Restoring Muribaculum intestinale –Derived Butyrate Mitigates Skeletal Muscle Loss in Cancer Cachexia

C2C12 cells were cultured in DMEM with 10% FBS and 1% P/S until confluent and then differentiated into myotubes in DMEM with 2% horse serum for 5–7 days. Colon‐26 carcinoma (C26) cells were cultured in RPMI 1640 medium with 10% FBS and 1% P/S. Lewis lung carcinoma (LLC) cells were cultured in DMEM with the same supplements. After 48 h, C26 and LLC supernatants were collected, filtered and co‐cultured with myotubes. Sodium butyrate (1 mM) was added 24 h before RNA and protein extraction. All cells were maintained at 37°C with 5% CO₂.

Specific pathogen‐free (SPF) male BALB/c and C57BL/6J mice, aged 6–8 weeks, were obtained from Nanjing Junke Biotechnology Co., Ltd. and housed under a 12h light/dark cycle at 22°C–24°C and 40%–60% humidity with free access to food and water. Cancer cachexia was induced by subcutaneous injection of 5 × 10⁵ C26 cells into BALB/c mice and 5 × 10⁵ LLC cells into C57BL/6J mice, respectively [22]. The inulin diet and control diet were prepared by Nanjing Junke Biotechnology Co., Ltd., based on published literature [23]. Details of different animal experiments are described in the Supporting Information.

Total RNA was extracted from frozen muscle tissue or cells using a total RNA extraction reagent. After determining the RNA concentration using a NanoDrop, 1 μg was reverse transcribed into cDNA. SYBR Green qPCR mix was used with β‐actin as an internal reference; primer sequences for all target genes are in Table S2.

Collected mouse faeces were immediately snap‐frozen in liquid nitrogen and preserved at −80°C. Sequencing and analysis were performed by Wuhan MetWare Metabolic Biotechnology Co., Ltd. [24]. Details are described in the Supporting Information.

Mouse faeces were collected and stored as described above. SCFAs content was determined by Wuhan MetWare Metabolic Biotechnology Co., Ltd. using the Agilent 7890B‐7000D GC–MS/MS platform. Details are described in the Supporting. Information

The total RNA extracted from mouse skeletal muscle was subjected to RNA sequencing on the Illumina Novaseq 6000 platform by LC Bio Technology Co., Ltd. Details of the bioinformatics are described in the. Supporting Information

Other methods such as Muribaculum intestinale culture, haematoxylin and eosin (H&E) staining, immunofluorescence staining, protein extraction, Western blotting, ATP content measurement and untargeted metabolomics of muscle are described in the Supporting Information.

The data are presented as the mean ± SEM. Two‐tailed Student's t tests were used for comparisons between two groups. One‐way ANOVA tests were performed to compare multiple groups. The Spearman's correlation analysis was used to compute the association between differentially abundant microbial taxa and differential SCFAs. The analyses were performed using GraphPad Prism 8.0.1 and R 4.3.0. Values of p < 0.05 were considered to be statistically significant and were presented as *p < 0.05, **p < 0.01, ***p < 0.001; ns indicates non‐significant.

To elucidate the impact of gut microbiota on cancer cachexia, faecal samples from two murine models exhibiting cachexia were subjected to comprehensive 16S rRNA gene amplicon sequencing analysis. In the C26 cachexia model, alpha diversity, including Chao1, Shannon, ACE and Simpson indexes, showed no significant variation between control and cachexia groups, implying stability in microbial richness and evenness in cachexia (Figure S1a). In contrast, beta diversity analysis using nonmetric multidimensional scaling (NMDS) based on the amplicon sequence variant (ASV) revealed distinct microbial community clustering between control and cachexia mice (Figure 1a), indicating substantial structural changes. Further, a detailed taxonomic composition analysis revealed significant shifts at the phylum level. There was an increase in the relative abundances of p_Firmicutes and p_Proteobacteria and a decrease in p_Bacteroidota in the cachexia model (Figure 1b). LEfSe analysis revealed differentially abundant microbial taxa in cachexia mice, with enrichment of s_Escherichia_coli, o_unidentified_Clostridia, f_Lachnospiraceae, f_Bacteroidaceae and c_Gammaproteobacteria, and reduction of f_Muribaculaceae, o_Bacteroidales, c_Bacteroidia and s_Muribaculum intestinale (Figure 1c). Tax4Fun2‐based functional predictions from ASV data indicated significant enrichment in cellular processes, environmental information processing and various other biological and disease pathways (Figure S1b). The 16S rRNA gene amplicon sequencing revealed pronounced decreases in Muribaculaceae and Muribaculum intestinale in the cachexia group compared to p_Bacteroidota and o_Bacteroidales (Figure S1c–f). Specifically, Muribaculaceae and Muribaculum intestinale abundances dropped by 76.0% (Figure S1c) and 82.0% (Figure S1d), respectively, while p_Bacteroidota and o_Bacteroidales showed smaller reductions of approximately 35.4% (Figure S1e) and 36.3% (Figure S1f), respectively. Diverse Muribaculaceae species display substantial V4 16S rRNA gene sequence variation (Figure 1d), with Muribaculum intestinale being one of the few identified (Figure 1e). Its reduced abundance in cancer cachexia implicates a significant role for Muribaculaceae, particularly Muribaculum intestinale, in cachexia pathogenesis.

