The cGAS–STING pathway contributes to cisplatin-induced skeletal muscle atrophy through altered proteostasis and myogenic signaling

Abstract

Background

Cisplatin chemotherapy is widely used for cancer treatment but frequently induces skeletal muscle atrophy, which compromises physical function and patient outcomes. The molecular mechanisms underlying this process remain incompletely understood. The cyclic GMP–AMP synthase–stimulator of interferon genes (cGAS–STING) signaling pathway, classically involved in innate immune responses, has recently been implicated in cellular stress and tissue dysfunction. Whether cGAS–STING signaling contributes to cisplatin-induced skeletal muscle atrophy remains unclear.

Methods

We employed both pharmacological and genetic approaches. Wild-type (WT) mice received a single intraperitoneal injection of the STING agonist DMXAA prior to cisplatin administration. Genetic models included global cGAS and STING knockout mice, as well as skeletal muscle–specific cGAS knockout mice. Cisplatin was administered intraperitoneally (3 mg/kg/day) for four consecutive days. Body weight, skeletal muscle mass, myofiber cross-sectional area (CSA), and fiber diameter were assessed. Molecular and transcriptional analyses were performed using Western blotting, quantitative polymerase chain reaction, and RNA sequencing.

Results

Pretreatment with the STING agonist DMXAA exacerbated cisplatin-induced body weight loss and skeletal muscle atrophy. In contrast, genetic deletion of cGAS or STING attenuated the loss of gastrocnemius and tibialis anterior muscle mass. Skeletal muscle–specific cGAS deficiency preserved muscle weight and myofiber diameter following cisplatin exposure. Although CSA was also assessed, no significant difference was observed between groups. Transcriptomic analysis identified 696 differentially expressed genes upon cGAS deletion, with enrichment in pathways related to inflammatory signaling, proteasome function, and autophagy. Further analyses in skeletal muscle–specific cGAS-deficient mice showed reduced expression of muscle atrophy–associated genes (FBXO32 and Murf1), together with preservation of key myogenic regulators after cisplatin treatment. Consistently, NF-κB signaling and interferon-stimulated gene expression were diminished, accompanied by altered Beclin1 responses and partial attenuation of selected autophagy-related genes.

Conclusions

These findings support a role for cGAS–STING signaling in cisplatin-induced skeletal muscle atrophy, associated with enhanced innate immune and inflammatory signaling, proteolytic and autophagy-related alterations, and impaired myogenic regulation. Targeting the cGAS–STING pathway may represent a potential therapeutic strategy to mitigate chemotherapy-associated skeletal muscle atrophy.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12964-026-02989-8.

Background

Skeletal muscle, one of the largest and most vital organs in humans and animals, plays a crucial role in locomotion, metabolism, and overall homeostasis [1, 2]. Skeletal muscle atrophy, characterized by a progressive loss of muscle mass and function, represents a debilitating complication frequently associated with chronic diseases, aging, and chemotherapeutic interventions [3, 4].

Among anticancer agents, cisplatin remains a cornerstone of chemotherapy for various malignancies [5, 6]. However, its clinical application is often limited by severe systemic toxicities, including pronounced skeletal muscle wasting, which contributes to cancer cachexia and significantly impairs patient quality of life [7, 8]. Consistent with previous findings, our prior research also demonstrated that cisplatin treatment induces marked skeletal muscle atrophy in mice [9]. Despite these observations, the molecular mechanisms underlying cisplatin-induced muscle atrophy remain incompletely understood, posing significant challenges for the development of targeted interventions to preserve muscle homeostasis during chemotherapy [10].

Maintenance of skeletal muscle mass is tightly regulated by the dynamic balance between protein synthesis and degradation, primarily orchestrated through the ubiquitin–proteasome system (UPS) and autophagy–lysosomal pathways [11, 12]. Dysregulation of proteostasis, indicated by upregulation of muscle-specific E3 ligases including FBXO32 (Atrogin-1/MAFbx) and MuRF1 (TRIM63), as well as alterations in autophagy-associated markers, accelerates protein degradation processes and contributes to skeletal muscle atrophy [13–15].

In addition to proteostasis disruption, emerging studies have highlighted the pivotal role of inflammation in exacerbating skeletal muscle atrophy [16, 17]. Among the mediators of cellular stress and inflammatory responses, the cyclic GMP–AMP synthase (cGAS) and its downstream effector, stimulator of interferon genes (STING), have garnered considerable attention. Originally characterized as cytosolic DNA sensors orchestrating innate immune responses via type I interferon production, the cGAS–STING pathway has recently been implicated in regulating broader aspects of cellular homeostasis, including metabolism, autophagy, and tissue repair processes [18–21]. However, the potential involvement of cGAS–STING signaling in skeletal muscle during chemotherapeutic stress remains poorly defined.

Accumulating evidence suggests that cisplatin-induced cellular stress can engage cytosolic DNA sensing pathways, including cGAS–STING signaling, in several tissues. Cisplatin has been shown to promote mitochondrial dysfunction and mitochondrial DNA (mtDNA) release into the cytosol, a well-recognized trigger of cGAS activation, thereby initiating the mtDNA–cGAS–STING signaling cascade and downstream inflammatory responses in models of cisplatin-induced organ injury [22]. Consistent with this mechanism, activation of STING-associated signaling has been reported in cisplatin-induced kidney injury [23] and cardiotoxicity [24]. These observations suggest that cytosolic DNA sensing pathways may contribute to the inflammatory responses associated with cisplatin toxicity and raise the possibility that similar mechanisms may operate in skeletal muscle during cisplatin treatment.

However, whether molecular signatures consistent with cGAS–STING pathway engagement are present in skeletal muscle during chemotherapeutic stress remains unclear. In particular, it remains unknown whether this pathway contributes to proteolysis, inflammatory signaling, or impaired myogenic regulation during cisplatin-induced muscle atrophy.

In the present study, we aimed to elucidate the role of cGAS–STING signaling in cisplatin-induced skeletal muscle atrophy. By utilizing both global and skeletal muscle-specific knockout mouse models, we systematically dissected the contribution of cGAS and STING to muscle mass regulation under chemotherapeutic challenge. Our findings reveal molecular signatures consistent with cGAS–STING pathway involvement and demonstrate that genetic disruption of this pathway mitigates cisplatin-induced skeletal muscle wasting, providing new insights into the mechanisms underlying chemotherapy-associated muscle atrophy and identifying the cGAS–STING axis as a potential therapeutic target.

