Inactivation of SERCA2 at Cys674 induces skeletal muscle atrophy by activating the TGFβ/Smad-S100a4 axis to promote inflammation

Abstract

Background & aims

Disuse-induced skeletal muscle atrophy is a major clinical challenge lacking effective targeted therapies. Sarco/endoplasmic reticulum Ca²⁺-ATPase 2 (SERCA2) is essential for calcium homeostasis in skeletal muscle, but whether its C674 active site contributes to muscle atrophy remains unclear. This study investigated the role of SERCA2 C674 inactivation in skeletal muscle atrophy and the underlying mechanisms.

Methods

Oxidized SERCA2 (C674–SO₃H) expression was assessed in atrophied muscles from bedridden patients and in a murine hindlimb suspension (HLS) model. A SERCA2 C674S knock-in (SKI) mouse model was generated to mimic irreversible oxidation in vivo. Transcriptomic and proteomic analyses, together with in vitro experiments in C2C12 myoblasts and in vivo interventions using Losartan or adeno-associated virus serotype 9 (AAV9)-mediated shRNA, were performed to define downstream signaling pathways and therapeutic potential.

Results

Total SERCA2 protein levels were unchanged in atrophied muscles from patients and HLS mice, whereas C674–SO₃H expression was significantly increased. SKI mice developed spontaneous skeletal muscle atrophy, with reduced myofiber cross-sectional area and impaired muscle strength. Multi-omics analyses showed that SERCA2 dysfunction activated the renin-angiotensin system (RAS)-dependent TGF-β/Smad pathway and markedly upregulated S100a4. Mechanistically, S100a4 interacted with Smad3 and acted as a key downstream effector promoting oxidative stress, inflammation, and muscle protein degradation-related pathways. In vivo treatment with Losartan or AAV9-mediated S100a4 knockdown alleviated local inflammation, fibrosis, and muscle atrophy in both SKI and HLS mice.

Conclusions

The oxidative inactivation of the SERCA2 C674 site constitutes a novel mechanism driving skeletal muscle atrophy. The SERCA2-RAS-TGF-β/Smad-S100a4 signaling axis emerges as a highly promising therapeutic target for mitigating disuse-induced skeletal muscle wasting.

Highlights

• •SERCA2 Cys674 oxidation is increased in atrophic skeletal muscle.
• •Cys674 inactivation drives muscle atrophy in SKI mice and C2C12 models.
• •SERCA2 dysfunction activates the RAS-TGFβ/Smad-S100a4 axis.
• •Losartan or S100a4 silencing alleviates atrophy, fibrosis, and inflammation.

1. Introduction

Skeletal muscle atrophy, particularly disuse atrophy induced by prolonged bed rest, immobilization, or mechanical unloading, represents a prevalent and formidable clinical challenge in orthopedics and rehabilitation medicine [1,2]. This progressive pathological condition is characterized by a rapid decline in muscle mass and strength, which significantly increases the risk of falls, fractures, and adverse clinical outcomes [[3], [4], [5]]. Although previous studies have implicated oxidative stress, excessive protein degradation via the ubiquitin-proteasome system, and inflammatory microenvironment remodeling in the progression of muscle wasting [6,7], precise and effective targeted pharmacological interventions remain elusive. Disruption of intracellular calcium (Ca²⁺) homeostasis is considered a critical trigger for these catabolic pathways [8]; mechanistically, sustained cytosolic Ca²⁺ accumulation can activate calpain-mediated myofibrillar proteolysis, disturb mitochondrial function, enhance reactive oxygen species production, and stimulate downstream catabolic programs, including FoxO-driven expression of MuRF1 and Atrogin-1 [[9], [10], [11]]. In addition, aberrant Ca²⁺-dependent signaling through calcineurin/NFAT, CaMK, and inflammatory pathways may further exacerbate protein degradation, oxidative stress, and muscle fiber degeneration [[12], [13], [14]]. However, the upstream core mechanisms linking mechanical unloading to Ca²⁺ dysregulation and subsequent muscle atrophy have yet to be fully elucidated [15,16].

The sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA) plays an indispensable role in maintaining intracellular calcium (Ca²⁺) homeostasis by actively pumping cytosolic Ca²⁺ back into the sarcoplasmic reticulum against its concentration gradient [17]. In skeletal muscle, the SERCA2 isoform-particularly SERCA2a in slow-twitch fibers-is pivotal for the regulation of muscle relaxation and Ca²⁺ signaling [18,19]. Notably, cysteine 674 (C674) of SERCA2 serves as a critical redox-sensitive residue [20]. Within the microenvironment of disused skeletal muscle, which is characterized by excessive reactive oxygen species (ROS) production, this site is highly susceptible to irreversible oxidative modification (forming C674–SO₃H), leading to the functional impairment of the SERCA2 pump [21,22]. While ROS-mediated SERCA2 inactivation has been well-documented in cardiovascular dysfunction [23,24], it remains unknown whether the specific oxidative inactivation of the SERCA2 C674 site directly drives the pathological progression of skeletal muscle atrophy.

Sustained intracellular Ca²⁺ accumulation and oxidative stress can aberrantly activate downstream pathological signaling networks. Among these, the transforming growth factor-beta (TGF-β)/Smad pathway serves as a recognized central hub that negatively regulates muscle mass and promotes both fibrosis and protein degradation [25]. Intriguingly, recent multi-omics analyses and studies on fibrotic diseases have highlighted the calcium-binding protein S100a4 as a potent downstream effector of the TGF-β/Smad axis [26,27]. S100a4 acts as a critical node in amplifying inflammatory responses (e.g., facilitating the release of TNF-α and IL-1β) and driving extracellular matrix remodeling [28,29]. Nevertheless, in the context of skeletal muscle atrophy, it remains unclear whether a hierarchical regulatory axis exists linking SERCA2 dysfunction, aberrant TGF-β/Smad activation, and S100a4-driven inflammation, thereby representing a major gap in current understanding.

Accordingly, the present study aims to comprehensively elucidate the role and underlying molecular mechanisms of SERCA2 C674 inactivation in disuse-induced skeletal muscle atrophy. We propose a theoretical framework in which SERCA2 C674 inactivation disrupts intracellular Ca²⁺ homeostasis and enhances oxidative stress, thereby activating RAS-dependent TGF-β/Smad signaling. Activated Smad3 may induce S100a4 expression, and increased S100a4 may further amplify NF-κB related inflammatory responses and pro-inflammatory cytokine production. This inflammatory milieu, in turn, can promote proteolytic remodeling by enhancing atrophy-related ubiquitin ligases such as Atrogin-1 and MuRF1, ultimately activating protein degradation-related pathways. Utilizing clinical skeletal muscle biopsy samples from bedridden patients and a murine hindlimb suspension (HLS) model, we first evaluated the expression profiles of oxidatively inactivated SERCA2. Subsequently, we engineered a novel SERCA2 C674S knock-in (SKI) mouse model to simulate the irreversible oxidation of this site in vivo, and integrated transcriptomic and proteomic analyses to delineate the downstream regulatory networks. We hypothesize that SERCA2 C674 inactivation exacerbates skeletal muscle atrophy by activating the TGF-β/Smad-S100a4 signaling axis, thereby amplifying the inflammatory response and activating protein degradation-related pathways. Elucidation of this previously unrecognized mechanism would provide a strong theoretical basis for identifying and developing novel therapeutic targets for disuse-induced skeletal muscle atrophy.

2. Materials and methods

2.1. Study design

This study was designed to determine whether oxidative inactivation of SERCA2 at Cys674 contributes to disuse-induced skeletal muscle atrophy and to define the underlying downstream mechanisms.

First, we examined skeletal muscle samples from bedridden patients and a murine hindlimb suspension (HLS) model to determine whether SERCA2 C674–SO₃H was increased during disuse-associated muscle atrophy. Second, to assess whether SERCA2 C674 inactivation is sufficient to induce muscle wasting, we generated and characterized SERCA2 C674S knock-in (SKI) mice and evaluated muscle mass, myofiber morphology, and muscle function. Third, to identify downstream molecular pathways, we performed RNA sequencing in C2C12 cells overexpressing a C674-inactivated SERCA2 mutant and quantitative proteomic analysis of skeletal muscle from WT and SKI mice. Fourth, candidate signaling pathways and effector molecules, including the RAS-TGF-β/Smad-S100a4 axis, were validated using qRT-PCR, western blotting, immunohistochemistry, immunofluorescence, molecular docking, and co-immunoprecipitation assays.

Finally, to evaluate the therapeutic relevance of this pathway, losartan-mediated RAS inhibition and AAV-mediated S100a4 knockdown were performed in SKI and HLS mice, followed by assessment of muscle function, muscle mass, histopathology, inflammation, fibrosis, and atrophy-related protein degradation markers.

