Interleukin‐4 administration improves muscle function, adult myogenesis, and lifespan of colon carcinoma‐bearing mice

Abstract Background Anorexia, body wasting, inflammation, muscle, and adipose tissue loss are hallmarks of cancer cachexia, a syndrome that affects the majority of cancer patients, impairing their ability to endure chemotherapeutic therapies and reducing their lifespan. In the last 10 years, alterations of protein turnover and impairment of adult myogenesis have been proposed as major contributing factors. Methods Muscle stem cells, including satellite cells, mesoangioblasts, and fibroadipogenic progenitors, were isolated and characterized from C26 colon carcinoma‐bearing (C26) mice. Circulating levels of interleukin‐4/13 (IL4/IL13) were analysed by ELISA, and the effects of IL4 on muscle mass and function, protein synthesis, muscle regeneration, and myogenic progenitor cell number were analysed at both functional (treadmill and grip test) and molecular levels (qRT–PCR, immunofluorescence analysis, surface sensing of translation, and western blot). The Kaplan–Meier test was used to analyse the survival curve of IL4‐treated and IL4‐untreated C26 mice. Results The administration of IL4 to C26 mice rescued muscle mass by increasing protein synthesis. The IL4 treatment improved performances and prolonged survival of C26 mice. IL4 administration re‐established both number and function of satellite cells and fibroadipogenic progenitors without affecting mesoangioblasts in C26 mice, rescuing myogenesis. Upon IL4 treatment, a high number of cytotoxic lymphocytes and type II macrophages were observed with a subsequent increase in necrotic areas of C26 tumours. Conclusions The results here presented shed new light on IL4 signalling during muscle wasting and early stages of muscle regeneration that explain the beneficial effect observed in IL4‐treated C26 mice. These findings might aid to develop therapeutic approaches to improve mobility and quality of life in cachectic patients.


Introduction
Cachexia is a multifactorial syndrome associated to many types of cancers, affecting mainly gastrointestinal tract and lungs, that compromises absorption of nutrients and decreases the possibility for the patients to efficiently breathe. 1 The occurrence of cachexia increases chemotherapy toxicity, complicates patient management, and enhances mortality rates, accounting for about 20% of cancer deaths worldwide (Cancer fact sheet, World Health Organization 2017; https:// www.who.int/news-room/fact-sheets/detail/cancer).
Loss of muscle, with or without loss of fat mass, is one of the main hallmarks of the cachectic phenotype, is not reversed by nutrient supplementation, and, at least in experimental model systems, is mainly associated with the hyperactivation of catabolic signals involving the main proteolytic systems. [2][3][4] In addition, protein synthesis is altered in favour of degradation and cannot be restored in experimental models by forcing the activity of anabolic pathways. 5 Cytokines and pro-inflammatory signals produced by the host immune system interact with the tumour, contributing to generate insulin and other growth factors resistance in the muscle of cancer patients. Consistently, drugs able to enhance muscle anabolism and to increase appetite and energy storage, such as several ghrelin mimetics, have entered in phase II clinical trials. 6 In the last decade, different groups have proposed that also impaired myogenesis could contribute to cancer-induced muscle wasting. In this regard, in the muscle of tumourbearing animals, satellite cells (SCs) increase in number and accumulate in the interstitial space among the fibres. 7,8 The same pattern has been observed also in cancer patients. 8,9 From the mechanistic point of view, activation of NF-κB in myogenic progenitors has been pointed as responsible for a persistent up-regulation of PAX7, thus blocking the expression of transcription factors important for SC differentiation. 8 This myogenic impairment can be partially removed by treatment with MEK inhibitors. 7 Another study has shown the occurrence of a defect in muscle stem cell proliferation and differentiation in cachectic animals, which has been associated with decreased number of macrophages and mesenchymal progenitors. 10 Several studies have clarified that, in addition to SCs, also other cell types can contribute to myogenesis, such as mesoangioblasts (MABs), vessel-associated muscle progenitors, and fibroadipogenic progenitors (FAPs). Consistently, recent data from our group have demonstrated that BMP-SMAD signalling blockade improves MAB myogenic differentiation. 11 As for FAPs, they have been firstly described as non-myogenic interstitial stem cells present in the adult muscle that upon injury can drive SC differentiation through secretion of interleukin-4 (IL4) and interleukin-13 (IL13). 12,13 Indeed, these cells have been reported to improve skeletal muscle regeneration in acute 14 and chronic conditions. 15 On the other side, IL4/IL4R and IL13/IL13R have been found highly up-regulated in the skeletal muscle after exercise, suggesting an anabolic role played by the two cytokines in this tissue. 16 IL4 has also been shown to recruit myoblasts to form mature myotubes, 17 and further studies have demonstrated that this cytokine can exert a pro-migratory effect in vitro. 18 Interestingly, in pancreatic cancer patients affected by cachexia, IL4 messengers are down-regulated in liver and muscle. 19 The aim of the present study is to investigate the mechanisms underlying the impaired myogenesis in cancer hosts, focusing on the role that interstitial cells and cytokines involved in muscle differentiation might play in contributing to cancerinduced muscle wasting.

