Selumetinib Attenuates Skeletal Muscle Wasting in Murine Cachexia Model through ERK Inhibition and AKT Activation

Cachexia is a serious pathologic condition that affects about 30% of all patients with cancer (1). It is associated with reduced treatment tolerance, therapeutic response, quality of life, and survival (2). Cachexia is characterized by weight loss and skeletal muscle wasting. Previous studies have shown that weight loss in cachexia was mainly due to ongoing loss of skeletal muscle mass (3, 4). Animal experiments suggested that maintaining skeletal muscle mass not only reversed the loss, but also dramatically prolonged survival, without affecting fat loss or tumor growth (5). However, skeletal muscle wasting cannot be reversed by appetite stimulants or nutritional support, leaving few approaches to preserve lean body mass (6). Therefore, the pharmacologic approach for the prevention of skeletal muscle proteolysis is a topic of great interest in the management of cancer cachexia.

The MAPK/MAPK kinase (MEK)/extracellular signal–regulated kinase (ERK) signaling cascade is an important pathway controlling cellular processes in muscle atrophy (7–9). Activation of MEK/ERK pathway was detected in the wasted muscle tissue of in vivo cancer cachexia model and in the differentiated myotubes of in vitro myofiber atrophy model (10–13). Progressive skeletal muscle wasting was mainly mediated via ubiquitin–proteasome pathway (14). MuRF1 and MAFbx are the two main E3-ubiquitin ligases responsible for skeletal muscle protein ubiquitination proteasomal degradation (15). Previous results suggested that sorafenib, a multitargeted anticancer drug, prevented tumor-induced muscle depletion and reversed the elevated MAFbx/atrogin-1 expression via inhibition of ERK activity (13). However, a placebo controlled study showed muscle loss was specifically exacerbated by sorafenib treatment (16). The conflicting results suggested that ERK was involved in skeletal muscle mass control and represented a target for pharmacologic intervention against cancer cachexia.

Selumetinib is a potent and highly specific MEK1 inhibitor. In vitro studies showed that it inhibited ERK1/2 phosphorylation (17–19). In a phase II study of patients with stage IV cholangiocarcinoma (20), skeletal muscle anabolism was a specific side effect of selumetinib treatment. Muscle gain was seen in 84.2% of patients with an average gain of 3.9 kg in lean body weight (21). In vitro studies showed that it inhibited the secretion of cytokines such as IL6, IL1β, and TNFα, which promote cancer cachexia (22, 23). However, the in vivo effect of selumetinib remains unproven and the specific mechanisms were not investigated.

In the current study, a classical cancer cachexia model was established by the transplantation of CT26 colon adenocarcinoma cells into BALB/c mice. The role of selumetinib in body weight, tumor growth, skeletal muscle depletion, food intake, and cytokines was analyzed. The effects of selumetinib on E3 ligases and MEK/ERK–related crosstalk were analyzed. These studies not only validated the selumetinib treatment of cancer cachexia, but also elucidated the precise mechanism of MEK/ERK inhibition in skeletal muscle anabolism.

The CT26 colon adenocarcinoma cells (ATCC CRL-2638) were obtained from ATCC and amplified on the basis of the manufacturer's recommendations. Briefly, the cells were cultured in RPMI1640 medium with l-glutamine (Biowest) containing 10% FBS (Biowest) and penicillin–streptomycin (100 U/mL and 100 mg/mL, respectively) in a humidified atmosphere with 5% CO₂ at 37°C. The cell line used in this study was authenticated by ATCC and thawed from early passage stocks. As the cell line was passaged for fewer than 6 months, no authentication was done by the authors. The antibodies targeting mouse MHC (ab51263), MuRF1 (ab77577), and Fbx32 (ab74023) were obtained from Abcam. The antibodies against mouse Erk (#9102), p-Erk (#8544), Akt (#2920), and p-Akt (#13038) were purchased from Cell Signaling Technology. The antibodies against mouse mTOR (sc-8319), p-mTOR (sc-101738), GSK3β (sc-9166), p-GSK3β (sc-11757), FoxO3A (sc-11351), and p-FoxO3 (sc-101683) were from Santa Cruz Biotechnology. The antibodies against mouse PCNA (R1306) were from Hangzhou Huaan Biotechnology Co., Ltd. The ELISA kit for proinflammatory cytokines IL6, TNFα, and IL1γ was purchased from Dakewe Biotech. Selumetinib was purchased from MedChem Express.

