent sarcolipin promoter activity

Since ATF6 inhibits, while calcineurin activates, the sarcolipin promoter (Figs. 2C and 3B)14 and ATF6 is known to suppress calcineurin-dependent signaling34,35, we hypothesized that the reciprocal control of the sarcolipin promoter by ATF6 and calcineurin was interconnected. To test this, we examined whether ATF6 prevented calcineurin from activating the sarcolipin promoter. Overexpressing constitutively active calcineurin (CnA) increased sarcolipin promoter activity (Fig. 3F)14; however, co-expressing active ATF6 (nATF6) with CnA decreased calcineurin-dependent sarcolipin promoter activation down to basal levels (Fig. 3F). These data indicate that nATF6 completely blocks calcineurin-stimulated activation of the sarcolipin promoter, placing ATF6 downstream of calcineurin signaling in this context.

In the heart, ATF6 inhibits calcineurin activation, possibly by increasing expression of the calcineurin inhibitor RCAN135. We and others have shown that RCAN1 blocks calcineurin-stimulated sarcolipin promoter activity (Fig. 2C)14. This would suggest that S1P controls sarcolipin expression via an ATF6-RCAN1 axis that targets calcineurin. In muscle, RCAN1 exists as multiple isoforms (i.e., RCAN1.1 and 1.4)14. We used two different siRNAs to deplete both RCAN1.1 and RCAN 1.4 isoforms in C2C12 cells, and measured sarcolipin promoter activity in the presence of nATF6. Despite depleting both RCAN1.1 and RCAN1.4 with two different siRNAs, active ATF6 still inhibited calcineurin-stimulated sarcolipin promoter activity (Supplemental Fig. 1C, D), suggesting that ATF6 disrupts calcineurin signaling independently of RCAN1.

Sarcolipin promoter activity is driven by CREB in vitro and in vivo

To determine which portion of the sarcolipin promoter is important for basal and calcineurin-stimulated promoter activity, we generated a series of truncation mutants of our sarcolipin luciferase reporter construct and examined their transcriptional activity at baseline and in response to constitutively active calcineurin (Fig. 4A, B). At baseline, all truncation mutants had reduced promoter activity compared to the full-length promoter (-1339) (Fig. 4B). In the presence of active calcineurin, sarcolipin promoter activity increased for all mutant promoters relative to their untreated controls, with mutant -19 exhibiting the lowest increase (Fig. 4B). These data suggest that nucleotides -989 through -19 of the sarcolipin promoter are essential for promoter activation both at baseline and in response to calcineurin stimulation.

Fig. 4: CREB binds to the sarcolipin promoter and is necessary to activate calcineurin-dependent sarcolipin transcription.
figure 4

A Schematic of sarcolipin promoter reporter construct deletions used in the study. Digits denote the number of base pairs upstream of the sarcolipin transcriptional start site. B Sarcolipin promoter activity in cells expressing the indicated truncated versions of the sarcolipin promoter reporter construct. Cells were transfected with EV or CnA. n = 5 per group. C Diagram of the predicted location of CREB DNA binding site on the mouse sarcolipin promoter. D Sarcolipin promoter activity in cells transfected with EV, CnA, or dominant negative CREB (ACREB) as indicated, n = 5, 6 per group. E Sarcolipin promoter activity in cells expressing the unaltered full-length sarcolipin promoter (CREB WT Luc) or full-length sarcolipin promoter with mutated CREB binding site (CREB Mut Luc) and transfected with either EV or CnA, as indicated, n = 5 per group. F Chromatin immunoprecipitation of gastrocnemius of WT and S1PsmKO mice quantified (graph, n = 3,4) and representative agarose gel of ChIP qPCR products (n = 3). G Model of S1P-regulated sarcolipin promoter inhibition. A, C, and G were created in BioRender. Brookheart, R. (2025) https://BioRender.com/. EV empty vector, CnA constitutively active calcineurin. Data are reported as ± SEM. Statistical significance was determined by unpaired t-test, unpaired t-test with Welch’s correction for unequal variances where necessary, and Wilcoxon two-sample test.

