Background: Cancer cachexia (CAC) reduces patient survival and quality of life. Developments of efficient therapeutic strategies are required for the CAC treatments. This long-term process could be shortened by the drug- repositioning approach which exploits old drugs approved for non-cachexia disease. Amiloride, a diuretic drug, is clinically used for treatments of hypertension and edema due to heart failure. Here, we explored the effects of the amiloride treatment for ameliorating muscle wasting in murine models of cancer cachexia. Methods: The CT26 and LLC tumor cells were subcutaneously injected into mice to induce colon cancer cachexia and lung cancer cachexia, respectively. Amiloride was intraperitoneally injected daily once tumors were formed. Cachexia features of the CT26 model and the LLC model were separately characterized by phenotypic, histopathologic and biochemical analyses. Plasma exosomes and muscle atrophy-related proteins were quantitatively analyzed. Integrative NMR-based metabolomic and transcriptomic analyses were conducted to identify significantly altered metabolic pathways and distinctly changed metabolism-related biological processes in gastrocnemius. Results: The CT26 and LLC cachexia models displayed prominent cachexia features including decreases in body weight, skeletal muscle, adipose tissue, and muscle strength. The amiloride treatment in tumor-bearing mice distinctly alleviated muscle atrophy and relieved cachexia-related features without affecting tumor growth. Both the CT26 and LLC cachexia mice showed increased plasma exosome densities which were largely derived from tumors. Significantly, the amiloride treatment inhibited tumor-derived exosome release, which did not obviously affect exosome secretion from non-neoplastic tissues or induce observable systemic toxicities in normal healthy mice. Integrative-omics revealed significant metabolic impairments in cachectic gastrocnemius, including promoted muscular catabolism, inhibited muscular protein synthesis, blocked glycolysis, and impeded ketone body oxidation. The amiloride treatment evidently improved the metabolic impairments in cachectic gastrocnemius. * Correspondence: firstname.lastname@example.org; email@example.com Research and Communication Center of Exercise and Health, Xiamen University of Technology, Xiamen 361024, China Key Laboratory for Chemical Biology of Fujian Province, MOE Key Laboratory of Spectrochemical Analysis & Instrumentation, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005, China Full list of author information is available at the end of the article © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Zhou et al. Skeletal Muscle (2021) 11:17 Page 2 of 16 Conclusions: Amiloride ameliorates cachectic muscle wasting and alleviates cancer cachexia progression through inhibiting tumor-derived exosome release. Our results are beneficial to understanding the underlying molecular mechanisms, shedding light on the potentials of amiloride in cachexia therapy. Keywords: Amiloride, Cancer cachexia, Muscle wasting, Exosome, Exosome-release inhibition Introduction cachectic muscle wasting, we speculated that amiloride Cachexia is a systemic metabolic syndrome defined by might have some effects for ameliorating muscle wasting involuntary body weight and skeletal muscle loss (with in cancer cachexia. or without fat loss) and cannot be fully reversed by con- In the present work, we sought to determine whether ventional nutritional supplementations . Driven by a amiloride was able to ameliorate muscle wasting in mur- complicated combination of endocrine and metabolic ine cachexia models. Furthermore, we addressed mo- disorders as well as central nervous system perturba- lecular mechanisms of tumor-derived exosomes tions, cachexia is characterized by several predominant promoting muscle wasting by integrative metabolomic features including reduced food intake, decreased muscle and transcriptomic analyses, which provides the mech- mass, excess catabolism, unnecessary energy expend- anistic rationale for exploiting clinical potentials of iture, and hyper-inflammatory response . Pathologic amiloride for improving the CAC treatments. Our re- mechanisms of cancer cachexia are closely related to ac- sults demonstrated that the amiloride treatment could tivation of proteolysis, autophagy, lipolysis, and inflam- significantly ameliorate muscle wasting in cancer cach- mation . exia and thus alleviate the CAC progression through Cachexia is usually associated with chronic and malig- inhibiting tumor-derived exosome release. nant diseases, including kidney disease, heart failure, neurological disease, chronic obstructive pulmonary dis- Methods ease, AIDS, and, especially, cancer . As one of the Patient blood sample collection leading causes of death worldwide, cancer accounts for Patient blood samples were firstly kept in anti- an estimated 9.6 million deaths in 2018, and nearly 80% coagulation (100 mM sodium citrate) tubes and then of cancer patients are affected by cachexia in different centrifuged (1000g, 10 min, 4 °C) to obtain platelet-free grades . Cancer cachexia (CAC) directly causes about plasma within 2 h after blood collection. The plasma was one third of fatalities in cancer patients with significant aliquoted and collected in cryovials and kept at − 80 °C skeletal muscle wasting . Efficient therapeutic strat- until used. egies for the CAC treatments are urgently needed . Recently, mounting researches indicated that tumor- Exosome isolation and characterization derived exosomes contribute to cancer cachexia through Exosomes in either patient/mouse plasma or culture mediating the cross-talk between tumors and distally lo- media of CT26/LLC cells were isolated by ultracentrifu- cated skeletal muscles, resulting in decreased muscle gation [14, 15]. A Beckman Coulter XE-90K Ultracentri- weight, impaired organismal function, suppressed thera- fuge equipped with an SW 41 Ti rotor was used for the peutical response, and reduced quality of life, as well as ultracentrifugation. Particle sizes of exosomes were ana- remarkably enhanced cancer-related mortality [7–10]. lyzed using ZetaView® Nanoparticle Tracking Analyzer These works might provide a new strategy for the CAC (Particle Metrix GmbH, Meerbusch, Germany) and FEI treatments based on inhibition of tumor-derived exo- Tecnai 20 transmission electron microscope (Thermo somes release. Fisher, USA) according to the manufacturer’s manual Amiloride is an old drug with potassium-sparing diur-  or the published protocol , respectively. + + + + etic function (upon inhibition of Na /H and Na /K ex- changers), which has been clinically employed in the Exosome density and purity measurements treatments of hypertension, hypokalemia, edema, and Exosome density and purity were analyzed with a two- congestive heart failure for decades . Most import- channel high-sensitivity nano flow cytometer (HSFCM) antly, amiloride can inhibit exosome release from cells according to the protocol described previously . The and reverse exosome-promoted pathogenic processes, HSFCM was developed by the Laboratory of Professor including autoimmune disease and immuno-suppressive XM Yan from College of Chemistry and Chemical En- regulation [12, 13]. However, no attempts have been re- gineering, Xiamen University, which has been commer- ported to ameliorate cachectic muscle wasting through cialized by NanoFCM Inc, China. To assess exosome inhibiting tumor-derived exosome release. Given that purities, 1% final volume of Triton X-100 (Sigma-Al- tumor-derived exosomes are implicated in mediating drich, USA) was added to exosome suspensions, and the Zhou et al. Skeletal Muscle (2021) 11:17 Page 3 of 16 HSFCM measurement was repeated after incubation of with amiloride dissolved in PBS at a dose of 2 mg/kg 30 min on ice. (AM mice). Furthermore, The Rab27 knock-down tumor cells were subcutaneously injected into the mice follow- Cell culture ing the same procedure (KD mice). Similarly, normal The LLC, CT26, and HEK 293T cells were purchased control mice were injected with PBS on day 0 (NOR from the China Center for Typical Culture Collection mice). Both the KD and NOR mice were intraperitone- (CCTCC). C2C12 cells were provided by Stem Cell ally injected with PBS daily from day 9. Bank, Chinese Academy of Sciences. LLC and CT26 cells Mouse body weights were monitored every 3-day post- were cultured in DMEM and RPMI-1640, respectively. tumor implantation, and food intakes were measured Both culture media were supplemented with 100 units/ every day. Tumor volumes were calculated every 3-day mL penicillin, 100 μg/mL streptomycin, and 10% fetal post-tumor implantation using the formula: tumor vol- 3 2 bovine serum (Hyclone, USA). All cells were cultured in ume (mm ) = 0.52 × length × width , in which the a humidified atmosphere