BALB/c mice (weighing about 20 g) were purchased from Shanghai SLAC Laboratory Animal Co. Ltd and maintained at a constant temperature and humidity, with free access to food and water. They were handled in compliance with the Policy on Humane Care and Use of Laboratory Animals. Experimental protocols were approved by the Animal Care Committee of our institution in accordance with the government guidelines for animal manipulations in China. The cancer cachexia model was established by subcutaneous transplantation of CT26.WT cells into the left flanks of male mice as described previously (24, 25).

In the first study, we assessed the whole course prophylactic and therapeutic effects of selumetinib on tumor-induced cachexia from procachexia to cachexia period. The experiment protocol is shown in Fig. 1. Briefly, after acclimatization for 1 week, 40 male mice (average body weight 20.76 g with SD 1.22 g) were randomly divided into four groups, and designated as normal control, selumetinib-treated control, cancer cachexia, and selumetinib-treated cancer cachexia groups. Food intake, body weight, and tumor size were monitored daily. The tumor was palpable on day 8. Tumor length and width were measured using a digital caliper daily. Starting with day 8, mice in the selumetinib treatment groups were orally administered 30 mg/kg/d selumetinib, which was stabilized with sodium carboxymethyl cellulose at a concentration of 6 mg/mL. The other mice were treated with vehicle. On day 25, the skin over the tumor was partly ulcerated, and the body weight was decreased intensively. All the mice were anesthetized with ether and the tumors were dissected, measured, and weighed. We calculated the tumor weight (g) using the formula, 0.51× tumor weight (cm) × length (cm)². The body weight was calculated by subtracting the tumor weight from the total animal weight. The dynamic differences in body weight in vivo were calculated. In this experiment, we found that the body weight was lost significantly starting on day 11.

The second study evaluated the therapeutic effects against tumor-induced cachexia and investigated its related mechanisms. Forty male mice (average body weight 21.27 g with SD 1.06 g) were randomly divided into four groups as in the first experiment. Selumetinib was also orally administered at a dosage of 30 mg/kg/d. The first dosage was administered starting with day 11. On day 19, the mice were anesthetized with ether. The blood was collected into tubes and the serum was prepared within 1 hour for measurement of proinflammatory cytokines. Tumor, epididymal fat, and organs were dissected and weighed. The gastrocnemius and tibialis anterior muscle from both the hind legs were quickly removed and weighed. One half was frozen in liquid nitrogen to study the mechanism and the other half was fixed in 10% neutral-buffered formalin for histopathologic analysis.

Total IL6, TNFα, and IL1γ levels in the serum were measured following the manufacturers' recommendations. The levels were quantified by measuring the absorbance at 450 and 630 nm using Bio-Tek Instruments (Bio-Tek). The lowest detectable level of these cytokines was 7.8 pg/mL.

Transverse serial sections of the gastrocnemius muscle and the tumor were fixed with 4% paraformaldehyde in 0.2 mol/L PBS for 5 minutes at room temperature, washed with PBS, and stained using hematoxylin and eosin (HE) staining (Sigma Chemical Co.). Images of muscle sections were recorded using Leica DFC295 Microsystems. The paraffin sections were also used for tumor IHC. After deparaffinization in xylene and rehydration in ethanol, slides were treated with hydrogen peroxide to inactivate endogenous peroxidases and placed in 10 mmol/L of citrate buffer. After incubation with the primary antibodies against MEK (1:100) and PCNA (1:100) for 1 hour, a secondary antibody was used to detect protein expression. Finally, a substrate–chromogen mix was used for the visualization of the immune reaction.