2.2. Clinical samples

After obtaining written informed consent, discarded skeletal muscle tissues were collected from male patients aged 30-70 years who underwent primary unilateral hip arthroplasty in the Department of Joint Surgery at the First Affiliated Hospital of Harbin Medical University. Human skeletal muscle samples were obtained from the tensor fasciae latae muscle during routine surgical exposure or necessary soft-tissue release. All specimens used in this study were surgically discarded tissues; no additional muscle biopsy was performed solely for research purposes, and tissue collection did not alter the surgical procedure, expand the operative field, or increase patient risk. Patients were classified into the disuse-associated skeletal muscle atrophy group or the non-disuse control group according to preoperative history, physical examination, and imaging assessment, as detailed in the Supplementary Methods. Immediately after collection, discarded muscle specimens were snap-frozen in liquid nitrogen and stored for subsequent protein and RNA extraction. The study protocol was approved by the Ethics Committee of the First Affiliated Hospital of Harbin Medical University (IRB approval No. IRB-AF/SC-12/03.0), and written informed consent was obtained from all participants.

2.3. Mouse Hindlimb Suspension (HLS) model

A mouse hindlimb suspension/unloading (HLS/HLU) model was used to induce disuse skeletal muscle atrophy, as originally described by Morey-Holton and Globus [30]. By unloading the hindlimbs under a 30° head-down tilt, this model mimics mechanical unloading associated with prolonged bed rest or microgravity. HLS was maintained for 4 weeks according to established protocols [31,32].

2.4. Hematoxylin and Eosin (H&E) staining

Skeletal muscle tissues were fixed in 4% paraformaldehyde, paraffin-embedded, and sectioned at 4 μm. After deparaffinization and rehydration, sections were stained with hematoxylin and eosin, mounted, and imaged using a Leica microscope. Quantitative analysis was performed in ImageJ.

2.5. Masson's trichrome staining

Paraffin-embedded skeletal muscle sections (4 μm) were subjected to Masson's trichrome staining using a commercial kit (Solarbio, G1340) according to the manufacturer's protocol. Collagen deposition was imaged with a Leica microscope and quantified in ImageJ as the collagen-positive area fraction. Three random fields per section and three sections per animal were analyzed in a blinded manner.

For Masson's trichrome analysis, stained muscle sections were imaged at 10× magnification. The entire 10× field containing skeletal muscle tissue was analyzed, while non-tissue background, tissue folds, tears, and obvious staining artifacts were excluded. Collagen-positive fibrotic area was identified based on blue staining and quantified using ImageJ. The fibrotic area fraction was calculated as collagen-positive area divided by the total skeletal muscle tissue area within the analyzed field. Multiple non-overlapping fields were analyzed for each animal, and the mean value was used as one biological replicate.

2.6. Immunohistochemistry (IHC)

Paraffin-embedded skeletal muscle sections (4 μm) were subjected to immunohistochemical staining after routine deparaffinization, rehydration, antigen retrieval, and blocking. Sections were incubated with primary antibodies overnight at 4°C, followed by the corresponding secondary antibodies. Signals were visualized using a DAB detection kit (AR1027-3; Boster, China) and counterstained with hematoxylin. Images were acquired with a Leica microscope, and quantitative analysis was performed using ImageJ.

2.7. Immunofluorescence (IF)

Cells grown on coverslips or in plates were processed for immunofluorescence using standard procedures. Following incubation with primary antibodies overnight at 4°C and fluorophore-conjugated secondary antibodies, nuclei were counterstained with DAPI and images were acquired by fluorescence microscopy.

2.8. Quantitative Real-Time PCR (qRT-PCR)

Total RNA was extracted and reverse-transcribed into cDNA using a commercial kit (FSQ101, TOYOBO). Quantitative PCR was performed with SYBR Green on an ABI 7500 Fast Real-Time PCR System. Gene expression was normalized to GAPDH and analyzed using the 2^^{−ΔΔ Ct} method. Primer sequences are listed in Table 1.

Table 1.

Primer sequences used for quantitative real-time PCR.

Gene	Sequence
Mouse Atrogin-1-F	TCTCCAGACTCTCTACACATCC
Mouse Atrogin-1-R	GAATGGTCTCCATCCGATACAC
Mouse MuRF1-F	GGACTACTTTACTCTGGACTTAGAAC
Mouse MuRF1-R	CAGCCTCCTCTTCTGTAAACTC
Mouse S100a4-F	GAGGCCCTGGATGTAATTGT
Mouse S100a4-R	CTCCTTGAGCTCTGTCTTGTT
Mouse Smad3-F	CCTGAGTGAAGATGGAGAAACC
Mouse Smad3-R	CTGGCTGTAGGTCCAAGTTATT
Mouse TGF-β1-F	CGAAGCGGACTACTATGCTAAA
Mouse TGF-β1-R	TCCCGAATGTCTGACGTATTG
Mouse αSMA-F	TCAGGGAGTAATGGTTGGAATG
Mouse αSMA-R	GGTGATGATGCCGTGTTCTA
Mouse AGT-F	TCCCACGCTCTCTGGATTTA
Mouse AGT-R	CAAGTTCATCTTCCACCCTGTC
Mouse AGTR1-F	GCCTCTGTGGGCAGTTTAT
Mouse AGTR1-R	CACTGGCGTAGAGGTTGAAA
Mouse IL-6-F	CTTCCATCCAGTTGCCTTCT
Mouse IL-6-R	CTCCGACTTGTGAAGTGGTATAG
Mouse TNFα-F	TTGTCTACTCCCAGGTTCTCT
Mouse TNFα-R	GAGGTTGACTTTCTCCTGGTATG
Mouse IL-1β-F	CCACCTCAATGGACAGAATATCA
Mouse IL-1β-R	CCCAAGGCCACAGGTATTT
Mouse MIP2-F	GACAGAAGTCATAGCCACTCTC
Mouse MIP2-R	GCCTTGCCTTTGTTCAGTATC
Mouse SKI-F	CCACAAATGGCTCTCAGGTT
Mouse SKI-R	CAGCTCTAGGCAGAGGGACT
Mouse GAPDH-F	TGGTGAAGCAGGCATCTGAG
Mouse GAPDH-R	TGAAGTCGCAGGAGACAACC

2.9. Western Blotting (WB)

Cells or skeletal muscle tissues were lysed and subjected to immunoblotting using standard protocols. Equal amounts of protein were resolved by SDS-PAGE, transferred to PVDF membranes, and incubated with the indicated primary antibodies and IRDye-conjugated secondary antibodies. Signals were visualized with an Odyssey® CLx system and quantified relative to GAPDH.

2.10. Protein-protein docking

Protein-protein docking was performed to predict the potential interaction and binding interface between Smad3 and S100A4. The three-dimensional structures of both proteins were obtained from the AlphaFold database and preprocessed using PyMOL 3.1 and PDB2PQR 3.7.1. Docking was then conducted with ZDOCK 3.0.2 under default settings. The top-ranked docking models were further analyzed and visualized in PyMOL to identify potential interface-contacting residues and characterize the predicted interaction pattern.

2.11. Co-immunoprecipitation (Co-IP)

Cell lysates were prepared in NP-40 buffer containing protease and phosphatase inhibitors. After pre-clearing, equal amounts of protein were incubated with the indicated antibodies or control IgG, followed by capture with Protein A/G magnetic beads. Immunoprecipitated proteins were then analyzed by western blotting. Reciprocal Co-IP assays were performed as indicated.

2.12. C2C12 cell culture and myogenic differentiation

C2C12 myoblasts were cultured in growth medium and induced to differentiate at 80–90% confluence by switching to DMEM supplemented with 2% horse serum. Differentiation was continued for 7 days, with myotube formation monitored by phase-contrast microscopy.

2.13. Lentiviral SERCA2a-C674S construct

The lentiviral SERCA2a-C674S construct, referred to as “2as” in this study, was generated using the pLV3-CMV-MCS-3×FLAG-CopGFP-Puro backbone. The construct encoded mouse Atp2a2 isoform 2a (SERCA2a) carrying a cysteine-to-serine substitution at residue 674 (C674S). The mutant SERCA2a coding sequence was driven by the cytomegalovirus (CMV) promoter and fused with a C-terminal 3×FLAG tag. The vector also contained CopGFP for monitoring transduction efficiency and a puromycin resistance cassette for stable cell selection. The corresponding empty pLV3-CMV-MCS-3×FLAG-CopGFP-Puro vector was used as the negative control. Expression of the SERCA2a-C674S construct was verified by western blotting using antibodies against SERCA2 and FLAG.