of 5% CO at 37 °C. Culture length and perpendicular width were measured with a media of LLC and CT26 cells were collected and centri- vernier caliper. Forelimb grip forces were measured with fuged (1000g, 5 min, 4 °C) after 48 h of culture. C2C12 a Grip Strength Meter (YLS-13A, Shandong Academy of myoblasts were cultured to 85–90% confluence in Medical Sciences, China). For each mouse, the grip DMEM growth medium. Myoblast differentiation was strength was defined as the average of five successive induced by incubation for 96 h in DMEM supplemented measurements. On day 30, the mice were sacrificed. with 2% heat-inactivated horse serum. C2C12 myoblasts Both tumors and gastrocnemius were removed, weighed, were used within ten generations of culture. and quickly frozen in liquid nitrogen for subsequent analysis. The mouse blood samples were firstly kept in Lentiviral expression of shRNA in tumor cells anti-coagulation (100 mM sodium citrate) tubes and The pLKO.1-puro lentivirus vector was used to express pro-coagulation tubes and then centrifuged (1000g,10 the shRNAs. The virus was generated by four co- min, 4 °C) to obtain platelet-free plasma and serum, re- transfected plasmids, including the lentiviral vector, spectively, within 2 h after blood collection. The plasma pMDLg/pRRE, pRSV-Rev, and pMD2 in HEK 293T exosomes were quantitatively analyzed. Serum levels of cells. At 48 h, virus-containing supernatants were col- TNF-α, IL-6, and IL-1β were measured by ELISA kit lected for transduction in CT26 and LLC cells. The (R&D Systems China) according to the manufacturer’s shRNAs against mouse Rab27a and Rab27b were 5′- instructions. GCTTCTGTTCGACCTGACAAA-3′ and 5′-GCTT CTGGACTTAATCATGAA-3′, respectively. Muscular toxicity evaluation We evaluated potential muscular toxicities of the amilor- Animal experiments ide treatment in the CT26 and LLC models using To assess the effects of amiloride for ameliorating C57BL/6 and BALB/c normal control mice, respectively. muscle atrophy in vivo, we constructed murine models Either 12 C57BL/6 mice or 12 BALB/c mice were di- of CT26 (colon) and LLC (lung) cancer cachexia (Fig. vided into 2 groups: NOR mice and NOR-AM mice, 6 S1). Adult (age 6–8 weeks) C57BL/6 and BALB/c male per group. PBS was subcutaneously injected into the mice were purchased from Shanghai SLAC Laboratory right flank of the NOR and NOR-AM mice on day 0. Animal Co., Ltd. Mice were individually housed, accli- From day 9, the NOR-AM mice were intraperitoneally mated to their cages and human handling for 1 week be- injected daily with amiloride at the same dose of 2 mg/ fore animal experiments, and maintained in conditions kg following the procedure described above, whereas the of constant temperature and 12-h light/12-h dark cycles. NOR mice with PBS continually. Both the NOR-AM Tumor cells were subcutaneously injected into the and the NOR mice were sacrificed on day 30, and toxic right flank of mice on day 0. In detail, BALB/c mice effects of amiloride were assessed by the differences in were inoculated with the CT26 cells (1.0 × 10 /100 μL) body weight, gastrocnemius muscle weight, and plasma to induce colon cancer cachexia, while C57BL/6 mice exosome density between the NOR-AM mice and the were inoculated with the LLC cells (7 × 10 /100 μL) to NOR mice. induce lung cancer cachexia. Both BALB/c and C57BL/6 mice showed palpable tumors (about 5 mm in diameter) Histology study on day 9 after the inoculation. Both CT26-bearing mice C2C12 myotubes were fixed with pre-cold methanol for and LLC-bearing mice were randomly divided into 2 30 s, stained with 0.1% crystal violet solution for 10 min, groups (n = 8 per group): one group of mice intraperito- and rinsed with distilled water before taking microscopic neally injected daily with PBS from day 9 (CAC mice), photographs. Myotube diameters were measured in a another group of mice intraperitoneally injected daily total of 200 myotubes from ≥ 10 random fields. Mouse Zhou et al. Skeletal Muscle (2021) 11:17 Page 4 of 16 gastrocnemius was collected and fixed in 4% PFA. Paraf- two criteria: fold change (FC) ≥ 1.5 or FC ≤ 0.67, false fin sections were stained with H&E, and stained slides discovery rate (FDR) ≤ 0.05. The Kyoto Encyclopedia of were assessed using phase-contrast microscopy. Myofi- Genes and Genomes (KEGG) database was employed to ber areas were quantified by using ImageJ. To produce conduct the pathway enrichment analysis based on the frequency distribution histograms, five view fields were identified DEGs. Pathways with Q value ≤ 0.01 were measured with about 200 myofibers per field in each identified to be significantly changed biological section. processes. Protein expression analysis General statistical analysis Proteins were extracted using RIPA lysis buffer contain- Experimental data were expressed as means ± SD. For ing protease inhibitor and phosphorylation protease in- the quantitative comparison between two groups, data hibitor cocktails (Thermo Fisher, USA). The were analyzed by Student’s t test analysis using Graph- homogenates were then sonicated for 35 s and centri- Pad Prism. For pairwise comparisons among three or fuged (11,000g, 10 min, 4 °C) to remove the debris. The more groups, data were analyzed by using one-way supernatants were collected, and protein concentrations ANOVA followed by Tukey’s multiple comparison test were determined by BCA Protein Assay Kit (Beyotime using the SPSS 19 software. Statistical significances were Biotechnology). Then, the denatured samples were sub- as follows: p > 0.05 (NS), p < 0.05 (*), p < 0.01 (**), p < jected to SDS-PAGE and transferred to PVDF mem- 0.001 (***), and p < 0.0001 (****). branes (GE Healthcare, USA) for immunoblotting analysis. After blocking with 5% non-fat milk in Tris- Antibodies and reagents buffered saline containing 0.1% Tween 20 for 1 h, the The detailed information about antibodies and reagents membranes were probed with corresponding antibodies. is displayed in Table S1. Proteins were visualized by enhanced chemilumines- cence using horseradish peroxidase-conjugated anti- Results bodies and band densities were quantified by the ImageJ Amiloride treatment ameliorated cachectic muscle software. atrophy in mice The processes of establishing mouse models are illus- NMR-based metabolomic analysis trated in Fig. S1. The CT26 and LLC cells or PBS were Aqueous metabolites were extracted from mouse gastro- subcutaneously injected into the mice on day 0, while cnemius for NMR-based metabolomic analysis according amiloride or PBS was intraperitoneally injected daily to the protocol described previously [18–20]. All NMR from day 9 once mice developed palpable tumors. Both experiments were performed at 25 °C on a Bruker mouse body weights and tumor volumes were monitored Avance III 850 MHz spectrometer (Bruker BioSpin, along the process of establishing the CT26 cachexia Germany) equipped with a TCI cryoprobe. Both the un- model or the LLC cachexia model (Fig. 1a, b; Fig. 2a, b). supervised principal component analysis (PCA) and su- On day 30 after tumor inoculation, the tumor-bearing pervised partial least-squares discriminant analysis (PLS- mice displayed significant CAC features including de- DA) were applied to compare metabolic profiles of creases in body weights (mean loss rate: CT26, 5.9%; gastrocnemius among the four groups of the NOR, LLC, 6.5%), tumor-free body weights (TFBWs, mean loss CAC, AM, and KD mice by using the SIMCA 14.1 soft- rate: CT26, 14.8%; LLC, 12.4%), hind limb muscle ware (MKS Umetrics AB, Sweden). The metabolic path- weights (HLMWs, mean loss rate: CT26, 27.1%; LLC, way analysis was performed to identify significantly 26.2%), gastrocnemius weights (mean loss rate: CT26, altered metabolic pathways (significant pathways) on the 33.1%; LLC, 27.4%), and soleus weights (mean loss rate: MetaboAnalyst 4.0 webserver (https://www. CT26, 26%; LLC, 21%) as well as muscle strengths com- metaboanalyst.ca). This webserver was also used to ob- pared to the NOR mice (Fig. 1d–h; Fig. 2d–h). Com- tain heat-map plots of relative metabolite levels, and the pared to the NOR mice, the CAC mice exhibited Pearson’s correlation coefficients between catabolic pro- obvious myasthenia with the dramatic decline (CT26, tein expressions and metabolite levels. 