Muscle tissues from animals were homogenized and solubilized in lysis buffer [20 mmol/L Tris (pH 7.5), 150 mmol/L NaCl, 1% Triton X-100, sodium pyrophosphate, β-glycerophosphate, Na₃VO₄, EDTA, and leupeptin] using a commercial kit (Beyotime Institute of Biotechnology). The extracts were centrifuged to remove insoluble matter. After evaluating the protein content using a Bicinchoninic Acid Protein Assay Kit (Pierce), protein was degenerated in sample loading buffer. A 30-mg protein sample was loaded onto each lane of a 10% SDS-polyacrylamide gel for the separation of low molecular weight proteins, in addition to the 8% SDS-polyacrylamide gel for MHC proteins (homogenized and solubilized in high-salt lysis buffer). Proteins were subsequently electrophoretically separated and transferred onto polyvinylidene difluoride membranes (Millipore Corporation) for Western blot analysis. Primary antibodies were prepared and incubated according to the manufacturer's specifications at a dilution of 1:1,000. The protein expression was detected using a peroxidase-conjugated secondary antibody (1:5,000 diluted in Tris-buffered saline with Tween-20) and chemiluminescence using ImageQuant software. The protein expression was quantified using Gel-Pro Analyzer program (Media Cybernetics). Equal distribution of protein loading was verified by probing the blots with an anti-β-actin antibody.

The relative expression of protein was determined by comparing the treatment values to control values after normalization to loading controls. Data were summarized by treatment group using mean and SD. Statistical significance was assessed by one-way ANOVA. Post hoc analysis was performed by Tukey multiple comparison test to test the significance of differences using the GraphPad Prism software. Significant changes are indicated by P values of <0.05.

As showed in Fig. 2A and B, the first animal experiment revealed ongoing loss in body weight of the tumor-bearing mice around day 9, warranting early intervention against cachexia (26). In the current study, selumetinib was administered on day 9. Consistent with our previous results, compared with the vehicle-treated control, the body weight of corresponding cachexia groups decreased more than 5% and the difference was significant on day 11 after the tumor implantation (24). These results revealed cancer cachexia syndrome on day 11. Administration of selumetinib partially prevented weight loss from day 12 to day 19. However, from day 20, the weight loss was accelerated in the vehicle-treated cachexia group and selumetinib therapy partially attenuated the ongoing weight loss. Within 25 days of implantation, the average loss of body weight was 8.50 g and the average tumor weight was 3.64 g. Selumetinib treatment resulted in an average increase of 2.13 g lean body weight, while the tumor weight only decreased by 0.55 g. The body weights of vehicle-treated cachexia group declined to about 34.2%, while selumetinib conserved 11.4% of the body weight. These comparisons indicated that the skeletal muscle anabolism of selumetinib treatment was not completely due to the reduction of tumor burden.

Early treatment of cachexia is not clinically feasible due to limitations of early diagnosis. Therefore, we performed a second optimized animal experiment in which selumetinib was administered during the typical cancer cachexia period. On the basis of the first study, the body weight was decreased dramatically from day 11 to 19, and the first dosage was administered on day 11. The corresponding mice were also orally treated for 9 days with selumetinib at a dosage of 30 mg/kg/d and all the mice were euthanized on day 19 with ether. As showed in Fig. 2C and D, selumetinib treatment resulted in a 16.09% body weight gain and 15.11% decrease in tumor weight. Prevention of body weight loss consistently with selumetinib confirmed its anticachexia effect.