2.14. Viral transduction, siRNA transfection, and generation of stable cell lines

C2C12 cells were subjected to viral transduction or siRNA transfection using standard procedures. Stable cell lines were established by lentiviral infection followed by puromycin selection, and gene silencing was achieved using target-specific siRNAs. The siRNA/shRNA sequences targeting S100a4 and Smad3 are presented in Table 2. Experimental efficiency was confirmed by fluorescence microscopy, qRT-PCR, and western blotting.

Table 2.

Sequences of siRNAs used for gene knockdown.

Gene	Sequence
House mouse-S100a4-siRNA	GGACAGAUGAAGCUGCAUU
House mouse-Smad3-siRNA	CCAGAGCAAUAUUCCAGAA

2.15. RNA sequencing and data analysis

C2C12 cells transduced with control or SERCA2 mutant adenovirus were subjected to transcriptomic profiling on an Illumina platform. Differential expression analysis was performed using DESeq2, with significance defined as adjusted P ≤ 0.05 and |log2 fold change| ≥ 1. Heat maps and volcano plots were generated for visualization. Each group included four biological replicates.

2.16. Dual-luciferase reporter assay

Dual-luciferase reporter assays were performed to evaluate the regulatory effect of Smad3 on S100a4 promoter activity. C2C12 cells and 293T cells were seeded into 24-well plates and cultured until they reached approximately 70-80% confluence. Cells were then co-transfected with the S100a4 promoter-driven firefly luciferase reporter plasmid and the Renilla luciferase plasmid pRL-TK, together with either the Smad3 overexpression plasmid or the corresponding empty vector. For Smad3 knockdown experiments, cells were co-transfected with the S100a4 promoter luciferase reporter plasmid, pRL-TK, and either siSmad3 or siNC. The pGL4-Basic vector, a promoterless luciferase reporter vector, was used as the negative control for basal reporter activity. In selected experiments, the SBE4-luciferase reporter plasmid was used as a positive reporter to confirm Smad3 transcriptional activity.

Transfections were performed using Lipofectamine 2000 according to the manufacturer's instructions. After 24 h of transfection, cells were lysed, and firefly and Renilla luciferase activities were measured using a dual-luciferase reporter assay system. Firefly luciferase activity was normalized to Renilla luciferase activity to control for transfection efficiency. The relative luciferase activity was calculated as the firefly/Renilla ratio and normalized to the indicated control group. Each experiment was performed with at least three independent biological replicates.

2.17. Generation of SERCA2 C674S knock-in mice

SERCA2 C674S knock-in (SKI) mice on a C57BL/6J background were obtained from The Jackson Laboratory. As previously described [33], in these mice, the codon encoding Cys674 in exon 14 of the SERCA2 gene was replaced with a serine codon to mimic oxidative inactivation of SERCA2 in vivo. Because homozygous mutants were embryonically lethal, only heterozygous SKI mice (SKI^+/−) were used in this study, with wild-type littermates serving as controls.

2.18. Proteomic analysis of skeletal muscle tissues

Gastrocnemius muscles from WT and SKI mice were analyzed by quantitative proteomics. Mass spectrometry was performed on a Bruker timsTOF Pro 2 system with PASEF acquisition, and protein identification and quantification were conducted using DIA-NN (v1.8.1). Differential expression was defined as FDR ≤0.05 and |log2 fold change| ≥ 0.38, with three biological replicates per group.

2.19. Animal experimental design

Two animal experiments were performed with six mice per group. The detailed experimental time framework for atrophy induction, observable muscle size changes, and intervention is provided in the Supplementary Methods. Losartan was used as a pharmacological inhibitor of the renin-angiotensin system (RAS). As a selective angiotensin II type 1 receptor (AGTR1) blocker, losartan inhibits Ang II-AT1R signaling and thereby attenuates downstream profibrotic and pro-inflammatory pathways, including TGF-β/Smad signaling. In the first, 2-month-old SKI mice received losartan (20 mg/kg/day, gavage), 0.9% NaCl, AAV-sh-S100a4, or AAV-sh-NC via tibialis anterior injection; dosing and viral administration were based on previous reports [[34], [35], [36], [37]]. WT littermates were used as controls. In the second, HLS-treated WT mice received losartan in drinking water (0.6 g/L) or tibialis anterior injection of AAV-sh-S100a4 or AAV-sh-NC. To minimize handling during unloading, losartan was administered in drinking water according to a previous study [38]. After 4 weeks, muscle function and atrophy-related changes were evaluated by grid hanging, wet weight measurement, protein analysis, and histopathology.

2.20. Four-limb wire hang test

Muscle strength and endurance were evaluated by the inverted wire hang test according to established protocols [39,40]. Mice were placed on a wire grid, which was inverted and suspended above a padded surface, and the latency to fall was recorded. Three trials were performed for each mouse, and the average hanging time was used for analysis.

2.21. Statistical analysis

Statistical analyses were performed using GraphPad Prism 10. Data are presented as mean ± SD unless otherwise indicated. For comparisons between two independent groups, an unpaired two-tailed Student's t-test was used. For comparisons involving more than two groups with one independent factor, one-way ANOVA followed by Tukey's multiple comparisons test was performed. For grouped comparisons involving two independent factors, ordinary two-way ANOVA followed by Sidak's multiple comparisons test was used. For repeated measurements over time, two-way repeated-measures ANOVA followed by Sidak's multiple comparisons test was used when appropriate. Correlations between SERCA2 C674–SO₃H levels and S100a4 expression were assessed using Spearman's rank correlation analysis. All t-tests and correlation analyses were two-tailed, and P < 0.05 was considered statistically significant.

3. Results

3.1. Increased expression of inactivated SERCA2 at Cys674 in disuse-atrophied skeletal muscle

To establish a model of disuse-induced skeletal muscle atrophy, mice were subjected to hindlimb suspension (HLS) for 4 weeks (Fig. 1A). HLS markedly reduced the relative mass of the gastrocnemius (GAS), soleus (SOL), and tibialis anterior (TA) muscles compared with controls, indicating robust muscle wasting (Fig. 1B). Histological analysis further showed evident myofiber shrinkage and disorganized fascicular architecture in HLS muscles, which was confirmed by a significant reduction in myofiber cross-sectional area in GAS, SOL, and TA (Fig. 1D–F). Consistent with these morphological changes, the atrophy-related genes Atrogin-1 (Fbxo32) and MuRF1 (Trim63) were significantly upregulated in HLS muscles (Fig. 1C). Immunohistochemistry further demonstrated enhanced TRIM63 staining in GAS and SOL from HLS mice (Fig. 1E). Together, these findings confirm the successful establishment of a stable disuse atrophy model at both morphological and molecular levels.

Fig. 1.

SERCA2 C674–SO₃H is increased in skeletal muscle from murine disuse atrophy models and human atrophic specimens. (A) Schematic illustration of the hindlimb suspension (HLS) mouse model. (B) Ratios of gastrocnemius (GAS), soleus (SOL), and tibialis anterior (TA) muscle mass to body weight in control (Con) and HLS mice (n = 6). (C) Relative mRNA expression levels of Fbxo32 and Trim63 in skeletal muscle (n = 3). (D) Representative hematoxylin and eosin (H&E) staining of GAS, SOL, and TA muscles from Con and HLS mice (n = 6). Scale bar, 50 μm. (E) Representative immunohistochemical staining of TRIM63 in GAS and SOL muscles (n = 6). Scale bar, 100 μm. (F) Quantification of myofiber cross-sectional area (CSA) in GAS, SOL, and TA muscles, and quantification of the TRIM63-positive area in GAS and SOL muscles. (G) Representative western blots and densitometric analysis of SERCA2, C674–SO₃H, FBXO32, and TRIM63 protein levels in skeletal muscle from Con and HLS mice (n = 3). (H) Representative western blots and densitometric analysis of SERCA2, C674–SO₃H, FBXO32, and TRIM63 protein levels in human muscle specimens (n = 6). (I) Representative immunostaining of SERCA2 and C674–SO₃H in SOL and GAS muscles, corresponding higher-magnification images, and quantification of the C674–SO₃H-positive area (n = 6). Scale bar, 50 μm.Data are presented as mean ± SD. Statistical significance was determined using unpaired Student's t-test. ns, not significant; ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; ∗∗∗∗P < 0.0001.

We next examined whether SERCA2 was altered during disuse-induced muscle atrophy. Western blot analysis showed that FBXO32 and TRIM63 protein levels were significantly increased in HLS muscles, consistent with activation of the ubiquitin-proteasome pathway (Fig. 1G). Notably, a similar pattern was observed in skeletal muscle samples from patients with disuse atrophy after prolonged bed rest, supporting the clinical relevance of the HLS model (Fig. 1H). Although total SERCA2 protein abundance remained unchanged, the inactivated form of SERCA2 at Cys674 (C674–SO₃H) was markedly increased in both HLS mouse muscles and human atrophic muscle samples (Fig. 1G and H). Consistently, immunohistochemical staining confirmed a significant accumulation of SERCA2 C674–SO₃H in the GAS and SOL of HLS mice (Fig. 1I). These data indicate that disuse-induced skeletal muscle atrophy is associated not with altered SERCA2 expression, but with enhanced oxidative inactivation of SERCA2 at Cys674, implicating this modification in the pathogenesis of disuse muscle wasting.