Methods

Animal models and experimental design

All animal experiments were approved by the Animal Experimental Ethics Inspection of Guangzhou Sport University (No. 2022 DWLL-18) and were conducted in accordance with US National Institutes of Health (NIH Publication No. 85–23, revised 1996) guidelines. o

Male C57BL/6 J mice (8 weeks old) were used to establish a cisplatin-induced muscle atrophy model. To elucidate the role of cGAS/STING signaling in this process, the following genetically modified strains were employed: cGAS knockout (cGAS KO) mice, STING knockout (STING KO) mice and skeletal muscle-specific cGAS KO mice. cGAS knockout mice (B6(C)-Cgas^{tm1d(EUCOMM)Hmgu}/J, JAX #026554) [25] and STING knockout mice (B6(Cg)-Tmem173^tm1.2Camb/J, JAX #025805) [26] were purchased from The Jackson Laboratory (Bar Harbor, ME, USA). cGAS Flox mice (Cgas^flox/flox, Strain No. S-CKO-05865) and Myf5-Cre mice (Myf5Cre, Strain No. C001451) were purchased from Cyagen Biosciences (Suzhou, China). Skeletal muscle-specific cGAS knockout mice were generated by crossing Cgas^flox/flox mice with Myf5-Cre mice. Genotyping of cGAS knockout, STING knockout, and cGAS mKO mice was performed by PCR analysis and is shown in Supplementary Fig. S1. To further assess Cre-mediated recombination of the floxed Cgas allele in skeletal muscle, genomic qPCR analysis was conducted using skeletal muscle DNA isolated from Flox and cGAS mKO mice. The relative abundance of the floxed Cgas genomic region is presented in Supplementary Fig. S2.

5,6-Dimethylxanthenone-4-acetic acid (DMXAA) is a STING agonist that is commonly used in experimental studies [27, 28]. Mice in the DMXAA-treated group received a single intraperitoneal injection of DMXAA (10 mg/kg; MedChemExpress) prior to cisplatin administration, as previously described [29].

All mice were assigned into control and treatment groups. Cisplatin was administered intraperitoneally at 3 mg/kg once daily for four consecutive days [30]. Control mice received equal volumes of saline. Body weight was recorded before and after cisplatin treatment. Mice were euthanized 1 day after the final injection. The tibialis anterior (TA), soleus (SO), and gastrocnemius (GC) muscles from both hindlimbs were carefully dissected, weighed, and recorded. For each muscle, the weights from both hindlimbs (left and right) were summed and used for muscle mass analysis, and subsequently normalized to tibial length. The harvested tissues were used for histological, molecular, and biochemical analyses. For molecular, histological, and RNA-seq analyses, a subset of samples was used due to tissue availability and technical considerations, and samples were randomly selected to represent each experimental group. For molecular analyses, muscle samples were rapidly frozen in liquid nitrogen immediately after dissection and stored at − 80 °C until further use. For histological analyses, tissues were fixed in 4% paraformaldehyde prior to processing.

Western blotting

Total protein was extracted from frozen skeletal muscle using RIPA buffer supplemented with protease and phosphatase inhibitors (Servicebio, Wuhan, China). Protein concentrations were determined by BCA assay. Equal amounts of protein were loaded per lane within each SDS–PAGE gel (typically 20–30 µg, depending on the target protein). GAPDH was analyzed in parallel with the corresponding target proteins within the same electrophoresis run, although on separate gels processed under identical experimental conditions. Band intensities were quantified by normalizing target protein signals to GAPDH from the corresponding samples within each experiment. After transfer to PVDF membranes, blots were blocked in 5% non-fat milk and incubated overnight at 4 °C with primary antibodies against FBXO32 (ABclonal, A3193, 1:1000), Beclin1 (ABclonal, A7353, 1:1000), MyoD (ABclonal, A23881, 1:1000), LC3-I/II (Servicebio, GBGB11124, 1:5000), and GAPDH (Servicebio, GB15002, 1:2000). Membranes were then incubated with HRP-conjugated secondary antibodies (Servicebio, Wuhan, China), and signals were detected by enhanced chemiluminescence. Band intensities were quantified using ImageJ. Uncropped immunoblots are provided in the Supplementary File.

Quantitative real-time PCR (qPCR)

Total RNA was extracted from tibialis anterior muscles using FreeZol Reagent (Vazyme, Nanjing, China). One microgram of RNA was reverse transcribed into cDNA using the HiScript II 1 st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China). Quantitative PCR was performed with SYBR Green Master Mix (Vazyme, Nanjing, China) on a standard real-time PCR system. Gapdh was used as the internal control, and relative expression levels were calculated using the 2^−ΔΔCt method. Primer sequences are listed in Table 1.

Table 1.

Primers used for RT-PCR

Target gene	Primer sequences
GAPDH	5’-GAAAGCTGTGGCGTGATGGC-3’
	5’-CACGTCAGATCCACGACGGA-3’
MyoD	5’-GAGCGCATCTCCACAGACAG-3’
	5’-AAATCGCATTGGGGTTTGAG-3’
Myogenin	5’-CCAGTACATTGAGCGCCTAC-3’
	5’-ACCGAACTCCAGTGCATTGC-3’
HGF	5’-AGGAACAGGGGCTTTACGTT-3’
	5’-GCTGCCTCCTTTACCAATGA-3’
IGF-1	5’-GCTTGCTCACCTTCACCAG-3’
	5’-CACTCATCCACAATGCCTGT-3’
Myf6	5’-CCTCAGCCTCCAGCAGTCTT-3’
	5’-TTCTCCACCACCTCCTCCAC-3’
Myostatin	5’-TGCAAAATTGGCTCAAACAG-3’
	5’-GCAGTCAAGCCCAAAGTCTC-3’
MuRF1	5’-ACACAACCTCTGCCGGAAGT-3’
	5’-ACGGAAACGACCTCCAGACA-3’
Fbxo32	5’-GCTGTCACTGGCTGGTAGAA-3’
	5’-GCTCCCACCCTAAACAGCAT-3’
Ifit3	5’-AGGACAACCGGAAGTGTGTC-3’
	5’-TTCTCCCCATAAGCAGCACT-3’
IFNβ	5’-GCGTTCCTGCTGTGCTTCTC-3’
	5’-TGTAGGTGAGGTTGATCTTTCCATTC-3’
Isg15	5’-TCCTGGTGTCCGTGACTAACTC-3’
	5’-AAGACCGTCCTGGAGCACTG-3’
Irf3	5’-ATCTGGCTATTGTTTCTGATCCTTCTC-3’
	5’-GGCGGTCACCTCGAACTCC-3’
Irf7	5’-TGAGCGAAGAGAGCGAAGAGG-3’
	5’-CGTACACCTTATGCGGATCAACTG-3’
Map1lc3b	5′-CCACCAAGATCCCAGTGATTAT-3′
	5′-TGATTATCTTGATGAGCTCGCT-3′
Gabarapl1	5′-TCCCTGATCTGGATAAGAGGAA-3′
	5′-AAAGAAGAATAAGGCGTCCTCA-3′
Sqstm1	5′-GAACACAGCAAGCTCATCTTTC-3′
	5′-AAAGTGTCCATGTTTCAGCTTC-3′