Cachexia is mainly characterized by the wasting of skeletal muscle and adipose tissue. We assessed whether selumetinib prevented muscle wasting and fat loss. Weights of the gastrocnemius muscle, tibialis anterior muscle, and epididymal fat in tumor-bearing mice were 24.65%, 22.28%, and 84.99% less than in the control groups, respectively, as showed in Fig. 3. Although selumetinib treatment of the cachexia group decreased weight loss by 26.53%, 12.67%, and 61.62% in the gastrocnemius muscle, tibialis anterior muscle, and epididymal fat, respectively, when compared with vehicle treatment control group, the weights were significantly higher than in the vehicle-treated cachexia group. Selumetinib treatment resulted in 10.76%, 12.37%, and 155.75% increase in weight of gastrocnemius muscle, tibialis anterior muscle, and epididymal fat. From the wasting ratio of muscle and fat, we could conclude that selumetinib ameliorated cancer cachexia syndrome partly through preventing the wasting of skeletal muscle and adipose tissue. As there was no significant difference of the expression and activation of adipose triglyceride lipase between healthy control and cancer cachexia mice in our preliminary experiment, the possible mechanisms of fat preserve were not analyzed in the current study. The HE-staining results of the gastrocnemius muscle confirmed myofibers were wasting in the cachexia group, while muscle atrophy was reversed by selumetinib treatment (Fig. 4A). The effect of selumetinib in attenuating muscle wasting in relation to the anticancer effect remains to be investigated. Although selumetinib reduced the tumor burden by 20.8%, immunohistochemical analysis of the tumor revealed that selumetinib had no influence on CT26 cell proliferation as the PCNA expressions between selumetinib and vehicle treatment group were comparable (Fig. 4B). However, selumetinib treatment decreased MEK expression when compared with the expression of PCNA in the tumor. The diminishing MEK expression of tumor and the comparable cell proliferation indicated that the selumetinib treatment resulted in the decrease of MEK^high expression cancer cell. These results indicated that the role of selumetinib in attenuating muscle atrophy was not completely dependent on reducing cancer burden.

The cumulative food intake of the four groups was similar. The cumulative food intakes for vehicle-treated control and cancer cachexia mice were 512.4 and 496.5 g, while cumulative food intakes for selumetinib treatment control and cancer cachexia mice were 506.7 and 489.6 g, respectively. The results were consistent with the previous results showing that the CT26 tumor and selumetinib had little influence on appetite (27). Excessive production of proinflammatory cytokines, such as TNFα, IL1β and IL6, triggered proteolytic degradation of skeletal muscle clinically and in experimental animals with cancer cachexia. The tumor induced proinflammatory cytokines that acted on the corresponding receptors to activate proteolytic signaling pathways in skeletal muscle tissues. Furthermore, inhibition of the dysregulated cytokines attenuated skeletal muscle wasting in experimental animals (28). Therefore, we tested whether selumetinib reversed the excessive production of cytokines. It was interesting that cachexia groups showed high levels of TNFα, IL1β, and IL6 (Fig. 5). However, there was no significant difference in serum cytokine concentrations between selumetinib and vehicle-treated cachexia group. The results indicated that selumetinib had limited role in the production and release of proinflammatory cytokines. The effect of selumetinib on tumor weight, appetite stimulation, and serum cytokine levels suggests that its role in attenuating muscle wasting was independent of tumor burden and food intake.

MuRF1 and Fbx32 are the main E3 ubiquitin ligases in skeletal muscle and in muscle proteolysis (15). Under conditions of cachexia, MHC is the main proteolytic component that is degraded via ubiquitination pathway. Loss of skeletal muscle mass is associated with physical disability, diminished quality of life, and reduced survival. In vivo studies suggest that the reversal of muscle wasting resulted in prolonged survival (5). As reported in previous studies, we found that MuRF1 and Fbx32 were significantly upregulated in experimental cachexia mice (29). Selumetinib treatment significantly inhibited the expression of two muscle-specific E3 ligases, MuRF1 and Fbx32, as well as the increased expression of MHC in gastrocnemius muscle (Fig. 6). These results indicated that selumetinib decreased ubiquitin-mediated protein degradation.

