Oxaliplatin triggers adipose and muscle wasting through cancer-independent metabolic and CNS pathways in mouse models Authors: Junwei Du, Rachel L Mintz, Leland C Sudlow, Rachael L Field, Christine Stander, Jonathan R Brestoff, Gwendalyn J Randolph, Mikhail Y Berezin Published: 2026 ABSTRACT AbstractOxaliplatin, a platinum-based chemotherapeutic, is a first-line treatment for colorectal and other cancers frequently associated with cachexia. The extent to which oxaliplatin induces cachexia independently of cancer and the mechanisms involved remain unclear. Here we show that treatment with a human-equivalent dosage of oxaliplatin leads to cachexia-like symptoms in mice commonly observed in cancer patients, including severe loss of body mass resulting from adipose tissue depletion and wasting of skeletal muscle. The mice experience alterations in whole-body metabolism, including decreased food intake, reduced ambulatory activity, lowered core body temperature, and altered respiratory exchange ratio. Histological analyses demonstrate marked muscle fiber atrophy and increased immune cell infiltration. Transcriptomics analyses performed on subcutaneous adipose tissue, skeletal muscle, and the hypothalamus identify metabolic rewiring as a dominant response associated with chemotherapy-induced cachexia. RNA-seq demonstrates signaling pathways associated with these processes across muscle, adipose, and hypothalamus. Specifically, adipokine genes are dysregulated in white adipose tissue and in the hypothalamus, while genes involved in inflammatory pathways are upregulated in muscle. Subject terms: Chemotherapy, Chemotherapy Introduction Cachexia is highly prevalent across cancer types, affecting as high as 87% in patients with pancreatic and gastric cancer, 61% in patients with colon, lung, prostate cancer, and non-Hodgkin lymphoma, and 40% in patients with breast cancer, sarcoma, leukemia, and Hodgkin lymphoma1 and contributing to approximately 20% of cancer-associated deaths2. It significantly reduces tolerance to treatment, and thus, suboptimal drug doses are often delivered, resulting in diminished therapeutic efficacy and lower survival rates3,4. Despite its clinical importance, the biological basis of cancer-associated cachexia remains only partly understood, and current care is largely limited to palliative care rather than disease-modifying interventions5. Emerging clinical observations suggest that cachexia is not entirely tumor-driven but can be aggravated or even initiated by chemotherapy itself. Platinum-based agents, including oxaliplatin, for example, have been linked to weight loss, muscle atrophy, metabolic changes, and systemic inflammation in patients6. Oxaliplatin, a widely used third-generation platinum compound for the treatment of colorectal and other gastrointestinal cancers, has received attention due to its association with significant off-target effects such as chemotherapy-induced peripheral neuropathy (CIPN)7,8. However, its potential role in directly driving cachexia, independent of tumor burden, and the mechanism of this process, remains largely unexplored. In our prior study on oxaliplatin-induced CIPN, we observed a significant loss of body weight in mice treated with oxaliplatin9, prompting further investigation into its role in driving cachexia. Several groups, including ours, have reported the activation of the inflammatory pathways in peripheral nerves and dorsal root ganglia in CIPN9,10. We therefore propose that a similar mechanism may underlie toxicity in other organs11 collectively driving the patient to cachexia. Oxaliplatin-induced cachexia has rarely been studied separately from cancer, and the actual literature on this topic is scarce12–14. Cachexia is typically attributed to the combined effects of cancer and its treatments15, and oxaliplatin is most often studied within multi-agent regimens, making it difficult to isolate drug-specific host response16,17. Moreover, existing studies largely focus on short-term exposures to oxaliplatin18,19 limiting the insight into chronic effects. Distinguishing chemotherapy-induced cachexia from tumor-driven cachexia is critical because the underlying pathways may differ and might call for different therapeutic strategies. Here, we investigate the effects of chronic oxaliplatin administration in a non–tumor-bearing mouse model to isolate drug-specific contributions. Using a longitudinal design, we examine changes in body composition, metabolism, and tissue-specific gene expression. RNA sequencing of skeletal muscle, white adipose tissue (WAT), and hypothalamus identified systemic and tissue-level responses. This approach, together with bioinformatic analysis, allowed us to identify candidate mechanisms and possible biomarkers of oxaliplatin-related cachexia. Results Oxaliplatin results in severe weight loss, reduced activity, and mortality in mice. To provide a physiological context, locomotor activity was recorded continuously, metabolic and functional parameters were assessed at both week 4 and week 8, whereas transcriptomic and histological analyses were performed at the experimental endpoint (week 8), when the cachexia-associated phenotype became detectable. With respect to treatment duration, we incorporated both mid-treatment (week 4) and late-treatment (week 8) time points throughout the study, allowing assessment of early versus progressive effects. The timeline of oxaliplatin treatment is shown in Fig. 1A. Male and female mice treated weekly with 10 mg/kg oxaliplatin exhibited significant weight loss beginning at week 2 compared with controls treated with 5% dextrose (Fig. 1B). By the end of treatment, oxaliplatin-treated mice lost ~12% of their weight, while controls gained ~11%. There was no difference in tibial length between the two groups, suggesting the oxaliplatin’s effects on body weight were not due to growth retardation (Supplementary Fig. S1A). No weight loss occurred at the subclinical 5 mg/kg dose, equivalent to half the human-equivalent dose (Fig. S1B). By week 8, mice given 5 mg/kg accumulated the same total drug as those treated with 10 mg/kg by week 4, but without clinical manifestations. Further experiments incorporated the 10 mg/kg dose. Oxaliplatin at a human-equivalent dosage (10 mg/kg) reduces body weight and activity. A Schematic timeline of the study (Created in BioRender. Berezin, M. (2026) https://BioRender.com/x55iq39). Mice received weekly i.p. injections of oxaliplatin for 8 weeks. Body composition and metabolic measurements were performed at baseline, mid-treatment, and at the end of treatment period. Voluntary wheel running was monitored continuously. B Body weights of male and female mice were significantly lower in the oxaliplatin-treated group compared with controls. n = 10 male in control group, n = 10 male oxaliplatin group, n = 8 female control group, n = 16 female oxaliplatin group. C Representative wheel-running activity shown as the number of revolutions per six-minute interval in control and oxaliplatin-treated mice daily, and weekly (Fig. S2A–B). The calculated daily revolutions and distance are provided in Fig. S2A–B. Oxaliplatin injections produced an acute, sharp reduction in running activity, followed by partial recovery over subsequent days. Cumulative revolutions per week. Overall activity levels remained lower in the oxaliplatin group. D Hourly and average ambulatory activity at week 4 (n = 8/group, female) and (E) week 8 (n = 8/group, female). Oxaliplatin-treated mice demonstrated reduced activity during both the light and dark phases. F Water and food intake during the dark phase remained unchanged at week 4 between control (n = 8) and oxaliplatin treated mice (n = 8) but decreased significantly at week 8 in oxaliplatin-treated mice (n = 8). G Water and food intake decreased significantly at week 8 in oxaliplatin-treated mice (n = 8) compared to control mice (n = 8). H Hydration ratio was significantly elevated in oxaliplatin-treated mice compared with controls at both weeks 4 (n = 8/group, female) and 8 (n = 13/group, 5 male and 8 female). *p < 0.05, **p < 0.01, ***p < 0.001. All data are presented as mean ± SD. Critically, mice continued losing weight after the final injection (Fig. S1C), reaching a refractory, irreversible cachexia stage ( > 5% unintended weight loss over six months clinically)20. The overall survival rate was lower than 50% at 12 days post the final injection (Fig. S1D), calculated based on mice that died before the study endpoint or qualified for euthanasia (loss of ~20% of initial body weight). Each 10 mg/kg oxaliplatin injection led to an immediate and acute decrease in voluntary wheel running activity (Fig. 1C), which only partially recovered post-injection but never reached control levels. Metabolic cage studies confirmed the reduced locomotor and ambulatory activity (Fig. 1D). Figure S1E represents the wheel-free movement of a single mouse over 24 h. Notably, the reduction in both locomotor and ambulatory activities was primarily observed during the dark periods, consistent with mice’s nocturnal nature. Control mice displayed sustained levels of wheel-running activity, covering approximately 15–27 km per week ( ~ 2–4 km/day) (Fig. S2). In contrast, oxaliplatin-treated mice showed a pronounced and persistent reduction in voluntary locomotion, running approximately 7–13 km per week ( ~ 1–1.4 km/day), which constitutes approximately 40–60% of that in controls, indicating a substantial impairment in voluntary physical activity. Food and water intake were initially unaffected by oxaliplatin treatment up to week 4, with a significant reduction by week 8, particularly during the dark phase (Fig. 1F–G). Interestingly, hydration ratios were significantly higher in oxaliplatin-treated mice than in the control group at weeks 4 and 8, independent of changes in water intake (Fig. 1H). Oxaliplatin alters mouse whole-body metabolism Oxaliplatin’s metabolic effects were assessed via core body temperature assessments and metabolic cages (Figs. S3–S4). At baseline, mice maintained a body temperature of 37.95 ± 0.41 °C. By week 4, animals treated weekly with 10 mg/kg oxaliplatin exhibited a significant reduction in core temperature (Fig. S3A), which further declined to 33.35 ± 1.74 °C by week 8 (Fig. S3D), indicating sustained hypothermia and impaired thermoregulatory control. Energy expenditure, assessed as heat generation, remained unchanged at week 4 (Fig. S3B) but was markedly reduced at week 8 during the dark (active) phase (Fig. S3E). Because mice normally exhibit their highest locomotor and thermogenic activity during the dark cycle, this phase-specific decline suggests a progressive deterioration in metabolic functions with chronic oxaliplatin exposure. Oxaliplatin also disrupted substrate utilization patterns at week 8, but not week 4, as reflected in altered respiratory exchange ratio (RER) values (Fig. S3C, S3F). Control mice showed normal circadian RER patterns: 0.78 ± 0.03 during the light phase (fat oxidation) and 0.84 ± 0.04 during the dark phase (carbohydrate utilization), while treated mice exhibited an altered, circadian-lacking RER of ~0.85 for both light and dark phases, suggesting a shift towards protein catabolism for ATP generation. Overall, these data reveal a late-onset collapse in thermoregulation, energy expenditure, and fuel-selection mechanisms following chronic oxaliplatin treatment. Oxygen consumption data used to compute heat production are shown in Fig. S4. Oxaliplatin induces muscle loss, fiber damage, and inflammation Muscle wasting, a hallmark of cachexia2, was assessed over the course of oxaliplatin treatment. Lean mass was monitored in female and male mice using EchoMRI at baseline, mid-treatment, and post-treatment. In female mice, oxaliplatin induced a progressive decline in lean mass beginning after the fourth injection, ultimately resulting in ~14% loss of lean mass and a corresponding ~12% reduction in total body weight by week 8 (Fig. 2A). Male mice displayed a similar trajectory, with significant reductions in lean mass evident by week 8 (Fig. 2B). These findings demonstrate that chronic oxaliplatin administration leads to substantial muscle wasting in both sexes. Oxaliplatin induces muscle loss and structural fiber damage as signs of atrophy. A Longitudinal changes in lean body mass in female mice over 8 weeks of treatment. Significant reductions were observed after four doses of oxaliplatin at week 4 (n = 8/group). B Lean mass of male (n = 5/group) and female (n = 8/group) mice at week 8, showing decreased muscle mass in both sexes following oxaliplatin administration. Ex vivo weights of tibialis anterior (n = 5/group, male) (C) and gastrocnemius (n = 8 control mice, n = 6 oxaliplatin mice, male) (D) muscles. Both muscle groups were significantly reduced in oxaliplatin-treated mice compared with controls. E H&E staining of the gastrocnemius muscle demonstrates fiber degradation (blue arrows). Images were acquired using a 10× objective with a 2× adapter; scale bar 50 μm. F Muscle fiber cross-sectional area in the gastrocnemius muscle is significantly reduced in oxaliplatin-treated mice compared to controls. Points represent individual mice; bars indicate mean ± SD. Control n = 4, oxaliplatin n = 5. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. All data are presented as mean ± SD. Endpoint measurements of hindlimb muscle groups revealed selective muscle loss following oxaliplatin treatment. The tibialis anterior (TA) and gastrocnemius muscles showed significant reductions in mass in treated mice compared with controls (Fig. 2C–D), while the soleus muscle remained unaffected (Fig. S5A). Histological analysis of the gastrocnemius sections stained with H&E showed prominent evidence of round and polygonal muscle fiber atrophy (Fig. 2E) consistent with prior studies showing chemotherapy-induced muscle atrophy in non-tumor-bearing animals21,22. Analysis of muscle fiber cross-sectional area demonstrated a marked decrease in oxaliplatin-treated mice (700 ± 167 µm2) relative to controls (1425 ± 110 µm²; p = 0.0007 (Fig. 2F), consistent with muscle atrophy. In addition to signs of fiber degradation, inflammatory cell infiltration was also observed within the muscle tissue (Fig. 3A), suggesting an accompanying inflammatory response. To confirm the inflammatory response, gastrocnemius muscle sections were stained for CD45, a pan-leukocyte marker. Oxaliplatin-treated mice exhibited a marked increase in CD45-positive cells within the muscle, primarily localized to interstitial regions and surrounding muscle fibers. In contrast, control (dextrose-treated) mice showed minimal CD45 staining, indicating low baseline immune cell presence (Fig. 3B–C). These findings suggest that chronic oxaliplatin treatment promotes infiltration of immune cells into skeletal muscle, consistent with an inflammatory component of the observed muscle pathology. Despite these structural abnormalities, functional assessment using grip strength tests at week 8 showed no statistically significant differences between oxaliplatin treated and control groups (Fig. S5B), indicating that muscle function, as measured by this test, remained preserved at this time point. Oxaliplatin induces an inflammatory response. A H&E staining of tibialis anterior muscle showing infiltration of inflammatory cells in oxaliplatin-treated mice (blue arrows); scale bar 50 μm. The scale bar in the insert is 10 μm. B Representative images of CD45 immunohistochemistry in gastrocnemius muscle sections