Abbreviations: GAPDH Glyceraldehyde-3-phosphate dehydrogenase, MyoD Myogenic differentiation 1, Myogenin Myogenic factor 4, HGF Hepatocyte growth factor, IGF-1 Insulin-like growth factor 1, Myf6 Myogenic factor 6, Myostatin Growth differentiation factor 8, MuRF1 Muscle RING finger protein 1, Fbxo32 F-box protein 32, Ifit3 Interferon-induced protein with tetratricopeptide repeats 3, IFNβ Interferon beta, Isg15 Interferon-stimulated gene 15, Irf3 Interferon regulatory factor 3, Irf7 Interferon regulatory factor 7, Map1lc3b Microtubule-associated protein 1 light chain 3 beta, Gabarapl1 GABA type A receptor-associated protein-like 1, Sqstm1 Sequestosome 1

Immunofluorescence staining

TA muscle tissues were fixed in 4% paraformaldehyde, embedded in paraffin, and sectioned at 4 µm. Sections were deparaffinized with xylene and rehydrated through a graded ethanol series. Antigen retrieval was performed by heating the sections in citrate buffer (pH 6.0). After cooling, sections were permeabilized with 0.3% Triton X-100 and blocked with 5% BSA. Slides were incubated overnight at 4 °C with primary antibody against dystrophin (Abcam, ab15277), followed by incubation with Cy3-conjugated secondary antibody (Servicebio, Wuhan, China). Nuclei were counterstained with DAPI (Servicebio, Wuhan, China). Fluorescence images were captured using a fluorescence microscope, and skeletal muscle fiber dimensions were quantified using Image-Pro Plus 6 (IPP6) software by manually delineating fiber boundaries, following methods previously described in detail [9].

Immunohistochemical staining

TA muscle tissues were fixed in 4% paraformaldehyde. Samples were embedded in paraffin and sectioned at 4 µm. Sections were deparaffinized in xylene and rehydrated through a graded ethanol series. Antigen retrieval was carried out in citrate buffer (pH 6.0) using microwave heating. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide for 10 min. Slides were then incubated with 5% bovine serum albumin (BSA) for 30 min at room temperature. After blocking, sections were incubated overnight at 4 °C with a primary antibody against NF-κB p65 (Servicebio, Wuhan, China, #GB11997-100). After washing, slides were incubated with an HRP-conjugated secondary antibody (Servicebio, Wuhan, China) for 50 min. DAB was used as the chromogenic substrate to visualize antibody binding, and nuclei were counterstained with hematoxylin. All sections were imaged under identical exposure settings using a bright-field microscope. Quantitative image analysis was performed using QuPath software (version 0.6.0, University of Edinburgh, UK). The percentage of NF-κB p65-positive nuclei was calculated as the number of DAB-positive nuclei divided by the total nuclei.

RNA sequencing and bioinformatics analysis

Total RNA was extracted from tibialis anterior muscles using TRIzol reagent (Invitrogen, CA, USA) according to the manufacturer’s protocol. RNA purity and concentration were determined using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific), and RNA integrity was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies, USA). Only samples with RNA integrity number (RIN) ≥ 7.0 were selected for subsequent library construction.

RNA-seq libraries were prepared using the VAHTS Universal V6 RNA-seq Library Prep Kit (Vazyme, China) following the standard protocol, which includes mRNA enrichment, fragmentation, first- and second-strand cDNA synthesis, end repair, A-tailing, adapter ligation, and PCR amplification. Sequencing was performed on an Illumina NovaSeq 6000 platform with 150 bp paired-end reads, yielding an average of > 40 million raw reads per sample.

Raw sequencing data were subjected to quality control using fastp (v0.20.0) to remove adapter sequences, poly-N tails, and low-quality reads. Clean reads were aligned to the mouse reference genome (GRCm38/mm10) using HISAT2 (v2.1.0) with default parameters. Gene expression levels were quantified as fragments per kilobase of transcript per million mapped reads (FPKM) using StringTie (v2.1.4), and raw counts for each gene were calculated using HTSeq-count. Raw read counts were used for differential expression analysis with DESeq2, whereas FPKM values were used for data visualization.

Differential gene expression analysis between groups was performed using DESeq2, and genes with an adjusted p-value (Benjamini–Hochberg false discovery rate [FDR]) < 0.05 and an absolute log₂ fold change > 1 were considered significantly differentially expressed. Over-representation analysis of differentially expressed genes (DEGs) was conducted using the clusterProfiler package in R, and significantly enriched KEGG pathways were identified based on adjusted p-values < 0.05. Gene set enrichment analysis (GSEA) was performed using clusterProfiler based on ranked gene lists generated from DESeq2-derived log₂ fold change values. Enrichment score (ES), normalized enrichment score (NES), nominal p-values, and false discovery rate (FDR) were calculated for each pathway. Heatmaps and volcano plots were generated using the R packages pheatmap and EnhancedVolcano, respectively.

The raw sequence data supporting this study have been deposited in the Genome Sequence Archive (accession number CRA025652) and are publicly available at https://ngdc.cncb.ac.cn/gsa.