Selumetinib is a potent, highly selective MEK1 inhibitor. MEK/ERK pathway activates ubiquitination during muscle proteolysis. MEK phosphorylates ERK and activates the cascade of cytoplasmic signaling proteins and transcription factors, which regulate the genes related to ubiquitin proteasome. Selumetinib treatment resulted in decreased phosphorylation of ERK1/2 (Fig. 7). Activation of MEK/ERK pathway enhanced ubiquitination. Therefore, the decrease in E3 ligases MuRF1 and Fbx32, rather than excessive production of proinflammatory cytokines, followed selumetinib treatment in the cachexia group.

Furthermore, it is interesting that the inhibition of MEK/ERK pathway activated PI3K/AKT pathway, as reflected by the increased AKT phosphorylation. ERK activation overlapped partially with that of the PI3K/AKT signal transduction pathway. Previous studies showed that AKT and its three downstream kinases regulate muscle hypertrophy. It was reported that the loss of mTOR in skeletal muscle could result in a severe myopathy (30). Moreover, AKT and mTOR were inhibited in various muscular atrophy models (31). In myofibers, mTOR acted as a central regulator protein synthesis by integrating signals from nutrients, growth factors, and energy status. In our study, we found AKT and mTOR were inhibited in cancer cachexia model. However, selumetinib treatment resulted in the enhanced activation of AKT and mTOR. It is widely reported that the reduction of AKT phosphorylation lead to FoxO3a activation and increased MuRF1 transcription, thus controlling cell survival and proliferation (14). FoxO3a also regulated crosstalk between the PI3K/Akt and MEK/ERK pathways. Previous study showed when p-ERK/ERK ratio was increased after EGF treatment, p-FOXO3a/FoxO3a ratio decreased in nuclear, while increased in cytoplasmic, (Yang Jer-Yen et al. "ERK promotes tumorigenesis by inhibiting FOXO3a via MDM2-mediated degradation" Nature cell biology 10 (2008): 138-148, DOI: 10.1038/ncb1676). In the present study, we found the inactivation of FoxO3a in wasting muscle of cachexia mice. Selumetinib treatment resulted in partial downregulation of FoxO3a and this might be responsible for the myofiber proliferation. The third significantly changed downstream signaling molecule was GSK3β. Reduced PI3K/Akt activity in muscle atrophy during cancer cachexia leads to the activation of the GSK3β. However, selumetinib treatment could inhibit the activation of GSK3β. GSK3β was mainly involved in the regulation of energy metabolism and cell growth (32). On the basis of the regulating effect of MEK/ERK and PI3K/AKT pathways, we proposed the mechanism for selumetinib treatment on the reversal of skeletal muscle wasting (Fig. 8). These findings indicate that the hypertrophic signaling cascades were activated by selumetinib in cancer cachexia–induced atrophy of skeletal muscle.

Cancer cachexia is a serious multifactorial syndrome. It is responsible for more than 20% of cancer-related deaths. Several extracellular growth factors and intracellular signaling pathways are involved in the growth and differentiation of skeletal muscle. The role of MEK/ERK signaling pathway in skeletal muscle wasting is controversial (10, 33). We found that ERK phosphorylation was enhanced in wasting skeletal muscle in a murine model of cancer cachexia. Furthermore, pharmacologic inhibition of MEK/ERK by selumetinib significantly alleviated weight loss and attenuated muscle wasting reflected by the decreased expression of E3 ligases MuRF1 and Fbx32. Interestingly, selumetinib treatment had little effect on excessive production of the proinflammatory cytokines IL6, IL1β, and TNFα.