from mice treated with vehicle (dextrose) and oxaliplatin. CD45-positive cells are visualized as brown staining (DAB), and nuclei are counterstained with hematoxylin (blue). Oxaliplatin-treated muscle shows increased CD45-positive cell infiltration within the interstitial spaces and between muscle fibers compared to minimal staining in control tissue. The insert shows the zoomed version of the tissue. C Quantification of CD45-positive tissue based on the percentage of area stained. Control sections n = 14, oxaliplatin sections n = 15. ***p < 0.001. All data are presented as mean ± SD. Oxaliplatin treatment decreased body fat Cachexia is characterized not only by muscle wasting but also by pronounced reductions in adipose tissue1. To evaluate the effects of oxaliplatin on fat stores, we monitored body fat composition longitudinally using EchoMRI and quantified white and brown adipose depots at the study endpoint. In female mice, oxaliplatin treatment produced a progressive decline in both body fat composition and adiposity (fat mass relative to body weight), with total body fat significantly reduced after four injections (4 weeks, 40 mg cumulative dose) and adiposity diverging further by the eighth treatment (8 weeks, 80 mg cumulative dose) (Fig. 4A). In contrast, body fat composition in control mice remained stable throughout the experiment. Male mice displayed a similar reduction in fat mass and adiposity over the eight-week treatment period (Fig. 4B), indicating that oxaliplatin-induced adipose loss occurs independently of sex. Oxaliplatin induces marked depletion of adipose tissue. A Longitudinal changes in body fat composition and adiposity in female mice over 8 weeks of treatment (n = 8 per group). Total body fat was significantly reduced after four injections of oxaliplatin (40 mg cumulative dose), with adiposity differences reaching significance by week 8. B Body fat composition and adiposity in male n = 5 per group and female n = 8 per group mice at week 8. Oxaliplatin reduced fat mass and adiposity in both sexes. Ex vivo weights of visceral fat depots. Omental (n = 11 control, n = 10 oxaliplatin, male) (C) and mesenteric (n = 4/group, male) (D) fat weights were significantly decreased in oxaliplatin-treated mice compared with controls. E H&E staining of mesenteric adipose tissue. Tissue from oxaliplatin-treated mice shows a marked loss of white adipocytes. Magnification: 10× objective with 2× adapter.; scale bar: 50 μm. *p < 0.05, **p < 0.01, ****p < 0.0001. All data are presented as mean ± SD. Ex vivo analysis of adipose depots revealed significant reductions in visceral omental and mesenteric fat weights in treated mice (Fig. 4C–D, Fig. S6A). Brown adipose tissue mass was also decreased, though less notably than white fat depots (Fig. S6B). Histological evaluation via H&E staining of mesenteric fat showed a loss of white adipocytes in oxaliplatin-treated mice (Fig. 4E), with comparable adipocyte loss observed in the omentum (Fig. S6C). These results indicate that chronic oxaliplatin exposure leads to substantial loss of both white and brown adipose tissues, consistent with a cachexia-like phenotype. Oxaliplatin leads to transcriptional changes associated with inflammation in muscle Oxaliplatin treatment induced robust transcriptional alterations in the tibialis anterior (TA) muscle. RNA-seq identified 245 differentially expressed genes (DEGs) out of 12,327 detected transcripts, with 194 genes exhibiting an absolute fold change greater than 1.5 ( | FC | > 1.5) (Fig. 5A). Unsupervised clustering of these DEGs, as shown in a heatmap, demonstrated clear separation between control and oxaliplatin-treated mice (Fig. S7A). Gene Ontology analysis of upregulated genes revealed significant enrichment of pathways associated with hypoxia (Fig. 5B). Transcriptional profiling of tibialis anterior (TA) muscle following oxaliplatin treatment. A Volcano plot showing 194 differentially expressed genes (DEGs), including 115 upregulated and 79 downregulated transcripts (FDR < 0.05, |FC | > 1.5). B Gene Ontology (GO) biological process enrichment analysis (STRING) of upregulated DEGs (FC > 1.5). The most significantly enriched pathways were related to vascular responses and hypoxia-associated processes. C STRING interaction network of DEGs strongly correlated with TA muscle weight. Angpt2, Ccl2, Icam1, and Pgf emerged as central nodes within the network. Only connected genes are displayed. Confidence score = 0.4. RNA sample size that passed quality control and used for sequencing: n = 4 for the control group, n = 5 for the oxaliplatin group, male. To determine whether specific transcriptional changes were linked to muscle loss, we performed a global correlation analysis between DEG expression levels and TA muscle weights using IVCAA23. This approach correlates gene counts with muscle mass on a per-animal basis. DEGs showing strong associations with muscle weight ( | r | > 0.75) are listed in Supplementary Data 1. STRING network analysis of these highly correlated genes (Fig. 5C) highlighted Angpt2 (FC = +2.47), Ccl2 (+4.99), Icam1 (+3.05), and Pgf (+2.89) as central nodes. These genes are well-known mediators of inflammatory and hypoxia-related responses and were negatively correlated with TA muscle weight, establishing a direct link between inflammation, hypoxic signaling, and oxaliplatin-induced muscle atrophy. A broader STRING analysis incorporating all 194 DEGs revealed several major interaction clusters (Fig. S7B–C). The largest cluster (Cluster #1) consisted predominantly of genes associated with inflammation and muscle atrophy. Key nodal genes included Ccl2, Icam1, and as well as Pecam1 (FC = +1.6). Ccl2 encodes a chemokine that promotes monocyte and macrophage recruitment and is commonly upregulated in inflamed tissues24. Icam1 encodes a protein that facilitates the movement of immune cells into damaged tissues25, and its overexpression suggests increased leukocyte, neutrophil, and monocyte infiltration in muscle. Pecam1 is critical for leukocyte transendothelial migration26. Consistent with these transcriptional signatures, histological analysis of TA muscle demonstrated prominent inflammatory cell infiltration, particularly around damaged fibers (Fig. 3A). Not surprisingly, Ccl2 expression correlates with both Icam1 and Pecam1 expression (r = 0.88 and 0.91, respectively), confirming their coordinated involvement in the inflammatory response. Cluster #1 also includes genes linked to increased nitrogen metabolism (Car9, FC = +2.54; Car14, + 2.22), reduced ability to remodel extracellular matrix (Mmp9, −8.77, Prtn3, −112.0), and enhanced cellular response to hypoxia (Egln3, + 4.49), further supporting a pattern of tissue remodeling and metabolic stress. Additionally, upregulation of Ifrd1 (FC = +2.0), a gene previously implicated in multiple muscle-wasting conditions27 provides genetic evidence consistent with ongoing muscle atrophy. Collectively, the tight connection between these genes demonstrates a functional relationship and shared biological pathways in inflammation, immune response, and catabolism that converge to drive oxaliplatin -induced muscle degeneration. Oxaliplatin disrupts adipogenesis, lipogenesis, and ATP production in fat RNA-seq analysis of white adipose tissue (WAT) also revealed substantial transcriptional alterations, with more than 2000 genes exhibiting absolute fold changes greater than 1.5 (Fig. 6A). Similar to skeletal muscle, the heatmap shows a clear treatment-dependent separation of DEGs (Fig. S7D). KEGG pathway analysis demonstrated that oxaliplatin profoundly affects metabolic and thermogenic pathways in WAT, consistent with the metabolic cage findings (Fig. 6B). Additional highly enriched pathways included oxidative stress and