Statistical analysis

Data are presented as mean ± SEM, with each data point representing an individual mouse. Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software, San Diego, CA, USA). Normality was assessed using the Shapiro–Wilk test and homogeneity of variance using Levene’s test. For comparisons between two groups, an unpaired two-tailed Student’s t-test was applied. For comparisons involving a single factor with more than two groups, one-way ANOVA followed by Tukey’s or Dunnett’s multiple-comparisons test was used, as appropriate. For experiments involving two independent variables, two-way ANOVA followed by Bonferroni’s or Tukey’s post hoc test was applied, as appropriate. When assumptions of normality or homogeneity of variance were not met, nonparametric tests were used. A p value < 0.05 was considered statistically significant.

Results

STING agonist DMXAA exacerbates cisplatin-induced muscle atrophy

To assess the impact of STING agonist treatment on cisplatin-induced skeletal muscle wasting, mice were treated with DMXAA followed by cisplatin administration, and body weight and skeletal muscle parameters were evaluated (Fig. 1A). Body weight was comparable between the vehicle and DMXAA groups under non-cisplatin-treated conditions. Following cisplatin treatment, mice receiving DMXAA exhibited significantly greater body weight loss compared with vehicle-treated mice (Fig. 1B). Consistent with these findings, the weights of the soleus (SO), gastrocnemius (GC), and tibialis anterior (TA) muscles were significantly lower in the DMXAA-treated group than in vehicle-treated mice after cisplatin administration (Fig. 1C–E). When normalized to tibial length (TL), the relative muscle mass of SO, GC, and TA showed a similar reduction (Fig. 1F–H), indicating that STING agonist DMXAA treatment exacerbates cisplatin-induced skeletal muscle atrophy.

Fig. 1.

STING agonist exacerbates cisplatin-induced skeletal muscle atrophy. A Experimental design for DMXAA treatment combined with cisplatin administration. B Body weight of vehicle- and DMXAA-treated mice under control conditions and after cisplatin treatment. C-E Weights of the soleus (SO), gastrocnemius (GC), and tibialis anterior (TA) muscles following cisplatin administration. F–H Muscle weights normalized to tibial length (TL). I Representative dystrophin immunofluorescence staining of TA muscle sections from vehicle- and DMXAA-treated mice after cisplatin treatment. Scale bar, 50 μm. J-M Quantification of muscle fiber cross-sectional area (CSA), maximum diameter, minimum diameter, and mean diameter. Data are presented as mean ± SEM. Each dot represents an individual mouse (n = 3–12 per group). *p < 0.05 compared to Vehicle mice. **p < 0.01 compared to Vehicle mice

Morphological analyses using dystrophin immunofluorescence staining of TA muscles further supported these observations. Compared with vehicle-treated mice, DMXAA-treated mice displayed significantly reduced muscle fiber cross-sectional area (CSA) as well as decreased maximum, minimum, and mean fiber diameters following cisplatin treatment (Fig. 1I–M, p < 0.05). Together, these results indicate that STING agonist exacerbates cisplatin-induced skeletal muscle atrophy.

Loss of cGAS or STING mitigates cisplatin-induced body weight loss and skeletal muscle atrophy

To investigate the role of cGAS–STING signaling in cisplatin-induced skeletal muscle atrophy, cGAS knockout (cGAS KO) and STING knockout (STING KO) mice were used (Fig. 2A). Under control conditions, body weight was comparable among wild-type (WT), cGAS KO, and STING KO mice (Fig. 2B). Following cisplatin treatment, WT mice exhibited a marked reduction in body weight, whereas both cGAS KO and STING KO mice showed significantly attenuated body weight loss compared with WT mice (p < 0.05) (Fig. 2B).

Fig. 2.

Loss of cGAS or STING mitigates cisplatin-induced skeletal muscle atrophy. A Experimental design using wild-type (WT), cGAS knockout (cGAS KO), and STING knockout (STING KO) mice. B Body weight of WT, cGAS KO, and STING KO mice under control conditions and after cisplatin treatment. C-E Weights of soleus (SO), gastrocnemius (GC), and tibialis anterior (TA) muscles under control and cisplatin-treated conditions. F–H Muscle weights normalized to tibial length (TL). J Representative dystrophin immunofluorescence staining of TA muscle sections from WT, cGAS KO, and STING KO mice under control and cisplatin-treated conditions. Scale bar, 50 μm. I, K-M Quantification of muscle fiber cross-sectional area (CSA), maximum diameter, minimum diameter, and mean diameter under control and cisplatin-treated conditions. Data are presented as mean ± SEM. Each dot represents an individual mouse (n = 3–9 per group). *p < 0.05 compared to WT mice. **p < 0.01 compared to WT mice

Consistent with these observations, muscle weight measurements after cisplatin administration revealed genotype-dependent differences. The weights of the GC and TA muscles were significantly higher in cGAS KO mice than in WT mice following cisplatin treatment (Fig. 2D-E). A similar pattern was observed in STING KO mice (Fig. 2D-E). When normalized to tibial length (TL), GC/TL and TA/TL ratios showed the same trend (Fig. 2G-H). In contrast, SO muscle weight did not differ among genotypes after cisplatin treatment (Fig. 2C, F). Overall, deletion of cGAS or STING partly preserved muscle mass after cisplatin injury. This protective effect was most evident in GC and TA, while SO muscle showed little change.

Morphological analyses using dystrophin immunofluorescence staining of TA muscles further supported these findings. Under control conditions, muscle fiber cross-sectional area (CSA) as well as maximum, minimum, and mean fiber diameters were comparable among WT, cGAS KO, and STING KO mice (Fig. 2J–M). In contrast, following cisplatin treatment, both cGAS KO and STING KO mice exhibited significantly larger muscle fiber CSA and fiber diameters compared with WT mice (Fig. 2I–M, p < 0.05). Together, these results indicate that loss of cGAS or STING partially protects against cisplatin-induced skeletal muscle atrophy.

cGAS deficiency is associated with reduced inflammatory and innate immune signaling in skeletal muscle following cisplatin treatment

To further dissect the molecular mechanisms by which cGAS regulates skeletal muscle homeostasis under chemotherapeutic stress, we performed transcriptomic profiling on TA muscle samples from WT and cGAS KO mice following cisplatin treatment.