3.2. SKI mice exhibit spontaneous skeletal muscle atrophy

To determine whether SERCA2 Cys674 inactivation is sufficient to drive skeletal muscle atrophy in vivo, we generated SERCA2 C674S knock-in (SKI) mice on a C57BL/6J background, in which Cys674 was replaced by serine to mimic functional inactivation of this critical residue (Fig. 2A). Successful generation of the mutant line was confirmed by genotyping, and wild-type (WT) littermates were used as controls (Fig. 2B).

Fig. 2.

Inactivation of the SERCA2 Cys674 redox site induces spontaneous skeletal muscle atrophy in SKI mice. (A) Schematic illustration of the generation of SERCA2 C674S knock-in (SKI) mice. (B) Representative PCR genotyping of wild-type (WT) and SKI mice. (C) Representative gross morphology of tibialis anterior (TA), soleus (SOL), and gastrocnemius (GAS) muscles from WT and SKI mice. (D) Representative hematoxylin and eosin (H&E) staining of TA and SOL muscle sections from WT and SKI mice. Scale bar, 50 μm. (E) Body weight (BW), muscle-to-body weight ratios of TA, SOL, and GAS, and quantification of myofiber cross-sectional area (CSA) in TA and SOL muscles from WT and SKI mice (n = 3-5). (F) Muscle function in WT and SKI mice assessed by the grip strength/hanging test (n = 9). Data are presented as mean ± SD. Statistical significance was determined using unpaired Student's t-test. ns, not significant; ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; ∗∗∗∗P < 0.0001.

Although body weight was comparable between SKI and WT mice, SKI mice exhibited a marked reduction in the size and relative mass of the tibialis anterior (TA), soleus (SOL), and gastrocnemius (GAS) muscles (Fig. 2C–E). Histological analysis further revealed a significant decrease in myofiber cross-sectional area in both TA and SOL muscles from SKI mice (Fig. 2D and E). Consistent with these structural abnormalities, grip strength was significantly reduced in SKI mice relative to WT controls (Fig. 2F). Together, these findings demonstrate that SERCA2 Cys674 inactivation is sufficient to induce spontaneous skeletal muscle atrophy and functional decline in vivo.

3.3. Transcriptomic profiling of adv and 2as-transduced C2C12 cells

To further investigate the mechanisms by which SERCA2 dysfunction contributes to skeletal muscle atrophy, we performed RNA-seq in C2C12 cells transduced with control adenovirus (Adv) or 2as, a construct overexpressing the Cys674-inactivated form of SERCA2, after confirming transduction efficiency (Fig. S1A).

Transcriptomic analysis identified 270 differentially expressed genes (DEGs), including 243 upregulated and 27 downregulated genes in the 2as group relative to Adv controls (Fig. 3A and B). Functional enrichment analysis revealed that these DEGs were predominantly associated with extracellular matrix remodeling and fibrotic responses. GO analysis highlighted terms related to collagen-containing extracellular matrix, extracellular structure organization, and collagen fibril organization, whereas KEGG analysis identified significant enrichment of the renin-angiotensin system and protein digestion and absorption pathways (Fig. 3C and D). Consistently, Reactome analysis further showed enrichment in collagen formation, collagen degradation, extracellular matrix organization, and angiotensin-related metabolic pathways (Fig. 3E). In parallel, GSEA demonstrated significant activation of inflammatory and immune-related programs, including interferon-α and interferon-γ responses, complement, and coagulation signaling (Fig. 3F).

Fig. 3.

SERCA2 dysfunction induces transcriptomic programs associated with extracellular matrix remodeling, RAS activation, and inflammatory signaling. (A) Volcano plot showing differentially expressed genes (DEGs) in the Adv versus 2as comparison; upregulated and downregulated genes are shown in red and blue, respectively. (B) Heatmap showing hierarchical clustering of DEGs between the Adv and 2as groups. (C) Circular plot summarizing Gene Ontology (GO) enrichment of DEGs across biological process (BP), molecular function (MF), and cellular component (CC) categories. (D) Bar plot showing the top enriched GO terms and KEGG pathways identified from DEG enrichment analysis. (E) Bubble plot showing Reactome pathway enrichment of DEGs; bubble size indicates gene count and color indicates adjusted P value. (F) Ridge plot showing gene set enrichment analysis (GSEA) of Hallmark pathways in the Adv versus 2as comparison.

Together, these data indicate that SERCA2 dysfunction induces a transcriptional program characterized by activation of renin-angiotensin, inflammatory, and extracellular matrix remodeling pathways, supporting a role for these processes in SERCA2 dysfunction-driven muscle atrophy.

3.4. SERCA2 dysfunction activates the renin-angiotensin system and downstream TGF-β/Smad signaling

Based on the RNA-seq results, we next examined whether SERCA2 dysfunction activates profibrotic signaling in vitro and in vivo. In C2C12 cells, overexpression of the Cys674-inactivated SERCA2 mutant (2as) increased Fluo-4 fluorescence, indicating elevated intracellular Ca²⁺ levels, and was accompanied by myotube atrophy, as reflected by reduced myotube diameter (Fig. 4A–C). In parallel, Klf15 expression was markedly decreased, whereas Tgfβ1, Agt, and Agtr1a were significantly increased, supporting activation of the renin-ngiotensin system (RAS) and profibrotic signaling (Fig. 4B–D).

Fig. 4.

SERCA2 dysfunction activates the RAS-TGFβ/Smad fibrotic axis. (A) Representative Fluo-4 staining and quantification of intracellular Ca²⁺ fluorescence intensity in adv and 2as-transduced C2C12 cells. (B) qRT-PCR analysis of Klf15 expression in adv and 2as C2C12 cells(n = 4). (C) Representative MYHC immunofluorescence staining and quantification of myotube diameter in adv and 2as-treated C2C12 cells. Scale bar, 100 μm. (D) qRT-PCR analysis of Tgfβ1, Acta2, Agt, and Agtr1a expression in adv and 2as C2C12 cells(n = 4). (E) Representative Masson's trichrome staining of TA and SOL muscles from WT and SKI mice and quantification of collagen fiber area fraction(n = 3). Scale bar, 50 μm. (F) qRT-PCR analysis of Tgfβ1, Acta2, Agt, and Agtr1a expression in skeletal muscle from WT and SKI mice(n = 5). (G) Representative immunohistochemical staining of αSMA in TA and SOL muscles from WT and SKI mice and quantification of αSMA-positive area(n = 3). Scale bar, 100 μm. (H) Representative western blots and quantitative analysis of SERCA2, AGTR1, Renin, TGFβ1, p-Smad2/3, and αSMA protein expression in adv and 2as-transduced C2C12 cells(n = 3). (I) Representative western blots and quantitative analysis of TGFβ1, p-Smad2/3, αSMA, and AGTR1 protein expression in skeletal muscle from WT and SKI mice(n = 3). (J) Representative western blots and quantitative analysis of TGFβ1, p-Smad2/3, and αSMA protein expression in control and clinical skeletal muscle atrophy specimens(n = 6). Data are presented as mean ± SD. n.s. P > 0.05, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. t-test for A-D, F, H–I. two-way ANOVA followed by Sidak's multiple comparisons test ANOVA for E, G, J.

Consistent with the in vitro findings, skeletal muscle from SKI mice exhibited increased collagen deposition in the TA and SOL, as shown by Masson's trichrome staining (Fig. 4E). qRT-PCR further demonstrated upregulation of Tgfβ1, Acta2 (αSMA), and RAS-related genes in SKI muscle (Fig. 4F). Immunohistochemistry confirmed an increased αSMA-positive area in both TA and SOL (Fig. 4G). At the protein level, western blotting in both 2as-transduced C2C12 cells and SKI muscle showed increased expression of AGTR1, renin, TGFβ1, phosphorylated Smad2/3, and αSMA, indicating activation of the RAS/TGF-β/Smad axis (Fig. 4H and I). Importantly, skeletal muscle samples from patients with atrophy showed similar increases in TGFβ1 and p-Smad2/3, with a trend toward higher αSMA expression (Fig. 4J).

Together, these findings demonstrate that SERCA2 dysfunction promotes RAS activation and downstream TGF-β/Smad signaling, thereby driving fibrosis-associated remodeling in skeletal muscle.