hypoxia, indicating widespread disruption of adipose tissue homeostasis. Transcriptional profiling of WAT, muscle, and hypothalamus in mice treated with oxaliplatin. A Volcano plot showing 2,094 differentially expressed genes (DEGs) in WAT (FDR < 0.05, |FC | > 1.5). B KEGG pathway enrichment analysis of DEGs in WAT with an absolute fold change greater than 1.5. C Area plot illustrating the relationship between fold change and the cumulative number of DEGs (FDR < 0.05, no FC threshold) in WAT (2,628 DEGs) vs TA muscle (245 DEGs) after oxaliplatin treatment. Each point represents a gene ranked from highest to lowest fold change. The plot highlights the substantially greater transcriptional impact of oxaliplatin on WAT compared with muscle. D Oxaliplatin induces dysregulation of adipokines and hypothalamic neuropeptides. Lep and Nrg4 expressions were strongly reduced in WAT (Lep: FC = ‒8.54, FDR = 0.13, Nrg4: FC = ‒3.0, FDR = 0.025), Gdf15 was markedly increased (FC = +7.99, FDR = 3.39 × 10⁻⁹). In the hypothalamus, Npy expression was upregulated following oxaliplatin treatment (FC = +5.88, FDR = 2.4 × 10⁻³). n = 3 control, n = 4 oxaliplatin, male. Interestingly, a greater number of DEGs with larger fold changes were identified in WAT compared to muscle (Fig. 6C), suggesting that oxaliplatin exerts an even stronger transcriptional impact on adipose tissue. In cancer patients, muscle and weight loss are often accompanied by increased lipolysis and reduced adipogenesis and lipogenesis28. Consistent with this pattern, several key regulators of adipocyte differentiation and lipid synthesis were downregulated in WAT. The mRNA levels of C/EBPα (FC = ‒2.11), Acly (‒2.21), and Fasn (‒4.95) were markedly reduced. Additional genes essential for adipocyte maturation and lipid metabolic pathways also showed suppression, including Srebf1 (FC = ‒1.97, FDR = 0.06) and Scd (FC = ‒10.65, FDR = 0.08), with other Scd family members (Scd2, Scd3, and Scd4) decreased as well. Furthermore, reduced expression of Lipe (FC = ‒1.92), a key triglyceride hydrolase29, suggests impaired generation of fatty acids required for adipose storage and ATP production. Parallel to the muscle and fat loss observed in cancer cachexia, ATP depletion is a well-recognized consequence of impaired mitochondrial function and limited availability of metabolic substrates30. We observed a similar pattern in oxaliplatin-treated mice. Key genes involved in fatty acid oxidation were markedly reduced, including Cpt1b (FC = ‒4.6) and Decr1 (FC = ‒1.9). Genes encoding components of the mitochondrial respiratory chain were also significantly downregulated. For example, Hadh, a critical enzyme in mitochondrial β-oxidation, was reduced 2.3-fold, and Mtln, a regulator of respiratory chain complex I activity31, decreased by 2-fold. Numerous genes associated with the mitochondrial proton-transporting ATP synthase complex and other ATP-related pathways were broadly suppressed (Supplementary Data 2). Collectively, these findings indicate that oxaliplatin treatment disrupts oxidative phosphorylation and ATP production in WAT, mirroring the metabolic deterioration typically seen in cancer-associated cachexia. Adipokine signaling is dysregulated with oxaliplatin treatment Oxaliplatin strongly affected the expression of multiple adipokines, signaling proteins secreted by adipocytes that regulate metabolism, inflammation, and immune function32. Notably, leptin, a key hormone controlling appetite and energy balance, was strongly downregulated (Lep, FC = ‒8.54, Fig. 6D). Leptin normally signals to the brain to suppress food intake and enhance energy expenditure in response to increased fat stores33. In oxaliplatin-treated mice, Lep expression was barely detectable ( ~ 100 counts in 3 out of 4 animals) compared to >16,000 counts in controls. Lep expression was strongly correlated with body weight (r = 0.97, p = 0.0038), indicating that animals with the greatest leptin suppression also exhibited the most substantial weight loss. Consistent with this, mice with relatively higher residual leptin expression appeared more resistant to oxaliplatin-induced weight loss. Leptin is functionally connected to neuropeptide Y (NPY, encoded by Npy) in the hypothalamus34, which became highly upregulated with oxaliplatin treatment (FC = +5.88, Fig. 6D). NPY is a potent appetite stimulant upregulated during energy deficit states or low leptin levels. The interaction between leptin and NPY is a key regulatory mechanism in the body’s energy balance. Decreased leptin levels typically promote NPY release, leading to increased feeding behavior. This feedback loop, essential for maintaining body weight and regulating metabolic processes, is severely disrupted in oxaliplatin-treated mice, as evidenced by reduced food intake. GDF15 and neuregulin 4 (NRG4) are other signaling molecules that act as adipokines. GDF15 is a known metabolic regulator, and its higher levels can lead to lower body weight35, while Nrg4 has been recently discovered in connection with lipodystrophy36. Oxaliplatin induces a strong upregulation of Gdf15 in adipose tissue (FC = +7.99, Fig. 6D). The expression of Gdf15 is strongly correlated with the animal’s weight (r = ‒ 0.86, p = 0.012). Similarly, strong downregulation of the Nrg4 showed statistically significant correlation with the body weight (r = + 0.77, p = 0.042). Furthermore, there is a strong correlation between these adipokines (i.e., Lep and Gfd15; r = ‒0.92, p = 0.035), suggesting a highly coordinated interplay among multiple adipokines in response to oxaliplatin. Oxaliplatin-induced muscle atrophy shares gene expression similarities with human and preclinical cancer models To compare the gene signatures of chemotherapy-induced cachexia and cancer-induced cachexia, we employed publicly available RNA-seq datasets from cancer patients diagnosed with cachexia and animal models of cancer-induced cachexia. A study comparing subcutaneous fat transcriptomes of pancreatic ductal adenocarcinoma (PDAC) patients with cachexia (based on body and muscle indices) to cancer-free individuals was selected37. Most PDAC patients (23/24) had not undergone chemotherapy, enabling the identification of DEGs induced by cancer alone. In this dataset, 1229 genes with |FC | ≥ 1.5 (FDR < 0.05) were identified in WAT. In our mice, 2094 DEGs met this threshold, with 139 overlapping genes (Fig. 7A). However, hypergeometric testing indicated that this overlap was not statistically significant (ph = 0.99). In contrast, the analysis of muscle tissues revealed a statistically significant overlap. Among 747 DEGs identified in the muscle of PDAC patients and 167 DEGs from oxaliplatin-treated mouse muscle, 17 genes were shared, yielding a significant hypergeometric p-value (ph = 0.025) (Fig. 7B). Venn diagram of differentially expressed genes (DEGs) in adipose (WAT) and muscle (TA) tissues across oxaliplatin-treated mice, PDAC-associated cachexia, and fasting models. Genes were filtered for FDR < 0.05 and |FC | > 1.5. A In WAT, 2,094 DEGs were identified out of 15,774 expressed genes in oxaliplatin-treated mice. A total of 139 genes overlapped with DEGs from the subcutaneous fat of PDAC patients with cachexia, but this overlap was not statistically significant (ph = 0.986). B In the TA muscle, 17 overlapping genes were identified between oxaliplatin-treated mice and PDAC patients, representing a statistically significant enrichment (ph = 0.025). Venn diagram of DEGs between oxaliplatin-induced wasting tissues and fasting models. Genes were filtered for FDR < 0.05 and |FC | > 1.5. C Comparison of WAT DEGs between oxaliplatin-treated mice and fasted mice revealed 394 overlapping genes, a highly significant enrichment (ph = 1.70 × 10−41). D In skeletal muscle, 35 genes were