Animal experiments
All animal procedures 20 were performed according to the guidelines of the Animal Welfare Committee of KU Leuven and Belgian/European legislation (P051/2017). Four-weekold Balb/c mice of approximately 20 g were obtained from Charles River Laboratories (Wilmington, MA, USA). Mice were maintained on a regular dark-light cycle (light from 8 h a.m. to 8 h p.m.), with free access to food and water. C26 colon-adenocarcinoma cells 5 were maintained in highglucose Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% foetal bovine serum (FBS), 1 mg/mL of sodium pyruvate, and 1 IU/mL of penicillin and streptomycin. The plates were incubated at 37°C, 5% CO 2 , 5% O 2 , and the medium was refreshed every 2 days. Tumourbearing mice were inoculated subcutaneously dorsally with 5 × 10 5 C26 carcinoma cells. The number of mice per group depends on the experiments as it is indicated in the legends of the respective assays. Where indicated, 1.3 μg per mouse intraperitoneal IL4 (Peprotech, Rocky Hill, NJ, USA) was administered every day, starting from Day 5 to the day of sacrifice. 14 Multiple experiments were performed with body weight, tumour size, and survival rates recorded. The blood was collected through intracardiac injection at the sacrifice, obtained through cervical dislocation. After collecting the blood in heparin, centrifugation at 845 g was performed to obtain the blood cells (fixed with Thermo Fisher Scientific, 1-step Fix/Lyse Solution for flow cytometry analysis), and the plasma was kept at À80°C for ELISA experiments. Animal weight and food intake were recorded daily. Tumour-bearing mice were sacrificed under anaesthesia 13-14 days after tumour transplantation. Several muscles and organs were rapidly excised, weighed, frozen in isopentane cooled with liquid nitrogen, and stored at À80°C. Peritoneal lavage was performed with 50 μM of EDTA in phosphate-buffered saline (PBS) to extract lymphocytes and macrophages. Tumour tissue was digested for 90′ in 1 mg/mL collagenase II and IV 1:1.
Four-month-old alpha-sarcoglycan (αSG)-null mice from the internal colony of Sampaolesi laboratory were generated by the group of K.P. Campbell (University of Iowa, IA, USA). 21 Due to the absence of αSG from the muscle membrane, these mice, starting from 2 to 3 months of age, undergo continuous cycle of muscle fibre degeneration and regeneration. For these reasons, these mice were used to test the ability of C-MABs and C26-MABs to contribute to muscle regeneration upon cyclosporine immunosuppression (as already described in Costamagna et al. 11 ).

Survival experiment
Mice were followed until the KU Leuven predetermined humane endpoint criteria were fulfilled. Mice were ultimately sacrificed when C26 cell inoculation determined body mass loss above 20%, combined with an overall health status of severe sufferance, food intake decrease, abnormal appearance and posture (lack of grooming, piloerection, and hunched posture), unnatural behaviour (impaired locomotion, inactivity, and reduced reactivity to external stimuli), and body temperature below 28°C. During the experiment, mice were separated to avoid self-mutilation.

Treadmill and muscle strength
Mice were introduced to the treadmill belt to test the conditions before the actual test (motor speed set to zero, for 5 min). Later on, the motor speed was set to 10 m/min, with a 1 m/min increase and an uphill inclination of 10°, till exhaustion and >10 s stop. 11 Grip strength was measured with digital grip strength metre (Columbus Instruments). Three measurements with 10 min interval between tests were completed for each mouse.

Muscle injury
Mice received 50 μL injection of 10 μM solution of cardiotoxin (CTX, Naja mossambica mossambica, Sigma-Aldrich) in the tibialis anterior, the gastrocnemius, and the quadriceps of one hind limb. The contralateral muscles were injected with the same volume of PBS as a control. Ten days after the in vivo regeneration experiments, the mice were sacrificed by cervical dislocation, and muscles were processed for histological analysis. For the experiment of muscle injury in parallel to the survival experiment (regenerating muscle shown in Supporting Information, Figure S6b, n = 5 C26; n = 5 C26 + IL4), mice were sacrificed when all the signs of severe sufferance were reached and body temperature was below 28°C.

Surface Sensing of Translation (SUnSET) experiment
In order to detect protein synthesis at the moment of the sacrifice, mice were administered intraperitoneally of 0.040 μmol/g puromycin (Sigma-Aldrich) and euthanized 30 min after injection. 22 Muscles were rapidly removed, weighted, and frozen in liquid nitrogen. Puromycin-conjugated peptides were visualized by western blotting (WB) on a gradient gel by using a mouse monoclonal anti-puromycin antibody ( Table 2). The density of each lane has been then measured by QuantityOne software and normalized per aTubulin ( Table 2) levels to determine the protein synthesis rate expressed as puromycin relative abundance.

Cell isolation and culture
Adult SCs were isolated by enzymatic digestion with 0.01% collagenase D and 0.06% pancreatine during 90′ digestion of minced murine muscle and pre-plated on 0.1% gelatin-coated dishes to remove fibroblasts. Cells were counted and plated on 0.1% collagen for experiments of cell proliferation in high-glucose DMEM supplemented with 20% FBS, 1 mg/mL of sodium pyruvate, 1 IU/mL of penicillin and streptomycin, and 1 mg/mL of non-essential amino acids, 50 μg/mL of gentamicin, and 1:100 chick embryo extract (Gentaur, Genprice Inc., CA, USA). The plates were incubated at 37°C, 5% CO 2 , 5% O 2 , and the medium was refreshed every 2 days. Myoblast differentiation was obtained with the same medium as for C2C12 differentiation. After tissue explant, adult Mesoangioblasts (MABs) were sorted for ALP+ and induced to express GFP (as for 11 ). After sorting, the GFP + cells were cultured on 0.1% collagen-coated dishes and maintained in high-glucose DMEM supplemented with 20% FBS, 1 mg/mL of sodium pyruvate, 1 IU/mL of penicillin and streptomycin, and 1 mg/mL of non-essential amino acids. The plates were incubated at 37°C, 5% CO 2 , 5% O 2 , and the medium was refreshed every 2 days. Intramuscular injections of 2 × 10 5 murine GFP + MAB or MAB cells pretreated for 48 h with 5 μM dorsomorphin (Sigma-Aldrich; DM-MABs) were performed in αSG-null mice (n = 5 for each group). During the entire treatment, mice were maintained with cyclosporine (Sandimmune cyclosporine, Novartis; 10 mg/kg); 21 days post-injection mice were sacrificed and compared with sham-operated mice (saline solution injected).
C2C12 myoblast cells were maintained at 37°C in 5% CO 2 in air-humidified chamber in high-glucose DMEM with glutamine, supplemented with 10% FBS, 1 mg/mL of sodium pyruvate, and 1% penicillin/streptomycin (Thermo Fisher Scientific). When the cells reached confluency, growth medium was replaced with differentiation medium, supplementing high-glucose DMEM with 2% horse serum, 1mg/mL of sodium pyruvate, and 1 IU/mL of IL4 counteracts cancer cachexia penicillin/streptomycin. The plates were incubated at 37°C, 5% CO 2 , 5% O 2 , and the medium was refreshed every 2 days.
In order to efficiently knockdown IL4Ra and IL13Ra, endoribonuclease-prepared siRNAs against IL4Ra and IL13Ra (MISSION esiRNA Mouse Il4ra and MISSION esiRNA Human IL13ra1; Sigma-Aldrich) were applied following the manufacturer's instructions. Briefly, C2C12 at Day 2 of myotube differentiation was transfected using Lipofectamine 2000 and Opti-MEM. At different time points after transfection (24,48, and 72 h), cells were lysed for RNA extraction or WB analysis. RNA Mini Kit was used for RNA isolation, and genomic DNA traces were removed with Turbo DNase, following the manufacturer's instructions; 500 μg of RNA were reverse-transcribed with the SuperScript III kit. The resulting cDNA was transferred in a 384-well plate pre-filled with 250 nM primers. Platinum SYBER green mix was added to a final volume of 10 μL per well. The RT-qPCR was performed for 40 cycles (95°C, 15 s; 60°C, 45 s) and read on a ViiA 7 qPCR plate reader. The list of murine primers used is available in Table 1.