3.5. SERCA2 dysfunction upregulates S100a4 expression

To identify downstream effectors of SERCA2 dysfunction, we performed proteomic profiling of gastrocnemius muscles from 2-month-old WT and SKI mice. A total of 31 differentially expressed proteins were identified, including 14 upregulated and 17 downregulated proteins in SKI muscle (Fig. 5A and B). Integration of these proteomic data with the RNA-seq dataset from 2as-transduced C2C12 cells identified 29 overlapping candidates, among which S100a4 emerged as a prominent shared target (Fig. 5C).

Fig. 5.

Multi-omics analysis identifies S100a4 as a downstream target upregulated by SERCA2 dysfunction. (A) Volcano plot showing differentially expressed proteins (DEPs) in gastrocnemius (GAS) muscle from WT and SKI mice. (B) Heatmap showing hierarchical clustering of DEPs between WT and SKI mice. (C) Venn diagram showing the overlap between differentially expressed genes (DEGs) from adv versus 2as transcriptomic analysis and significant DEPs from WT versus SKI proteomic analysis, with the overlapping candidate molecules listed on the right. (D) qRT-PCR analysis of S100a4 mRNA expression in GAS muscle from WT and SKI mice and in adv and 2as-treated C2C12 cells, together with quantitative analysis of S100a4 protein expression(n = 3-4). (E, F) Representative western blots of S100a4 in GAS muscle from WT and SKI mice and in adv and 2as-treated C2C12 cells, respectively. The corresponding bar plots show S100a4 abundance identified by proteomic analysis, as well as Western blot quantification of S100a4 expression in skeletal muscle tissue and C2C12 cells (n = 3-5). (G) Representative Western blot and quantitative analysis of SERCA2 overexpression efficiency in adv and 2as-treated C2C12 cells(n = 5). (H) Representative immunofluorescence staining of SERCA2 and S100a4 in adv and 2as-treated C2C12 cells and quantification of S100a4 fluorescence intensity. Scale bar, 10 μm. n.s. P > 0.05, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Data are shown as the mean ± SD. t-test for D-H.

We therefore examined S100a4 expression in vivo and in vitro. In SKI muscle, S100a4 was significantly increased at both the mRNA and protein levels compared with WT controls (Fig. 5D and E). Consistently, 2as transduction in C2C12 cells markedly elevated S100a4 expression at both the transcript and protein levels (Fig. 5D–F). Immunofluorescence further confirmed enhanced S100a4 signal intensity in 2as-transduced cells (Fig. 5H). Notably, Western blot analysis verified increased SERCA2 expression in the 2as group, consistent with successful overexpression of the mutant construct (Fig. 5G). Together, these results identify S100a4 as a downstream effector associated with SERCA2 dysfunction.

To further determine the association between SERCA2 C674 oxidation and S100a4 expression, we performed correlation analyses using matched measurements from the same samples. In clinical skeletal muscle samples, S100A4 protein expression was significantly increased in the long-term disuse atrophy group compared with the control group (Fig. S2E). However, Spearman correlation analysis showed no statistically significant correlation between S100A4 protein expression and SERCA2 C674–SO₃H levels in clinical samples (Fig. S2F; ρ = 0.406, P = 0.191). In mouse gastrocnemius muscle, immunohistochemical analysis showed that S100a4-positive area was markedly increased after hindlimb suspension (Fig. S2G). Moreover, S100a4-positive area was significantly and positively correlated with SERCA2 C674–SO₃H-positive area in mouse gastrocnemius muscle sections (Fig. S2H; ρ = 0.783, P = 0.003). These results suggest that S100a4 upregulation is associated with SERCA2 C674 oxidation in the murine disuse muscle atrophy model.

3.6. S100a4 upregulation in SKI skeletal muscle shows a mixed cellular distribution

Because S100a4 can be expressed by multiple cell populations in skeletal muscle, we next performed double immunofluorescence staining to determine the cellular distribution of S100a4 immunoreactivity in SKI muscle. S100a4 was co-stained with Desmin, F4/80, or Vimentin to identify myofiber-associated regions, macrophage-enriched areas, and fibroblast-like regions, respectively. Compared with WT muscle, SKI muscle showed a marked increase in S100a4-positive area across all three staining conditions (Fig. 6A–D). S100a4 signals were detectable in Desmin-positive myofiber regions, indicating that part of the increased S100a4 immunoreactivity was associated with myofibers (Fig. 6A). In addition, S100a4 signals were also observed in F4/80-positive macrophage-enriched regions and Vimentin-positive fibroblast-like areas (Fig. 6B and C).

Fig. 6.

S100a4 upregulation in SKI skeletal muscle shows a mixed cellular distribution. (A-C) Representative double immunofluorescence staining of S100a4 with lineage-associated markers in skeletal muscle sections from WT and SKI mice. S100a4 was co-stained with Desmin, a myofiber marker (A), F4/80, a macrophage marker (B), or Vimentin, a fibroblast-like cell marker (C). Enlarged regions and separated fluorescence channels are shown to illustrate the spatial distribution of S100a4 signals relative to each lineage-associated marker. Nuclei were counterstained with DAPI. Scale bars, 50 μm. (D) Quantification of S100a4-positive area in WT and SKI skeletal muscle. Data were normalized to the corresponding WT group within each staining condition. (E) Quantification of S100a4⁺ lineage-marker⁺ area in SKI skeletal muscle, showing the relative extent of S100a4 signal overlapping with Desmin, F4/80, or Vimentin-positive regions. n.s. P > 0.05, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Data are presented as mean ± SD. For grouped comparisons in (D), two-way ANOVA followed by Sidak's multiple comparisons test was used. For comparisons among lineage-marker-associated S100a4-positive areas in (E), one-way ANOVA followed by Tukey's multiple comparisons test was used.

Quantitative analysis further showed that S100a4-positive area was significantly increased in SKI muscle compared with WT muscle when normalized to the corresponding WT group within each staining condition (Fig. 6D). In SKI muscle, analysis of S100a4⁺ lineage-marker⁺ area showed detectable overlap of S100a4 signal with Desmin, F4/80, and Vimentin-positive regions (Fig. 6E). These results indicate that S100a4 upregulation in SKI skeletal muscle exhibits a mixed cellular distribution, involving myofiber-associated regions as well as macrophage-enriched and fibroblast-like compartments.

3.7. S100a4 induction is dependent on TGF-β/Smad signaling and associated with Smad3

Across both basal C2C12 conditions and 2as adenovirus-mediated SERCA2 dysfunction, S100a4 expression and its induction depend on TGFβ/Smad signaling, and S100a4 physically associates with Smad3. In parental C2C12 cells, exogenous TGFβ1 stimulation increased S100a4 expression in a dose-dependent manner, with S100a4 transcript and protein levels rising significantly as the TGFβ1 concentration increased (Fig. S1G and H). We first assessed the knockdown efficiency of the Smad3-targeting siRNAs (Fig. S1E and F). Among the sequences tested, si3 achieved a significant reduction in Smad3 protein levels and was therefore used for Smad3 silencing in all subsequent experiments. Smad3 knockdown significantly reduced basal S100a4 transcription (Fig. S1I) and markedly attenuated the TGFβ1-induced upregulation of S100a4 mRNA and protein; consistently, IF showed reduced S100a4-positive signal and decreased fluorescence intensity under Smad3 knockdown (Fig. 7A–C). To facilitate subsequent S100a4 knockdown under conditions of C674 site inactivation, we generated and validated a 2as overexpression stable cell line (Fig. S1D). In these 2as stable cells, S100a4 was significantly increased at both the transcript and protein levels, whereas Smad3 knockdown similarly reversed the 2as-induced S100a4 upregulation and reduced the S100a4 IF signal (Fig. 7D–F). These observations prompted the hypothesis that S100a4 and Smad3 may directly interact.

Fig. 7.

S100a4 is regulated by TGFβ/Smad signaling and interacts with Smad3 under conditions of SERCA2 dysfunction. (A, B) qRT-PCR and immunofluorescence analyses of S100a4 expression in C2C12 cells transfected with control siRNA or Smad3 siRNA, with or without TGFβ1 stimulation(n = 3-5). (C) Representative western blots and quantitative analysis of S100a4 protein expression in control and Smad3-silenced C2C12 cells in the presence or absence of TGFβ1(n = 3). (D, E) qRT-PCR and immunofluorescence analyses of S100a4 expression in LV-vec and LV-2as stable C2C12 cells transfected with control siRNA or Smad3 siRNA(n = 3-5). (F) Representative western blots and quantitative analysis of S100a4 protein expression in LV-vec and LV-2as stable C2C12 cells with or without Smad3 knockdown(n = 3). (G) Protein-protein docking model predicting the interaction between Smad3 and S100a4, with an enlarged view of the putative binding interface and key contact residues. (H, I) Co-immunoprecipitation assays in 2as cells using anti-S100a4 or anti-Smad3 antibodies for immunoprecipitation, followed by immunoblotting for S100a4 and Smad3; IgG served as a negative control. (J) Immunofluorescence co-localization analysis of Smad3 and S100a4 in control C2C12 cells and LV-2as stable cells, with Pearson's correlation coefficients and line-scan intensity profiles shown for quantitative evaluation. n.s. P > 0.05, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Data are shown as the mean ± SD. two-way ANOVA for A-F.