29.6%; LLC, 27.1%) in grip strength (Fig. 1 h; Fig. 2 h). In both the CT26 model and the LLC model, AM mice ex- Transcriptomic analysis hibited alleviated cachexia features as reflected by main- The RNA-seq experiments were performed by Gene tained body weights, TFBWs, HLMWs, and Denovo Biotechnology Co. (Guangzhou, China). The gastrocnemius weights as well as improved muscle NOISeq R/Bioc package was used to identify differen- strengths compared with the CAC mice. Besides, the tially expressed genes (DEGs) from pairwise compari- amiloride treatment significantly increased soleus sons among the groups of mouse gastrocnemius with weights in the CT26 cachexia model, but not observably Zhou et al. Skeletal Muscle (2021) 11:17 Page 5 of 16 Fig. 1 Amiloride alleviated cachexia progression in the CT26 murine model. a, b Body weight and tumor volume growth curves of the mice during the processes of establishing the animal models. c–h Characterization of cachexia features in the mice on day 30 ( = 6-8). c, d Tumor weights and tumor-free body weights of the mice. e Hindlimb muscle weights normalized to tibia length. f Gastrocnemius weights normalized to tibia length. g Soleus weights normalized to tibia length. h Grip strengths normalized to bodyweight. Statistical significances: p > 0.05, NS; p < 0.05, *; p < 0.01, **; p < 0.001, ***. NOR, C57BL/6 or BALB/c normal control mice; CAC, CT26 cachexia mice; AM, amiloride-treated mice; KD, mice inoculated with Rab27-knockdown CT26 cells changed those in the LLC cachexia model (Fig. 1; Fig. 2). Interestingly, expressions of MyoD1 and Myogenin were The cross-section area analysis revealed a decreased myo- not statistically significantly changed both of which regu- fiber size in gastrocnemius of the CAC mice (cachectic lated muscle cell growth and differentiation (Fig. 3e, f). In gastrocnemius), indicating the muscle dysfunction condi- general, the amiloride treatment preserved muscle weights tion in the CT26 and CT26 cachexia models (Fig. 3a–d). and muscle strengths in tumor-bearing mice. As is known, skeletal muscle mass is influenced by the counter-balance between muscle degradation and myo- Amiloride treatment alleviated cachexia-related features genesis. In cachectic gastrocnemius, the decreased ratio of in mice p-FoxO3a/FoxO3a (Fig. S2) distinctly upregulated the ex- Cachexia is usually accompanied by fat loss. We ana- pressions of Atrogin-1 and MuRF-1 (Fig. 3e, f). The lyzed epididymal adipose tissues in the CT26 model and amiloride treatment significantly downregulated the levels the LLC model, especially white adipose tissues which of Atrogin-1 and MuRF-1 in gastrocnemius (Fig. 3e, f). are responsible for the organismal energy storage. Fig. 2 Amiloride alleviated cachexia progression in the LLC murine model. a, b Body weight and tumor volume growth curves of the mice during the processes of establishing the animal models. c–h Characterization of cachexia features in the mice on day 30 (n = 6-8). c, d Tumor weights and tumor-free body weights of the mice. e Hindlimb muscle weights normalized to tibia length. f Gastrocnemius weights normalized to tibia length. g Soleus weights normalized to tibia length. h Grip strengths normalized to bodyweight. Statistical significances: p > 0.05, NS; p < 0.05, *; p < 0.01, **; p < 0.001, ***. NOR, C57BL/6 or BALB/c normal control mice; CAC, LLC cachexia mice; AM, amiloride-treated mice; KD, mice inoculated with Rab27-knockdown LLC cells Zhou et al. Skeletal Muscle (2021) 11:17 Page 6 of 16 Fig. 3 Amiloride profoundly ameliorated gastrocnemius atrophy in the CT26 model and the LLC model. Cancer cachexia features of gastrocnemius were analyzed for the NOR, CAC, AM, and KD mice on day 30. a, b Representative microscope pictures of hematoxylin and eosin staining for gastrocnemius of the CT26 model (a) and the LLC model (b) (scale bar = 50 μm). c, d Myofiber size distributions of gastrocnemius of the CT26 model (c) and the LLC model (d) (n = 5, 200 myofibers were used). e Expressions of MuRF-1, Atrogin-1, MyoD1, and Myogenin proteins in gastrocnemius of the CT26 and LLC models. f Quantification of the expressed proteins (n = 4). Statistical significances: p > 0.05, NS; p < 0.05, *; p < 0.01, **; p < 0.001, ***; p < 0.0001, ****. NOR, C57BL/6 or BALB/c normal control mice; CAC, CT26/LLC cachexia mice; AM, amiloride-treated mice; KD, mice inoculated with Rab27-knockdown CT26/LLC cells Compared with the NOR mice, the CAC mice displayed corresponding to the profoundly upregulated serum IL-6 observable loss of epididymal adipose tissues, which was level. Significantly, the expression of phosphorylated partially restored in the AM mice (Fig. 3a, b; Fig. 4a, b). Stat3 was downregulated in gastrocnemius of the AM On the other hand, cachexia is mostly associated with mice (AM gastrocnemius; Fig. S2). In addition, given systemic inflammatory response. In both the CT26 that the p38 kinase participates in the cellular response model and the LLC model, the CAC mice showed a pro- to multiple stresses and inflammatory cytokines, we ana- foundly upregulated serum level of IL-6 and almost lyzed expressions of phosphorylated p38 and cleaved identical serum levels of TNF-α and IL-1β compared caspase-3, and observed significant enhancements of p38 with the NOR mice (Fig. S3c-e; Fig. S4c-e). Remarkably, kinase and caspase-3 activities in the CAC gastrocne- the AM mice displayed declined serum level of IL-6 but mius, which was partially reversed in the AM gastrocne- nearly unchanged serum levels of TNF-α and IL-1β mius (Fig. S2). compared to the CAC mice in both models (Fig. S3c-e; Considering that anorexia is a crucial factor of cach- Fig. S4c-e). Furthermore, the expression of down-stream exia, we continuously monitored food intakes across the phosphorylated Stat3 was remarkably upregulated in four groups of mice. In the CT26 model, the CAC mice gastrocnemius of the CAC mice (CAC gastrocnemius), showed a slight decrease in average food intake Zhou et al. Skeletal Muscle (2021) 11:17 Page 7 of 16 Fig. 4 Amiloride inhibited tumor-derived exosome release both in vitro and in vivo. a Gastrocnemius weights of the NOR and NOR-AM mice in the CT26/LLC models (n = 6). b Plasma exosome densities of the NOR and NOR-AM mice in the CT26/LLC models (n = 6). c Body weights of the NOR and NOR-AM mice in the CT26/LLC models (n = 6). d Exosome densities in culture media of the CT26, CT26-AM, LLC, and LLC-AM cells (n = 4). Cells were treated by amiloride at 10 μM for 6 h. e Plasma exosome densities of the NOR, CAC, AM, and KD mice in the CT26 model (n = 8). f Plasma exosome densities of the NOR, CAC, AM, and KD mice in the LLC model (n = 8). Statistical significances: p > 0.05, NS; p < 0.05, *; p < 0.01, **; p < 0.001, ***. NOR, C57BL/6 or BALB/c normal control mice; NOR-AM, amiloride-treated NOR mice; CAC, CT26/LLC cachexia mice; AM, amiloride-treated CT26/LLC mice; KD, mice inoculated with Rab27-knockdown CT26/LLC cells; CT26-AM, amiloride-treated CT26 cells; LLC-AM, amiloride-treated LLC cells compared with the other three groups of mice (Fig. S3f). exosomes, LLC-AM exosomes). Significantly, the treat- Differently, in the LLC model, the CAC mice showed an ment of amiloride at 10 μM for 6 h evidently inhibited average food intake similar to the other three groups of exosome release from the CT26 and LLC tumor cells mice (Fig. S4f). Besides, both CT26 and LLC models did (Fig. 4d), indicating that the amiloride treatment pro- not exhibit observable differences in heart weight among foundly decreased exosomes produced by the tumor the four groups of mice (Fig. S3g; Fig. S4g). cells. Furthermore, we isolated exosomes from plasma of Amiloride treatment ameliorated muscle wasting through the four groups of mice in the CT26 model and the LLC inhibiting tumor-derived exosome release model and quantified plasma exosome densities. Not- We further investigated molecular mechanisms under- ably, the CAC mice showed dramatically increased lying the favorable anti-cachexia effects of the amiloride plasma exosome densities compared to the NOR mice in treatment. The AM mice exhibited similar tumor both models (Fig. 4e, f). More importantly, the AM mice weights to the CAC mice in both the CT26 model and exhibited distinctly reduced plasma exosome densities in LLC model (Fig. 1c; Fig. 2c). Furthermore, the weight of the two cachexia models (Fig. 4e, f), indicating that the the NOR-AM gastrocnemius did not show statistically amiloride treatment efficiently inhibited tumor-derived significant change relative to the NOR gastrocnemius in exosome release. Note that the amiloride treatment did both models (Fig. 4a). not significantly influence the normal tissue-derived exo- Given that amiloride can exert inhibitory effects on some release as indicated by the statistical comparison the cellular exosome release [12, 13], we treated the of plasma exosome density between the NOR-AM mice CT26 and LLC cells with amiloride at various concentra- and the NOR mice (Fig. 4b). Besides, the NOR-AM mice tions (1–200 μM) for 6 h (plasma half-life of amiloride is did not show observable muscular toxicities as reflected about 6–9 h) and did not observe statistically significant by basic unchanged body weights and gastrocnemius changes in the viabilities of both tumor cells (Fig. S5). weights compared to the NOR mice (Fig. 4a, c). We isolated exosomes from culture media