Full size image

Rotter et al. identified two predicted CREB binding sites on the sarcolipin promoter, one of which is positioned between nucleotides -989 and -19 (Fig. 4C)14. We hypothesized that CREB is an important component of sarcolipin promoter activity and may control the promoter at this predicted CREB site. To test this hypothesis, we first examined whether a dominant negative CREB (ACREB) can inhibit sarcolipin promoter activation at baseline and in the presence of active calcineurin37. At baseline, expression of ACREB in C2C12 cells decreased sarcolipin promoter activity by 50% compared to cells expressing empty vector (Fig. 4D). In the presence of active calcineurin, ACREB inhibited sarcolipin promoter activity by 48% compared to cells expressing active calcineurin alone (Fig. 4D). These data indicate that a functional CREB protein is a necessary factor for activating the sarcolipin promoter both at baseline and in response to calcineurin.

We next examined whether the CREB binding site between -989 and -19 identified by Rotter et al. is required for sarcolipin promoter activiation14. We mutated this CREB binding site in our full-length sarcolipin luciferase reporter construct and examined promoter activity at baseline and in the presence of active calcineurin (Fig. 4C). Mutation of the CREB binding site inhibited basal promoter activity by 87% relative to control cells expressing an intact full-length promoter construct (Fig. 4E). In the presence of active calcineurin, mutation of the CREB binding site inhibited promoter activity by 85% compared to calcineurin-treated cells expressing the intact full-length promoter (Fig. 4E). These data suggest that the CREB binding site on the sarcolipin promoter is required for promoter activity both at baseline and in response to calcineurin. We extended these studies in vivo and examined the presence of CREB on the endogenous sarcolipin promoter of S1PsmKO and WT gastrocnemius muscles. Using chromatin immunoprecipitation, we detected CREB bound to the sarcolipin promoter in both WT and S1PsmKO gastrocnemius; however, higher levels of CREB were observed on the sarcolipin promoter of S1PsmKO muscle relative to WT muscles (Fig. 4F). Together, our in vitro and in vivo data identify sarcolipin as a CREB target gene in skeletal muscle and that calcineurin is a driver of CREB-dependent sarcolipin promoter activation.

Discussion

In the present study, we define several novel and complex regulatory circuits controlling the expression of sarcolipin in skeletal myocytes. We show that S1P negatively regulates sarcolipin expression in skeletal muscle and define the mechanisms at play (Fig. 4G). Interestingly, we report that S1P’s influence on sarcolipin transcription does not impact contractile function or SERCA-dependent SR calcium flux at baseline. This latter finding suggests that the increased level of sarcolipin expression in S1PsmKO muscle is insufficient to impact SERCA activity–suggesting that a higher level of sarcolipin protein is required to alter calcium flux and impact muscle contractile function. Alternatively, our findings also suggest that sarcolipin activity is controlled by mechanisms beyond transcriptional regulation or protein level. Indeed, post-translational modifications are a known regulatory mechanism for controlling sarcolipin and SERCA38,39,40. One group reported that phosphorylation of sarcolipin by the kinase STK16 inhibits its ability to suppress SERCA40. Thus, it is possible that the increased level of sarcolipin protein observed in S1PsmKO skeletal muscles remains inactive due to the presence of an inhibitory protein modification—a possibility that we are exploring.

Several studies have reported that increased sarcolipin levels in the soleus muscle correlate with defective SR calcium uptake and contractile dysfunction24,25,26,27. Interestingly, reports using a diverse range of mouse models have shown that sarcolipin levels are increased in the soleus (a muscle composed primarily of oxidative muscle fibers), but not in mixed muscle groups like the gastrocnemius, which has a combination of oxidative and glycolytic muscle fibers41,42. S1PsmKO mice have increased sarcolipin in the gastrocnemius and quadriceps, another mixed fiber-type muscle group (Fig. 1C, D). Yet, we observed no increase in sarcolipin expression in the soleus and tibialis anterior of S1PsmKO mice – muscles composed of mainly oxidative or glycolytic fibers, respectively (Supplementary Fig. 1A, B). It is well documented that the type of muscle group dictates the function, gene expression, cellular calcium flux, and metabolism of the muscle25,43,44,45,46,47. Indeed, our recent publication showed that S1P controls muscle gene expression and size in a muscle-type-dependent manner21. Our data, coupled with those from previous groups, suggest that the regulation of sarcolipin expression and its influence on SR calcium flux is muscle-type dependent41,42. Current work is focused on teasing apart the mechanisms by which gene expression and muscle function are differentially regulated between mixed muscle groups versus groups composed primarily of one fiber type.

