NAD ⁺ Enhanced Mesenchymal Stromal Cells Effect on Muscle Atrophy by Improving SIRT1‐Mediated Mitochondrial Function via NAMPT

MSCs were harvested from the fresh umbilical cords of healthy newborns with informed parental consent, as previously reported [23]. MSCs were cultured in α‐MEM medium (Gibco, NY, USA) supplemented with 10% foetal bovine serum (FBS; Gibco), along with 100 U/mL penicillin and 100 μg/mL streptomycin (Gibco). The third to fifth passages of the cells were used for flow cytometry, induction of differentiation, administration and co‐culture using the Transwell system. NAD⁺‐MSCs were obtained by NAD⁺ stimulation (150 μmol/L) of MSCs for 48 h.

To explore the underlying mechanism of NAD⁺ action, MSCs were transfected with small interfering RNA (siRNA) targeting nicotinamide phosphoribosyltransferase (NAMPT) prior to NAD⁺ exposure using Lipofectamine 2000 transfection reagent (Invitrogen, CA, USA) following the manufacturer's instructions. MSCs were seeded in six‐well plates and then shifted to Opti‐MEM I–reduced serum medium (Gibco) containing NAMPT siRNA (125 nM) for 6 h. Next, the α‐MEM medium was introduced and MSCs were subjected to NAD⁺ treatment for 48 h. The siRNA oligonucleotides were synthesised by GenePharma Co. Ltd. (Shanghai, China). The sequences for the negative control (NC) siRNA were outlined as follows: sense 5′‐UUCUCCGAACGUGUCACGUTT‐3′ and antisense 5′‐ACGUGACACGUUCGGAGAATT‐3′. The sequences for NAMPT siRNA were as follows: sense 5′‐GCAGAACACAGUACCAUAATT‐3′ and antisense 5′‐UUAUGGUACUGUGUUCUGCTT‐3′.

Human embryonic lung fibroblasts (HELFs) were acquired from the China Cell Culture Center (Shanghai, China) and maintained in high‐glucose Dulbecco's modified eagle's medium (DMEM, Gibco), enriched with 10% FBS, 100 U/mL penicillin and 100 μg/mL streptomycin at 37°C in a 5% CO₂ incubator. Mouse C2C12 myoblasts were purchased from China Infrastructure of Cell Line Resource (Beijing, China) and cultured in high‐glucose DMEM supplemented with 10% FBS and antibiotics. Upon reaching 80%–90% confluence, a differentiation medium comprising DMEM and 2% heat‐inactivated horse serum was introduced for 4 days. Fully differentiated myotubes were subsequently stimulated with D‐gal (20 mg/mL; MCE, Shanghai, China) for 48 h, followed by treatment with MSCs, HELFs and NAD⁺‐MSCs via a Transwell system (Corning, NY, USA) for 24 h.

To validate the role of SIRT1 in MSCs, C2C12 myoblasts were transfected with short hairpin↗ RNA (shRNA, GenePharma Co. Ltd.) using a lentivirus (LV) according to the manufacturer's protocols. The transfected cells were fully differentiated for 4 days and exposed to MSCs/NAD⁺‐MSCs for 24 h. The shRNA oligonucleotides were synthesised by GenePharma Co. Ltd. The sequences for NC shRNA were 5′‐TTCTCCGAACGTGTCACGT‐3′, while the sequences for SIRT1 shRNA were 5′‐TTAACAACCTCTTGATCCC‐3′.

Six‐week‐old male C57BL/6 J mice were obtained from Jiangsu Huachuang Sino PharmaTech Co. Ltd. (Suzhou, China) and maintained on a standard diet. The mice were housed in an environment with a controlled 12 h light/dark cycle at 22°C–25°C with 55% ± 5% humidity. Following a 1‐week period of accommodative feeding, mice received daily subcutaneous injections of D‐gal at a dose of 100 mg/kg for 12 weeks. MSCs or NAD⁺‐MSCs (1 × 10⁶ cells/mouse) in PBS were injected into the quadriceps (QUAD), tibialis anterior (TA) and gastrocnemius (GAS) muscles of the hind limbs every 7 days for six cycles. The controls received an injection of PBS at the same volume, devoid of MSCs.