In the LLC cachexia model, alpha diversity indices showed increased microbial diversity (Figure S1g), and beta diversity highlighted intergroup differences (Figure S1h). Phylum‐level analysis revealed increased Firmicutes and a decrease in Bacteroidota (Figure S1i). LEfSe analysis indicated enrichment of g_Bacteroides, s_Bacteroides_sartorii and f_Bacteroidaceae and a reduction in f_Muribaculaceae, s_Faecalibaculum_rodentium, s_Muribaculum intestinale and s_Bifidobacterium_pseudolongum in LLC cachexia mice (Figure S1j). Functional annotation clustering confirmed gene enrichment in key pathways, including those for cellular processes and environmental information processing (Figure S1k). Our results reveal a significant alteration in gut microbiota in mouse models of cachexia, characterized primarily by a substantial decrease in Muribaculaceae and Muribaculum intestinale.

We evaluated the impact of Muribaculum intestinale, a Muribaculaceae family member, on a cachectic mouse model. Anaerobically cultured to a concentration of 10⁹ CFU/mL, the strain was administered at 100 μL via gavage. Treatment with MI significantly improved cachexia, reducing weight loss (Figure S2a), preserving lean mass (Figure 2a), hindlimb muscles (Figure S2b) and quadriceps weight (Figure S2c) and enhancing grip strength (Figure 2b), with no effect on tumor size (Figure S2d).

MI treatment downregulated mRNA (Figure 2c) and protein (Figures 2d and S2e) levels of muscle atrophy markers Atrogin‐1 (Fbxo32 in mRNA) and MuRF1 (Murf1 in mRNA) in cachectic mice. This was paralleled by a reduction in p‐STAT3 (Figures 2d and S2e), suggesting an anti‐inflammatory effect and amelioration of muscle atrophy. H&E staining indicated muscle fibre preservation, with an increased cross‐sectional area (CSA) following treatment (Figures 2e and S2f). 16S rRNA gene amplicon sequencing confirmed elevated Muribaculum intestinale levels post‐administration (Figure 2f).

In the LLC cachexia model, MI treatment led to significant preservation of lean mass (Figure S2g) and hindlimb muscles (Figure S2h), and enhanced grip strength (Figure S2i), with no change in tumour weight (Figure S2j). H&E staining confirmed muscle fibre protection, as evidenced by increased CSA (Figure S2k). qPCR and Western blotting results mirrored those from C26 mice, showing reduced expression of muscle atrophy markers (Figure S2l, S2m). The 16S rRNA gene amplicon sequencing revealed increased faecal Muribaculum intestinale levels post‐treatment in LLC mice (Figure S2n). These findings suggest that Muribaculum intestinale, as a potential probiotic, can positively affect cachectic mouse models.

To investigate the potential consequences of the observed gut dysbiosis, we performed SCFAs analysis on faecal samples from C26 cachexia mice and healthy controls, and observed a significant reduction in SCFAs in the cachectic group: acetic acid (AA) by 52.5%, propionic acid (PA) by 60.9%, and valeric acid (VA) by 51.1%. Notably, butyric acid (BA) showed the most substantial decrease, at 65.3%. Levels of isovaleric acid (IVA), isobutyric acid (IBA), caproic acid (CA) and 2‐methylbutyric acid (2‐BA) remained unchanged (Figures 3a and S3a). Muscle tissue metabolomics revealed a clear separation between the NC and C26 cachexia groups in principal component analysis (PCA) (Figure S3b, S3c) and distinct clustering in metabolite heatmaps (Figure S3d,S3e), accompanied by downregulation in butanoate metabolism pathways (Figure 3b), suggesting a systemic impact of altered SCFA profiles. This decrease in butanoate metabolism suggests its critical role in cachexia, emphasizing its importance in the condition's pathophysiology.