In the heart, ATF6 suppresses calcineurin signaling through the protein RCAN1 – RCAN1 is also expressed in skeletal muscle (Supplementary Fig. 2B)14,35. Our data suggest that ATF6 inhibits calcineurin-stimulated sarcolipin promoter activity independently of RCAN1. Our approach used siRNAs to deplete cells of both RCAN1.1 and RCAN1.4; it is possible that the remaining amounts of RCAN1 or other RCAN isoforms were sufficient for ATF6 to block calcineurin signaling and prevent sarcolipin promoter activity. We observed no differences in total calcineurin A levels in the gastrocnemius of S1PsmKO mice compared to WT muscle (Supplementary Fig. 2A), suggesting that ATF6’s influence on calcineurin in this context is not dependent on altering calcineurin A protein levels. In addition to inhibiting calcineurin, ATF6 may also influence the sarcolipin promoter through a mechanism that involves CREB, since ATF6 also functions as an inhibitor of CREB DNA-binding48. These reports, coupled with the data we present here, suggest that S1P activates ATF6, which is then able to inhibit calcineurin and CREB, leading to decreased sarcolipin expression (Fig.4G).

CREB is known to be phosphorylated on several residues by a plethora of kinases49,50,51. These posttranslational modifications regulate the binding of CREB to DNA as well as the activation of CREB-occupied promoters52,53,54,55. Moreover, coactivators of CREB (i.e., CRTC) can also play key roles in CREB-dependent promoter activation56. Whether these diverse regulatory mechanisms are at play in the control of sarcolipin transcription is not clear.

In conclusion, our study uncovered a complex series of regulatory players that control sarcolipin expression in skeletal muscle. We show that S1P inhibits sarcolipin promoter activity through its activation of the transcription factor ATF6 and that ATF6 suppresses the sarcolipin promoter by blocking calcineurin-dependent promoter activation (Fig. 4G). We also identified a functional CREB-binding site on the sarcolipin promoter and showed that CREB binds to and is an activator of this promoter. This work contributes to our understanding of the dual function that transcription factors can play as both activators and suppressors of gene promoters. Sarcolipin is implicated in the pathogenesis of DMD and metabolic disease, where it is suggested to play conflicting roles—high levels of sarcolipin are associated with negative DMD disease outcome, while high levels of sarcolipin correlate with improved muscle function in obese mice, but not in humans with obesity7,8,9,10,11,12. The nature of this discrepancy is not clear, but understanding how sarcolipin expression and activity are controlled in skeletal muscle will provide the foundational knowledge upon which we begin to understand this gene’s importance in human disease.

Methods

Animal studies

All mouse studies were approved by the Institutional Animal Care and Use Committee of Washington University in St. Louis. We have complied with all relevant ethical regulations for animal use. S1P floxed and S1P skeletal muscle-specific knockout mice in the C57BL/6J background were previously described17,21. Briefly, C57BL/6J HSA-Cre79 (B6.Cg-Tg(ACTA1-cre)79Jme/J; Stock No. 006149; Jackson Laboratory) mice were crossed with S1P floxed mice to generate skeletal-muscle-specific S1P knockout mice. Littermates not expressing Cre recombinase (S1P floxed mice) were used as controls (WT) for experiments. Mice were genotyped for Cre recombinase and floxed S1P alleles using gene-specific primers, as described previously21. Studies were performed using male mice aged 12–14 weeks of age. Studies were conducted using mice from different litters and across different cages. Mice were fed ad libitum a standard laboratory chow diet and water, and group housed on a 12 h light/dark cycle. Animals were routinely monitored for signs of discomfort, including changes in weight, fighting with littermates, lethargy, or being unkept. Any animals that appeared lethargic or unwell were humanely omitted from studies as per the approved protocol.