Unsupervised hierarchical clustering of differentially expressed genes clearly separated cGAS KO and WT samples, indicating distinct transcriptional profiles associated with cGAS deficiency (Fig. 3A). Volcano plot analysis identified 696 differentially expressed genes (DEGs), including 293 upregulated and 403 downregulated genes in cGAS KO compared with WT mice (adjusted p < 0.05, |log₂FC|> 1) (Fig. 3B). The complete list of DEGs is provided in Supplementary Table 1.

Fig. 3.

Transcriptomic profiling reveals altered gene expression and pathway enrichment in cGAS-deficient skeletal muscle following cisplatin treatment. A Heatmap of the top 30 differentially expressed genes (DEGs) in tibialis anterior muscle from cGAS KO and WT mice after cisplatin treatment, including the top 15 upregulated and top 15 downregulated genes ranked by log₂ fold change. DEGs were defined by adjusted p < 0.05 and |log₂ fold change|> 1. Expression values are shown as row-wise Z-scores. B Volcano plot illustrating the distribution of DEGs between cGAS KO and WT mice. Significantly upregulated genes (red) and significantly downregulated genes (blue) were defined by an adjusted p-value < 0.05 and |log₂ fold change|> 1. Representative genes involved in innate immune signaling and muscle remodeling, including Cgas, Sting1 (Tmem173), Fbxo32, Trim63, Myod1, Myog, and Isg15, are indicated in the plot. C–F Gene set enrichment analysis (GSEA) plots showing enriched pathways in WT versus cGAS KO muscle, including NF-κB signaling (C), mitophagy (D), NOD-like receptor signaling (E), and IL-1 signaling (F). ES, NES, nominal p values, and FDR are shown in each panel. n = 3 per group

To further define pathway-level alterations, gene set enrichment analysis (GSEA) was performed. Several inflammation- and stress-related pathways were significantly enriched in WT relative to cGAS KO muscle, including NF-κB signaling, NOD-like receptor signaling, and IL-1 signaling (Fig. 3C–F). Mitophagy-related pathways were also enriched, suggesting altered mitochondrial quality control. Notably, these pathways showed negative normalized enrichment scores (NES), indicating their suppression in cGAS KO muscle. The full list of significantly enriched pathways (FDR < 0.05) is provided in Supplementary Table 2.

Together, these results indicate that cGAS deficiency attenuates inflammatory signaling and stress-response pathways while reshaping transcriptional programs in response to cisplatin-induced muscle atrophy.

Skeletal muscle-specific deletion of cGAS attenuates cisplatin-induced muscle atrophy

To further examine the role of cGAS in skeletal muscle atrophy, we generated skeletal muscle-specific cGAS knockout mice (cGAS mKO) by crossing cGAS^flox/flox (Flox) mice with Myf5-Cre transgenic mice (Fig. 4A). Under control conditions, there were no significant differences in body weight between cGAS mKO and Flox control mice (Fig. 4B). Following cisplatin treatment, Flox mice exhibited a marked reduction in body weight, whereas cGAS mKO mice showed significantly attenuated body weight loss (Fig. 4B, p < 0.05).

Fig. 4.

Skeletal muscle-specific deletion of cGAS attenuates cisplatin-induced skeletal muscle atrophy. A Schematic of the skeletal muscle-specific cGAS knockout (cGAS mKO) experimental design. B Body weight of the cGAS Flox and cGAS mko mice under control conditions and after cisplatin treatment. C–E Weights of soleus (SO), gastrocnemius (GC), and tibialis anterior (TA) muscles under control and cisplatin-treated conditions. F–H Muscle weights normalized to tibial length (SO/TL, GC/TL, and TA/TL). K Representative dystrophin immunofluorescence staining of TA muscle sections from Flox and cGAS mKO mice under control and cisplatin-treated conditions. Scale bar, 50 μm. I, J, L, M Quantification of muscle fiber cross-sectional area (CSA), maximum diameter, minimum diameter, and mean diameter. Data are presented as mean ± SEM. Each dot represents an individual mouse (n = 3–12 per group). *p < 0.05 compared to cGAS Flox mice. **p < 0.01 compared to cGAS Flox mice

Muscle weight measurements after cisplatin administration revealed that the weights of SO, GC, and TA muscles were significantly higher in cGAS mKO mice than in Flox mice (Fig. 4C-E). When normalized to tibial length (TL), SO/TL, GC/TL, and TA/TL ratios showed the same trend (Fig. 4F–H). These findings suggest that skeletal muscle-specific deletion of cGAS alleviates cisplatin-induced muscle atrophy.

To assess muscle fiber morphology, dystrophin immunofluorescence staining of TA muscles was performed. Representative images are shown in Fig. 4 K. Under control conditions, muscle fiber cross-sectional area (CSA), as well as minimum, maximum, and mean fiber diameters, were similar between cGAS mKO and Flox mice (Fig. 4 I–M). Following cisplatin treatment, while both groups exhibited significant reductions in fiber CSA and fiber diameter, these parameters remained significantly greater in cGAS mKO mice compared with Flox controls (Fig. 4 I–M, p < 0.01). Collectively, these findings indicate that skeletal muscle-specific deletion of cGAS attenuates cisplatin-induced skeletal muscle atrophy.

Skeletal muscle-specific deletion of cGAS attenuates cisplatin-induced upregulation of atrogenes

The expression of Fbxo32 (encoding Atrogin-1), a critical E3 ubiquitin ligase implicated in muscle proteolysis, was examined by western blotting and quantitative real-time PCR (Fig. 5A–E). In untreated control groups, both FBXO32 protein and mRNA levels were similar between cGAS mKO and Flox mice, indicating that cGAS deletion does not affect baseline expression. Upon cisplatin treatment, FBXO32 levels were markedly elevated in skeletal muscles of cGAS Flox mice compared to their untreated controls (p < 0.01). In contrast, this cisplatin-induced upregulation was significantly blunted in cGAS mKO mice at both the protein (p < 0.05) and transcript levels (Fig. 5B, D, p < 0.01). Consistently, mRNA levels of Murf1 were also significantly increased in cGAS Flox mice following cisplatin administration (p < 0.01). However, this induction was significantly attenuated in cGAS mKO mice (p < 0.05), suggesting a broader suppression of the muscle catabolic program (Fig. 5C, E). These findings demonstrate that skeletal muscle-specific cGAS deletion preserves muscle homeostasis under chemotherapeutic stress by attenuating the cisplatin-induced activation of key atrogenes, including Fbxo32 and Murf1.