To determine the correlation between microbial taxa shifts and metabolite profiles, we performed a Spearman correlation hierarchical clustering analysis. This analysis disclosed significant associations between gut microbiota at various taxonomic levels and SCFAs. At the phylum level, significant reductions in Bacteroidota, Euryarchaeota, Altiarchaeota and Crenarchaeota were positively correlated with decreases in specific SCFAs. In contrast, Myxococcota and Firmicutes showed negative correlations with SCFA levels (Figure S3f). At the class and order levels, we highlighted taxa such as Thermoplasmata, Limnochordia and Thermococci, which exhibited decreased abundance and positive correlations with certain SCFAs. Conversely, the abundance of classes such as Clostridia and Negativicutes increased, showing negative correlations with SCFA levels (Figure S3g). Similarly, at the order level, Eubacteriales and Bacteroidales significantly decreased and positively correlated with several SCFAs, while unidentified taxa within the unidentified Clostridia and Kapabacteria orders were negatively correlated with these metabolites (Figure S3h). At the family level, a significant decrease in Muribaculaceae, along with reduced abundances of unidentified Eubacteriales and Lactobacillaceae, was positively correlated with several differential SCFAs, particularly butyric acid (Figure 3c). Increases in families such as Chthonomonadaceae, Sphingobacteriaceae, Lachnospiraceae and Oscillospiraceae were associated with negative correlations to SCFA levels (Figure 3c). At the genus and species levels, decreases in taxa such as Anaerofustis sp. Marseille‐P2832, Muribaculum intestinale and Christensenella minuta were significantly correlated with declines in several differential SCFAs. In contrast, increases in unidentified Lachnospiraceae, mouse gut metagenome, Acutalibacter muris and Pasteurellaceae bacterium were negatively correlated with these metabolites (Figure 3d). These findings indicate a profound disruption in both the composition and functional capacity of the gut microbiota in cachexia, characterized by a marked decline in SCFA levels, particularly BA. The strong correlation between Muribaculaceae and BA levels, as well as between Muribaculum intestinale and BA levels highlights the significance of these bacteria in SCFA production. Interestingly, MI supplementation in two cachectic mouse models led to significant changes in faecal SCFA profiles, particularly an increase in butyrate levels (Figure 3e, 3f). These results suggest that Muribaculum intestinale may be pivotal in modulating the host butyrate concentrations.

Given the depletion of SCFAs, especially BA, in cachexia, we explored the therapeutic effects of sodium butyrate (NaB) supplementation in the C26 cachexia model (Figure 4a). Treatment resulted in amelioration of the cachexia phenotype, including reduced body weight loss (Figure S4a). Although on Day 13, there was an upward trend in body weight in the CNaB group of mice, no significant difference was observed compared to Day 8 (Figure S4b) or the NaB control group on Day 13 (Figure S4c). Lean body mass and hindlimb muscle were preserved (Figures 4b and S4d), despite a reduction in tumour weight (Figure S4e). Molecular analyses demonstrated reduced mRNA levels of atrophy markers Fbxo32 and Murf1 in NaB‐treated cachectic mice (Figure 4c). Western blotting confirmed decreased protein expression of the cachexia‐specific markers Atrogin‐1 and MuRF1 (Figures 4d and S4f), suggesting effective alleviation of muscle atrophy. H&E staining showed increased CSA of muscle fibres and adipose tissues in NaB‐treated cachectic mice, indicating a reduction in tissue degradation (Figures 4e and S4g). In the LLC cachexia mouse model, NaB supplementation increased the muscle fibre CSA (Figure S4h) and suppressed muscle atrophy markers (Figure S4i, S4j). Unlike the C26 model, tumour weight remained unchanged in the LLC model with NaB treatment (Figure S4k, S4l), suggesting that its effects on tumorigenesis may vary between cachexia models.

We also validated this in the C2C12 myotube. NaB treatment decreased atrophy markers (Figure 5a–5c) and protected myotubes from C26 conditional medium, as shown by immunofluorescence (Figure 5d). The protective effect was consistent when myotubes were treated with LLC conditional medium, showing a notable decrease in Murf1 expression (Figure 5e) and protection against myotube shrinkage (Figure 5f). The findings show that NaB not only reduces cachexia‐related muscle and fat loss in vivo, but also directly affects muscle atrophy pathways.