Plantarflexor torque testing

Contractile function of the plantarflexor muscle group (which consists of 70.6% gastrocnemius muscle by mass) was assessed via stimulation of the tibial nerve as previously described28,57. Briefly, hindlimbs of anesthetized mice were shaved, and a small ~2 mm long skin incision was made lateral to the knee where the sciatic nerve bifurcates into the peroneal and tibial branches. The peroneal branch was visualized by resecting the overlying musculature and then severed to avoid recruitment of the ankle dorsiflexors during stimulation. The skin was then closed with glue (Vetbond), and the mouse was transferred to a physiology testing rig (1300 A Aurora Scientific), where it was positioned supine on a heated platform with the knee clamped and the ankle secured to a footplate attached to a dual-mode ergometer. Two platinum needle electrodes were then inserted subcutaneously on either side of the tibial nerve. Optimal stimulation current was determined using a current sweep, and then a twitch contraction was acquired to assess twitch kinetics (peak twitch torque, time to peak torque and half-relaxation time). Peak tetanic torque was determined as the average of three contractions at 150 Hz with 2 min rest times between. Following these measurements, a fatiguing protocol was delivered with 150 Hz tetanic contractions every 2 s for 2 min. Time to fatigue was calculated as the time until peak torque dropped to 50% of the initial value.

Constructs

pSLN-Luciferase (pSLN-Luc) construct was generated by GENEWIZ in which the first 1339 base pairs upstream of the mouse sarcolipin start site were cloned into pGL3-Basic (Promega) using NheI and BglIII restriction enzyme sites. Truncated and CREB mutant pSLN-Luc constructs were generated using QuikChange Lightening Site-Directed Mutagensis Kit (Agilent, Cat# 210518) as per manufacturer’s instructions and custom designed primers (Supplemental Table 1). Custom primers were designed with the Agilent QuikChange Primer Design Program (https://www.agilent.com/store/primerDesignProgram.jsp). Additional constructs used in this study are pCGN-ATF6 (1-373) (a gift from Ron Prywes to Addgene plasmid # 2717358), pEGFP-ATF6-(S1P-) (a gift from Ron Prywes to Addgene plasmid # 3295636); RCAN1 (a gift from Jeffery Molkentin to Addgene plasmid # 6541359); CnA60; ACREB37; S1P-KDEL (a gift from Peter Espenshade61); and pMAXGFP (Lonza).

Cell culture

C2C12 cells were grown at 37 °C with 5% CO2 in DMEM supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin. For plasmid expression studies, cells were plated in 6-well plates at 1.2 × 105cells per well. After 24 h, cells were transfected with the indicated constructs using Lipofectamine 2000 (Life Technologies) as per manufacturer instructions. For siRNA studies, cells were transfected with RNAiMax (Life Technologies) as per manufacturer’s instructions. After 4 days post-transfection, cells were harvested using RIPA buffer (Sigma) for protein isolation or RNA-STAT 60 (Tel-TEST Inc) for RNA isolation.

Luciferase assays

Cells were plated in 12-well plates at 5 × 104 cells per well. After 24 h, cells were transfected with pSLN-Luc (either full length or truncated), pMAX-GFP, and additional constructs indicated in the figures using Lipofectamine 3000 (Life Technologies) as per manufacturer’s instructions. For studies using both DNA constructs and siRNAs, cells were transfected 24 h after plating with the indicated constructs and siRNAs using Lipofectamine 2000 (Life Technologies) following manufacturer’s instructions for simultaneous transfection of cells with DNA and siRNAs. Four days post-transfection, luciferase assays were performed using One-Glo Luciferase Assay System (Promega, Cat # E6120) as per manufacturer’s instructions. Luminescence followed by GFP fluorescence was quantified with a BioTek plate reader.