Fig. 5.

Skeletal muscle-specific deletion of cGAS attenuates cisplatin-induced activation of atrogenes. A Representative Western blot images showing FBXO32 protein levels in tibialis anterior muscles from cGAS Flox and cGAS mKO mice under control conditions and following cisplatin treatment. GAPDH served as the loading control. B Quantification of FBXO32 protein abundance. C–D Relative mRNA expression of Murf1 and Fbxo32 in tibialis anterior muscle assessed by qPCR. Data are presented as mean ± SEM. Each dot represents an individual mouse (n = 4–5 per group). ns: not significant; *p < 0.05; **p < 0.01

Skeletal muscle-specific cGAS deletion attenuates the suppression of myogenic regulatory signals under cisplatin stress

To determine whether cGAS affects myogenic signaling, we assessed MyoD protein by western blot. We also measured several myogenic and growth-related transcripts by qPCR, including myogenin, myostatin, Myf6, Igf1 and Hgf, which are involved in the regulation of skeletal muscle regeneration and muscle atrophy [31, 32]. Under control conditions, MyoD protein and mRNA were comparable between Flox and cGAS mKO mice (Fig. 6A, E). Myogenin, myostatin, Myf6, Igf1 and Hgf mRNA also showed no genotype-dependent differences. Cisplatin reduced MyoD expression in both groups, but the decrease was significantly smaller in cGAS mKO mice (Fig. 6A, B). MyoD protein and mRNA remained higher than in Flox controls after cisplatin treatment. Myostatin and Myf6 were only mildly affected by cisplatin and did not differ between genotypes (Fig. 6C, E). By contrast, Myogenin and Igf1 expression levels were markedly reduced in Flox mice, whereas this decline was attenuated in cGAS mKO muscle (Fig. 6F, H). Hgf showed a similar pattern, with partial preservation in the cGAS mKO group (Fig. 6G). Taken together, these findings indicate that skeletal muscle-specific cGAS deletion attenuates the cisplatin-induced suppression of MyoD, Myog, Hgf, and Igf1 expression, suggesting a partial preservation of myogenic regulatory signaling under chemotherapeutic stress.

Fig. 6.

Skeletal muscle–specific deletion of cGAS preserves myogenic regulatory factor expression following cisplatin treatment. A Representative Western blot images showing MyoD protein levels in tibialis anterior muscles from cGAS Flox and cGAS mKO mice under control conditions and after cisplatin treatment. GAPDH served as the loading control. B Quantification of MyoD protein abundance. C–D Relative mRNA expression of myostatin and myf6. E–F Relative mRNA expression of MyoD and Myogenin. G–H Relative mRNA expression of HGF and IGF-1. Data are presented as mean ± SEM. Each dot represents an individual mouse (n = 4–7 per group). ns: not significant; *p < 0.05; **p < 0.01

Deletion of cGAS in skeletal muscle attenuates cisplatin-induced NF-κB activation and interferon-stimulated gene expression

Using immunohistochemistry, we assessed the nuclear translocation of the NF-κB p65 subunit, a hallmark of NF-κB pathway activation. Under basal conditions, the proportion of NF-κB p65–positive nuclei was low and comparable between skeletal muscle–specific cGAS mKO mice and Flox controls (Fig. 7A-B). Following cisplatin treatment, an increased nuclear localization of NF-κB p65 was observed in both groups, indicating enhanced NF-κB activation in response to cisplatin-induced stress (Fig. 7A–B). However, the percentage of NF-κB p65-positive nuclei was significantly lower in cGAS mKO mice compared to Flox mice following cisplatin treatment (Fig. 7A–B).

Fig. 7.

Skeletal muscle-specific deletion of cGAS attenuates cisplatin-induced NF-κB activation and interferon-stimulated gene expression. A Representative immunohistochemical images showing NF-κB p65 nuclear localization in tibialis anterior muscle from cGAS Flox and cGAS mKO mice under control and cisplatin-treated conditions. Representative p65-positive nuclei are indicated by arrows. Dashed boxes denote regions selected for higher-magnification views. Scale bar: 50 μm. B Quantification of NF-κB p65-positive nuclei. C–G Relative mRNA expression of interferon-related genes, including Irf3, Isg15, IFNβ1, Irf3, and Irf7. Data are presented as mean ± SEM. Each dot represents an individual mouse (n = 3–4 per group). ns: not significant; *p < 0.05; **p < 0.01

To further assess downstream signaling associated with cGAS–STING activation, we measured the expression of several interferon-stimulated genes, which represent canonical transcriptional outputs of this pathway. Cisplatin treatment markedly induced the expression of Ifit3, IFN-β, Isg15, Irf3, and Irf7 in Flox mice (Fig. 7C–G). In contrast, these responses were significantly attenuated in cGAS mKO mice. In the absence of cisplatin, basal levels of these genes remained comparable between the two groups (Fig. 7C–G). Together, these findings indicate that skeletal muscle-specific cGAS deletion suppresses both NF-κB activation and the associated interferon response following cisplatin exposure. These changes are consistent with known downstream outputs of cGAS–STING signaling and support the modulation of pathway activity in the genetic models used in this study.

Deletion of cGAS in skeletal muscle attenuates cisplatin-induced Beclin1 upregulation and autophagy-related gene expression

We next examined autophagy-related markers in skeletal muscle, as previous studies have reported that cisplatin treatment is associated with alterations in autophagy-related pathways in skeletal muscle [33, 34]. Under control conditions, Beclin1 protein levels were similar between cGAS Flox and cGAS mKO mice. Following cisplatin treatment, Beclin1 expression increased markedly in Flox mice, whereas the increase was significantly lower in cGAS mKO mice (Fig. 8A–B, p < 0.05), indicating an attenuated Beclin1 upregulation. We further assessed LC3-I and LC3-II protein levels in skeletal muscle. Although LC3-II/LC3-I ratios showed modest changes following cisplatin treatment, no statistically significant differences were observed between groups (Fig. 8C–D). Consistent with these findings, the expression of Microtubule-associated protein 1 light chain 3 beta (Map1lc3b) remained unchanged across groups, showing no significant effect of genotype or treatment (Fig. 8E). In contrast, GABA type A receptor-associated protein like 1 (Gabarapl1) expression was increased following cisplatin treatment in Flox mice, whereas this response was attenuated in cGAS mKO mice (Fig. 8F). Similarly, Sequestosome 1 (Sqstm1) expression was significantly elevated after cisplatin administration, and this increase was partially reduced in cGAS mKO mice (Fig. 8G). Collectively, these findings indicate that skeletal muscle-specific deletion of cGAS attenuates cisplatin-induced changes in autophagy-related gene expression and Beclin1 upregulation, but do not provide direct evidence for alterations in autophagic flux.