Flow cytometry analysis
Staining was performed with primary antibody ( Table 2) during 30′ at room temperature, and it was followed by 30′ secondary antibody staining (just for Ki67, Alexa Fluor 647, Thermo Fisher Scientific) or from final washes. Fluorescence minus one and compensations were included when necessary for appropriate gating. Cells were analysed and quantified by flow cytometry (BD FACS Canto HTS; BD Biosciences), and FlowJo Software was used to analyse the data (FlowJo V10 LLC).

ELISA assay
IL4 and IL13 serum levels were detected by a commercially available mouse ABTS-ELISA kit, used according to the manufacturer's instructions (Peprotech). Serum from each animal (50 μL) was assayed in triplicate. Quantitative calibration was obtained performing a standard curve with recombinant mouse IL4 and IL13.
Immunofluorescence Tissues (7 μm muscle tissue and 5 μm tumour tissue) and cells were fixed with 4% paraformaldehyde (Polysciences Europe GmbH) in PBS and permeabilized with 0.2% Triton X-100 in PBS containing 1% (w/v) bovine serum albumin (BSA). A blocking solution containing donkey serum (1:10 diluted in PBS; VWR) was applied, and primary antibodies (Table 2) were incubated overnight in PBS supplemented with 1% BSA. Secondary Alexa Fluor donkey antibodies were diluted at 4 μg/mL in PBS supplemented with 1% BSA, and the nuclei were counterstained with 10 μg/mL Hoechst. Fluor Save reagent (Millipore, Chemicon) was used as mounting medium. Images were acquired using an Eclipse Ti inverted microscope (Nikon).

Statistical analysis
All results were expressed as mean ± standard deviation, with the exception of gene expression (mean ± SEM). Representative WBs were selected from independent experiments. When data distribution approximated normality and two groups were compared, a Student's t-test was used. When three or more groups were compared, a one-way analysis of variance was used. In case of non-normal distribution of the data pools, such as for statistical comparison of the differences between percentage values, a Kruskal-Wallis one-way analysis of variance by rank test was used. All statistical tests were performed via Prism software (GraphPad).

Results
Mice bearing the C26 tumour represent a well-characterized system to study cancer cachexia. 5 As previously reported, 14 days after C26 cell implantation, body weight in tumourbearing mice was about 75% of that observed in control animals (C; body weight at Day 0 = 20.41 ± 1.89 g; Figure 1A). A comparable reduction (about 25% lower than C values) was observed for several muscles, namely, tibialis anterior, gastrocnemius, and quadriceps ( Figure 1B).