To clarify NaB's role in alleviating cachexia‐associated muscle atrophy, we performed RNA‐Seq on quadriceps muscle. PCA showed the NaB‐treated group's gene expression was closer to normal controls than the untreated group (Figure S5a). The heatmap revealed significant gene expression differences among groups, indicating substantial transcriptomic changes from NaB (Figure S5b). NaB downregulated key atrophy‐related genes like Trim63, Fbxo32, Acot1, Acot2, Pdk4 and Il6ra (Figure 6a), which are typically upregulated in cachectic muscle (Figure S5c), suggesting a direct impact on muscle preservation. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis showed that in the cachexia group, pathways such as mitophagy and FoxO signalling were upregulated, while glycolysis and glycine, serine and threonine metabolism were downregulated (Figure S5d). Remarkably, NaB reversed these trends by reducing activity in autophagy‐related pathways and enhancing pathways associated with arginine and proline metabolism, as well as glycosaminoglycan biosynthesis (Figure 6b). Gene ontology (GO) enrichment analysis revealed significant alterations in biological processes, including the regulation of nervous system development, axon growth and wound response, underscoring the extensive systemic impact of NaB (Figure S5e). Regarding cellular components and molecular functions, there was notable enrichment in extracellular matrix structures and calcium ion binding activities, among other aspects (Figure S5f,S5g). Gene set enrichment analysis (GSEA) showed that in cachexia, TNF‐α, inflammatory and JAK‐STAT3 pathways were upregulated, and myogenesis pathways were downregulated (Figure S5h). NaB significantly suppressed TNF‐α and inflammatory pathways, mildly affected the JAK‐STAT3 pathway, and enhanced muscle generation pathways (Figure 6c). It also activated oxidative phosphorylation (Figure 6d). These findings were associated with coordinated transcriptional changes, including the downregulation of Pdk4 (Figure 6e) and autophagy‐related genes (Figures 6f and S5i), collectively highlighting NaB's multifaceted modulation of inflammatory signalling and energy metabolic rewiring in cachectic muscle.

To validate key findings, we investigated NaB's effects on inflammation and autophagy signalling, both implicated in cachexia‐induced muscle wasting. Western blotting results revealed a significant reduction in p‐STAT3 levels in cachectic mice's skeletal muscles and myotubes treated with NaB (Figure 7a). qPCR analysis revealed that NaB reduced the elevation of Il6 mRNA levels in myotubes induced by C26 and LLC conditional medium (Figure 7b). The multiplex assay kit assessed plasma inflammatory cytokines in both cachexia mouse models. In the C26 model, NaB significantly reduced TNF‐α, IL‐6, IL‐1β and MCP‐1 (Figure S6a–S6d). In the LLC model, IL‐6 and MCP‐1 decreased, but TNF‐α and IL‐1β showed no significant change (Figure S6e–S6h). These findings indicate that NaB can mitigate cachexia‐associated systemic inflammation.

Further qPCR analysis showed that NaB treatment significantly downregulated mRNA levels of autophagy and mitophagy‐specific genes, indicating reduced catabolic activity (Figure 7c). Western blotting confirmed this finding, showing a decreased LC3 II/I ratio and reduced autophagy activity in muscle tissues (Figure 7d). The qPCR results showed a significant decrease in Pdk4, which is involved in inhibiting oxidative phosphorylation (Figures 7e and S6i). Interestingly, in the previous experiment, MI supplementation led to increased butyrate levels (Figure 3e,3f), and qPCR analysis of muscle samples from cachectic mice supplemented with MI showed significant suppression of Pdk4 expression (Figure S6j), similar to direct NaB supplementation. Increased ATP content confirmed the enhancement of oxidative phosphorylation in muscles following NaB treatment (Figure 7f).

Regarding clinical application potential, inulin might have an advantage in safety, transport and patient agreement. Thus, we employed an inulin fructan‐enriched prebiotic diet that is known to increase the abundance of Muribaculaceae [15, 23]. To expand the protective effect against cancer cachexia, we administered an inulin fructan‐enriched diet to the cachexia model mice (Figure S7a). 16S rRNA gene amplicon sequencing revealed that inulin supplementation improved cachexia‐associated microbial dysbiosis (Figure S7b). At the phylum level, Firmicutes in cachectic mice decreased following inulin intervention, while the proportion of Bacteroidota increased (Figure S7c). At the class level, Bacteroidia and Clostridia showed an upward trend (Figure S7d). At the order level, Bacteroidales and Lachnospirales also increased (Figure S7e). At the family level, Muribaculaceae increased significantly (Figures S7f and 8a). At the species level, Muribaculum intestinale increased significantly (Figure 8b). These changes suggest that inulin may improve gut microbiota balance in cachexia, benefiting intestinal health.

The prebiotic intervention notably improved the cachexia phenotype, including reduced body weight loss (Figure S7g), preserving lean mass (Figure 8c) and quadriceps muscles (Figure S7h) and hindlimb muscle (Figure S7l). Importantly, these benefits occurred without altering tumour weight (Figure S7i), adipose tissue mass (Figure S7j, S7l) and heart weight (Figure S7k), possibly indicating muscle‐specific effects. H&E staining showed increased muscle fibre CSA in the quadriceps (Figure 8d), with no significant change in adipose tissue CSA (Figure S7m), indicating an effective counteraction of muscle atrophy by the inulin diet. Then qPCR of muscle RNA showed reduced expression of Fbxo32 and Murf1 in the cachexia model mice fed the inulin diet (Figure 8e). Western blotting confirmed decreased protein levels of Atrogin‐1 and MuRF1, and a reduced LC3II/I ratio (Figure 8f). These results indicate that inulin effectively normalized gut microbiota composition, enhanced Muribaculaceae and Muribaculum intestinale abundance and significantly improved key cachexia markers.