commonly altered between oxaliplatin treatment and fasting (ph = 5.79 × 10−05). Across both tissues, approximately 80% of DEGs were uniquely affected by oxaliplatin, demonstrating a substantial chemotherapy-specific transcriptional signature. As reduced food intake was observed, we further explored the relationship between oxaliplatin-induced weight loss and fasting. To do this, we compared transcriptional changes between an external dataset from WAT and skeletal muscle from wild-type mice that fasted for 48 hours38 and the corresponding RNA-seq data from our study. In WAT, the external fasting dataset identified 1,562 DEGs (FDR < 0.05, |FC | > 1.5), a number comparable to the 2094 DEGs induced by oxaliplatin. Approximately 20% of these genes overlapped, and this overlap was highly significant based on hypergeometric testing (ph = 2.37 × 10⁻⁴¹, Fig. 7C). In skeletal TA muscle, genes were also mutually affected by oxaliplatin and fasting. Fasting affected 1307 genes in TA muscle compared to 167 in oxaliplatin-treated mice, with 35 genes shared between the two conditions (Fig. 7D) and such overlap was also statistically significant (ph = 5.79 × 10⁻⁵). These comparisons suggest that reduced food intake contributes to aspects of muscle and fat loss. However, most transcriptional changes arise from oxaliplatin treatment itself rather than from fasting alone. While fasting produced relatively balanced effects on muscle and fat, oxaliplatin exerted a disproportionately strong impact on adipose tissue. Together, these findings indicate that oxaliplatin shares mechanistic features with both cancer and fasting but also induces a distinct gene expression signature, particularly in adipose tissue, consistent with a unique form of chemotherapy-associated cachexia. Discussion Cachexia is a multifactorial metabolic syndrome characterized by involuntary weight loss driven by skeletal muscle wasting, often accompanied by adipose tissue depletion, and commonly associated with chronic inflammatory conditions such as cancer39,40. It is now recognized as a progressive continuum encompassing pre-cachexia, cachexia, and refractory cachexia, with important differences in underlying biology and therapeutic responsiveness41. While late-stage cachexia is largely resistant to intervention, earlier stages remain potentially modifiable, highlighting a critical window for therapeutic intervention15. Using a non–tumor-bearing model, we demonstrate that oxaliplatin alone is sufficient to induce a cachexia-like, largely irreversible phenotype independent of tumor-derived metabolic and inflammatory signals. This approach allowed us to isolate chemotherapy-specific effects that are otherwise obscured in tumor-bearing models, where cancer induces strong metabolic and inflammatory remodeling39,40. Chronic oxaliplatin administration resulted in progressive loss of body weight, lean mass, and adipose tissue, accompanied by reduced physical activity, impaired metabolic function, and increased mortality, closely recapitulating key features of human cachexia. Repeated oxaliplatin administration in mice caused progressive behavioral, metabolic, and physical changes9 resembling human cachexia-like symptoms. While some studies showed that oxaliplatin had a limited impact on mice’s water and food consumption and exercise capacity42,43, we observed a significant reduction in food and water intake, voluntary wheel running, and ambulatory activity with repeat administration of oxaliplatin at cumulative dosages. Reduced physical activity likely contributes to muscle loss, as decreased movement and increased sedentary behavior exacerbate muscle wasting and reduce energy expenditure. Pronounced CD45-positive immune cell infiltration together with the upregulation of inflammatory genes that negatively correlated with muscle mass (including Ccl2, Icam1, Angpt2, and Pecam1) strongly support that muscle atrophy is driven by inflammation rather than by disuse. Although oxaliplatin is classically described as a DNA-adduct–forming agent that suppresses proliferation in rapidly dividing cells, our multi-tissue transcriptomic analysis indicates that in post-mitotic tissues such as muscle and adipose, its dominant effects are associated with metabolic remodeling and inflammation rather than activation of cell-cycle–related programs (Fig. S7). In contrast, highly proliferative tissues, such as bone marrow, are characterized primarily by cell-cycle arrest and apoptosis pathways44. Thematic gene ontology (thematicGO) analysis45 of the RNA-seq DEGs also revealed that the most prominently affected pathways in both tissues are related to metabolic rewiring (Fig. S8). Cell-cycle and apoptosis pathways that are canonically associated with the cytotoxic action of oxaliplatin show a comparatively smaller contribution. This suggests that, in these tissues, oxaliplatin exerts a stronger impact on metabolic processes than on proliferation-related pathways. Our observation is consistent with the biology of adipocytes and myocytes, which are terminally differentiated and largely non-proliferative, and therefore less dependent on cell-division pathways. In highly proliferative tissues, such as bone marrow, the cell-cycle and apoptosis pathways become the dominant response44. The sustained reduction in core body temperature observed with chronic oxaliplatin, as well as in other treatments46, further supports a systemic metabolic impairment. While hypothermia can be associated with systemic inflammation or sepsis47, our data do not support this interpretation. Features typically associated with sepsis in mice, such as acute lethargy, ruffled fur, or diarrhea, were not observed. The markers of sepsis, IL-6 and CCL248,49 were not elevated as we showed earlier44. Instead, the observed selective increase in TNF-α, together with reduced TGF-β (a key regulator of immune homeostasis, tissue remodeling, and muscle maintenance50) indicated a pro-catabolic inflammatory shift associated with metabolic dysfunction51,52. Consistent with this interpretation, thematicGO45 analysis of RNA-seq data of fat tissue identified thermogenesis as strongly associated with metabolic rewiring (Figure S8A). Theme–theme overlap network analysis further demonstrated strong connectivity between metabolic rewiring, fatty-acid oxidation, and thermogenesis pathways, indicating coordinated disruption of mitochondrial energy metabolism in adipose tissue (Fig. S9). Adipose tissue appeared particularly sensitive to oxaliplatin, exhibiting more pronounced transcriptional and metabolic alterations than skeletal muscle. RNA-seq revealed that oxaliplatin impaired lipid storage capacity by decreasing lipogenesis and adipogenesis, and reducing ATP production, thus shifting the energy source toward protein degradation and affecting muscle function. The loss of fat was in line with a profound alteration in adipokine signaling. Leptin levels in adipose tissue were strikingly reduced, similar to what has been previously observed in cancer patients53. A low level of leptin was paralleled by increased hypothalamic Npy expression, which normally induces appetite, leading to increased food intake and restoring body fat mass. However, in the case of oxaliplatin induced cachexia, this mechanism seems to be broken. Apparently, factors such as cytokines and tumor products interfere with the normal hypothalamic signaling pathways, leading to persistent anorexia and weight loss despite the body’s attempts to compensate54–56. Our data also revealed marked dysregulation of adipokines in WAT, particularly Gdf15 and Nrg4, both of which act as an alternative to the leptin pathway and contribute to appetite suppression and weight loss through their respective receptors in the central nervous system (CNS). GDF15 