Muscle stem cells isolated from C26 hosts: in vitro proliferation and differentiation
The number of SCs isolated from muscles of C26 hosts (hereafter C26-SCs) was higher than that obtained from C muscles (C-SCs; Figure 1C). Consistently, 4 days after plating, C26-SC cultures displayed a higher number of clones than C-SC ones (Figure 1D), demonstrating that the proliferative ability was higher in C26-SCs than in C-SCs. In immunofluorescence analysis, both C26-SCs and C-SCs showed positivity for PAX7 and MyoD, confirming in vitro their myogenic lineage ( Figure 1E, left  panels). When exposed to differentiation medium, these cells were able to activate the differentiation programme, fusing into multinucleated myotubes and expressing both MyoD and MyHC ( Figure 1E, right panels), or ACTN2 (Supporting Information, Figure S1a). The expression of terminally differentiated myotube markers let us conclude that ex vivo C26-SCs behave comparably with C-SCs. Similarly, MABs sorted as alkaline phosphatase (ALP)-positive cells ( Figure 1F) were isolated from the skeletal muscle of C26 hosts in higher number than in C mice ( Figure 1F, quantification). To evaluate the ability of MABs to contribute to muscle differentiation, C-MABs and C26-MABs were transfected to express GFP ( Figure 1G, left panels) and cocultured with C2C12 myoblasts. After 5 days of differentiation, GFP-expressing multinucleated myotube-like structures appeared ( Figure 1G, central panels). MyHC + and GFP + myotubes were detected in both C-MABs and C26-MABs co-cultures, demonstrating that GFP + MABs were able to fuse with C2C12 myotubes ( Figure 1G, right panels; Supporting Information, Figure S1b and S1c). In order to test if C-MABs and C26-MABs display the same ability in vivo, these cells were injected intramuscularly into αSG-null mice. 11 Twenty-one days after C-MAB and C26-MAB injection, the muscle of αSG-null mice showed the presence of GFP + fibres expressing αSG, clearly originated from the fusion of the injected C-MABs and C26-MABs ( Figure  1H, left panels). Moreover, as previously described, 11 both engraftment and differentiation were further improved when injected C-MABs and C26-MABs were pretreated in culture for 48 h with 5 μM dorsomorphin (DM), in order to inhibit type I BMP receptor complex and obtain a proliferation arrest, finally promoting differentiation ( Figure 1H, right panels and quantification). Taken together, these results confirm that the myogenic properties of MABs isolated from C26 hosts are not impaired.
Finally, the majority of interstitial muscle cells, isolated from the muscle of C26-bearing mice and put in culture, differentiated to adipocyte-like cells, a pattern that was not observed in cultures prepared with cells isolated from C muscles. Increased PLIN1 expression of Oil-Red-O-positive cells ( Figure 1I, right panel) and the positive bright field staining for Oil-Red-O ( Figure 1I, left) confirmed the acquisition of an adipocyte-like phenotype in vitro.
These data demonstrate that the impaired myogenesis occurring in the C26 hosts does not result from a stem cell autonomous defect but is rather related to the microenvironment these cells are exposed to. The hypothesis is that tumour-induced modulations of muscle microenvironment do not allow stem cell differentiation. On the basis of the adipogenic drift observed when cells isolated from the muscle of C26 hosts were put in culture, the possibility that myogenic relevant cytokines 14 could be perturbed was assessed. Data available in literature have stressed the importance of IL4 and IL13 during the first phases of muscle regeneration; thus, we evaluated their expression in both muscle and blood from C and C26 hosts. Quantitative qRT-PCR showed in muscle tissues from C26-bearing mice a significant increase of IL4Ra and a decrease in IL13 transcript levels, with respect to control values, while no significant differences could be observed for both IL4 and IL13R mRNAs (Supporting Information, Figure S1d). The low availability of efficient antibodies for WB let however any conclusion on the expression of these molecules still at a speculative level. Finally, while circulating IL4 was comparable in both C26 hosts and C mice ( Figure 1J), IL13 plasma levels were significantly lower in the former than in the latter ( Figure 1K).

Interleukin-4 treatment improves cachexia in tumour-bearing mice
The data reported earlier suggested that IL4/IL13 signalling was affected in the C26 hosts. The next step of the study was to investigate if potentiating such signalling could improve impaired regeneration and cachexia in the C26 hosts. The choice to use IL4 instead of IL13, as it could be expected looking at their circulating levels in the C26 hosts, was based on the observation that IL4 exerted its bioactivity engaging both its own receptor and the IL13 receptor (reviewed in Wu et al. 23 , thus providing a rather complete spectrum of action on the signalling pathways depending on both cytokines. Along this line, C26-bearing mice were treated with IL4 after having excluded, by in vitro experiments, the possibility that IL4 could exert a direct effect on tumour cells (Supporting Information, Figure S2a). Starting from Day 5 after C26 inoculation, IL4 was administered intraperitoneally daily. IL4 absorption curve in treated control mice confirmed high circulating levels of the cytokine during the first 3 h after administration, followed by a reduction to control values, approximately 8 h after IL4 injection (Supporting Information, Figure S2b). Consistently, after 10 days of IL4 daily treatment, circulating levels of the cytokine were significantly increased in both C and C26 hosts ( Figure 2B). To rule out the possibility of a local response of the immune system to IL4 administration, intraperitoneal lavage and flow cytometry analysis were performed. The results obtained showed, differently from what observed in positive controls (LPS-injected mice; Supporting Information, Figure S2e and S2h), comparable lymphocyte (Supporting Information, Figure S2f and S2g) and macrophage (Supporting Information, Figure S2i and S2j) percentages and characteristics in IL4-treated and IL4-untreated C26 bearers. No differences were as well reported comparing the peritoneal lymphocytes and macrophages of IL4-treated mice, confirming no effect at a peritoneal level after IL4 treatment (data not shown).
As expected, body weight in the C26 hosts was lower than in C (Day 0: 21.57 ± 1.07 g; 25% reduction), while it remained comparable with Day 0 in tumour-bearing mice receiving IL4 (Figure 2A). In contrast, the reduction of food intake occurring in the C26-bearing mice was not affected by IL4 treatment (mouse food intake on Day 13: C = 4.33 g; IL4 = 4.65 g; C26 = 1.45 g; C26 + IL4 = 1.88 g). Similarly, epididymal fat pad was completely depleted in both C26 and C26 + IL4 groups with respect to controls (C = 2.01 ± 0.57; IL4 = 2.63 ± 0.78 mg/100 g i.b.w.; C26 = ND; C26 + IL4 = ND; data are expressed as mg/100 g i.b.w., P < 0.22 C vs. IL4). IL4 treatment improved C26-induced muscle wasting in terms of both muscle mass and cross-sectional area (CSA) stabilizing this latter to the level of C ( Figure 2C-2E and Supporting Information, Figure S2d). Because improved muscle mass or myofiber CSA do not necessarily mean better muscle function, an additional experiment was performed to obtain functional data. The results showed that grip strength was higher in IL4-treated than in IL4-untreated C26 hosts reaching values comparable with C mice ( Figure 2F). Such improvement was confirmed by an exhaustion treadmill assay, showing that C26 hosts receiving IL4 run for longer time and distance than untreated C26 mice, similarly to C mice ( Figure 2G and 2H).
Finally, a survival experiment demonstrated that IL4treated C26-bearing mice were able to survive longer than untreated tumour hosts (31 days for C26 + IL4 vs. 13 days for C26; Figure 2I).