of the CT26 We isolated and characterized exosomes from plasma cells, CT26-AM cells, LLC cells, and LLC-AM cells and of the CAC patients and the Non-CAC patients (CAC characterized the four groups of tumor-derived exo- exosomes, non-CAC exosome; Fig. S6-8). Based on the somes (CT26 exosomes, CT26-AM exosomes, LLC limited patient samples used in this study, we found that Zhou et al. Skeletal Muscle (2021) 11:17 Page 8 of 16 exosome densities in patient plasma varied even by 20- the NOR, AM and KD gastrocnemius (Fig. 5a). Similarly, fold among individuals (Fig. S8b). In addition, the CAC the supervised partial least-squares discriminant analysis patients did not display significant statistical differences (PLS-DA) scores plots display significant metabolic dis- in plasma exosome density from the Non-CAC patients tinctions in gastrocnemius between the NOR and CAC (Fig. S8b). mice, the CAC and AM mice, and the CAC and KD On the other hand, the exosomes derived from both mice (Fig. 5b). Significant metabolites were identified the CT26/LLC culture media and the CAC patient from the PLS-DA models (Fig. S12). Furthermore, we plasma induced apparent myotube atrophy in vitro, as performed univariate analysis to quantitatively compare indicated by the obviously decreased myotube diameters metabolite levels among the four groups of gastrocne- (Fig. S8a-d, h-i) and distinctly enhanced expressions of mius and identify differential metabolites (Table S3). Atrogin-1 and MuRF-1 relative to PBS controls (Fig. S9). Combining the significant metabolites with the dif- Our observation that plasma exosomes of the CAC pa- ferential metabolites, we determined characteristic tients profoundly promoted myotube atrophy, support- metabolites for the four groups of mouse gastrocne- ing the previous studies [7, 21]. mius (Table S4). Totally, 18, 20, and 22 characteristic To further confirm the experimental observation that metabolites were identified in gastrocnemius for CAC tumor-derived exosomes significantly induced muscle vs. NOR, AM vs. CAC, and KD vs. CAC, respectively. atrophy in cancer cachexia, we constructed Rab27 Among 17 shared characteristic metabolites, 10 me- knock-down CT26 and LLC cell lines (CT26-KD cells, tabolites (glucose, IMP, glycine, creatine, methylmalo- LLC-KD cells) as positive controls (Fig. S8e, f). Expect- nate, niacinamide, aspartate, glutamate, fumarate and edly, knock-down of Rab27a and Rab27b potently tyrosine) were increased in the CAC gastrocnemius inhibited exosome release from tumor cells without but consistently decreased in the AM and KD remarkably changing viabilities of the tumor cells gastrocnemius, and 7 metabolites (lactate, taurine, in- (Fig. S8g; Fig. S10). The exosomes isolated from cul- osine, alanine, 2-phosphoglycerate, ATP and glutam- ture media of the CT26-KD and LLC-KD tumor cells ine) were decreased in the CAC gastrocnemius but displayed markedly reduced abilities to induce myo- consistently increased in the AM and KD gastrocne- tube atrophy relative to their corresponding wild-type mius (Table S4). counterparts (Fig. S8h, i; Fig. S9b).Asexpected,the Furthermore, significant pathways were identified by KD mice inoculated with the Rab27 knock-down performing the metabolic pathway analysis for the four tumor cells displayed alleviated cachexia features, as groups of gastrocnemius (Fig. 5c). Notably, both the indicated by increased body weights, TFBWs, amiloride treatment and Rab27 knock-down significantly HLMWs, and gastrocnemius weights as well as im- altered the identical metabolic pathways. Interestingly, 8 proved muscle strengths compared with the CAC of the 11 significant pathways were closely related to mice in the two models (Fig. 1;Fig. 2). Furthermore, amino acid metabolism. The heat-map plot of relative theCAC,AM, andKDmiceof bothmodelsshowed metabolite levels illustrates that a large number of amino similar tumor weights on day 30 (Fig. 1c; Fig. 2c), in- acids were increased in the CAC gastrocnemius, includ- dicating that amiloride inhibited tumor-derived exo- ing alanine, glutamine, aspartate, and glycine (Fig. 6a). some release without significantly affecting tumor The upregulated levels of phosphorylated AMPK-Tyr112 growth in vivo. These results indicated that the and downregulated levels of phosphorylated Akt1 indi- amiloride treatment could profoundly ameliorate cated a strongly activated catabolism in cachectic gastro- cachectic gastrocnemius atrophy and thereby alleviate cnemius (Fig. 6b, c). Moreover, the raised ratio of LC3II/ cancer cachexia through inhibiting tumor-derived exo- LC3I and declined levels of MHC and MLC were indica- some release. tive of promoted autophagy in cachectic gastrocnemius (Fig. 6b, c). Additionally, the levels of upregulated amino Amiloride treatment improved hyper-catabolism in acids were positively correlated with expressions of cata- gastrocnemius bolic proteins, but negatively correlated with expressions To mechanistically understand metabolomic features of of anabolic proteins in the CAC gastrocnemius (Fig. 6d). skeletal muscles in the four groups of mice, we conducted Transcriptomic profiling identified 1466 upregulated NMR-based metabolic profiling of gastrocnemius. A total genes and 1941 downregulated genes in the CAC gastro- of 32 metabolites were identified (Fig. S11a; Table S2). cnemius relative to the NOR gastrocnemius, of which 753 The resonance assignments were confirmed by using 2D differentially expressed genes (DEGs) were shared by the 1 13 H- C HSQC spectra (Fig. S11b). pairwise comparisons of CAC vs. NOR, AM vs. CAC, and The unsupervised principal component analysis (PCA) KD vs. CAC (Fig. S13a; Fig. S14a). The KEGG enrichment scores plot illustrates that the metabolic profile of the analysis based on the identified DEGs screened out dis- CAC gastrocnemius is distinctly different from those of tinctly changed metabolism-related processes including Zhou et al. Skeletal Muscle (2021) 11:17 Page 9 of 16 Fig. 5 Metabolomic analyses of the NOR, CAC, AM and KD gastrocnemius. a PCA scores plot of 1D H NMR data obtained on aqueous extracts derived from gastrocnemius of the NOR, CAC, AM and KD mice (n = 7). b PLS-DA scores plots and cross-validation plots of 1D H-NMR data obtained from gastrocnemius of the NOR, CAC, AM and KD mice. Top: CAC mice vs. NOR mice; Middle: AM mice vs. CAC mice; Bottom: KD mice vs. CAC mice. The PLS-DA models were cross-validated to evaluate their robustness with random permutation tests (200 cycles). c Metabolic pathway analyses of CAC vs. NOR, AM vs. CAC, KD vs. CAC, using the Pathway Analysis module provided by MetaboAnalyst 4.0. Numbers in the three panels represent significantly altered metabolic pathways, which were identified with -ln(p) > 2.995 (corresponding to p < 0.05) and pathway impact value > 0.2: 1, taurine and hypotaurine metabolism; 2, nicotinate and nicotinamide metabolism; 3, pyruvate metabolism; 4, phenylalanine metabolism; 5, glycine, serine and threonine metabolism; 6, alanine, aspartate and glutamate metabolism; 7, glutathione metabolism; 8, D-glutamine and D-glutamate metabolism; 9, phenylalanine, tyrosine and tryptophan biosynthesis; 10, histidine metabolism; 11, starch and sucrose metabolism. NOR, C57BL/6 or BALB/c normal control mice; CAC, CT26/LLC cachexia mice; AM, amiloride-treated mice; KD, mice inoculated with Rab27-knockdown CT26/LLC cells Zhou et al. Skeletal Muscle (2021) 11:17 Page 10 of 16 Fig. 6 Metabolomic analyses showed that the amiloride treatment attenuated hyper-catabolism in cachectic gastrocnemius. a Heat-map plot of relative levels of the identified metabolites (n = 6). b Expressions of AMPK, p-AMPK, LC3, MHC, MLC, Akt1, p-Akt1(T308), and p-Akt1(S473) proteins analyzed by using western blot. c Quantification of the expressed proteins (n = 4). d Heat-map plot of the correlations between the catabolic/ anabolic protein expressions and identified metabolite levels in the CAC gastrocnemius. The gradient red/blue colors indicate that the positive/ negative correlations, and significant correlations were identified with the criterion of |r| > 0.576 (n = 6). Statistical significances: p > 0.05, NS; p < 0.05, *; p < 0.01, **; p < 0.001, ***; p < 0.0001, ****. NOR, C57BL/6 or BALB/c normal control mice; CAC, CT26/LLC cachexia mice; AM, amiloride- treated mice; KD, mice inoculated with Rab27-knockdown CT26/LLC cells; BCAAs, branch-chain amino acids; IMP, inosine monophosphate; 2PG, 2- phosphoglycerate; 3-HB, 3-hydroxybutyrate; MLC, myosin light chain; MHC, myosin heavy chain carbohydrate metabolism, lipid metabolism, and amino gastrocnemius (Fig. 7b, c; Table S5), which facilitated acid metabolism (Fig. S13b; Fig. S14b). Consistently, the mobilization and oxidation of adipose tissues. loss of epididymal adipose tissues in cachexia mice Overall, the amiloride treatment substantially