One week after the last injection of MSCs/NAD⁺‐MSCs, the exhaustive running distance of mice was evaluated on a treadmill (Xinrun Information Technology Co. Ltd., Shanghai, China). Prior to testing, the mice underwent a two‐day acclimation period. Exercise endurance was assessed by measuring the distance they could run until exhaustion was reached. Initially, the mice underwent a 10‐min running session at a speed of 8 m/min. Subsequently, the treadmill's speed was escalated by 0.2 m/min. Exhaustion was determined when the hind limbs of the mice remained on the electric grid for more than 10 consecutive seconds. The grip strength was measured using an electronic dynamometer (Shanghai Xinrun Information Technology Co. Ltd.). The mice were trained to hold onto a horizontal grid linked to the dynamometer using all four limbs and pulled back horizontally with a steady force. The force exerted on the grid each time a mouse lost its grip was recorded. This assessment was conducted three times for each mouse and the obtained measured values underwent an averaging process.

Following 1 week of in vivo muscle performance evaluation, the mice were euthanised using CO₂. Blood samples were collected and plasma was extracted to measure lipid levels. Subsequently, the bilateral muscles, including QUAD, TA, GAS and soleus (SO), were dissected. The TA and SO muscle weights were normalised to the body weight. One of the fresh QUAD muscle samples was used for RNA‐seq, untargeted metabolomics and ATP content assay, while the other was snap‐frozen in liquid nitrogen and stored at −80°C for protein extraction and oil red O staining. TA muscles were preserved in 4% paraformaldehyde for subsequent histology staining, including haematoxylin–eosin (H&E), sirius red and Masson staining.

The fixed TA muscles were encased in paraffin and sectioned to a thickness of 5 μm at the maximum cross‐section. After deparaffinisation, sections were subjected to standard H&E staining, sirius red staining and Masson staining. The frozen QUAD muscles were sliced to a 7‐μm thickness at the maximum cross‐section and used for oil red O staining following standard procedures. Images were captured using a microscope (BX53; Olympus, Japan) and quantified using Image‐Pro Plus software. Intramuscular lipid content was represented as the ratio of oil red O–positive area to muscle fibre area.

Total RNA was isolated from QUAD muscles and MSCs using TRIzol reagent (Invitrogen), following the manufacturer's instructions. Subsequent library preparation and sequencing were performed by Beijing Tsingke Biotechnology Co. Ltd. DESeq2 (v1.26.0) was used to perform differential expression analyses. Gene ontology (GO) enrichment analysis of differentially expressed genes (DEGs) was conducted using the GOseq R package, which relies on the Wallenius non‐central hypergeometric distribution. KOBAS software was employed to assess the statistical enrichment of DEGs within KEGG pathways.

Then, 1 μg of RNA was converted into cDNA through reverse transcription using the Prime Script RT Reagent Kit (Cat. No. RR047A; Takara, Japan). Primers were chemically synthesised by Tsingke Biotechnology Co. Ltd. and the sequences of them were as follows: Mus‐Gapdh, sense 5′‐AAGGGCTCATGACCACAGTC‐3′ and antisense 5′‐CAGGGATGATGTTCTGGGCA‐3′; Mus‐p16, sense 5′‐GCTCAACTACGGTGCAGATTC‐3′ and antisense 5′‐GCACGATGTCTTGATGTCCC‐3′; Mus‐p21, sense 5′‐CCTGGTGATGTCCGACCTG‐3′ and antisense 5′‐CCATGAGCGCATCGCAATC‐3′; Homo‐Gapdh, sense 5′‐ACAACTTTGGTATCGTGGAAGG‐3′ and antisense 5′ ‐GCCATCACGCCACAGTTTC‐3′; Homo‐p16, sense 5′‐GATCCAGGTGGGTAGAAGGTC‐3′ and antisense 5′‐CCCCTGCAAACTTCGTCCT‐3′; Homo‐p21, sense 5′‐CGATGGAACTTCGACTTTGTCA‐3′ and antisense 5′‐GCACAAGGGTACAAGACAGTG‐3′; Homo‐IL‐1β, sense 5′ ‐ATGATGGCTTATTACAGTGGCAA‐3′ and antisense 5′‐GTCGGAGATTCGTAGCTGGA‐3′ and Homo‐IL‐6, sense 5′‐ACTCACCTCTTCAGAACGAATTG‐3′ and antisense 5′‐CCATCTTTGGAAGGTTCAGGTTG‐3′. Real‐time PCR analysis was conducted utilising the SYBR Green PCR kit (Cat. No. RR420A; Takara). Variations in gene expression were evaluated by employing the comparative CT (2^−ΔΔCt) approach, and the results were quantified through normalisation against Gapdh, which served as the reference control.