Interleukin-4 treatment increases intratumour cytotoxic lymphocytes and macrophages
The administration of IL4 to the C26 hosts resulted in increased tumour mass (Supporting Information, Figure S2c) that however does not necessarily reflect an increased number of tumour cells. Indeed, the H&E staining revealed that more than 30% of tumour area in the IL4-treated C26-bearing mice (sacrificed at Day 31 after C26 cell injection) was necrotic (eosin positive) probably surrounded by small inflammatory cell infiltrate, a pattern that could not be observed in tumours isolated from C26 hosts (Day 13 after C26 implantation; Figure  3A and 3B). 24 A possible limitation of these results is that the histological analysis compares Day 13 C26 with Day 31 C26 + IL4 tumours. However, because in the absence of IL4 treatment the C26 hosts die at Day 13 after tumour implantation, it was not possible to have untreated tumours at Day 31. Future investigation will assess if the peculiar histological aspect of C26 + IL4 tumours at Day 31 is already evident at Day 13.

Interleukin-4 administration improves protein synthesis in tumour-bearing mice
The increased muscle mass and myofiber CSA could result from both reduced protein degradation and increased protein synthesis rates. An experiment was performed to assess this latter, by treating the animals with puromycin 30 min before sacrifice. 22 The results showed that muscle protein synthesis is significantly improved in IL4-treated than in IL4-untreated C26 hosts ( Figure 4A and 4B). Such improvement was associated with enhanced phosphorylation of Akt and p70, as shown by WB analysis (Figure 4C-4E). Moreover, the levels of P-JNK and P-ERK, this latter previously shown to correlate with muscle wasting in C26 hosts, 6 decreased in the muscle of IL4-treated C26-bearing mice ( Figure 4C, 4F, and 4G). On the other side, no significant differences among treated and untreated C26 hosts could be observed as for P-p38 levels (Supporting Information, Figure S4a and S4b) and the state of activation of STAT3 and STAT6, both transcription factors reported to transduce the signals dependent on IL4Ra 25,26 (Supporting Information, Figure S4c-S4e). In addition, the main markers of proteolysis, such as Atrogin 1 and Trim 63 for the ubiquitin proteasome system (Supporting Information, Figure S4f-S4h) and Beclin1 and LC3B for the autophagic system (Supporting Information, Figure S4i-S4l), were not affected by IL4 treatment. Differently, the levels of p62-SQSM1 were lower in IL4-treated C26 mice when compared with the untreated ones and similar to controls (Supporting Information, Figure S4i and S4l-S4n).

Interleukin-4 treatment impinges on satellite cell accumulation in the muscle of C26 hosts
To investigate if the beneficial effects induced by IL4 on muscle mass and performance could also result from improved myogenesis, we employed flow cytometry and established gating strategies to identify the myogenic stem cells present in the muscles of IL4-treated and IL4-untreated C26-bearing mice. The total number of cells isolated from the muscle of IL4-treated C26 hosts was comparable with controls and reduced with respect to untreated C26-bearing mice (Supporting Information, Figure S5). As for the different muscle stem cell populations, the results showed that SC number was lower in C26 hosts administered with IL4 than in untreated tumour-bearing animals (C26 = 17.1 ± 5.2%, C26 + IL4 = 10.9 ± 2.0%; Figure 5A and 5B). This reduction was paralleled by a significant decrease in PAX7 and myogenin protein levels (Myog; Figure 5C-5E) and by an increased number of type II macrophages ( Figure 5F-5H). While no differences were appreciated for MABs (C26 = 8.8 ± 2.6%, C26 + IL4 = 8.8 ± 2.0%; Figure 5I and 5J), a trend to decrease, although not significant, was shown for FAPs (C26 = 28.9 ± 5.1%, C26 + IL4 = 21.7 ± 5.3%; Figure 5K and 5L). Similarly, immunofluorescence analysis for PDGFRa (known markers for FAP cells 13 ) and Ki67 staining in C26 + IL4 muscle sections showed a lower accumulation of cells when compared with C26 ones (Supporting Information, Figure S5b). Consistently, the number of PLIN1 and Oil-Red-O-positive cells was lower in primary cultures obtained from the muscle of IL4-treated vs. IL4-untreated C26-bearing mice ( Figure 5M), suggesting that interstitial FAPs in the former are less aberrant than in the latter.

Interleukin-4 administration affects skeletal muscle regeneration in C26-bearing mice
Previous studies reported that muscle regeneration is delayed in tumour-bearing mice. 7,8 On the other side, IL4 treatment was reported to improve muscle regeneration in glycerol-injured animal model. 17,27 Along this line, the myogenic response following CTX-induced muscle injury in the C26 hosts, in the presence or in the absence of IL4 treatment, was assessed. The results showed that regeneration is improved by treatment with the cytokine (Figure 6A-6C). H&E staining and quantification showed a decreased myofiber CSA in CTX-C26 hosts and CTX-IL4 mice compared with CTX-C ( Figure 6D and 6F; Supporting Information, Figure S6a). In CTX-C26 hosts treated with IL4, myofibers were smaller than in the CTX-C26-bearing mice. This reduction in size, however, is counterbalanced by an increase in number of newly generated fibres with low CSA (diameter between 0 and 1000 μm 2 :   Figure 6D and 6F and Supporting Information, Figure S6a). Consistently, in muscles from CTX-C26 hosts exposed to IL4, the levels of embryonic myosin heavy chain (eMyHC) were similar to CTX-C or CTX-IL4 muscles and lower than observed in muscles from CTX-C26 hosts ( Figure 6E, 6G, and 6H). Finally, in the injured muscle of C26 hosts treated with IL4 and analysed around 25-31 days after injury, the number of mature fibres characterized by peripheral nuclei was higher than in untreated injured C26-bearing mice (that died 13 days after tumour transplantation and 8 days after CTX injury; Supporting Information, Figure S6b).
No differences could be observed in both PAX7 and Myog expressions in both IL4-treated and IL4-untreated injured C26 hosts, although Myog levels showed a trend to decrease that did not reach significance ( Figure 6I-6K). Finally, muscle type II macrophages were higher in CTX-C26 hosts than in CTX-C26-bearing mice exposed to IL4 ( Figure 6L-6N).