signals through its dedicated receptor GFRAL, expressed in the brainstem, while NRG4 engages ERBB4 receptors, which are highly expressed in the hypothalamus. Elevated GDF15 levels have been strongly linked to cancer cachexia in both preclinical models57–59 and clinical studies60. Conversely, reduced circulating NRG4 has been associated with adipose tissue loss in patients36. Together, these findings emphasize GDF15 and NRG4 as potential mediators of oxaliplatin-induced metabolic deterioration. Male cancer patients generally exhibit a higher prevalence of cachexia (i.e., 60% vs 40%), greater muscle wasting, and worse clinical outcomes compared with females61. Consistent with this, preclinical cancer cachexia models have shown partial protection against muscle loss in females62,63 although often accompanied by greater adipose depletion62. Interestingly, clinical observations indicate that female patients experience worse side effects from chemotherapy across a variety of drugs, including oxaliplatin64,65. In the present study, however, we did not observe significant sex-dependent differences in cachexia outcomes. Apparently, chronic treatment with oxaliplatin in mice engages mechanisms that are less dependent on sex specific factors (i.e., hormones). Several limitations should be considered when interpreting our findings. First, the study was performed in non–tumor-bearing mice treated with oxaliplatin alone, allowing isolation of chemotherapy-specific effects. However, future studies incorporating tumor models and combination therapies will be required to better reflect clinical settings and to determine how these additional variables influence the development of cachexia. Second, although both sexes were included, the study was not powered to detect subtle sex-dependent differences, and larger studies will be required to address this question. Third, the experimental design focused primarily on the endpoint at week 8 when the cachexia-like phenotype became detectable, limiting insight into early versus secondary effects of treatment. Longitudinal studies with higher temporal resolution, particularly during early treatment cycles, will be necessary to define the progression of cachexia and identify early predictive markers. While histological observations, immunohistochemistry, and transcriptomic analyses suggested immune cell infiltration and inflammatory activation in skeletal muscle, additional validation and more granular information about the type of immune cells will be performed to better define the contribution of immune responses to tissue remodeling and wasting. Finally, while our data indicate immune infiltration and inflammatory activation in muscle, further characterization of immune cell populations and longitudinal systemic and transitional assessment of adipokines such as GDF15 and NRG4 will be important to clarify their roles and evaluate their potential as biomarkers. In summary, our study provides strong evidence that oxaliplatin, a broadly used chemotherapeutic agent, can independently drive the development of cachexia-like phenotypes, including diminished physical activity, reduced metabolic function, and irreversible lethal weight and muscle loss. Adipose tissue exhibited greater sensitivity to oxaliplatin-induced changes than skeletal muscle and may involve communication with the CNS through an adipokine–brain axis. The study demonstrates a significant overlap in the molecular mechanisms between oxaliplatin-induced cachexia and cancer-induced cachexia, yet it also identifies distinct pathways uniquely activated by chemotherapy. Identification of chemotherapy-specific drivers at the early stages is essential to prevent progression to severe, treatment-refractory cachexia without compromising anti-tumor efficacy. Methods Animal models We have complied with all relevant ethical regulations for animal use. All animal studies were performed under Washington University in St. Louis IACUC-approved protocols 21-0384 and 24-0297 and have therefore been conducted in accordance with all ethical standards and National Institutes of Health (USA) guidelines. Male and female 10-week-old C57BL/6 mice (Charles River Laboratory, Wilmington, MA, USA) were housed at room temperature and maintained on a 12:12-h light/dark cycle with free access to water and food. Mice were randomly assigned to treatment or control groups using simple randomization based on cage and sex. Where possible, littermates were distributed across experimental groups to minimize potential confounding effects of cage or genetic background. Mice were injected with either 5 mg/kg or 10 mg/kg oxaliplatin in dextrose (Sandoz Inc., Princeton, NJ, USA) or an equal quantity of vehicle control (5% dextrose). Drugs were administered intraperitoneally (i.p.) for 8 weeks. The body weights of the mice and physiological parameters were monitored. The oxaliplatin dose (10 mg/kg) was calculated based on the clinical regimen (85 mg/m² biweekly) and converted to a mouse equivalent dose (mg/kg) using body surface area calculations, then adjusted to each mouse’s weight to ensure clinical relevance66. For all experiments, the experimental unit was a single mouse. Oxaliplatin-treated mice were compared with vehicle-treated control mice of the same sex and age. At the end of the experiments, usually within 7–8 days after the last injection, mice were anesthetized via isoflurane induction (2% at 1.5 L O2/min) and euthanized by cervical dislocation. Body composition analyses Mice were weighed immediately before body composition analyses using an EchoMRI 100H 2-in-1 (EchoMRI LLC, Houston, TX, USA) with a horizontal probe configuration to measure whole-body lean, fat, total water, and free water masses. The instrument was calibrated daily with canola oil before measurements. The Hydration Ratio was calculated as follows from Eq. 1: Where Total Water is the overall water content in the body (free + bound water); Free Water is the mobile, unbound water, typically extracellular. Lean Mass is a fat-free body mass of mice, including muscle, bone, and other non-fat components. Adiposity was calculated as the ratio of fat mass to body weight of the mouse using Eq. 2: where Fat Mass is the total fat mass, and Body Weight is the measured weight of the mouse. Metabolic cages assessment As metabolic cage assessment requires single housing of animals, only female mice were assigned for the test to avoid fighting after being re-housed in groups. Briefly, mice were single-housed in a 16-cage Comprehensive Laboratory Animal Monitoring System with an environmental enclosure set at 22.2 °C (Columbus Instruments, Inc., Whitehall, PA, USA). Mice were acclimated for 16 h, and data were recorded for 24 h starting at 06:00 the next day. The lights were turned on at 06:00 and off at 18:00 for a 12 h:12 h light-dark cycle. Mice had free, unlimited access to food and water. Hanging feeders and water dispensers on load cells were used to measure food intake and water consumption, respectively. The airflow rate was set at 0.9 L/min. O2 and CO2 levels in the air were measured with the OxyFast gas sensor system in each cage, in series, with 18 seconds of air bleed time followed by 2 seconds of measurement. After the 16th hour, air from the enclosure was sampled. The differentials between the enclosure and in-cage O2 and CO2 were used to calculate the volume of O2 consumption (VO2) and CO2 production (VCO2). The respiratory exchange ratio (RER) was defined as VCO2 divided by VO2. Total energy expenditure (heat) was calculated using the standard equation and normalized to body mass. Ambulatory activity levels were monitored using an infrared laser array at the level of the cage’s bedding on the x- and y-axes and are expressed as total