Molecular docking predicted the potential for a stable structural interface between S100a4 and Smad3, and representative high-scoring conformations were visualized (Fig. 7G). This predicted interaction was subsequently supported by IF co-localization and immunoprecipitation assays, which demonstrated an association between S100a4 and Smad3 in 2as stable cells (Fig. 7H–J).

Collectively, these data indicate that, in both basal C2C12 cells and the 2as-mediated SERCA2 dysfunction context, S100a4 expression and its induction are TGFβ/Smad-dependent, supporting S100a4 as a key downstream effector of this pathway.

3.8. Smad3 promotes S100a4 promoter activity and nuclear-associated S100a4 accumulation in C2C12 cells

To further determine whether Smad3 functionally regulates S100a4 transcriptional activity, we performed dual-luciferase reporter assays in C2C12 cells. The efficiency of Smad3 overexpression was first confirmed at both the mRNA and protein levels (Supplementary Fig. S1J and K). A schematic illustration of the reporter assay is shown in Fig. 8A. Smad3 overexpression significantly increased SBE4-luciferase reporter activity, confirming enhanced Smad3-dependent transcriptional activity in C2C12 cells (Fig. 8B). We next examined the effect of Smad3 on S100a4 promoter activity. Compared with the pGL4-Basic control, the S100a4 promoter reporter showed detectable basal activity, which was further markedly increased by Smad3 overexpression (Fig. 8C). Conversely, Smad3 knockdown significantly reduced S100a4 promoter activity (Fig. 8D). These results indicate that Smad3 positively regulates S100a4 promoter activity in C2C12 cells.

Fig. 8.

Smad3 promotes S100a4 promoter activity and nuclear-associated S100a4 accumulation in C2C12 cells. (A) Schematic illustration of the dual-luciferase reporter assay used to evaluate the regulatory effect of Smad3 on S100a4 promoter activity. Activation of Smad3 signaling was expected to increase reporter gene transcription, whereas Smad3 inhibition was expected to reduce reporter activity. (B) SBE4-luciferase reporter assay showing increased Smad3 transcriptional activity in C2C12 cells transfected with Smad3 overexpression plasmid compared with vector control cells. Firefly luciferase activity was normalized to Renilla luciferase activity. (C) Dual-luciferase reporter assay showing that Smad3 overexpression significantly increased S100a4 promoter activity in C2C12 cells. pGL4-basic served as the promoterless negative control. (D) Knockdown of Smad3 significantly reduced S100a4 promoter activity in C2C12 cells. (E) Representative confocal immunofluorescence images of Smad3 and S100a4 in C2C12 cells transfected with vector or Smad3 overexpression plasmid. Nuclei were counterstained with DAPI. Scale bar, 10 μm. (F) Quantification of the nuclear/cytoplasmic fluorescence intensity ratio of S100a4, showing increased nuclear-associated S100a4 signal after Smad3 overexpression. n.s. P > 0.05, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Data are presented as mean ± SD. For two-group comparisons, unpaired two-tailed Student's t-test was used. For multi-group comparisons, one-way ANOVA followed by Tukey's multiple comparisons test was used.

To further validate this regulatory effect in an independent cell system, we also performed dual-luciferase reporter assays in 293T cells. Consistent with the findings in C2C12 cells, Smad3 overexpression markedly enhanced S100a4 promoter activity in 293T cells, whereas Smad3 knockdown reduced S100a4 promoter activity (Supplementary Fig. S2C and D). These results support that Smad3-mediated activation of the S100a4 promoter is reproducible across different cellular contexts.

Consistently, confocal immunofluorescence analysis showed that Smad3 overexpression increased S100a4 fluorescence intensity in C2C12 cells and enhanced the nuclear-associated S100A4 signal (Fig. 8E). Quantitative analysis further showed that the nuclear/cytoplasmic fluorescence intensity ratio of S100A4 was significantly increased after Smad3 overexpression (Fig. 8F). Together, these findings suggest that Smad3 enhances S100a4 transcriptional activity and promotes nuclear-associated accumulation of S100a4 in C2C12 cells.

3.9. S100a4 knockdown attenuates SERCA2 dysfunction-induced inflammatory responses

Given the close link between skeletal muscle atrophy and inflammatory activation, we next investigated whether S100a4 contributes to the inflammatory response induced by SERCA2 dysfunction. In 2as-stable C2C12 cells, S100a4 knockdown significantly reduced the mRNA expression of Il-6, Tnfα, and Il-1β, with a similar downward trend observed for Cxcl2 (Fig. 9A). Consistently, Western blot analysis showed that the elevated levels of COX-2, NLRP3, and IL-1β induced by SERCA2 dysfunction were attenuated after S100a4 silencing (Fig. 9B).

Fig. 9.

S100a4 knockdown attenuates inflammatory and oxidative stress responses induced by SERCA2 dysfunction. (A) qRT-PCR analysis of Il-6, Tnfα, Il-1β, and Cxcl2 mRNA expression in LV-vec, LV-2as, LV-2as + siNC, and LV-2as + siS100a4 C2C12 cells(n = 3). (B) Representative western blots and quantitative analysis of COX-2, NLRP3, and IL-1β protein expression in the indicated groups(n = 3). (C) Representative western blots and quantitative analysis of p-p65/p65 expression, together with the nuclear/cytoplasmic distribution of NF-κB p65. (D) Representative immunofluorescence staining of p65 and DAPI in the indicated groups, showing nuclear translocation of p65. Scale bar, 100 μm. (E) Representative DHE staining and quantification of relative fluorescence intensity in the indicated groups. Scale bar, 100 μm. n.s. P > 0.05, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Data are shown as the mean ± SD. two-way ANOVA for A-C, E.

We further examined NF-κB signaling, a central inflammatory pathway. SERCA2 dysfunction increased p65 phosphorylation, whereas S100a4 knockdown reduced the p-p65/p65 ratio (Fig. 9C). Immunofluorescence analysis further showed enhanced p65 activation in 2as cells, which was markedly diminished following S100a4 silencing (Fig. 9D). In parallel, DHE staining revealed that SERCA2 dysfunction markedly increased intracellular oxidative stress, whereas S100a4 knockdown significantly decreased DHE fluorescence intensity (Fig. 9E).

Together, these findings identify S100a4 as an important mediator of the inflammatory and oxidative stress responses triggered by SERCA2 dysfunction.

3.10. Inhibition of the renin-angiotensin system or S100a4 knockdown alleviates spontaneous skeletal muscle atrophy and fibrotic remodeling in SKI mice

To evaluate potential interventions targeting SERCA2 Cys674 inactivation-induced muscle pathology, SKI mice were treated with either losartan or AAV-sh-S100a4 according to the experimental design shown in Fig. 10A. Efficient knockdown of S100a4 in skeletal muscle was confirmed by immunofluorescence, qRT-PCR, and western blotting (Fig. 10C–E, F).

Fig. 10.

Losartan treatment or muscle-targeted S100a4 knockdown ameliorates skeletal muscle atrophy and fibrotic remodeling in SKI mice. (A) Schematic of the in vivo experimental design. (B) Representative gross morphology of tibialis anterior (TA) and gastrocnemius (GAS) muscles from the indicated groups. (C) Representative IF staining of S100a4 in GAS and TA muscles from SKI mice treated with AAV-shNC or AAV-shS100a4. Scale bar, 200 μm; enlarged images of the dashed boxed areas are shown below (scale bar, 50 μm). (D) Quantification of TA/body weight (TA/BW) and GAS/body weight (GAS/BW) ratios(n = 6). (E, F) Representative Western blot and qRT-PCR/Western blot quantification of S100a4 knockdown efficiency in skeletal muscle from AAV-shNC and AAV-shS100a4 mice(n = 6). (G, H) Representative H&E and Masson's trichrome staining of TA muscles from the indicated groups. (I) Quantification of grip strength, myofiber cross-sectional area (CSA), and collagen-positive area in TA muscles(n = 6). n.s. P > 0.05, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Data are shown as the mean ± SD. t-test for E, F. two-way ANOVA for D, I.

Compared with WT littermates, SKI mice exhibited marked spontaneous muscle atrophy, as evidenced by reduced TA and GAS size, decreased relative muscle mass, impaired grip strength, and smaller myofiber cross-sectional area (Fig. 10B–D, G, I). Masson's trichrome staining further revealed pronounced interstitial collagen accumulation, indicating fibrosis-like remodeling (Fig. 10H and I).