Interleukin-4 directly stimulates myocyte differentiation
To better understand the response of muscle stem cells to IL4, we first analysed the effects of IL4 on C2C12 myoblasts in proliferating conditions. Staining for Ki67 revealed an antiproliferative effect when 10 ng/mL IL4 were administered for 48 h to C2C12 myoblast cultures ( Figure 7A). These results were confirmed by flow cytometry analysis and quantification ( Figure 7B and 7C). In addition, IL4 was also able to reduce the stimulation of C2C12 myoblast proliferation induced by treatment with 100 ng/mL TNFα, a pro-inflammatory cytokine known to inhibit C2C12 differentiation ( Figure 7B and 7C). These observations suggested that IL4 could impact on C2C12 differentiation. 15 To clarify this point, IL4/IL4Ra and IL13/IL13R mRNA levels were assessed at Days 0 and 5 of myogenic differentiation (Supporting Information, Figure S7a). As for protein, while IL4Ra levels were not detectable at Day 2 of differentiation (Supporting Information, Figure S7b), at Day 5, when completely differentiated myotubes were expressing MyHC, IL4Ra became detectable ( Figure 7D-7F). When myotubes were exposed to 100 ng/mL TNFα during 5 days, a decrease in MyHC and in IL4Ra was reported. Such changes were not rescued by co-treating myotubes with both TNFα and IL4. This pattern was maintained even if TNFα levels were reduced (50 ng/mL) and IL4 concentrations were increased (100 ng/mL IL4, Supporting Information, Figure S7c and S7d; 200 ng/mL IL4, Supporting Information, Figure S7e and S7f).
To better understand the role of the IL4-dependent signalling in C2C12 differentiation, loss-of-function experiments were performed. IL4Ra was transiently silenced in differentiating C2C12. Interestingly, qRT-PCR data showed that the loss of IL4Ra can be compensated by IL13Ra. Indeed, when IL4Ra was down-regulated, the levels of IL13Ra increased, likely compensating the missing IL4Ra ( Supporting Information, Figure S7g). The other way around, when down-regulating IL13Ra, IL4Ra levels increased (Supporting Information, Figure  S7h), confirming our interpretation. IL4Ra silencing decreased transcript levels of genes promoting myoblast fusion, [28][29][30] such as Myomixer (Mymx), Myomaker (Mymk), and Myomerger-Minion (here indicated as Gm7325, analysed as short version and long version at the transcriptional level), to values lower than that of control cultures (Supporting Information, Figure S7i). Therefore, to obtain a more effective abolishment of IL4/IL13 signalling pathway, we induced a transient IL4Ra and IL13Ra esiRNA double down-regulation in C2C12 induced to differentiate. The results confirmed that while in control C2C12 myotubes reached a complete differentiation, as shown by MyHC expression, transfected cells did not ( Figure  7G-7J). Moreover, the protein levels of Myomerger (known also as embryonic stem cell and germ cell-specific protein, ESGP) were reduced, confirming that the lack of IL4/IL13 signalling impaired myoblast fusion ( Figure 7K-7M). Finally, WB analysis for ESGP protein levels in regenerating muscles from C26-bearing mice did show similar levels in both IL4-treated and IL4-untreated CTX-C26 hosts (Supporting Information, Figure S7j and S7k), suggesting that the delayed regenerative response observed in IL4-treated vs. IL4-untreated C26 hosts does not rely on impaired fusion.