im- (Fig. S2a, b; Fig. S3a, b) was accompanied by upregu- proved hyper-catabolism in cachectic gastrocnemius, lated expressions of fatty acid translocase CD36 and and Rab27 knock-down showed a similar improvement Acyl-coenzyme A thioesterase 1 (Acot1) in the CAC of hyper-catabolism (Fig. 6; Fig. 7). Zhou et al. Skeletal Muscle (2021) 11:17 Page 11 of 16 Fig. 7 Transcriptomic analyses exhibited that the amiloride treatment improved glycolysis and ketone body oxidation in cachetic gastrocnemius. a Heat-map plot of relative transcription levels of muscular atrophy-related genes (n = 4). b Expressions of ACOT1, CD36, OXCT1, BDH2, and ACAT1 proteins. c Quantification of the expressed proteins (n = 4). Statistical significances: p > 0.05, NS; p < 0.05, *; p < 0.01, **; p < 0.001, ***; p < 0.0001, ****. NOR, C57BL/6 or BALB/c normal control mice; CAC, CT26/LLC cachexia mice; AM, amiloride-treated mice; KD, mice inoculated with Rab27-knockdown CT26/LLC cells; ACOT1, acyl-coenzyme A thioesterase 1; CD36, fatty acid translocase; OXCT1, 3-oxoacid CoA transferase 1; BDH2, 3-hydroxybutyrate dehydrogenase 2; ACAT1, acetyl coenzyme A acetyltransferase 1 Amiloride treatment improved blocked glycolysis and transporters (MCT1 and MCT4; Fig. S2) and the domin- impeded ketone body oxidation in gastrocnemius ant organismal ketone body—3-hydroxybutyrate (3-HB; Glycolysis is one of the fundamental energy sources in Table S3). These results were indicative of blocked gly- muscle cells, and ketone bodies are alternative energy colysis and impeded ketone body oxidation in the CAC sources under harsh conditions. Glucose was signifi- gastrocnemius. Significantly, both the AM and KD cantly increased in the CAC gastrocnemius relative to gastrocnemius displayed profoundly improved glycolysis the NOR gastrocnemius, contrarily, the glycolysis end and ketone body oxidation via the inhibition of tumor- product—pyruvate—was decreased (Table S3). The ex- derived exosome release (Fig. 7; Fig. S15). pressions of multiple glycolytic catalyzing enzymes were downregulated, but the expression of glycolytic inhib- Discussion ition enzyme (pyruvate dehydrogenase kinase isoenzyme Cancer cachexia evidently reduces patient survival and 4, Pdk4) was upregulated in cachectic gastrocnemius quality of life due to its high incidence and mortality rate relative to normal control (Fig. 7a; Table S5). Further- [4, 22]. Developments of efficient therapeutic strategies more, the CAC gastrocnemius exhibited more than 2- are urgently required for the CAC treatments. Previous fold decreases in expressions of ketone body oxidation studies have demonstrated that amiloride possesses enzymes 3-oxoacid CoA transferase 1 (OXCT1) and 3- potassium-sparing diuretic function, which has been hydroxybutyrate dehydrogenase 2 (BDH2; Fig. 7b,c) but clinically used in the treatments of hypertension and substantially unchanged expressions of ketone body edema due to heart failure [11, 23–25]. Moreover, Zhou et al. Skeletal Muscle (2021) 11:17 Page 12 of 16 amiloride can inhibit exosome release from cells and re- metabolic mechanisms underlying the CAC progression. verse exosome-promoted pathogenic processes [26, 27]. Previously, we established a mouse model of gastric can- In this study, we established CT26/LLC-induced mouse cer cachexia by orthotopically implanting BGC823 cells models of lung/colorectal cancer cachexia, assessed the and identified significantly impaired metabolic pathways effects of the amiloride treatment for alleviating muscle in cachectic gastrocnemius . The two works of ours atrophy in the two cachexia models, and addressed the unanimously confirm that aberrant catabolism in cach- underlying molecular mechanisms. Our results reveal ectic gastrocnemius is triggered primarily by upregulated that amiloride is a potential therapeutic drug capable of E3 ligases, which exhibit profoundly increased levels of ameliorating muscle wasting in cancer cachexia through amino acids (isoleucine, leucine, valine, glutamate). inhibiting tumor-derived exosome release (Fig. 8). Interestingly, the CT26/LLC cachexia models show Both the CT26 model and the LLC models showed re- blocked glycolysis in gastrocnemius, while the BGC823 markable cachexia features and significant metabolic im- cachexia model displays promoted glycolysis. The meta- pairments in gastrocnemius. Significantly, the amiloride bolic distinction might be dictated by several factors in- treatment prevented the losses of body weight, skeletal cluding differences in food intake, inflammatory muscle, and fat mass, which did not obviously affect cytokines, tumor type, and stage of cachexia. Addition- tumor growth and induce observable systemic toxicities ally, the discrepancy between orthotopic and subcutane- in normal control mice, as indicated by basically un- ous cachexia models might be associated with discrepant changed body weights and gastrocnemius weights of the metabolic features in gastrocnemius, potentially contrib- NOR-AM mice relative to the NOR mice. More import- uting to the metabolic distinction, even though these antly, multiple cachexia features were improved, as evi- models undergo similar muscular atrophic processes. denced by downregulated expressions of muscular In the past decades, strategies for cachexia therapy atrophic proteins, partially restored muscle strength and mainly focused on the development of new drugs, in- neutralized systemic inflammation. The further mechan- cluding ghrelin and ghrelin receptor agonists, myostatin istic study revealed that the amiloride treatment pro- antagonists, inflammatory cytokine neutralizing anti- foundly inhibited tumor-derived exosome release and bodies, and natural product extracts [28–37]. Further- attenuated hyper-catabolism, significantly improved the more, recent studies have identified several key metabolic impairments in cachectic gastrocnemius, molecules (namely cachectin) and correspondingly ex- thereby alleviating the CAC progression. plored inhibitory chemicals for alleviating the CAC pro- By integrative metabolomic and transcriptomic ana- gression. However, most of these efforts have not lyses, we identified significantly impaired metabolic obtained satisfactory clinical trial results . The pri- pathways in cachectic gastrocnemius relative to normal mary reason might attribute to the heterogeneity of control, including promoted muscular catabolism, inhib- cachectin resulting from either different tumor types or ited muscular protein synthesis, blocked glycolysis, and being aroused by inherent reprogramming processes impeded ketone body oxidation. Expectedly, the im- within identical tumor types, which requires more com- paired metabolic pathways potentially contribute to prehensive investigation. Besides, it is time-consuming Fig. 8 Graphic model of the amiloride treatment ameliorating cachectic muscle wasting through inhibiting tumor-derived exosome release Zhou et al. Skeletal Muscle (2021) 11:17 Page 13 of 16 for clinical trials and further approval of newly devel- tumor tissues. The amiloride treatment (2 mg/kg/day) oped drugs. In contrast, applying existing drugs for new did not statistically change plasma exosome densities de- indications could be a more feasible alternative strategy. rived from normal tissues in the normal control mice, as Given that most of the existing drugs are usually associ- indicated by the observation that the NOR-AM mice ated with well addressed pharmacokinetic and pharma- and the NOR mice did not show statistically significant codynamic properties and toxicity profiles, the alterative difference in plasma exosome density. Thus, it could be strategy might greatly reduce the time required to de- expected that the decreases in plasma exosome density velop novel drugs for the CAC treatment. in the amiloride-treated CAC mice might reflect the re- Amiloride has been clinically used for nearly three de- ductions in exosomes produced by the tumor cells. Fur- cades in the treatments of hypertension, edema and con- thermore, the quantitative analysis of exosome densities gestive heart failure [23–25, 39]. In the present study, we in culture media of the CT26/LLC tumor cells indicated demonstrated for the first time the therapeutic potentials that the amiloride treatment (10 μM, 6 h) profoundly de- of amiloride in the treatments of cancer cachexia. Previ- ceased exosome release from the tumor cells but not ob- ously, Cameron et al. showed that increased Na contents servably affected viabilities of the tumor cells. These in multiple tissues of H6 hepatoma-induced cachexia mice experimental results reveal that the amiloride treatment can be partially reversed by the amiloride treatment . significantly reduces exosome production by the tumor Their study mainly determined amiloride-induced cells. Nevertheless, more experiments should be