Both losartan treatment and S100a4 knockdown substantially ameliorated these pathological changes. Specifically, both interventions improved gross muscle morphology and relative muscle mass, increased grip strength and myofiber CSA, and markedly reduced collagen deposition in skeletal muscle (Fig. 10B–D, H, I). These findings indicate that pharmacological inhibition of the renin–angiotensin system or targeted suppression of S100a4 effectively attenuates skeletal muscle atrophy and fibrotic remodeling driven by SERCA2 Cys674 functional impairment.

3.11. Inhibition of the renin-angiotensin system or S100a4 knockdown alleviates HLS-induced muscle atrophy, fibrosis, and inflammatory proteolytic signaling

We next examined whether inhibition of the renin-angiotensin system (RAS) or S100a4 silencing could mitigate disuse-induced muscle atrophy in the HLS model (Fig. 11A). Compared with weight-bearing controls, HLS mice exhibited marked reductions in TA and GAS size, hanging time, relative muscle mass, and myofiber cross-sectional area, together with increased interstitial collagen deposition, confirming severe muscle atrophy and fibrosis-like remodeling (Fig. 11B–F).

Fig. 11.

Losartan or S100a4 silencing alleviates disuse-induced muscle atrophy and inflammatory remodeling in HLS mice. (A) Schematic of the in vivo experimental design. C57BL/6J mice were subjected to hindlimb suspension (HLS) and treated with losartan or intramuscular AAV-mediated S100a4 knockdown, followed by functional testing and tissue collection. (B) Representative gross morphology of tibialis anterior (TA) and gastrocnemius (GAS) muscles from the indicated groups. (C) Quantification of muscle function by the grip force/hanging test(n = 6). (D) Quantification of TA/body weight (TA/BW) and GAS/body weight (GAS/BW) ratios, myofiber cross-sectional area (CSA), and collagen-positive area in TA muscles(n = 6). (E, F) Representative H&E and Masson's trichrome staining of TA muscles from the indicated groups(n = 6). Scale bar, 50 μm. (G) Representative immunohistochemical staining and quantification of TNFα, FBX32, and TRIM63 expression in skeletal muscle from the indicated groups(n = 6). Scale bar, 50 μm. n.s. P > 0.05, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Data are shown as the mean ± SD. two-way ANOVA for C-G.

Both losartan treatment and AAV-sh-S100a4 improved the HLS-induced phenotype. Losartan broadly ameliorated muscle weakness, loss of muscle mass, and histopathological injury, whereas S100a4 knockdown significantly improved grip strength and TA-related morphological indices and exerted a pronounced anti-fibrotic effect, although its effect on GAS/BW did not reach significance (Fig. 11C–F). At the molecular level, immunohistochemistry showed that HLS markedly increased TNFα, FBXO32, and TRIM63 expression in skeletal muscle, indicating activation of inflammatory and proteolytic pathways (Fig. 11G). Both losartan and S100a4 silencing significantly suppressed these changes.

Collectively, these findings demonstrate that RAS inhibition or S100a4 knockdown attenuates disuse-induced skeletal muscle atrophy and fibrosis, at least in part by limiting local inflammation and activation of protein degradation pathways.

4. Discussion

Our study elucidates a pathway wherein SERCA2 dysfunction drives skeletal muscle atrophy via a RAS-TGFβ/Smad-S100a4 axis. In healthy muscle, SERCA2 pumps cytosolic Ca²⁺ back into the sarcoplasmic reticulum to maintain calcium homeostasis during excitation-contraction coupling [18,[41], [42], [43]]. Prior work shows that impaired SERCA activity leads to cytosolic Ca²⁺ accumulation, activating proteases, apoptosis, and other catabolic pathways that damage myofibers and promote atrophy [[44], [45], [46]]. Of particular importance, the Cys674 residue in the SERCA2a isoform is a redox-sensitive site: excessive ROS can irreversibly oxidize Cys674 and inactivate the pump [47,48]. Since ROS levels rise markedly in disuse atrophy [49,50], SERCA2 impairment likely links oxidative stress to muscle wasting. In this context, antioxidant administration may theoretically influence SERCA2 C674–SO₃H status by reducing the oxidative burden in disused skeletal muscle. By lowering ROS and reactive nitrogen species, antioxidant strategies may limit further oxidation of the redox-sensitive Cys674 residue and help preserve SERCA2 activity. However, C674–SO₃H represents an overoxidized and largely irreversible modification of cysteine, and conventional antioxidants are unlikely to directly reduce or reverse SERCA2 C674–SO₃H once it has formed. Therefore, the potential benefit of antioxidant treatment would be expected to occur mainly through prevention or attenuation of ongoing Cys674 overoxidation rather than direct reversal of established SERCA2 C674–SO₃H. Future studies using antioxidant interventions at different stages of disuse atrophy will be required to determine whether reducing oxidative stress can prevent SERCA2 C674–SO₃H accumulation and thereby ameliorate downstream Ca²⁺ dysregulation and muscle wasting.

We found that in a disuse atrophy model, total SERCA2 protein in muscle was unchanged, but the oxidized, inactive form (Cys674-SO₃H) was significantly elevated. This suggests that SERCA2-related pathology in atrophy arises not from expression change but from oxidation at the C674 site and consequent pump dysfunction. Consistent with this, SERCA2 C674S knock-in (SKI) mice-which mimic permanent C674 oxidation-developed spontaneous muscle weakness and atrophy (reduced fiber cross-sectional area and hanging endurance) without unloading. Thus, Cys674 inactivation alone suffices to induce muscle atrophy, indicating SERCA2 dysfunction is a primary upstream event, not merely a bystander of disuse. This aligns with reports that boosting SERCA activity can partially counteract unloading-induced weakness and atrophy [51,52], underscoring the importance of preserving SERCA2 function for muscle health.

We then investigated downstream pathways. Transcriptomics revealed that SERCA2 C674S triggers RAS system upregulation and strongly activates TGFβ/Smad signaling. Excess RAS signaling is well known to drive inflammation and fibrosis in heart and kidney [53,54], and evidence links RAS hyperactivation to muscle wasting in aging [[55], [56], [57]]. TGFβ/Smad is also a key pro-atrophy axis: activated Smad3 induces expression of atrophy genes (Atrogin-1, MuRF1) [58,59], inhibits protein synthesis, and promotes fibrosis, whereas inhibiting this pathway can slow disuse atrophy [60,61]. Indeed, SKI mice showed marked increases in muscle TGFβ1 and p-Smad2/3, paralleled in human atrophic muscle, suggesting SERCA2 dysfunction promotes atrophy via RAS-TGFβ/Smad activation. This connects oxidative Ca²⁺ dysregulation to classic inflammatory、fibrotic signals in atrophy.

Proteomic analysis and experiments identified S100a4 as a critical effector of this axis. S100a4, a calcium-binding protein, is known to amplify inflammation-fibrosis loops: it regulates cytoskeleton and, when secreted, activates NF-κB signaling to boost TNFα, IL-1β and other cytokines [[62], [63], [64], [65], [66]]. We observed that SERCA2 dysfunction upregulates S100a4 in muscle cells and SKI mice. Exogenous TGFβ1 dose-dependently increased S100a4, while Smad3 knockdown reduced it, implying S100a4 is downstream of TGFβ/Smad. Moreover, molecular docking, co-IP, and colocalization showed S100a4 interacts with Smad3, suggesting it may amplify this pathway. Functionally, S100a4 knockdown in vitro blunted the rise of COX-2, IL-1β, TNFα and NF-κB activity, and in vivo muscle-specific S100a4 silencing (AAV9) in HLS mice reduced these cytokines, collagen deposition, and MuRF1/Atrogin-1 expression, while improving fiber size and strength. Therefore, S100a4 serves as a key link between SERCA2 dysfunction and inflammation amplification in muscle atrophy, acting to enhance both inflammatory and fibrotic responses.

Finally, we tested interventions along this axis. Blocking RAS with losartan or knocking down S100a4 each significantly ameliorated muscle atrophy in SKI and HLS models: muscle mass and fiber area were restored, collagen deposition decreased, and hanging time improved. That both upstream (RAS blockade) and downstream (S100a4 silencing) interventions were effective confirms the RAS-TGFβ/Smad-S100a4 axis as central to disuse atrophy, and suggests multiple potential therapeutic targets within this pathway. Although we observed increased expression of atrophy-related ubiquitin ligases, including FBXO32 and TRIM63, actual muscle protein breakdown rates were not directly measured in this study. Therefore, our findings support the activation of atrophy-related proteolytic signaling rather than directly demonstrating increased protein degradation kinetics. Future studies using protein turnover assays, such as amino acid release assays, stable isotope tracing, or protein half-life measurements, will be required to determine whether SERCA2 C674 inactivation directly accelerates muscle protein breakdown.