Fig. 8.

Skeletal muscle-specific deletion of cGAS attenuates autophagy-related responses following cisplatin treatment. A Representative Western blot images of Beclin 1 protein levels in tibialis anterior muscles from cGAS Flox and cGAS mKO mice under control and cisplatin-treated conditions. GAPDH served as the loading control. B Quantification of Beclin 1 protein abundance. C Representative Western blot images of LC3-I and LC3-II protein levels in tibialis anterior muscles. GAPDH served as the loading control. D Quantification of the LC3-II/LC3-I ratio. E–G Relative mRNA expression of autophagy-related genes Map1lc3b, Gabarapl1, and Sqstm1. Data are presented as mean ± SEM. Each dot represents an individual mouse (n = 4–7 per group). ns, not significant; *p < 0.05; **p < 0.01

Discussion

In this study, we identify a previously underappreciated role of the cGAS–STING pathway in skeletal muscle under chemotherapeutic stress. Cisplatin causes rapid muscle loss, but the upstream drivers remain unclear. Our data show that cGAS or STING deletion markedly reduces this loss. Our results also indicate that cisplatin exposure induces interferon-related gene expression patterns that are consistent with activation of the cGAS–STING signaling pathway, as evidenced by increased NF-κB activation and elevated expression of interferon-stimulated genes, including IFNβ, ISG15, Ifit3, Irf3, and Irf7. These molecular signatures are commonly used indicators of cGAS–STING pathway activation in response to cellular stress. Previous studies suggest that cGAS deficiency can alter basal inflammatory transcriptional signatures [35]. Because our RNA-seq analysis was performed in cisplatin-treated skeletal muscle, genotype-dependent baseline differences cannot be entirely excluded. This activation may be associated with cellular stress induced by cisplatin treatment, which has been reported to trigger cGAS–STING signaling in various pathological contexts. The protective effect appears in both global and muscle-specific knockout models, indicating a muscle-intrinsic function. We observed that cGAS deletion suppresses key atrogenes such as Fbxo32 and Murf1. It also limits NF-κB activation and interferon-stimulated gene expression. Autophagy-related markers, including Beclin1, were altered following cisplatin treatment. However, these changes do not provide direct evidence for altered autophagic flux. Together, these findings suggest that cGAS–STING signaling coordinates several pathological responses to cisplatin, including proteolysis, inflammation, autophagy-related alterations, and impaired myogenic signaling. The pathway may therefore represent an important regulator of muscle vulnerability during chemotherapy. A schematic summary of this integrative mechanism is presented in Fig. 9.

Fig. 9.

Proposed model of cGAS–STING signaling in cisplatin-induced skeletal muscle atrophy. Cisplatin treatment is associated with inflammatory signaling changes consistent with engagement of the cGAS–STING pathway in skeletal muscle, including activation of NF-κB signaling, induction of atrophy-related genes, and suppression of myogenic regulators, collectively contributing to muscle atrophy. Genetic deletion of cGAS or STING attenuates these responses and partially protects against cisplatin-induced skeletal muscle wasting

Chemotherapy-induced skeletal muscle atrophy, exemplified by cisplatin treatment, represents a major clinical challenge, significantly impairing patient prognosis and quality of life [36]. Previous studies, including our own, have demonstrated that cisplatin administration results in substantial reductions in muscle mass by activating multiple pathological signaling cascades [9, 30, 37, 38]. Consistent with these observations, our data further suggest that cGAS functions as an upstream regulator that contributes to the exacerbation of muscle atrophy. Traditionally, cGAS is recognized for its role in antiviral innate immunity [39]. However, emerging evidence implicates aberrant activation of the cGAS–STING pathway in non-infectious tissue injury and chronic inflammation. This has been observed in pathological contexts such as myocardial infarction [40], ischemic stroke [41, 42], and aging-related degeneration [43].

Mechanistically, we observed that cisplatin treatment significantly upregulated Fbxo32 (Atrogin-1), a key E3 ubiquitin ligase mediating proteasomal degradation in skeletal muscle [44, 45]. This finding aligns with previous studies identifying Fbxo32 as a pivotal contributor to muscle wasting in both cancer cachexia [46] and disuse atrophy [47, 48]. Importantly, skeletal muscle-specific deletion of cGAS markedly attenuated cisplatin-induced Fbxo32 upregulation, suggesting that cGAS may influence upstream regulators of proteolytic pathways. Moreover, the reduction in Fbxo32 expression observed in cGAS-deficient mice is in line with previous studies reporting that Daidzein treatment can attenuate skeletal muscle atrophy [49], supporting a potential role for cGAS ablation in modulating molecular pathways associated with muscle catabolism.

Autophagy is a key cellular process involved in maintaining skeletal muscle quality control. Controlled activation of autophagy facilitates the clearance of damaged proteins and organelles, thereby preserving muscle integrity [13, 50]. Although autophagy has been implicated in muscle loss under certain conditions [51, 52], our results indicate that cisplatin treatment was associated with increased Beclin1 expression and changes in several autophagy-associated genes. Importantly, skeletal muscle-specific deletion of cGAS attenuated this cisplatin-induced Beclin1 upregulation. However, LC3-II/LC3-I ratios showed only modest, non-significant changes between groups, and therefore our findings do not provide direct evidence for altered autophagic flux. In addition, Xiong et al. found that activation of the cGAS–STING pathway suppressed autophagy in cardiac tissue [53], indicating that cGAS–STING signaling may influence autophagy-related pathways in a context- and tissue-dependent manner. Together, these findings suggest that cGAS may contribute to autophagy-related responses in skeletal muscle under cisplatin-induced stress, although the precise relationship between cGAS signaling and autophagic flux requires further investigation.