Immunoblotting

Skeletal muscle protein lysates were generated by homogenizing tissues in lysis buffer (20 mM Tris, 15 mM NaCl, 1 mM EDTA, 0.2% NP-40, and 10% glycerol) containing 2x Protease Complete cocktail tablet (Roche) and 1x Phosphatase Inhibitors (Roche) with 5 mm stainless steel beads in a TissueLyzer II (Qiagen). Protein lysates were rotated for 45 min at 4 °C, centrifuged at 15,000 ×  g for 15 min at 4 °C. For cultured cells, protein lysates were generated using RIPA buffer supplemented with 2x Protease Complete cocktail tablet (Roche) and 1x Phosphatase Inhibitors (Roche). Proteins were quantified by bicinchoninic acid assay (BCA, Pierce Biotechnology). Equal amounts of protein (20 µg) were resolved on a 4–12% BIS-Tris gradient gel (Invitrogen) with 1x MES running buffer and transferred to 0.2 µm nitrocellulose membrane (Bio-Rad). Blots were probed with appropriate primary and secondary antibodies and proteins visualized by LI-COR Odyssey imaging system. The following antibodies were used in our studies at a 1:1000 dilution: alpha-tubulin (T5168, Sigma); SERCA1 (ab2819, Abcam); SERCA2 (ab3625, Abcam); sarcolipin (ABT13, Sigma); His-tag (2365S, Cell Signaling Technology); calcineurin (2614S, Cell Signaling Technology); RYR1 (8153 T, Cell Signaling Technology). Densitometry analysis was performed using Bio-Rad Image Lab version 6.1.

Real time quantitative PCR

Total RNA was isolated from C2C12 cells and skeletal muscle with RNA STAT-60 (Tel-Test Inc) as per manufacturer’s instructions. For tissues, RNA was isolated by disrupting tissue in RNA STAT-60 using 5 mm steel beads (Qiagen) and a TissueLyser II (Qiagen). RNA was reverse transcribed into cDNA using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Quantitative real-time PCR was performed using Power SYBR green (Applied Biosystems) and transcripts quantified on an ABI QuantiStudio three sequence detection system (Applied Biosystems). Data was normalized to 36B4 expression and results analyzed using the 2−ΔΔCt method and reported as relative units to controls. Primer sequences are listed in Supplemental Table 1.

Free calcium flux studies

SERCA-mediated calcium uptake assay was performed as described25,62. Briefly, snap frozen mouse gastrocnemius muscles were cryopulverized, then homogenized with a cold steel bead in a Qiagen Tissuelyser II for 3 min at 30 Hz in 10 μL of Homogenizing buffer (250 mM sucrose, 5 mM HEPES, 0.2 mM PMSF, and 0.2% NaN3, pH 7.5) for each mg of sample. Samples were aliquoted and frozen to avoid more than two freeze-thaw cycles. Buffers were always kept on ice, and samples were kept on ice until the initial read. Samples were incubated with 1 μL of 2 mM Fluo-4 (diluted in molecular grade water; ThermoFisher #F14200) in Ca2+ uptake buffer (200 mM KCl, 20 mM HEPES, 10 mM NaN3, 5 μM TPEN, 15 mM MgCl2, pH 7.0) for 10 min before reading on BioTek plate reader at 485 nm excitation and 528 nm emission. After the initial read, 4 μL 250 mM ATP (pH 7.0) was simultaneously added to each well using a multichannel pipettor to initiate the kinetic reaction and avoid row-dependent variance. Maximum fluorescence values were acquired by repeatedly adding 80 μL of 100 mM CaCl2 and reading until values stayed consistent. Minimum fluorescent values were acquired by repeatedly adding 30 μL 50 mM EGTA (pH 7.0) to samples then reading until reaching plateau. CPA (2.2 μL 10 mM, Sigma Aldrich #C1530) or equivalent volume of vehicle (DMSO) was added to each sample followed by the addition of Fluo-4 and ATP and read as above to specifically determine SERCA-mediated calcium uptake. Total sample protein amounts were quantified by Micro BCA Protein Assay Kit (ThermoFisher #23235). All initial and kinetic readers were normalized to total sample protein concentration, then to the WT initial read value (at t = 0 s).