Discussion
The present study shows for the first time that cancer-induced muscle wasting is associated with increased number of different muscle stem cell populations that cannot complete differentiation in vivo despite maintaining their intrinsic myogenic potential. Indeed, they are able to form mature myotubes when isolated from the cancer host micro-environment. These results agree with previous data showing the accumulation of SCs in C26-bearing mice. 7,8 The present results extend the analysis to MABs that, similarly to SCs, increase in number in the muscle of C26 hosts. These cells maintain the ability to fuse with existing myofibers when exposed to permissive regenerating conditions such as those occurring in the αSGnull mice and are still able to respond to dorsomorphin. 11 Taken together, these results highlight that stem cell number and their intrinsic myogenic differentiation ability are not the limiting factor contributing to muscle wasting in cachectic tumour-bearing mice. In contrast, muscle interstitial stem cells seem to miss the input to differentiate. In this regard, myogenic differentiation is physiologically regulated by FAPs, a muscle stem cell population dependent on eosinophil granulocyte secretion of IL4 and IL13. 14 When isolated from the muscle of the C26 hosts and put in culture, FAP cells degenerate to adipocyte-like cells showing an adipogenic shift able to persist in vitro and probably suggesting an intrinsic modification of IL4 counteracts cancer cachexia  IL4 counteracts cancer cachexia these cells. Modulations of muscle IL4/IL4R and IL13/IL13R, as well as of IL4 and IL13 plasma levels, showed that the signalling dependent on these cytokines is likely altered in the C26bearing mice, supporting the possibility of improper FAP behaviour and functions. In this regard, an intriguing hypothesis is that FAP degeneration to adipocyte-like phenotype could contribute to myosteatosis, a common feature in patients affected by chronic inflammation and cachexia. 31,32 A detailed analysis of muscle biopsies obtained from cancer patients reported intramyocellular lipid droplets and large accumulation of neutral lipids in connective tissues in the rectus abdominis. 33 Further studies are required to understand the nature of this phenomenon and to clarify if related to the excessive adipogenic differentiation of FAP cells that we report in C26 mice. In this regard, the lack of myosteatosis in the C26 hosts could be due to the short time interval between tumour implantation and animal sacrifice. Finally, altered FAPdependent signalling might also explain the impaired myogenesis occurring in tumour-bearing mice (see succeeding text). Along this line, the results of the present study show for the first time that IL4 administration to mice bearing the C26 tumour improves cachexia in terms of body weight, muscle mass, and function. Moreover, IL4-treated C26-bearing mice can survive 20 days longer than untreated C26 hosts. Similarly, increased survival was previously obtained by injecting intraperitoneally a plasmid encoding for IL4 in mice bearing an ascitic form of the C26 tumour. 34 The effectiveness of IL4 administration in C26 hosts relies, partially at least, on the promotion of protein synthesis, as shown by puromycin incorporation and by sustained phosphorylation levels of Akt and p70. These kinases were shown to be unchanged, or even up-regulated in the muscle of the C26 hosts, 4,5 likely due to an attempt to counteract the tumourinduced hypercatabolic drift. In this regard, the data reported in the present study suggest that IL4 treatment is able to successfully push forward such compensatory response. The observation that no changes in both protein synthesis and anabolic kinase phosphorylation could be observed in IL4treated controls likely reflects the lack of effect in physiological states, where protein metabolism is not altered. Moreover, IL4 treatment results in decreased phosphorylation of some MAP kinases usually hyperactivated during cancer cachexia and shown to play a role in promoting muscle protein degradation. 4,7 This last observation is as well supported by lower levels of p62 accumulation in IL4-treated than IL4untreated C26 hosts.
Interleukin-4 administration to C26-bearing mice also results in decreased SC number, while a trend towards reduction can be observed in FAPs. These latter, however, show a markedly reduced degeneration towards adipocyte-like cells when isolated from the muscle of IL4-treated C26 hosts and put in culture, supporting the idea that a restored IL4/IL13 signalling can decrease the degeneration of muscle stem cells in C26-bearing mice. These findings are in accordance with the reduced muscle atrophy and fibrosis obtained in mdx mice by preventing FAP degeneration. 15,35 The delay in muscle regeneration occurring in the C26 hosts is partially corrected by IL4 treatment, as shown by the increased number of myofibers with central nuclei, that slowly reach the mature phenotype, characterized by low expression of eMYHC. Interestingly, these fibres are not blocked in their growth, but they rather undergo a temporary delay, showing, after a longer time period, several fully matured fibres with peripheral nuclei. In line with our results, a recent study showed that IL4-conjugated gold nanoparticles injected into injured murine skeletal muscle improved muscle histology and performance associated with two-fold increase of type II macrophages. 36 The idea that IL4 treatment in the C26 hosts acts on the muscle by stimulating myogenic differentiation, likely through improved FAP activity, is supported by the results obtained on C2C12 cultures exposed to the cytokine. Indeed, IL4 not only is able to inhibit myoblast proliferation prompting them to differentiation but also counteracts, partially at least, the inhibition of differentiation induced by TNFα. In this regard, TNFα maintains C2C12 cultures in a proliferative setting, 37 which could partially explain the accumulation of SCs observed in the muscle of tumour-bearing mice. Interestingly, the inhibition of C2C12 differentiation by TNFα is associated with reduced levels of IL4Ra. Thus, increasing IL4 concentrations do not rescue MyHC levels. The ability of IL4 to guide myogenesis is in line with previous data obtained in human muscle cells. 16,17 The involvement of IL4 in myogenesis is further supported by the observation that loss of function of both IL4Ra and IL13Ra during differentiation impairs the ability of myoblasts to fuse. As for the mechanisms by which the lack of IL4-dependent signalling may inhibit differentiation, down-regulation of Myomerger (ESGP) protein levels as well as of other genes relevant to fusion has been shown to occur. These results demonstrate that physiologic expression of IL4Ra is necessary to determine myocyte fusion, in C2C12 cultures at least, and it is not dispensable to complete myotube differentiation.
A possible limitation in the present study is that increased tumour mass occurs when IL4 is administered to C26-bearing mice. In this regard, in colorectal cancer stem cells, protection from apoptosis is achieved by autocrine production of IL4 through up-regulation of anti-apoptotic mediators. 38 These observations could introduce a note of care to the possibility that IL4 treatment can be proposed as a strategy to counteract cancer cachexia. However, despite being increased in size, tumours at Day 31 in IL4-treated C26 hosts present with conspicuous necrotic areas and with a high number of infiltrating inflammatory/immune cells (CD8 + lymphocytes and type II macrophages), which could be involved in tumour cytotoxicity and necrosis. These results suggest that IL4 administration to C26-bearing mice, far from stimulating cancer growth, increases the immune response against the tumour. However, this observation is in apparent contrast with the well-known anti-inflammatory action of IL4. In this regard, further experiments are required to clarify if systemic IL4 administration might exert different actions at the tumour and muscle levels, respectively. Consistently, the observations here reported show that in vitro IL4 is able to modulate myogenesis but not to affect C26 tumour cell growth curve, suggesting that while the effects on muscle are directly mediated by IL4, those on the tumour could depend on additional mediators.
In conclusion, these results demonstrate the importance of IL4 as a cytokine able to impinge on the inflammatory system, promoting an anti-inflammatory muscle micro-environment, and to raise the state of activation of anabolic kinases involved in the maintenance of muscle protein synthesis. Of interest, IL4 shares at least some beneficial actions with exercise, previously shown to reduce systemic inflammation 39 and to increase muscle protein synthesis. 21 In this regard, the hypothesis that IL4 could work as a sort of exercise-mimetic drug could be proposed, although further analyses aimed at assessing muscle energy metabolism after IL4 administration are required to clarify this point. On the whole, the results shown in the present study highlight that IL4/IL13 signalling is effective in keeping the correct equilibrium between anabolic and catabolic cues. Physiologic levels of IL4/IL13 are probably responsible for the normal homeostasis of muscle tissue; when dysregulated, they can contribute to muscle atrophy. Further efforts should be performed to determine if the administration of IL4 could ameliorate cancer-induced muscle atrophy when associated to chemotherapeutic drugs and perhaps extend this treatment to other disorders where muscle precursor accumulation occurs, such as muscular dystrophies (at least in the initial phases of the disease) or myotonic dystrophies, also characterized by chronic inflammation.

