Aerobic exercise, but not metformin, prevents reduction of muscular performance by AMPk activation in mice on doxorubicin chemotherapy

2.1 Animals

C57BL/6 male mice aged 8–10 weeks were used. The animals were kept in a room with a light–dark cycle of 12–12 hr and at a temperature of 22 ± 2°C, with a normal diet (Nuvital ration of Nuvilab, Colombo, PR, Brazil) and with food and water ad libitum. All the procedures of this study followed the ethical principles of animal experimentation and were submitted to the Committee of Ethics in Animal Experimentation of the University of São Paulo.

Part of these animals received DOX chloridrate (Eurofarma Laboratory, Campinas, Brazil), 2.5 mg/kg body weight, twice a week (DOX), whereas the control group (CT) received the same volume of saline solution. Animals that received DOX were still further divided into groups of animals that performed treadmill exercise (DOX+T) or received metformin (Sigma-Aldrich, St. Louis, MO), 300 mg/kg of body weight daily by gavage (DOX+MET).

After 6 weeks, mice were fasted for 6 hr and euthanized for the collection of blood and tissue samples. Gastrocnemius (GC) muscle and adipose tissue cushions were weighed and stored for the further analysis of messenger RNA and protein expression.

2.2 Exercise protocol

All groups performed 1 week of treadmill adaptation, in which they ran at 10 m/min for 5 days, before beginning the treatment protocols. From 6 weeks, the DOX+T group was submitted to an aerobic training protocol, 60 min at 60% of the maximal speed run (MS), 5 days/week. The MS consisted of a warm-up (5 min at 10 m/min) phase and from the sixth minute, the treadmill speed increased 3 m/min until the animals had run mechanical alterations, featuring exhaustion (n = 13–14).

2.3 Serum analysis

Fasting blood glucose was assessed using a colorimetric assay. Serum insulin (Millipore Corp., Bedford, MA), adiponectin (DuoSet ELISA, R&D Systems, Minneapolis, MN), testosterone, and corticosterone (Assay Designs, Inc., Ann Arbor, MI) were quantified using the enzyme-linked immunosorbent assay.

2.4 Cross-sectional area analysis

This method was adapted from Baehr et al. (2016). Briefly, frozen muscles were cut 10 μm thick from the medial portion of the GC using a Leica CM 3050S cryostat (Leica Microsystems, Nussloch, Germany). To determine fiber cross-sectional area (CSA; μm²), sections were stained with anti-laminin (1:1,000; Sigma-Aldrich) and visualized using the 3,3′-Diaminobenzidine (DAB) detection method according to the manual. Laminin images were acquired using a bright field Nikon E1000 microscope (Melville, NY). Gastrocnemius muscle images at 200× total magnification were captured using a Nikon DMX1200 digital camera (Nikon Corporation, Melville, NY). For each muscle, four nonoverlapping regions were used, and at least 200 fibers were analyzed per animal. Images were quantified using Image Pro Plus analysis software (Image Pro Plus software, NIH). The same four regions were analyzed across all mice. All analyses were conducted by a single observer blinded to the mouse’s identity. Data are reported as means ± standard error of the mean based on individual animal mean values (n = 5 – 6).

2.5 Enzyme assay

For determination of citrate synthase activity, soleus muscles were lysed using complete protease inhibitor cocktail (Roche) and CelLytic^TM MT Cell Lysis Reagent (cat no. C3228; Sigma-Aldrich; n = 6 – 7). Citrate synthase activity was assessed using 8 μg protein with the Citrate Synthase Assay Kit (cat no. CS0720; Sigma-Aldrich).

2.6 Puromycin assay

Protein synthesis was measured using the surface sensing of translation method, as described by Baehr et al. (2016). After dissolving the puromycin (cat. no. 5540222; EMD Millipore, Billerica, MA) in sterile saline, animals received an intraperitoneal (i.p.) injection (0.02 μmol/g body weight) 30 min before starting muscle collections. Thirty minutes after i.p. injection, the gastrocnemius muscle was excised, flash frozen in liquid N₂, and stored at –80°C. Gastrocnemius was powdered and homogenized in sucrose lysis buffer (50 mM Tris pH 7.5, 250 mM sucrose, 1 mM EDTA [ethylenediaminetetraacetic acid], 1 mM ethylene glycol-bis(β-aminoethyl ether)-N,N,N',N'-tetraacetic acid, 1% Triton X 100, and 50 mM NaF). After a 10 min centrifugation at 8,000g, protein concentrations were determined using the Bradford method (Bio-Rad^®, Hercules, CA). Twenty micrograms of protein was separated on 10% acrylamide gels by sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE) and transferred to the nitrocellulose membrane (Amersham Biosciences, Piscataway, NJ). The membrane was stained with Ponceau S and imaged to see total protein. Membranes were washed in 1× Tris-buffered saline with 0.1% Tween-20 (TBST) three times (5 min each wash), then blocked in 3% nonfat dairy milk for 1 hr. After a wash, the membranes were then probed with a primary antibody against puromycin (cat. no. MAEBE343; EMD Millipore, Temeluca, CA) overnight at 4°C. After a wash, membranes were incubated with peroxidase-conjugated secondary antibody at the concentration of 1:10,000 for 1 hr at room temperature, washed, and incubated in the Western ECL substrate detection system for 5 min before image acquisition (Clarity TM; Bio-Rad). Image acquisition and band quantification were performed using the AI600 (GE) System and Image Studio^TM software version 5×.