Untargeted metabolomics of the QUAD muscles was performed at Beijing Tsingke Biotechnology Co. Ltd. The dataset comprising peak numbers, sample identifiers and normalised peak areas was integrated into the SIMCA16.0.2 software (Sartorius Stedim Data Analytics AB, Umea, Sweden) for comprehensive multivariate statistical analysis. A supervised method known as orthogonal projections to latent structure‐discriminant analysis (OPLS‐DA) was used to facilitate the visual assessment of group segregation and identification of significantly altered metabolites. Additionally, the variable importance in the projection (VIP) scores from the primary principal components of the OPLS‐DA model was extracted. Metabolites exhibiting VIP > 1 and p < 0.05 (Student's t‐test) were classified as significantly altered metabolites.

Upon reaching 80%–90% confluency, the MSCs were subjected to serum‐free medium, followed by a 24‐h incubation period. Subsequently, the supernatant was harvested and concentrated 20‐fold using 3‐kDa molecular weight cutoff ultrafiltration membranes (Millipore), which was subsequently utilised for proteomic analysis.

Proteomics was conducted using the Vanquish Neo UHPLC system for sample separation. Data‐independent acquisition (DIA) was performed using the Vanquish Neo system (Thermo Fisher Scientific) for chromatographic separation. Raw MS data were analysed using DIA‐NN (v1.8.1) with a library‐free method. A spectral library was generated using the uniprotkb_proteome_UP000005640_human_82 493_20240528.fasta database (82 493 sequences), leveraging advanced neural network algorithms for deep learning. To build a spectral library from DIA data, the Match Between Runs feature was applied, which was subsequently utilised for reanalysis. The false discovery rate of the search outcomes was adjusted to < 1% at both the protein and precursor ion levels. Only the remaining valid identifications were employed for subsequent quantitative analysis.

Upon reaching 80%–90% confluence, the MSCs/NAD⁺‐MSCs were cultured in serum‐free medium for 24 h, and the cell supernatants were collected. The cell supernatants were centrifuged at 10,000 × g for 1 h and filtered through a 0.22‐μm filter to remove cellular debris. Subsequently, the medium was ultracentrifuged at 100,000 × g for 70 min at 4°C to obtain extracelluar vesicles (EVs) used for protein extraction.

QUAD muscle, C2C12 myotubes, MSCs and EVs were subjected to lysis with a radioimmunoprecipitation assay (RIPA) buffer (P0013B; Beyotime, Shanghai, China). The extracted proteins were subsequently separated and transferred onto polyvinylidene difluoride (PVDF) membranes (IPVH00010 0.45 μm; Millipore, MA, USA). Subsequent to blocking with a 5% skim milk solution at room temperature for 1 h, the membranes underwent overnight incubation with specific primary antibodies at 4°C. After incubation with horseradish peroxidase (HRP)–conjugated secondary antibodies for 1 h at room temperature, protein detection was carried out via enhanced chemiluminescence (ECL) analysis. Band quantitation was performed using the ImageJ software, with normalisation to GAPDH as an internal control.

The following primary antibodies were used: Atrogin 1 (1:5000; Cat. No. 67172‐1‐Ig; Proteintech, IL, USA), MuRF1 (1:1000; Cat. No. 55456‐1‐AP; Proteintech), GAPDH (1:5000; Cat. No. AB0037; Abways, Shanghai, China), SIRT1 (1:1000; Cat. No. 13161‐1‐AP; Proteintech), PGC‐1α (1:1000; Cat. No. ab191838; Abcam, MA, USA), PPAR‐α (1:1000; Cat. No. ab126285; Abcam), acyl‐coenzyme A dehydrogenase medium‐chain (ACADM; 1:1000; Cat. No. A1873; ABclonal, Wuhan, China), acyl‐coenzyme A dehydrogenase long chain (ACADL; 1:1000; Cat. No. A1266; ABclonal), NADH dehydrogenase (ubiquinone) 1 beta subcomplex 8 (NDUFB8; 1:1000; Cat. No. 14794‐1‐AP; Proteintech), succinate dehydrogenase complex subunit B (SDHB; 1:5000; Cat. No. 10620‐1‐AP; Proteintech), ubiquinol‐cytochrome c reductase core protein II (UQCRC2; 1:1000; Cat. No. 14742‐1‐AP; Proteintech), cytochrome c oxidase II (MTCO2; 1:1000; Cat. No. 55070‐1‐AP; Proteintech), ATP synthase, H⁺ transport, mitochondrial F1 complex, alpha subunit 1 (ATP5A1; 1:2000; Cat. No. 14676‐1‐AP; Proteintech), heat shock protein 90 (HSP90; 1:1000; Cat. No. A5027; ABclonal) and NAMPT (1:2000; Cat. No. A0256; ABclonal), CD63 (1:5000; Cat. No. 67605‐1‐Ig; Proteintech).