Several limitations should be acknowledged. First, the clinical component of this study was based on a relatively small cohort of hip arthroplasty-derived muscle specimens, and inter-individual heterogeneity in age, immobility duration, comorbidities, disease background, and medication history may have contributed to the modest effect size and variability in SERCA2 C674–SO₃H immunoreactivity. Therefore, the human findings should be interpreted as supportive translational observations rather than definitive evidence of clinical causality or biomarker-level significance. Second, although SERCA2 C674–SO₃H immunoreactivity was increased in human and mouse atrophic muscle, direct site-specific biochemical confirmation by LC-MS/MS or differential alkylation was not performed in the present study. Third, the 2as cell model represents overexpression of C674S-mutant SERCA2a and cannot fully separate the effects of C674 substitution from those of exogenous SERCA2a expression. Finally, direct SERCA pump activity and muscle protein breakdown rates were not measured, and future studies will be required to further define the functional consequences of SERCA2 C674 modification during disuse-induced skeletal muscle atrophy.

Although SKI mice exhibited reduced skeletal muscle mass without a significant change in total body weight, body composition was not directly assessed in the present study. Therefore, we cannot exclude the possibility that changes in fat mass or other non-muscle compartments partially compensated for the loss of skeletal muscle mass. Future studies using DEXA, MRI, or adipose tissue depot measurements will be required to determine whether SERCA2 C674 functional inactivation alters whole-body composition in addition to skeletal muscle mass.

We discovered that oxidative inactivation of SERCA2 at a specific cysteine (Cys674) triggers a cascade involving the renin-angiotensin system and TGFβ/Smad signaling, leading to upregulation of the inflammatory protein S100a4. This cascade amplifies inflammation and muscle proteolytic signaling, driving skeletal muscle wasting. Importantly, drugs blocking RAS or lowering S100a4 levels were able to significantly reduce muscle loss in mice.

5. Conclusion

In summary, our study identifies oxidative inactivation of SERCA2 at Cys674 as a mechanistic link between redox stress and skeletal muscle atrophy. SERCA2 dysfunction drives inflammatory and atrophic remodeling through activation of the RAS-TGF-β/Smad-S100A4 axis, whereas losartan treatment or S100A4 silencing confers significant protection both in vitro and in vivo, as shown in Fig. 12. These findings establish S100a4 as a key downstream effector of redox-sensitive SERCA2 signaling and a potential therapeutic target for disuse-induced skeletal muscle atrophy.

Fig. 12.

Proposed mechanism by which the SERCA2-TGFβ/Smad-S100a4 axis drives skeletal muscle atrophy. Disuse-associated oxidative stress impairs SERCA2 function through Cys674 oxidation, leading to Ca²⁺ overload and activation of RAS/TGFβ/Smad signaling. This induces S100a4 expression, promotes inflammatory cytokine production and fibrotic remodeling, and ultimately drives skeletal muscle atrophy. Losartan or S100a4 silencing interrupts this pathogenic cascade and attenuates muscle wasting.

Data availability

Data will be made available on reasonable request.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

During the preparation of this manuscript, the authors used ChatGPT (OpenAI) for translation and language polishing. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.

Funding

This work was supported by the Natural Science Foundation of Heilongjiang Province (Grant No. PL2024H045) and the Medical Scientific Research Development Fund Project (Grant No. YXKY-HX2002). This study was also supported by National Natural Science Foundation of China [82470473 to X.T.].

CRediT authorship contribution statement

Fei Nan: Data curation, Project administration, Software, Visualization, Writing – original draft, Writing – review & editing. Siyao Liu: Data curation, Software. Shunyi Lei: Software, Visualization. Yu Peng: Formal analysis, Software. Hailong Zhang: Data curation, Software. Xun Chen: Data curation, Software. Shixin Jin: Methodology, Software. Yuanfu Mao: Formal analysis, Software. Dan Lin: Data curation, Software. Xiaoyong Tong: Conceptualization, Funding acquisition, Supervision, Writing – review & editing. Yanlong Qu: Conceptualization, Funding acquisition, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Abbreviations

AAV: adeno-associated virus; ATP: adenosine triphosphate; Atrogin-1: muscle atrophy F-Box protein; 2as: SERCA2a-C674S lentiviral construct encoding mouse Atp2a2 isoform 2a with cysteine 674 substituted by serine; BP: biological process; C674: cysteine 674; CC: cellular component; Fbx32: F-box protein 32; GAS: gastrocnemius muscle; GSEA: Gene set enrichment analysis; GO: gene ontology; HLS: hindlimb suspension; IL-1β: interleukin-1β; IL-6: interleukin-6; KEGG: Kyoto Encyclopedia of Genes and Genomes; LV-vec: lentiviral vector; LV-2as: lentiviral SERCA2a C674–SO₃H; MuRF1: muscle RING finger 1; MIP2: macrophage inflammatory protein-2; MF: molecular function; NF-κB: nuclear factor kappa-B; qRT-PCR: Quantitative real-time PCR; RAS: Renin-angiotensin system; ROS: reactive oxygen species; SERCA: sarcoplasmic/endoplasmic reticulum calcium ATPase; SKI: SERCA2 C674S mutant knock-in; S674: serine 674; SOL: soleus muscle; siRNA: small interfering RNA; shRNA: short hairpin RNA; TA: Tibialis anterior.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Supplementary Fig 1.

Experimental validation of SERCA2 C674S overexpression and TGFβ/Smad3/S100A4 signaling modulation. (A) Representative western blots and quantitative analysis of SERCA2 expression in adv and 2as-transduced C2C12 cells. (B, C) qRT-PCR and western blot analyses of S100a4 expression after transfection with three independent siRNAs, identifying the most effective sequence for subsequent experiments. (D) Representative western blots and quantitative analysis confirming SERCA2 overexpression in LV-vec and LV-2as stable C2C12 cells. (E, F) qRT-PCR and western blot analyses of Smad3 expression after transfection with the most effective Smad3 siRNA for subsequent experiments. (G, H) Representative western blots and qRT-PCR analyses showing that recombinant TGFβ1 increases S100a4 expression in a dose-dependent manner. (I) qRT-PCR analysis showing that Smad3 knockdown reduces basal S100a4 expression in C2C12 cells. (J) Representative western blot image and quantitative analysis showing increased Smad3 protein expression in C2C12 cells transfected with the Smad3 overexpression plasmid compared with vector control cells. GAPDH was used as the loading control. (n = 3). (K) qRT-PCR analysis showing increased Smad3 mRNA expression in Smad3-overexpressing C2C12 cells compared with vector control cells. (n = 3). n.s. P > 0.05, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Data are shown as the mean ± SD.t-test for A, I. one-way ANOVA for B–H.

Supplementary Fig 2.

Additional clinical and experimental evidence supporting Smad3-mediated S100A4 upregulation and its association with SERCA2 C674 oxidation. (A) Comparison between the affected and contralateral limbs; (B) Reduced lower limb circumference on the affected side compared with the contralateral side. (C) Dual-luciferase reporter assay in 293T cells showing that Smad3 overexpression significantly increased S100a4 promoter activity. pGL4-basic served as the promoterless negative control, and Firefly luciferase activity was normalized to Renilla luciferase activity. (D) Knockdown of Smad3 significantly reduced S100a4 promoter activity in 293T cells. (E) Representative western blot images and quantitative analysis of S100A4 protein expression in skeletal muscle samples from control subjects and patients with skeletal muscle atrophy (SMA). GAPDH was used as the loading control(n = 6). (F) Spearman correlation analysis between S100A4 protein expression and SERCA2 C674–SO₃H levels in clinical skeletal muscle samples. Although both proteins were increased in the SMA group, no statistically significant correlation was observed between S100A4 and SERCA2 C674–SO₃H expression in clinical samples (Spearman ρ = 0.406, P = 0.191). (G) Representative immunohistochemical staining images and quantitative analysis of S100a4-positive area in gastrocnemius (GAS) muscle sections from control and hindlimb suspension (HLS) mice(n = 6). Scale bar, 50 μm. (H) Spearman correlation analysis between S100a4-positive area and oxidized SERCA2 C674–SO₃H-positive area in mouse GAS muscle sections. A significant positive correlation was observed between S100a4 expression and oxidized SERCA2 C674–SO₃H levels in mouse skeletal muscle tissues (Spearman ρ = 0.783, P = 0.003). n.s. P > 0.05, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. Data are presented as mean ± SD. t-test for E, G. For multi-group comparisons, one-way ANOVA followed by Tukey's multiple comparisons test was used. For two-group comparisons, unpaired two-tailed Student's t-test was used.