Author contributions
D.C. participated in conception and design, collection and assembly of data, data analysis and interpretation, and manuscript writing; R.D. participated in flow cytometry collection and assembly of data, data analysis and interpretation; F.P. participated in data interpretation; D.N. participated in data interpretation; P.C. participated in conception and design, data analysis and interpretation, manuscript writing, and final approval of the manuscript; M.S. participated in conception and design, data analysis and interpretation, manuscript writing, and final approval of the manuscript.

Online supplementary material
Additional supporting information may be found online in the Supporting Information section at the end of the article. Figure S1: Characterization of C-and C26-SCs and MABs. a) IF of SCs from C and C26 muscles stained for ACTN2 (red) and MyoD (green; at day 2 of differentiation). Nuclei were stained with HOECHST (blue). Scale bar: 500 μm. WB for MyHC and aTub on b) GFP + C-and c) GFP + C26-MABs in co-culture with C2C12 at day 0 and 5 of myotube differentiation. d) qRT-PCR for the expression of IL4, IL4Ra1, IL13 and IL13R in C and C26 muscles normalized for the housekeeping genes Gapdh, HPRT, TBP. Significance of the differences: *p < 0.05, ****p < 0.0001 vs C. Figure S2: IL4 treatment. a) C26 number of cells treated for 72 h in vitro with 10, 100 and 500 ng/ ml IL4. b) Absorption curve of IL4 in Balb/c mice at different time points after 1.3 ug IL4 administration (n = 3 per time point). c) Tumor weight in C26 and C26 + IL4 at the day of the sacrifice (14 days after tumor cell injection; n = 4 per group). d) Quadriceps muscle weight of C, IL4, C26 and C26 + IL4 mice (C: 410 ± 33 mg/100 g i. b. w.). Representative flow cytometry analysis on cells extracted after peritoneal lavage from (e, h) 1 mg/kg LPS injected mice (as positive control), or C26 and C26 + IL4 peritoneal lavage was analyzed for the main markers of (f) lymphocytes and (g) quantified or for (i) macrophage populations and (j) quantified. Significance of the differences **p < 0.01, ***p < 0.001, ****p < 0.0001 vs C; + IL4 counteracts cancer cachexia +p < 0.01, +++p < 0.001 vs IL4; $p < 0.05, $$$p < 0.001 vs C26. Figure S3: Flow cytometry analysis of circulating immune cells. Representative flow cytometry analysis of the circulating immune cells from C26 and C26 + IL4 mice for the main markers and quantifications of (a) lymphocytes and (b) macrophages. Significance of the differences $$$p < 0.001 vs C26. Figure S4: analysis of other important protein expressions in muscle tissue. a) WB for p38 MAP-Kinase and (b) quantification. c) WB for nuclear extract of P-STAT3 and P-STAT6 with (d, e) respective quantifications. f) WB for Atrogin1 and TRIM32 with (g, h) respective quantification. i) WB for Beclin1, LC3B, p62 and (j -l) respective quantifications. m) WB for the comparison of p62 levels between C26 and C26 + IL4 and (n) quantification. All the values were normalized for the total protein levels and on aTub or GAPDH expression levels. Significance of the differences is reported as *p < 0.05, **p < 0.01, ***p < 0.001 vs C; ++p < 0.01, +++p < 0.001 vs IL4; $p < 0.05 vs C26. Figure S5: Numbers of cells extracted from muscle of C, IL4, C26 and C26 + IL4 normalized for the muscle mass. a) Cells were counted immediately after tissue digestion. b) IF on muscle slides for the expression of PDGFRa (green) and Ki67 (magenta) from C, IL4, C26, C26 + IL4 muscles. HOECHST (blue) was used to stain nuclei. The arrows indicate double positive nuclei surrounded by a green signal. *p < 0.05 vs C; ++p < 0.01 vs IL4; $$$ p < 0.001 vs C26. Scale bar: 100 μm. Figure S6: IL4-treated muscles do not show impairment in regeneration after CTX injury. a) Average of fibre CSA in muscles after 10 days of CTX injection. b) H&E staining of muscle sections taken at longer time points after CTX injury was induced. In particular, for CTX-C26 (n = 5), 9 to 12 days after CTX injury, and for CTX-C26 + IL4 mice (n = 5), 16 to 30 days after CTX injury. Significance of the differences ***p < 0.001 vs C; +++p < 0.001 vs IL4; $$$p < 0.001 vs C26. Figure S7: Increasing concentrations of IL4 dos not rescue 50 ng/ ml TNFa-impaired myotube differentiation. a) qRT-PCR for the expression of IL4, IL4Ra, IL13 and IL13R in C2C12 at day 0 and 5 of myotube differentiation normalized for the housekeeping genes Gapdh, HPRT, TBP. b) C2C12 myoblasts at day 2 of differentiation were analyzed for IL4R expression on the levels of aTUB. c) MyHC levels by WB were normalized for aTub and quantified in C2C12 treated with low concentration of TNFa (50 ng/ml) and (d) 100 ng/ ml IL4 or (e-f) 200 ng/ml IL4. g) Silencing for IL4R1a and (h) for IL13R1a in myotubes was measured by qRT-PCR during the three last days of differentiation (esi24h = day 3, esi48h = day 4, esi72h = day 5 of differentiation medium). i) Levels of Myomixer (Mymx), Myomaker (Mymk) and Myomerger (Gm7325s and Gm7325l) were analyzed on the IL4R1a silenced cells by qRT-PCR normalized for the housekeeping genes Gapdh, HPRT, TBP. j) WB and (k) quantification for ESGP (Myomerger) on the levels of GAPDH in CTX-C CTX-IL4, CTX-C26 and CTX-C26 + IL4 muscle protein extract. Significance of the differences *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 vs C; +p < 0.05, +++p < 0.001 vs IL4. £p < 0.05 vs CTX-C;^p < 0.05 vs CTX-IL4.