Regarding inflammation, we confirmed that cisplatin treatment activated the NF-κB pathway, as indicated by increased nuclear localization of NF-κB p65. This finding is consistent with previous reports showing that chemotherapy-induced inflammation contributes to skeletal muscle atrophy [9, 36, 54]. Importantly, skeletal muscle-specific deletion of cGAS significantly attenuated NF-κB activation following cisplatin exposure. This is consistent with the work by Yum et al., who demonstrated that cGAS deficiency markedly impairs TBK1 recruitment and subsequent NF-κB signaling activation, as evidenced by reduced phosphorylation of p65 and IκBα following STING agonist stimulation [55]. Together, these results highlight a critical role for cGAS–STING signaling in orchestrating NF-κB-driven inflammatory responses. Consistent with our findings, NF-κB activation has been closely associated with skeletal muscle atrophy, and inhibition of NF-κB signaling has been shown to effectively mitigate muscle wasting across various pathological conditions [17, 56].

Furthermore, MyoD is a key transcription factor that governs myogenic differentiation and plays a critical role in the maintenance of skeletal muscle mass [57, 58]. In addition to MyoD, several other factors analyzed in this study are also involved in the regulation of skeletal muscle mass. Myogenin and Myf6 belong to the myogenic regulatory factor family and participate in the control of myogenic transcriptional programs [59]. In contrast, myostatin is a well-established negative regulator of skeletal muscle growth and a key mediator of muscle wasting under various pathological conditions [60]. Growth factors such as IGF-1 and HGF also contribute to the maintenance of muscle mass through anabolic signaling pathways [61]. Dysregulation of these factors has been widely associated with skeletal muscle atrophy in disease and chemotherapy settings. Consistent with this notion, our results showed that several of these regulators were suppressed following cisplatin exposure, whereas cGAS deletion partially preserved their expression. Following cisplatin exposure, MyoD expression was significantly downregulated, consistent with previous observations reported in skeletal muscle atrophy models [62, 63]. Importantly, skeletal muscle-specific deletion of cGAS preserved MyoD expression at both the mRNA and protein levels after cisplatin treatment. This suggests that cGAS–STING signaling contributes to cisplatin-induced muscle atrophy, at least in part, by suppressing key myogenic regulatory pathways. Given that impaired myogenesis aggravates muscle mass loss and delays recovery during pathological stress, the preservation of MyoD expression in cGAS-deficient mice likely underpins their resistance to cisplatin-induced skeletal muscle atrophy. Together with our findings on proteolytic activation and inflammatory signaling, these results highlight that cGAS–STING activation exacerbates muscle atrophy through a multifaceted mechanism involving enhanced proteolysis, heightened inflammation, and impaired myogenic maintenance.

Compared with prior strategies aimed at mitigating chemotherapy-induced muscle loss, such as exercise [64], pharmacological antioxidants [65, 66], and nutritional supplementation [67], targeting cGAS may represent a complementary approach by modulating multiple pathological processes, including proteolysis, autophagy-related alterations, and inflammation.

Nonetheless, several limitations should be acknowledged. First, although our data indicate modulation of key signaling pathways, the cellular source contributing to cGAS–STING pathway engagement in skeletal muscle remains unclear. Mitochondrial DNA (mtDNA) release, a recognized trigger of cGAS signaling [68], may contribute to this process and warrants further investigation. The cGAS–STING pathway activates TBK1–IRF3 signaling and induces interferon-stimulated genes. In this study, several interferon-related genes (IFNβ, Isg15, Ifit3, Irf3, and Irf7) were analyzed. However, direct assessment of TBK1 and IRF3 phosphorylation would further clarify the activation status of cGAS–STING signaling in skeletal muscle under cisplatin treatment. Second, while our study focused on muscle-intrinsic mechanisms, the contributions of immune cells and systemic cytokines to muscle atrophy were not fully dissected. In particular, infiltration of immune cells such as neutrophils and macrophages was not examined in the present study and may contribute to the inflammatory microenvironment during chemotherapy-induced muscle atrophy. Third, the long-term effects of cGAS deletion on muscle atrophy and overall metabolism remain to be explored.

These limitations also highlight several directions for future investigation. Pharmacological inhibition of cGAS (e.g., RU.521) [69] or STING (e.g., small-molecule antagonists) [70, 71], may represent a potential therapeutic strategy to mitigate muscle atrophy in chemotherapy settings. In addition, investigating the interplay between cGAS–STING signaling, mitochondrial dynamics [72], and cellular senescence [43] in skeletal muscle may further uncover novel therapeutic targets.

Conclusion

In conclusion, our study identifies a role for cGAS–STING signaling in cisplatin-induced skeletal muscle atrophy. Genetic deletion of cGAS or STING preserved muscle mass and fiber size, accompanied by reduced expression of proteolysis-related genes, attenuation of NF-κB–associated inflammatory signaling, and partial preservation of myogenic regulators. These findings suggest that cGAS–STING signaling is involved in coordinating proteostasis disruption, inflammatory responses, and impaired myogenic regulation during chemotherapeutic stress. Targeting this pathway may represent a potential strategy to mitigate chemotherapy-associated muscle wasting. Further studies evaluating pharmacological inhibition of cGAS–STING signaling in preclinical models will be important to assess its therapeutic potential.

Authors’ contributions

Conceptualization: XL, LZ; Methodology: LZ; Formal analysis: XL, MX, LZ, HW1, HW2, HW3, WF, ML, JH, FC, HW4, YY; Investigation: XL, MX, LZ, HW1, HW2, HW3, WF, ML, JH, FC, HW4, YY; Writing–original draft: XL, MX; Writing–review & editing: XL, MX, LZ.

Funding

This work was supported by the National Natural Science Foundation of China (No. 32300964), the Guangdong Province Key-Area Research Special Fund for General Higher Education Institutions (No. 2025ZDZX2040) and Guangdong Basic and Applied Basic Research Foundation (No. 2022A1515111105).

Data availability

All data needed to evaluate the conclusions in the paper are present in the paper and the Supplementary Materials. Additional data or original files related to this paper may be requested from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.