per- changes of Na and other ions contents in liver cach- formed to further support the conclusion that amiloride exia mice, but did not further examine whether cach- ameliorates muscle wasting through inhibiting tumor- exia symptoms were relieved. derived exosome release and thereby alleviates cancer The most outstanding characteristic of amiloride is the cachexia. Potentially, the alleviation of cachexia could be + + + + efficient inhibitory effects on the Na /H and Na /K due to several factors including reduced circulating IL-6, transporters [11, 23, 26, 41]. Under normal conditions, maintenance of cardiac function, and direct effects on + + + the Na /H transporter mediates H efflux from the skeletal muscle, etc. These factors are worthy of further cells in exchange for Na influx. Further works indicated exploration in future studies. + + that blocking Na influx/ H efflux can inhibit cell A previous study exhibited that the amiloride treat- growth, and tumor cells are more vulnerable to the ment at two doses of 10 mg/kg and 15 mg/kg did not ex- blocking of H efflux than normal cells which might hibit significant toxic effects in a multiple myeloma endow amiloride antineoplastic effects [26, 27, 41–44]. xenograft murine model, as indicated by basically un- In this study, we observed that the treatment of changed body weights . Consistently, the amiloride amiloride at various concentrations (1–200 μM) for 6 h treatment at 2 mg/kg used in our study did not show ob- did not observably change cell viabilities of tumor cells servable muscular toxic effects, as indicated by basically Moreover, both the CT26 model and the LLC model unchanged body weights and gastrocnemius weights in showed that the amiloride treatment did not statistically the NOR-AM mice relative to the NOR mice. However, significantly affect tumor growth in the CAC mice. our limited results only reflect that amiloride has not These results suggest that the antineoplastic effects of significant muscle toxicity in healthy mice. A systemic amiloride do not significantly contribute to amiloride- toxicity test should be conducted in further studies. mediated alleviation of the CAC progression. Nevertheless, the amiloride treatment may be beneficial On the other hand, intracellular Na contents also to the amelioration of cachectic muscle wasting and thus regulate the trafficking of extracellular vesicles including to the alleviation of the CAC progression. + 2+ exosomes . The Na /Ca antiporter mediates the Previous studies documented that cancer cachexia can + 2+ appropriate Na efflux in exchange for Ca influx. The be induced by multiple factors [35, 38, 46, 47], including 2+ increased cytoplasmic Ca content is a prerequisite for cytokines, hormones, tumor factors, and gut microbes. multi-vesicular bodies (MVBs) generation and following More significantly, the present study displayed that the exosome biogenesis [40, 45]. Thus, amiloride probably exosomes isolated from plasma of cancer cachexia pa- 2+ decreases intracellular Ca content through mediating tients and culture media of the CT26/LLC tumor cells Na efflux by the antiporter, and obstructs cellular exo- induced remarkable myotube atrophy, well confirming some release. Note that we could not currently confirm that tumor-derived exosomes can induce muscle wasting this speculation as this study had not measured the con- in cancer cachexia. Furthermore, it has been demon- 2+ + tents of Ca and Na . strated that several individual components in tumor- Both the CL26 and LLC murine models showed dra- derived exosomes can significantly contribute to cancer matic decreases in plasma exosome densities of the AM cachexia progression, such as miR-21 , heat shock mice relative to the CAC mice. Note that plasma exo- proteins , and metal ions . Expectedly, exploration somes could also be derived from normal tissues besides of other key components in tumor-derived exosomes Zhou et al. Skeletal Muscle (2021) 11:17 Page 14 of 16 would be greatly beneficial to further understand the for extrahepatic tissues (mainly in the brain and skeletal molecular mechanisms of exosome-induced muscle muscles). We detected downregulated expressions of key wasting and early diagnosis of cancer cachexia. Such enzymes (BDH2 and OXCT1) for ketone body oxidation studies are worthy of being conducted in the future. in cachectic gastrocnemius, implying that ketone body The present study revealed that the amiloride treat- oxidation was potentially impeded. Consistently, the in- ment can significantly inhibit tumor-derived exosome hibition of tumor-derived exosome release by the release, and thereby profoundly ameliorate muscle wast- amiloride treatment can alleviate the impediment of ke- ing and alleviate the CAC progression, indicating clinical tone body oxidation in cachectic gastrocnemius. We potentials of amiloride for treatments of the CAC thus speculate that the impeded ketone body oxidation patients. Similar to amiloride, some other drugs or might attribute to a potential protective mechanism for chemical inhibitors, such as GW4869 [49–51], omepra- ensuring the preferential supply of ketone bodies to the zole , chlorpromazine , and statins [53–55], also brain. Other alternative energy sources such as amino possess the effects of inhibiting cellular exosome release, acids and acetyl-CoAs are available for other organs to and could be exploited as potential drugs against cancer sustain energy production and biomolecules synthesis. cachexia too. Here, we could not confirm this speculation as we did Cachectic muscle atrophy primarily results from an not assess the ketone body utilization in the brain of imbalance of catabolism and anabolism . Amiloride- cachexia mice. Nevertheless, our study could be an in- mediated inhibition of tumor-derived exosome release novative supplementation for clarifying the molecular significantly improves metabolic impairments in cachec- mechanisms of skeletal muscle atrophy in cancer cach- tic gastrocnemius. As is known, both activation of the exia. Expectedly, it is of great value to exploit ketone AMPK signaling cascade and inhibition of the Akt path- body metabolism-related enzymes as novel targets for way are responsible for muscle wasting in cancer cach- improving ketone body utilization and thereby amelior- exia. Moreover, activation of the apoptosis pathway ating cachectic muscle wasting. It seems that enhancing mediated by the p38 kinase also contributes to myofi- ketone body utilization rather than simply serving cach- brillar protein degradation and muscle dysfunction. ectic mice with ketogenetic diets , might be a more Cachectic gastrocnemius exhibited a more than 5-fold efficient way to ameliorate muscle wasting in cancer increase in the ratio of p-AMPK (Tyr112)/AMPK and cachexia. significant decreases in the ratio of p-Akt1/Akt1, indicat- ing the promoted catabolism and inhibited anabolism in Conclusions cancer cachexia. The metabonomic analysis of the CAC We have demonstrated that amiloride is a potential drug gastrocnemius exhibited upregulated amino acid levels capable of ameliorating muscle wasting in cancer cach- relative to normal control, further confirming the pro- exia. Our results reveal that the amiloride treatment sig- moted degradations and inhibited syntheses of muscular nificantly improves metabolic impairments in cachectic proteins. The accumulated amino acids are capable of gastrocnemius through efficiently inhibiting tumor- acting as supplementary sources for both TCA cycle derived exosome release, including promoted muscular anaplerosis and glycolysis, fulfilling increased energy de- catabolism, inhibited muscular protein synthesis, mand in cancer cachexia. Furthermore, the transcrip- blocked glycolysis, and impeded ketone body oxidation. tomic analysis of the CAC gastrocnemius showed Our results are beneficial to mechanistic understanding downregulated expressions of five glycolytic catalyzing the effects of the amiloride treatment for ameliorating enzymes and the upregulated expression of a glycolytic muscle wasting in cancer cachexia and alleviating the inhibiting enzyme relative to the NOR gastrocnemius, CAC progression. Our study sheds light on the poten- indicating that blocked glycolysis significantly promoted tials of amiloride in cachexia therapy. Further studies are muscle wasting in cancer cachexia. More significantly, needed both to validate the practical universalities of the amiloride-mediated inhibition of tumor-derived exosome amiloride treatment for other cancer cachexia models, release enhances the promoted muscular proteolysis, and to explore clinical potentials of amiloride for im- inhibited muscular protein synthesis, and blocked gly- proving the CAC treatment. colysis in cachectic gastrocnemius. In addition, we found that tumor-derived exosomes Abbreviations CAC: Cancer cachexia; NOR: Normal control; AM: Amiloride treated; are also involved in the regulation of ketone body me- KD: Rab27 knock-down; MVBs: Multi-vesicular bodies; CM: Conditioned tabolism in skeletal muscle. In harsh energy conditions media; HSFCM: High-sensitivity nano flow cytometer; FID: Free induction (continuous intensive exercise training; constant hunger, delay; DEGs: Differentially expressed genes; FC: Fold change; FDR: False discovery rate; TFBWs: Tumor-free body weights; HLMWs: Hind limb muscle etc.), ketone bodies are mostly transported across the weights; PCA: Principal components analysis; PLS-DA: Supervised partial least- blood-brain barrier to fuel the brain. In cachexia mice, squares discriminant analysis; OXCT1: 3-Oxoacid CoA transferase 1; BDH2: 3- hepatogenic ketone bodies are available energy sources Hydroxybutyrate dehydrogenase 2; FFA: Free fatty acid; FAA: Free amino Zhou et al. Skeletal Muscle (2021) 11:17 Page 15 of 16 acid; KB: Ketone body; α-KG: α-Ketoglutarate; AcAc: Acetoacetic acid; F-1,6- 361005, China. Research and Communication Center of Exercise and Health, BP: Fructose-1,6-bisphosphate; G3P: Glyceraldehyde-3-phosphate; Xiamen University of Technology, Xiamen 361024, China. High-field NMR DHAP: Dihydroxyacetone phosphate; 3PG: 3-Phosphoglycerate; 2PG: 2- Center, College of Chemistry and Chemical Engineering, Xiamen University, Phosphoglycerate; PEP: Phosphoenolpyruvate; 3-HB: 3-Hydroxybutyrate; Xiamen 361005, China. IMP: Inosine monophosphate Received: 17 January 2021 Accepted: 23 June 2021 Supplementary Information The online version contains supplementary material available at https://doi. org/10.1186/s13395-021-00274-5. References 1. Fearon K, Strasser F, Anker SD, Bosaeus I, Bruera E, Fainsinger RL, et al. Definition and classification of cancer cachexia: an international consensus. Additional file 1: Supplementary Fig. S1-S15 and Tables S1-S5. Lancet Oncol. 2011;12(5):489–95. https://doi.org/10.1016/S1470-2045(1 Additional file 2: The DEGs identified in the transcriptomics study. 0)70218-7. 2. Porporato PE. Understanding cachexia as a cancer metabolism syndrome. Additional file 3: The metabolomics data obtained from the present Oncogenesis. 2016;5(2):e200. https://doi.org/10.1038/oncsis.2016.3. study. 3. Baracos VE, Martin L, Korc M, Guttridge DC, Fearon KCH. Cancer-associated cachexia. Nat Rev Dis Primers. 2018;4(1):17105. https://doi.org/10.1038/ Acknowledgements nrdp.2017.105. We thank Prof. Xiaomei Yan from College of Chemistry and Chemical 4. Lok C. Cachexia: The last illness. Nature. 2015;528(7581):182–3. https://doi. Engineering, Xiamen University, for kindly providing the apparatus of HSFCM. org/10.1038/528182a. 5. World Health Organization. [https://www.who.int/en/news-room/fact- Authors’ contributions sheets/detail/cancer]. Accessed 16 Jan 2021. L.Z., D.L., and C.H. conceived this project. L.Z., T.Z., R.L., S.L., H.Z., and H.L. 6. Argilés JM, Busquets S, Stemmler B, López-Soriano FJ. Cancer cachexia: performed the experiments. H.Z. and W.L. provided clinical samples. L.Z., C.H., understanding the molecular basis. Nat Rev Cancer. 2014;14(11):754–62. W.S. B.J, Q.L., and D.L. performed the data analyses and helped with the https://doi.org/10.1038/nrc3829. discussions. L.Z., D.L., and C.H. wrote this manuscript. All authors commented 7. Chitti SV, Fonseka P, Mathivanan S. Emerging role of extracellular vesicles in on the manuscript. The authors read and approved the final manuscript. mediating cancer cachexia. Biochem Soc Trans. 2018;46(5):1129–36. https:// doi.org/10.1042/BST20180213. Funding 8. Sagar G, Sah RP, Javeed N, Dutta SK, Smyrk TC, Lau JS, et al. Pathogenesis of This work was supported by the National Natural Science Foundation of pancreatic cancer exosome-induced lipolysis in adipose tissue. Gut. 2016; China (No. 31971357) and the Open Research Fund of State Key Laboratory 65(7):1165–74. https://doi.org/10.1136/gutjnl-2014-308350. of Cellular Stress Biology, Xiamen University (SKLCSB2020KF002). 9. Zhang G, Liu Z, Ding H, Zhou Y, Doan HA, Sin KWT, et al. Tumor induces muscle wasting in mice through releasing extracellular Hsp70 and Hsp90. Availability of data and materials Nat Commun. 2017;8(1):589. https://doi.org/10.1038/s41467-017-00726-x. Transcriptome datasets can be found with a GEO accession number: 10. Wu Q, Sun S, Li Z, Yang Q, Li B, Zhu S, et al. Tumour-originated exosomal GSE173250. Other datasets supporting the conclusions of this article are miR-155 triggers cancer-associated cachexia to promote tumour included within the article and its additional files. progression. Mol Cancer. 2018;17(1):155. https://doi.org/10.1186/s12943-018- 0899-5. Declarations 11. Tang CM, Presser F, Morad M. Amiloride selectively blocks the low threshold The authors declare that the research was conducted in the absence of any (T) calcium channel. Science. 1988;240(4849):213–5. https://doi.org/10.1126/ commercial or financial relationships that could be construed as a potential science.2451291. conflict of interest. 12. Chalmin F, Ladoire S, Mignot G, Vincent J, Bruchard M, Remy-Martin JP, et al. Membrane-associated Hsp72 from tumor-derived exosomes mediates Ethics approval and consent to participate STAT3-dependent immunosuppressive function of mouse and human All animal studies were performed in Xiamen University Laboratory Animal myeloid-derived suppressor cells. J Clin Invest. 2010;120(2):457–71. https:// Center, according to protocols approved by the Institutional Animal Care doi.org/10.1172/JCI40483. and Use Committee of Xiamen University. The study protocol of collecting 13. Fitzner D, Schnaars M, van Rossum D, Krishnamoorthy G, Dibaj P, Bakhti M, blood samples of cancer patients was approved by the Ethics Committee of et al. Selective transfer of exosomes from oligodendrocytes to microglia by Affiliated Zhongshan Hospital of Xiamen University, China. Patients with macropinocytosis. J Cell Sci. 2011;124(3):447–58. https://doi.org/10.1242/jcs. colorectal cancer, lung cancer, and gastric cancer were enrolled, which were informed and wrote the consent. 14. He WA, Calore F, Londhe P, Canella A, Guttridge DC, Croce CM. Microvesicles containing miRNAs promote muscle cell death in cancer Consent for publication cachexia via TLR7. Proc Natl Acad Sci U S A. 2014;111(12):4525–9. https:// Not applicable doi.org/10.1073/pnas.1402714111. 15. Tian Y, Ma L, Gong M, Su G, Zhu S, Zhang W, et al. Protein profiling and Competing interests sizing of extracellular vesicles from colorectal cancer patients via flow The authors declare that they have no competing interests. cytometry. ACS Nano. 2018;12(1):671–80. https://doi.org/10.1021/acsnano. 7b07782. Author details 16. Dragovic RA, Gardiner C, Brooks AS, Tannetta DS, Ferguson DJ, Hole P, et al. Key Laboratory for Chemical Biology of Fujian Province, MOE Key Laboratory Sizing and phenotyping of cellular vesicles using Nanoparticle Tracking of Spectrochemical Analysis & Instrumentation, College of Chemistry and Analysis. Nanomedicine. 2011;7(6):780–8. https://doi.org/10.1016/j.nano.2011. Chemical Engineering, Xiamen University, Xiamen 361005, China. Xiamen 04.003. Cardiovascular Hospital, Xiamen University, Xiamen 361000, China. 17. Rikkert LG, Nieuwland R, Terstappen L, Coumans FAW. Quality of Department of Oncology, Institute of Gastrointestinal Oncology, Zhongshan extracellular vesicle images by transmission electron microscopy is operator Hospital, Xiamen University, Xiamen 361004, China. State Key Laboratory of and protocol dependent. J Extracell Vesicles. 2019;8(1):1555419. https://doi. Cellular Stress Biology, School of Life Sciences, Xiamen University, Xiamen org/10.1080/20013078.2018.1555419. 361102, China. Department of Medical Oncology, Xiang’an Hospital of 18. Cui P, Shao W, Huang C, Wu CJ, Jiang B, Lin D. Metabolic derangements of Xiamen University, Xiamen, China. Department of Gastrointestinal Surgery, skeletal muscle from a murine model of glioma cachexia. Skelet Muscle. The Affiliated Zhongshan Hospital, Xiamen University, Xiamen 361004, Fujian, 2019;9(1):3. https://doi.org/10.1186/s13395-018-0188-4. China. Collaborative Innovation Center of Chemistry for Energy Materials, 19. Cui P, Huang C, Guo J, Wang Q, Liu Z, Zhuo H, et al. Metabolic profiling of College of Chemistry and Chemical Engineering, Xiamen University, Xiamen tumors, sera, and skeletal muscles from an orthotopic murine model of Zhou et al. Skeletal Muscle (2021) 11:17 Page 16 of 16 gastric cancer associated-cachexia. J Proteome Res. 2019;18(4):1880–92. 39. Andersen H, Hansen PB, Bistrup C, Nielsen F, Henriksen JE, Jensen BL. https://doi.org/10.1021/acs.jproteome.9b00088. Significant natriuretic and antihypertensive action of the epithelial sodium 20. Beckonert O, Keun HC, Ebbels TM, Bundy J, Holmes E, Lindon JC, et al. channel blocker amiloride in diabetic patients with and without Metabolic profiling, metabolomic and metabonomic procedures for NMR nephropathy. J Hypertens. 