ambulatory beam breaks for the indicated phases of the light:dark cycle. Grip test Four-limb grip strength testing was measured using a BIO-GS4 grip strength meter (Bioseb, France) with a mesh adapter. After the mouse gripped the mesh with all four limbs, it was gradually lifted with force guidance using Bioseb BIO-CIS software. The force applied during mesh release was measured as maximal grip strength. Voluntary wheel running Mice were individually housed in cages equipped with a wheel connected to a magnetic activity sensor and a data collection system with remote access (Coulburn Instruments). They were maintained under standard 12-hour day/night light conditions with unlimited access to food and water. Before treatment, mice were acclimated for one week. Oxaliplatin was administered weekly at 2 pm, and voluntary wheel-running activity was continuously recorded for 144 hours, starting at 6 pm on the day of treatment, over an 8-week chemotherapy course. Cages were cleaned every seventh day and restocked with fresh food and water as needed. Wheel rotations were recorded automatically in 6-minute intervals as the number of revolutions. The running distance was calculated from the wheel circumference. The wheel diameter was 4.5 inches (11.43 cm), corresponding to a circumference of 35.9 cm (C = π × D), with each revolution corresponding to 0.359 m of running distance. The distance traveled in each recording interval was calculated by multiplying the number of revolutions by the wheel circumference, and the total running distance was obtained by summing the distances across all intervals. Tissue extraction and histology To avoid the confounding effects of voluntary wheel running, all histological evaluations were performed on tissues collected from a separate cohort of mice without access to running wheels. The tibialis anterior (TA) muscle, gastrocnemius muscle, mesentery, omentum, and inguinal subcutaneous white adipose tissues were quickly dissected and weighed four days after the last oxaliplatin injection, following the completion of all behavioral assessments. Interscapular brown adipose tissue was collected, the surrounding muscle layers were removed, and the tissue was weighed. The TA muscle and inguinal fat pads were immersed in ice-cold RNAlater (ThermoFisher Scientific, Waltham, MA, USA) solution and stored at -20°C before RNA extraction. The gastrocnemius muscles, omentum, and mesentery from both groups were fixed in 10% NBF overnight at 4°C, dehydrated in a series of 30%, 50%, and 70% ethanol, and stored in 0.1% sodium azide PBS until paraffin embedding. Fixed tissues were then cut into 10 μm sections for H&E staining. Stained tissue slides were imaged with the Olympus BX51 microscope. Blinded investigators analyzed the slides. Immunofluorescence staining for CD45 Formalin-fixed, paraffin-embedded (FFPE) gastrocnemius muscle tissues were prepared by the Washington University Musculoskeletal Research Center (NIH P30 AR074992). Tissue blocks were sectioned at 5 μm thickness using a microtome, with serial sections collected at 50 μm intervals, and mounted onto glass microscope slides. For immunohistochemical analysis, unstained sections were deparaffinized and rehydrated through a graded ethanol series. Heat-induced antigen retrieval was performed in antigen retrieval buffer pH 6.0 (Diva Decloaker, DV2004, Biocare Medical, Pacheco, CA, USA) for 25 min at 100 °C using a pressure cooker. Following antigen retrieval, sections were incubated overnight at 4 °C with a rabbit monoclonal anti-CD45 antibody (clone EPR20033, ab208022, Abcam, Cambridge, UK) at a dilution of 1:500. After washing, sections were incubated with a goat anti-rabbit horseradish peroxidase (HRP)-conjugated polymer secondary antibody (ab214880, Abcam). Signal was visualized using 3,3′-diaminobenzidine (DAB) chromogen (ImmPACT DAB, SK-4105, Vector Laboratories, Newark, CA, USA), resulting in a brown precipitate at sites of antigen localization. Sections were counterstained with hematoxylin (7221, Epredia, Portsmouth, NH, USA,) for 1 min, dehydrated, and coverslipped using Cytoseal mounting medium (8312, Epredia). All staining procedures were performed at the Animal Research Histology Core (Department of Developmental Biology, Washington University). Slides were imaged using a Revolution microscope (model RON-K; Discover Echo Inc., San Diego, CA, USA) equipped with Revolution software (version 2.0.8.1). Images were acquired under identical settings across experimental groups. Regions of interest in the images of gastrocnemius muscles of both control and oxaliplatin treated mice were established in ImageJ (v1.54), which excluded areas of marked hematoxylin staining. The images were converted to 8-bit gray scale, and thresholds were applied to those images of the gastrocnemius muscles. Pixels in stained versus pixels in unstained regions were tallied. The percentage of the area that was stained was calculated from the following equation, Eq. 3: RNA sequencing and analysis To avoid the confounding effects of voluntary wheel running, RNA-seq analyses were performed on tissues collected from a separate cohort of mice without access to running wheels. The RNA-seq sample sizes were as follows: adipose tissue, n = 3 control and n = 4 oxaliplatin-treated mice; tibialis anterior skeletal muscle, n = 4 control and n = 5 oxaliplatin-treated mice; hypothalamus, n = 3 control and n = 4 oxaliplatin-treated mice. The differences in group size reflect the RNA quality control criteria required for sequencing, as several samples did not meet the threshold for library preparation (RIN > 8) and were therefore excluded. RNA-seq followed our established protocol11. Total RNA was extracted with the magnetic bead-based MaxwellRSC 48 instrument using the Maxwell RSC simply RNA Tissue kit (Promega Corporation, Madison, WI, USA). The RNA quality was assessed using the Agilent TapeStation 4200 System (Agilent Technologies, Santa Clara, CA, USA). Samples with an RNA integrity number (RIN) greater than 8 were used for sequencing. cDNA libraries were prepared following the manufacturer’s protocol. The library was indexed, pooled, and sequenced on an Illumina NovaSeq 6000 (Illumina, San Diego, CA, USA). RNA-seq reads were aligned to the Ensembl release 101 primary assembly using STAR (RRID:SCR_004463. version 2.5.1a). Aligned reads were imported into Partek Flow (RRID:SCR_011860, Partek Inc., St. Louis, MO, USA) for downstream analysis. Gene-level counts were derived from the number of uniquely aligned, unambiguous reads and normalized using the median ratio method. DEGs were identified with the DESeq2 package (RRID:SCR_015687) implemented in Partek Flow. DEGs with an FDR < 0.05 and absolute fold change > 1.5 were used for downstream analyses including KEGG pathway (RRID:SCR_018145). Gene clustering and pathway analysis were performed using the STRING online tool (RRID:SCR_005121 https://string-db.org/). In addition, Gene Ontology enrichment was analyzed using a custom-designed thematicGO framework45, which aggregates GO terms into higher-level biological themes to facilitate interpretation of complex transcriptomic responses Correlation analysis Correlation analysis between genes and pathways was performed using an IVCCA bioinformatics analysis tool23,67. IVCCA aims to identify and quantify patterns of correlation by analyzing the variability of individual components, such as genes, across datasets. Hypergeometric distribution The hypergeometric distribution calculated the probability of having a certain number of overlapping genes from the two sets of genes according to the following formula for the hypergeometric probability mass function (PMF) (Eq. 4) where N is the total number of genes (the population size), K is the number