Chromatin Immunoprecipitation (ChIP)

Snap-frozen mouse gastrocnemius tissue was pulverized and fixed with 1% formaldehyde solution containing phosphatase and protease inhibitors and rotated for 20 min at room temperature. Crosslinking was quenched with 2.5 M glycine solution with rotation for 5 min at room temperature. Samples were pelleted and washed twice in 1x PBS containing protease and phosphatase inhibitors. Samples were resuspended in 0.5% Igepal and 0.25% Triton X-100 lysis buffer with protease and phosphatase inhibitors followed by Dounce homogenization. Samples were pelleted and washed in Tris-HCl lysis buffer with protease and phosphatase inhibitors, spun down, and the resulting pellet was resuspended in low salt lysis buffer containing protease and phosphatase inhibitors, followed by sonication to shear chromatin DNA. Optimal DNA shearing was confirmed by 1% agarose gel electrophoresis (250 bp–1 kb smears). After adding 10% Triton X-100 to the sonicated lysate, samples were centrifuged to pellet cellular debris, and 50 µL of the supernatants were collected to serve as total genomic inputs. Samples were incubated with appropriate antibodies (1–10 µg of antibody per 25 µg of DNA)–normal IgG Rb (2729S, Cell Signaling Technology) or CREB antibody (9197S, Cell Signaling Technology) for 1 h at 4 °C with rotation. Protein A/G magnetic beads (80105 G, Invitrogen) were washed three times in 0.5% BSA blocking solution and resuspended in blocking solution with salmon sperm DNA (Sigma) with rotation for 30 min at room temperature. Beads were added to chromatin-antibody sample complexes and incubated overnight at 4 °C with rotation. Samples were washed seven times with RIPA wash buffer containing protease and phosphatase inhibitors. Protein-DNA complexes were eluted from beads by incubating at 65 °C for 15 min in Tris-HCl Elution buffer and further incubated at 65 °C overnight to reverse protein-DNA crosslinks along with total genomic inputs. Samples were incubated with RNase A and then Proteinase K supplemented with 300 mM CaCl2 solution. DNA was isolated using the Qiagen Dneasy Blood and Tissue Kit (69594, Qiagen) as per manufacturer’s instructions. DNA was quantified by NanoDrop, and qPCR was carried out as described above, and products ran on an agarose gel. Efficiency of sarcolipin DNA primers was confirmed by running a qPCR primer efficiency curve. Primer sequences are listed in Supplemental Table 1.

RNA-sequencing studies

RNA-sequencing studies of mouse gastrocnemius were performed as reported previously in ref. 21. Accession number for the RNA-Seq data presented in this study was previously deposited at NCBI GEO under accession number GSE199014 located at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE199014.

Statistics and reproducibility

Normality of data distribution was examined with graphical inspection and the Shapiro-Wilk Test. Normally distributed data were analyzed by unpaired t-test, and Welch’s Correction for unequal variances was applied when the homogeneity of variance assumption was violated. For data with unequal distribution, groups were compared using a Mann-Whitney Test. Data with small sample sizes (groups containing ≤ 4 observations) were analyzed with the Wilcoxon Test. Analysis was performed with GraphPad Prism Version 10.2.3 for Windows (Dotmatics). A p value  ≤ 0.05 was considered statistically significant. Data are reported as ± SEM with no exclusions. Statistical details and the number of samples used in each study are indicated in the figure legends and in individual data points for bar graphs. Sample sizes were determined based on prior research21,22. For studies using mouse tissues, biological replicates are defined as the number of individual mice used for each experimental group. For cell culture-based studies, biological replicates are noted as the number of times the experiment was performed from start to finish by using separate plates and performing separate treatments and transfections for each replicate. During animal studies, personnel conducting the experiment were blinded to which samples corresponded to which group. Mice were housed in the same mouse facility room and rack.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.