2016;34(8):1621–9. https://doi.org/10.1097/HJH. spectroscopy of urine, plasma, serum and tissue extracts. Nat Protoc. 2007; 0000000000000967. 2(11):2692–703. https://doi.org/10.1038/nprot.2007.376. 40. Cameron IL, Hunter KE. Effect of cancer cachexia and amiloride treatment on the intracellular sodium content in tissue cells. Cancer Res. 1983;43(3): 21. Rong S, Wang L, Peng Z, Liao Y, Li D, Yang X, et al. The mechanisms and 1074–8. treatments for sarcopenia: could exosomes be a perspective research 41. Kim KM, Lee YJ. Amiloride augments TRAIL-induced apoptotic death by strategy in the future? J Cachexia Sarcopenia Muscle. 2020;11(2):348–65. inhibiting phosphorylation of kinases and phosphatases associated with the https://doi.org/10.1002/jcsm.12536. P13K-Akt pathway. Oncogene. 2005;24(3):355–66. https://doi.org/10.1038/sj. 22. Schmidt SF, Rohm M, Herzig S, Berriel DM. Cancer cachexia: more than onc.1208213. skeletal muscle wasting. Trends Cancer. 2018;4(12):849–60. https://doi.org/1 42. Harguindey S, Pedraz JL, García Cañero R, Pérez de Diego J, Cragoe EJ, Jr. 0.1016/j.trecan.2018.10.001. Hydrogen ion-dependent oncogenesis and parallel new avenues to cancer 23. Oxlund CS, Buhl KB, Jacobsen IA, Hansen MR, Gram J, Henriksen JE, et al. prevention and treatment using a H(+)-mediated unifying approach: pH- Amiloride lowers blood pressure and attenuates urine plasminogen related and pH-unrelated mechanisms. Crit Rev Oncog. 1995, 6:1-33. activation in patients with treatment-resistant hypertension. J Am Soc 43. Iorio J, Duranti C, Lottini T, Lastraioli E, Bagni G, Becchetti A, et al. K(V)11.1 Hypertens. 2014;8(12):872–81. https://doi.org/10.1016/j.jash.2014.09.019. Potassium channel and the Na(+)/H(+) antiporter NHE1 modulate adhesion- 24. Fuchs SC, Poli-de-Figueiredo CE, Figueiredo Neto JA, Scala LC, Whelton PK, dependent intracellular pH in colorectal cancer cells. Front Pharmacol. 2020; Mosele F, et al. Effectiveness of chlorthalidone plus amiloride for the 11:848. prevention of hypertension: the PREVER-prevention randomized clinical trial. 44. Tang JY, Chang HW, Chang JG. Modulating roles of amiloride in irradiation- J Am Heart Assoc. 2016;5(12). https://doi.org/10.1161/JAHA.116.004248. induced antiproliferative effects in glioblastoma multiforme cells involving 25. Fuchs FD, Scala LCN, Vilela-Martin JF, Whelton PK, Poli-de-Figueiredo CE, Akt phosphorylation and the alternative splicing of apoptotic genes. DNA Pereira ESR, et al. Effectiveness of chlorthalidone/amiloride versus losartan in Cell Biol. 2013;32(9):504–10. https://doi.org/10.1089/dna.2013.1998. patients with stage I hypertension and diabetes mellitus: results from the 45. Savina A, Furlán M, Vidal M, Colombo MI. Exosome release is regulated by a PREVER-treatment randomized controlled trial. Acta Diabetol. 2020. calcium-dependent mechanism in K562 cells. J Biol Chem. 2003;278(22): 26. Chang WH, Liu TC, Yang WK, Lee CC, Lin YH, Chen TY, et al. Amiloride 20083–90. https://doi.org/10.1074/jbc.M301642200. modulates alternative splicing in leukemic cells and resensitizes Bcr- 46. Zhou J, Liu B, Liang C, Li Y, Song YH. Cytokine signaling in skeletal muscle AblT315I mutant cells to imatinib. Cancer Res. 2011;71(2):383–92. https://doi. wasting. Trends Endocrinol Metab. 2016;27(5):335–47. https://doi.org/10.101 org/10.1158/0008-5472.CAN-10-1037. 6/j.tem.2016.03.002. 27. Rojas EA, Corchete LA, San-Segundo L, Martínez-Blanch JF, Codoñer FM, 47. Argilés JM, Stemmler B, López-Soriano FJ, Busquets S. Inter-tissue Paíno T, et al. Amiloride, an old diuretic drug, is a potential therapeutic communication in cancer cachexia. Nature Reviews Endocrinology. 2019; agent for multiple myeloma. Clin Cancer Res. 2017;23(21):6602–15. https:// 15(1):9–20. https://doi.org/10.1038/s41574-018-0123-0. doi.org/10.1158/1078-0432.CCR-17-0678. 48. Yang J, Zhang Z, Zhang Y, Ni X, Zhang G, Cui X, Liu M, Xu C, Zhang Q, Zhu H, 28. Hong Y, Lee JH, Jeong KW, Choi CS, Jun HS. Amelioration of muscle et al. ZIP4 promotes muscle wasting and cachexia in mice with orthotopic wasting by glucagon-like peptide-1 receptor agonist in muscle atrophy. J pancreatic tumors by stimulating RAB27B-regulated release of extracellular Cachexia Sarcopenia Muscle. 2019;10(4):903–18. https://doi.org/10.1002/ vesicles from cancer cells Gastroenterology. 2019, 156:722-734.e726. jcsm.12434. 49. Hu W, Ru Z, Xiao W, Xiong Z, Wang C, Yuan C, et al. Adipose tissue 29. Huang WC, Kuo KT, Bamodu OA, Lin YK, Wang CH, Lee KY, Wang LS, Yeh browning in cancer-associated cachexia can be attenuated by inhibition of CT, Tsai JT. Astragalus polysaccharide (PG2) Ameliorates Cancer Symptom exosome generation. Biochem Biophys Res Commun. 2018;506(1):122–9. Clusters, as well as Improves Quality of Life in Patients with Metastatic https://doi.org/10.1016/j.bbrc.2018.09.139. Disease, through Modulation of the Inflammatory Cascade. Cancers (Basel). 50. Trajkovic K, Hsu C, Chiantia S, Rajendran L, Wenzel D, Wieland F, et al. 2019, 11. Ceramide triggers budding of exosome vesicles into multivesicular 30. Chen JA, Splenser A, Guillory B, Luo J, Mendiratta M, Belinova B, et al. endosomes. Science. 2008;319:1244–7. Ghrelin prevents tumour- and cisplatin-induced muscle wasting: 51. Luberto C, Hassler DF, Signorelli P, Okamoto Y, Sawai H, Boros E, et al. characterization of multiple mechanisms involved. J Cachexia Sarcopenia Inhibition of tumor necrosis factor-induced cell death in MCF7 by a novel Muscle. 2015;6(2):132–43. https://doi.org/10.1002/jcsm.12023. inhibitor of neutral sphingomyelinase. J Biol Chem. 2002;277(43):41128–39. 31. Del Fabbro E, Inui A, Strasser F. Managing cancer cachexia. In: Cancer https://doi.org/10.1074/jbc.M206747200. Cachexia. Tarporley: Springer Healthcare Ltd; 2012. p. 51–72. 52. Feng D, Zhao WL, Ye YY, Bai XC, Liu RQ, Chang LF, et al. Cellular 32. Busquets S, Serpe R, Sirisi S, Toledo M, Coutinho J, Martínez R, et al. internalization of exosomes occurs through phagocytosis. Traffic. 2010;11(5): Megestrol acetate: its impact on muscle protein metabolism supports its 675–87. https://doi.org/10.1111/j.1600-0854.2010.01041.x. use in cancer cachexia. Clin Nutr. 2010;29(6):733–7. https://doi.org/10.1016/j. 53. Svensson KJ, Christianson HC, Wittrup A, Bourseau-Guilmain E, Lindqvist E, clnu.2010.06.003. Svensson LM, et al. Exosome uptake depends on ERK1/2-heat shock protein 33. Argilés JM, Anguera A, Stemmler B. A new look at an old drug for the 27 signaling and lipid Raft-mediated endocytosis negatively regulated by treatment of cancer cachexia: megestrol acetate. Clin Nutr. 2013;32(3):319– caveolin-1. J Biol Chem. 2013;288(24):17713–24. https://doi.org/10.1074/jbc. 24. https://doi.org/10.1016/j.clnu.2013.01.004. M112.445403. 34. Zhou X, Wang JL, Lu J, Song Y, Kwak KS, Jiao Q, et al. Reversal of cancer 54. Sanvee GM, Panajatovic MV, Bouitbir J, Krähenbühl S. Mechanisms of insulin cachexia and muscle wasting by ActRIIB antagonism leads to prolonged resistance by simvastatin in C2C12 myotubes and in mouse skeletal muscle. survival. Cell. 2010;142(4):531–43. https://doi.org/10.1016/j.cell.2010.07.011. Biochem Pharmacol. 2019;164:23–33. https://doi.org/10.1016/j.bcp.2019.02.02 35. Tisdale MJ. Reversing cachexia. Cell. 2010;142(4):511–2. https://doi.org/10.101 6/j.cell.2010.08.004. 55. Palus S, von Haehling S, Flach VC, Tschirner A, Doehner W, Anker SD, et al. 36. Chen X, Wu Y, Yang T, Wei M, Wang Y, Deng X, et al. Salidroside alleviates Simvastatin reduces wasting and improves cardiac function as well as cachexia symptoms in mouse models of cancer cachexia via activating outcome in experimental cancer cachexia. Int J Cardiol. 2013;168(4):3412–8. mTOR signalling. J Cachexia Sarcopenia Muscle. 2016;7(2):225–32. https:// https://doi.org/10.1016/j.ijcard.2013.04.150. doi.org/10.1002/jcsm.12054. 56. Koutnik AP, Poff AM, Ward NP, DeBlasi JM, Soliven MA, Romero MA, 37. Cheung WW, Hao S, Wang Z, Ding W, Zheng R, Gonzalez A, et al. Vitamin D Roberson PA, Fox CD, Roberts MD, D'Agostino DP. Ketone bodies attenuate repletion ameliorates adipose tissue browning and muscle wasting in wasting in models of atrophy. J Cachexia Sarcopenia Muscle. 2020. infantile nephropathic cystinosis-associated cachexia. J Cachexia Sarcopenia Muscle. 2020;11(1):120–34. https://doi.org/10.1002/jcsm.12497. 38. Argilés JM, López-Soriano FJ, Stemmler B, Busquets S. Novel targeted Publisher’sNote therapies for cancer cachexia. Biochem J. 2017;474(16):2663–78. https://doi. Springer Nature remains neutral with regard to jurisdictional claims in org/10.1042/BCJ20170032. published maps and institutional affiliations.
Skeletal Muscle – Springer Journals
Published: Jul 6, 2021
Keywords: Amiloride; Cancer cachexia; Muscle wasting; Exosome; Exosome-release inhibition