The TA muscles were dissected and promptly immersed in a fixative solution containing 1% phosphate‐buffered osmium tetroxide and 2.5% glutaraldehyde. Following the embedding process, the muscles were sectioned and stained with uranyl acetate and lead citrate. Subsequently, transmission electron microscope (JEM‐1200EX II, JEOL, Tokyo, Japan) was employed to capture electron photomicrographs illustrating the ultrastructural details of the TA muscle.

Fresh QUAD muscles were used to detect ATP content using the ATP Content Assay Kit (Cat. No. BC0300; Solarbio, Beijing, China) following the manufacturer's instructions. The ATP content of the muscles was normalised to muscle weight.

MSCs/NAD⁺‐MSCs were used to detect intracellular NAD⁺ content and NAD⁺/NADH ratio using the NAD⁺/NADH Assay Kit with WST‐8 (Cat. No. S0175; Beyotime, Shanghai, China) according to the manufacturer's instructions.

The assessment of oxygen consumption rate (OCR) was conducted utilising the Mito Stress Test Kit (Cat. No. 103015‐100; Agilent Technologies, CA, USA) following the manufacturer's instructions. C2C12 myoblasts were seeded at a density of 1 × 10⁴ cells/well into an XF96 cell culture microplate and allowed to fully differentiate. Subsequently, the cultured cells were placed in XF assay medium formulated with Seahorse XF DMEM (pH 7.4), containing 1 mM pyruvate, 10 mM glucose and 2 mM pyruvate glutamine, and treated as indicated. The oligomycin, carbonyl cyanide‐4‐(trifluoromethoxy) phenylhydrazone (FCCP) and antimycin A/rotenone concentrations were 1.5, 1.5 and 0.5 μM, respectively. The OCR was evaluated using a Seahorse XF96 Analyser (Agilent Technologies).

All data were expressed in the form of mean ± SEM Differences between groups were analysed using unpaired Student's t‐test or one‐way analysis of variance (ANOVA), followed by Tukey's test. These statistical analyses were performed using GraphPad Prism 8 software. p < 0.05 was deemed statistically significant.

In this study, we explored the effects of MSCs/NAD⁺‐MSCs on D‐gal–induced skeletal muscle atrophy (Figure S1A). Flow cytometric analysis revealed that the MSCs exhibited positive results for CD105 and CD73 (> 95%) and negative results for CD34 and HLA‐DR (< 2%) (Figure S1B). MSCs have the potential for multi‐directional differentiation in adipogenesis and osteogenesis, as characterised by oil red O and Alizarin red S staining, respectively (Figure S1C,D). NAD⁺ intervention did not change the surface marker characteristics or adipogenic and osteogenic differentiation abilities of MSCs (Figure S1B–D). However, NAD⁺ reduced the mRNA levels of p16, p21, IL‐1β and IL‐6 in MSCs, which indicated alleviation of MSC aging (Figure S1E).

The grip strength and exhaustive running distance tests indicated a decline in muscle strength in the D‐gal + PBS group compared with that in the control + PBS group, which was elevated by MSC injection and further enhanced by NAD⁺‐MSC intervention (Figure 1A). MSCs did not influence body weight (Figure S2A) but increased TA and SO muscle mass, which was even more pronounced in the D‐gal + NAD⁺‐MSCs group (Figures 1B and S2B,C). H&E staining showed lower levels of muscle fibre CSA in the TA muscles of D‐gal + PBS mice compared with that in control + PBS mice, which were all increased by MSC treatment and further elevated by NAD⁺‐MSCs (Figure 1C). Furthermore, MSCs suppressed muscle atrophy‐associated upregulation of the E3‐ubiquitin ligases, Atrogin 1 and MuRF1, which were lower in the D‐gal + NAD⁺‐MSCs group (Figure 1D). Additionally, muscle fibrosis in D‐gal–induced mice was also decreased by MSCs and further ameliorated by NAD⁺‐MSCs (Figure S2D). Meanwhile, in C2C12 myotubes, MSC intervention downregulated D‐gal–induced Atrogin 1/MuRF1 expression and increased the myotube diameter (Figures 1E,F and S2E). NAD⁺ pretreatment further enhanced the effect of MSCs on the muscle atrophy protein Atrogin 1/MuRF1 and myotube diameter (Figures 1E,F and S2F). These results indicate that MSCs improve D‐gal–induced muscle atrophy and NAD⁺ pretreatment can further enhance the therapeutic effect of MSCs.