Previous studies emphasized the inhibition of synthesis and release of proinflammatory cytokines IL6, IL1β, and TNFα in cancer cachexia (34). Our experiment also showed elevated serum levels of these cytokines in mice with cancer cachexia. However, the MEK inhibitor selumetinib showed no effect on elevated serum cytokines. The excessive production of proinflammatory cytokines led to MEK phosphorylation and activation of MEK/ERK pathway. TNFα treatment upregulated the expression of atrogin1/MAFbx in C2C12 myotubes as well as the wasting skeletal muscle in adult mice (12, 35). However, our data showed in wasting skeletal muscle the downregulation of E3 ligase (MuRF1 and Fbx32) expression by selumetinib treatment was not mediated via proinflammatory cytokines.

The detailed mechanism was investigated by measuring the phosphorylation of ERK. Inhibiting the activation of ERK by selumetinib treatment resulted in the downregulated transcription and translation of downstream MuRF1 and Fbx32. These two E3 ligases are responsible for the ubiquitin proteolytic degradation in the wasting skeletal muscle during cancer cachexia. Furthermore, the associated AKT and downstream signaling pathways were investigated. Our data suggested that the MEK activity in skeletal muscle was elevated, while ATK activity was decreased in catabolic states of cancer cachexia. Pharmacologic inhibition of MEK following selumetinib treatment not only inhibited the activation of ERK, but also enhanced AKT activation. The MEK/ERK and PI3K/Akt signaling pathways interact with each other (36, 37). Differentiated myofibers were simultaneously regulated in response to the same stimulant, but exerted opposing effects (33). During muscle atrophy, the MEK/ERK pathway was activated while Akt activation was inhibited in the differentiated myotubes (34). Blocking MEK chemically activated the PI3K/AKT pathways via cross-inhibition (38). Activation of MEK/ERK pathway was associated with degradation of the muscle protein, while PI3K/AKT pathway promoted the differentiation and hypertrophy of myofibers (39). Consistent with previous reports, pharmacologic inhibition of MEK via selumetinib treatment attenuated ERK phosphorylation and enhanced AKT activation (40). The phosphorylation of AKT by selumetinib treatment resulted in the activation of mTOR, as well as inhibitions of FoxO3a and GSK3β. Phosphorylation of mTOR would promote protein synthesis. However, inhibition of GSK3β and FoxO3a would regulate myofiber energy metabolism and proliferation.

Taken together, the current study showed that selumetinib efficiently attenuated ongoing muscle wasting in a murine cachexia model. It inhibited ERK phosphorylation and enhanced AKT activation in the gastrocnemius muscle without decreasing the excessive production of inflammatory cytokines IL6, IL1β, and TNFα. It should be noted that skeletal muscle wasting was refractory and there is no treatment clinically recommended for muscle atrophy. Early nutritional intervention was recommended for the prevention of muscle wasting. However, no clinical biomarkers are available for early diagnosis and drug treatment. Selumetinib is recommended for clinical application if monitored via appropriate biomarkers. Furthermore, MEK/ERK pathway is an important pathway controlling extensive cellular processes. The extensive pharmacologic activities and nontargeted effects of selumetinib need in-depth research. These current results only validated the potential of selumetinib in the prevention and treatment of skeletal muscle wasting during cancer cachexia.

No potential conflicts of interest were disclosed.

Conception and design: Y. Quan-Jun, G. Run

Development of methodology: Y. Quan-Jun, H. Yong-Long, W. Li-Li, L. Jie, H. Jin-Lu, G. Run

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Quan-Jun, L. Jin, G. Run

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Quan-Jun, H. Yong-Long, C. Peng-Guo, G. Run

Writing, review, and/or revision of the manuscript: Y. Quan-Jun, H. Yong-Long, G. Run

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y. Quan-Jun, H. Yan, G. Run

Study supervision: H. Yan, G. Run

Other (animal experiment): G. Run

This work was supported by grants from the Natural Science Foundation of China, (81503155; awarded to Q. Yang).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.