of DEGs in the dataset 1, n is the number of DEGs in the dataset 2, k is the number of overlapping genes. Kk represents the binomial coefficient, calculated according to Eq. 5: The hypergeometric distribution was calculated in MATLAB (RRID:SCR_001622) using hygecdf.m function. We considered p < 0.05 as statistically significant. ThematicGO analysis Gene Ontology (GO) enrichment was further analyzed using a thematic grouping strategy that consolidates enriched GO terms into biologically meaningful categories based on a curated, keyword-driven assignment approach (thematicGO). A complete list of themes and associated keywords is provided in Supplementary Data 3. ThematicGO was performed on DEGs identified in adipose tissue (n = 2094 DEGs; FDR < 0.05, |FC | > 1.5) and tibialis anterior (TA) muscle (n = 167 DEGs; FDR < 0.05, |FC | > 1.5). Bar plots represent cumulative enrichment scores ( − log₁₀ p-values) for grouped biological themes derived from significantly enriched GO terms (p < 0.05) (Fig. S8A-B). GO terms were assigned to predefined themes based on keyword matching. Theme–theme gene overlap networks generated using the thematicGO framework is shown in Fig. S9. In these networks, nodes represent biological themes and are scaled by the number of associated genes, while edges represent shared genes and are weighted and color-coded according to the extent of overlap (blue to red). Statistics and reproducibility No animals were excluded from analysis unless they met predefined humane endpoint criteria, including loss of approximately 20% of initial body weight or signs of severe distress, at which point animals were euthanized in accordance with IACUC guidelines. Investigators were blinded to the treatment group during metabolic cage recording, histological analysis, and image quantification; blinding was not feasible during drug administration. Primary outcome measures included changes in body weight, lean mass, fat mass, and survival. Secondary outcomes included food and water intake, locomotor activity, core body temperature, metabolic parameters, histological changes, and tissue-specific gene expression. The number of mice used in each experiment is indicated in the figure legends or corresponding text. All results are expressed as mean ± SD. Statistical analysis on non-RNA-seq data was performed in GraphPad Prism 10 (RRID:SCR_002798, GraphPad Software, San Diego, CA, USA) using a two-tailed unpaired t-test or two-way ANOVA with multiple comparisons. A p-value < 0.05 was considered statistically significant. For RNA-seq, differentially expressed genes were considered statistically significant at FDR (false discovery rate) < 0.05. For pathway analyses, statistical significance was defined as p-value < 0.05. Physiological measurements, including body weight, metabolic parameters, locomotor activity, and body composition, were obtained longitudinally from the same animals. In contrast, endpoint analyses such as histology and RNA-seq were performed on independently collected tissue samples. RNA-seq experiments were conducted on a separate cohort of animals to avoid confounding effects from prior behavioral or metabolic assessments. No additional covariates were included in statistical analyses. Data were assumed to be normally distributed, but this was not formally tested. For analyses involving multiple comparisons, appropriate post hoc tests were applied as implemented in GraphPad Prism. Effect sizes were not calculated. To ensure reproducibility and transparency, all experiments were independently replicated, and the key findings were confirmed across multiple animals. Detailed descriptions of the experimental design, including sample sizes and statistical methods, are provided in the corresponding subsections in the Methods section and in the figure captions. No data points were excluded unless predefined criteria were met, as described above. All reagents, instruments, and software used in this study are specified with identifiers and references where applicable. Supplementary information Supplemental information file Description of Additional Supplementary Files Supplementary Data 1 Supplementary Data 2 Supplementary Data 3 Supplementary Data 4 Supplementary Data 5 Supplementary Data 6 Supplementary Data 7 Supplementary Data 8 Supplementary Data 9 Acknowledgements We thank Dr. John E. Morley (Saint Louis University) for the discussion that led to this research. We thank the Siteman Cancer Center (SCC) and the Institute of Clinical and Translational Sciences (ICTS) at Washington University in St. Louis for using the Genome Technology Access Center. The SCC is supported in part by an NCI Cancer Center Support Grant #P30 CA091842, and the ICTS is funded by the National Institutes of Health’s NCATS Clinical and Translational Science Award (CTSA) program grant #UL1 TR002345. We also thank the Washington University Center for Cellular Imaging (WUCCI) for imaging studies. Author contributions J.D., R.L.M., C.S. and R.L.F. conducted experiments and analyzed data. J.D., R.L.M., M.Y.B., J.R.B. and G.J.R. designed the experiments. L.C.S. and J.E.M. provided essential expertise in executing the experiments. M.B., J.E.M., J.R.B. and GJR provided supervision, training, regulatory compliance, and funding. J.D., R.L.M., and M.Y.B. authored the manuscript, with editing contributions from all the authors. Peer review Peer review information Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editor: Dario Ummarino. Funding M.Y.B. discloses support for the research of this work from NIH/NCI (R01CA208623, R21CA269099) and NIH/NINDS (1R01NS139461, R21NS135646). R.L.M. discloses support for the research from NIH/NCI F30CA281124, J.R.B. discloses support for the research of this work from NIH DP5OD028125 and Burroughs Welcome Fund (CAMS #1019648). Data availability The bulk RNA-seq datasets generated in this study have been deposited in the Gene Expression Omnibus (GEO) repository https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi under accession number GSE285906 (2026). Included files are: 1. GSE285906_Fat_raw_counts.txt.gz, 2. GSE285906_Muscle_raw_counts.txt.gz 3. GSE285906_Hypothalamus_raw_counts.txt.gz Competing interests The authors declare the following competing interests: M.Y.B. is a consultant for Sarya LLC and Daxor Inc and a founder of HSpeQ LLC. J.R.B. is a member of the Scientific Advisory Board of LUCA Science, Inc.; receives research support from LUCA Science and Edgewise Therapeutics; is a consultant for Columbus Instruments; has consulted for DeciBio within the past 12 months (discontinued); receives royalties from Springer Nature Group; is an inventor on technology licensed to Columbus Instruments with royalty rights; and is an inventor on pending patent applications related to the treatment of metabolic diseases (63/625,555), allergic diseases (US20210128689A1) and mitochondria transfer (018984/US). The other authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Junwei Du, Rachel L. Mintz. Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s42003-026-10441-3. References Associated Data Supplementary Materials Supplemental information file Description of Additional Supplementary Files Supplementary Data 1 Supplementary Data 2 Supplementary Data 3 Supplementary Data 4 Supplementary Data 5 Supplementary Data 6 Supplementary Data 7 Supplementary Data 8 Supplementary Data 9 Data Availability Statement The bulk RNA-seq datasets generated in this study have been deposited in the Gene Expression Omnibus (GEO) repository https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi under accession number GSE285906 (2026). Included files are: 1. GSE285906_Fat_raw_counts.txt.gz, 2. GSE285906_Muscle_raw_counts.txt.gz 3. GSE285906_Hypothalamus_raw_counts.txt.gz