To investigate the mechanism underlying MSC‐mediated alleviation of muscle atrophy, we conducted RNA‐seq analysis of the QUAD muscle from D‐gal–exposed mice injected with PBS or MSCs. We identified 960 DEGs; 525 genes were upregulated in MSCs versus PBS, and 435 genes were downregulated (Figure S3A). GO analysis suggested that the DEGs upregulated during biological processes were enriched in fatty acid catabolic processes and the respiratory electron transport chain (Figure 2A). KEGG pathway analysis revealed that the upregulated DEGs were enriched in fatty acid degradation, fatty acid metabolism, oxidative phosphorylation and PPAR signalling (Figure 2B). Cluster of orthologous groups of proteins (COG) function classification showed that the upregulated DEGs were mainly related to energy production and lipid metabolism (Figure S3B). Meanwhile, the downregulated DEGs were mainly associated with muscle fibrosis, including complement coagulation cascades, MAPK, PI3K‐AKT and TGF‐β signalling (Figure S3C–E). Furthermore, we performed untargeted metabolomics of the QUAD muscle and found that lipid‐associated metabolites accounted for approximately one‐third of all metabolites (Figure S3F). Compared with the D‐gal + PBS group, 19 metabolites were increased and 54 metabolites were decreased in the D‐gal + MSCs group (Figure S3G), among which there were significant differences in lipid metabolites (Figure S3H). These results suggest that MSCs mainly regulate mitochondrial function and fatty acid metabolism in muscles.

SIRT1 plays a vital role in aging, mitochondrial function regulation and fatty acid oxidation [14]. Therefore, we verified the SIRT1 pathway and related indicators of mitochondrial function and fatty acid oxidation in vivo. Lower SIRT1 and downstream PGC‐1α levels were detected in the muscles of D‐gal + PBS mice than those in control + PBS mice, whereas MSC injection upregulated SIRT1/PGC‐1α levels (Figure 2C). Meanwhile, MSC intervention improved the mitochondrial structure of muscle tissue, including alleviation of swollen mitochondria and destroyed mitochondrial cristae (Figure S4A), promoted the expression of mitochondrial complexes NDUFB8, SDHB, UQCRC2, MTCO2 and ATP5A1 (Figure 2D) and increased ATP content (Figure 2E). These indicators were further improved in the NAD⁺‐MSCs group (Figure 2C–E, S4A). Moreover, MSCs decreased circulating triglyceride and low‐density lipoprotein levels and increased high‐density lipoprotein levels, while not influencing total cholesterol levels (Figure S4B). Additionally, MSC intervention reduced lipid deposition in muscle (Figure S4C) and promoted the expression of PPAR‐α, ACADM and ACADL, which are related to fatty acid oxidation (Figure 2F). NAD⁺‐MSCs further improved plasma lipid levels and fatty acid oxidation in muscles (Figures 2F and S4B,C). In conclusion, MSCs/NAD⁺‐MSCs upregulated muscle SIRT1 expression and improved mitochondrial function and fatty acid oxidation. These results demonstrate that MSCs activate SIRT1/PGC‐1α signalling pathways, alleviate mitochondrial dysfunction and promote fatty acid oxidation in skeletal muscles, which are enhanced by NAD⁺ pretreatment.

To comprehensively explore the direct cell‐autonomous effects of MSCs on mitochondrial function in muscles, an in vitro C2C12‐differentiated myotube‐based model was utilised. HELFs were used as controls. MSCs increased SIRT1/PGC‐1α levels (Figure 3A) and elevated the expression of mitochondrial complexes, including NDUFB8, SDHB and MTCO2 (Figure 3C), which were further elevated by NAD⁺‐MSCs (Figure 3B,D). Moreover, MSCs/NAD⁺‐MSCs promoted the expression of PPAR‐α, ACADM and ACADL (Figure 3E,F). These results corroborate the direct cell‐autonomous effects of MSCs on myotubes and suggest that MSCs/NAD⁺‐MSCs regulate SIRT1/PGC‐1α signaling and mitochondrial function in vitro.

To ascertain the necessity of SIRT1 in mediating the therapeutic effects of MSCs/NAD⁺‐MSCs, C2C12 myotubes were pretreated with LV‐shRNA targeting SIRT1. Western blotting showed that MSC/NAD⁺‐MSC–mediated upregulation of SIRT1/PGC‐1α signalling and mitochondrial complexes was partially weakened by sh‐SIRT1 (Figure 4A,B). Simultaneously, seahorse analysis showed that SIRT1 knockdown diminished the effects of MSCs/NAD⁺‐MSCs on oxidative phosphorylation (OXPHOS), including increased basal respiration, maximal respiration and spare respiratory capacity (Figure 4C) in myotubes. Meanwhile, MSC/NAD⁺‐MSC–mediated upregulation of PPAR‐α, ACADM and ACADL was impaired by sh‐SIRT1 (Figure 4D). Consequently, sh‐SIRT1 weakened the MSC/NAD⁺‐MSC–dependent decrease in Atrogin 1/MuRF1 levels (Figure 4E) and increase in myotube diameters (Figures 4F and S4D). These results indicate that SIRT1 mediates the beneficial effects of MSCs/NAD⁺‐MSCs in myotubes.

To investigate the mechanism by which NAD⁺ enhances the therapeutic effect of MSCs, we first detected intracellular NAD⁺ levels in MSCs following NAD⁺ treatment. The results showed that NAD⁺ treatment increased intracellular NAD⁺ content and NAD⁺/NADH ratio in MSCs (Figure S5A), which suggested that NAD⁺ supplementation altered cellular NAD⁺ metabolism. Next, we performed RNA‐seq analysis of MSCs with and without NAD⁺ treatment (Figure 5A). GO enrichment analysis showed that the upregulated DEGs were mainly associated with calcium‐dependent exocytosis (Figure 5B), indicating that NAD⁺ treatment affected the paracrine effects of MSCs. Therefore, we conducted a proteomic analysis of culture supernatants derived from MSCs and NAD⁺‐MSCs. The results showed that cytoplasmic proteins exhibited the most obvious changes (Figure S5B). KEGG enrichment showed that differential proteins were enriched in nicotinate and nicotinamide metabolism (Figure 5C), and the level of NAMPT was significantly increased by NAD⁺ (Figure 5D,E). We further verified that the intracellular NAMPT expression of MSCs was upregulated after NAD⁺ treatment (Figure 5F), which laid the foundation for further mechanistic verification using siR‐NAMPT. Additionally, NAD⁺ treatment also upregulated NAMPT levels in EVs derived from MSCs (Figure 5F) without influencing the total protein content of EVs (Figure S5C), indicating that NAD⁺ increased NAMPT concentration in EVs.

To further determine whether NAD⁺ enhances the efficacy of MSCs through NAMPT, we treated MSCs/NAD⁺‐MSC with NAMPT‐targeted siRNA and co‐cultured them with D‐gal–stimulated myotubes. Western blotting showed that siR‐NAMPT effectively reduced the expression of NAMPT in MSCs, and NAD⁺‐mediated upregulation of NAMPT was eliminated by siR‐NAMPT (Figure S5D). After NAMPT knockdown, the improvement effects of MSCs/NAD⁺‐MSCs on SIRT1/PGC‐1α signalling, mitochondrial complex expression, oxidative phosphorylation and fatty acid oxidation were significantly reduced (Figure 6A–D). Consequently, siR‐NAMPT diminished the MSC/NAD⁺‐MSC–dependent decrease in Atrogin 1/MuRF1 levels (Figure 6E) and the increase in myotube diameter (Figures 6F and S5E). These results suggested that NAD⁺ enhanced the therapeutic effect of MSCs on D‐gal–induced muscle atrophy by promoting NAMPT secretion.

The RNA‐seq data can be available in the Gene Expression Omnibus (GEO) datasets (GSE304351↗). Other data generated or analysed during this study are included in this published article.