2.7 Western blotting

The gastrocnemius muscles were carefully homogenized in the extraction buffer containing protease and phosphatase inhibitors. After proper centrifugations, the proteins were determined by the Bradford assay (Bio-Rad^®, Hercules, CA). Aliquots of each sample with the same concentration of total protein (20–40 µg) were diluted in Laemmli buffer. The samples were then subjected to electrophoresis on SDS-PAGE, transferred to polyvinylidene difluoride membrane, which was incubated first with primary antibodies at a 1:1,000 dilution against phospho-AMPKα Thr172 (cat. no. 2531), AMPKα (cat. no. 2532), pS6 Thr389 (cat. 9205), 4EBP1 Thr37/46 (cat. no. 2855), 4EBP1 (cat. no. 4644), Beclin (cat. no. 3495), LC3β (cat. no. 12741), ATG7 (cat. no. 8558)—all obtained from Cell Signaling Technology (Danvers, MA)—Puromycin (cat. no. MAEBE343; EMD Millipore, Temeluca, CA), and PGC-1α (cat. no. AB54481; Abcam, Cambridge, MA) and then incubated with an anti-IgG antibody conjugated with a peroxidase-conjugated secondary antibody at a 1:10,000 dilution for 1 hr at room temperature then washed three times per 5 min in TBST. After the incubations and washes, these membranes were incubated with the peroxidase substrate (ECL kit; Bio-Rad^®) and quantified by optical densitometry. Ponceau S staining was used as a loading control (Gilda & Gomes, 2013; n = 4–6).

2.8 Proteasome activity

The activities of 20S and 26S proteasome were performed using muscle powder from whole gastrocnemius muscle, as previously described by Baehr et al. (2016). The chymotrypsin-like (β5) activity assays were carried out in a total volume of 100 μl in 96-well opaque plates. The 26S ATP-dependent assays were performed in a homogenization buffer with the addition of 100 μM ATP. The 20S ATP-independent assays were carried out in the assay buffer containing 25 mM 4-(2-hydroxyethyl)-1piperazineethanesulfonic acid (HEPES), 0.5 mM EDTA, and 0.001% SDS (pH 7.5). Proteasome activities were determined using 100 μM of Z-Leu-Leu-Glu-7-AMC (Peptide Institute), Boc-Leu-Ser-Thr-Arg-7-AMC (Bachem), or succinyl-Leu-Leu-Val-Tyr-7-amido-4-methylcoumarin (LLVY-AMC; Bachem) for the β5 subunit. Each assay was conducted in the absence and presence of the proteasome inhibitor bortezomib at a final concentration of 2 mM. Released AMC was measured by a Synergy™ H1 hybrid Multi-Mode Microplate Reader (Biotek Highland Park, Winooski, VT) at an excitation wavelength of 390 nm and an emission wavelength of 460 nm. Fluorescence was measured at 15 min intervals for 2 hr. The activity of the 20S and 26S proteasome was measured by calculating the difference between fluorescence units recorded with or without the inhibitor in the reaction medium. (n = 4 – 6).

2.9 Quantitative real-time polymerase chain reaction

The expression of skeletal muscle genes was assessed by quantitative real-time polymerase chain reaction (qRT-PCR) with SYBR Green marker. Total RNA from the gastrocnemius muscle was extracted with Trizol reagent (Invitrogen Life Technologies, Carlsbad, CA). The complementary DNA (cDNA) was synthesized from 2 μg of total RNA extracted using reverse transcriptase with the High-Capacity cDNA kit (Applied Biosystems, Warrington, UK). RNA was measured by a spectrophotometer at 260 nm, and purity was determined by the ratio 260/280 nm. Gene expression was evaluated by the RT-PCR using a Rotor Gene (Qiagen) and SYBR Green as the fluorescent dye. Quantification was performed using the comparative Ct method. Primer sequences are shown in Supporting Information Table 1 (n = 6–7).

2.10 Statistical methods

The data are presented as mean ± standard deviation and analyzed by one-way analysis of variance, followed by the the Bonferroni post test. Analyses were performed using GraphPad Prism 6.0 software. Differences were considered significant when p < 0.05.

3.1 Body weight and food intake

All groups treated with DOX had decreased body weight (Figure 1a), but without change in food intake (Figure 1b), indicating that neither metformin nor aerobic training are able to restore the decreased body weight.

The loss of body weight is due to loss of both fat and lean body mass. The chemotherapy treatment caused extensive reduction of adipose tissue and skeletal muscle. The pharmacological treatment with metformin or exercise did not recovery the weight loss of fat pad (subcutaneous, epididymal, retroperitoneal, and mesenteric) and gastrocnemius muscle caused by DOX (Table 1). There was no change in brown adipose tissue weight (Table 1).

3.2 Effects on systemic parameters

On the basis of the reductions in adipose tissue weight caused by DOX, we evaluated the concentration of adiponectin in serum. No changes were observed (Table 2). Also, both adipose tissue and skeletal muscle play a key role in insulin-mediated glucose uptake. The chemotherapy decreased basal glycemia, without change in insulin concentration (Table 2). Adjuvant treatments did not alter this condition. The balance between anabolic and catabolic hormones is important in skeletal muscle homeostasis. Therefore, we investigated whether the treatments would cause changes in the circulating concentration of testosterone and corticosterone. Corticosterone was increased in the DOX group and both treatments groups (metformin and exercise) were able to reduce them (Figure 2a). The testosterone concentration was not altered (Figure 2b).

3.3 Aerobic running capacity

As expected, the final maximum velocity after 6 weeks of aerobic training increased in the DOX+T group, whereas in the group treated just with DOX, there was a decrease in this physical capacity (Figure 3a). These differences were not observed in the DOX+MET group. Furthermore, there was no change in the citrate synthase activity for any group (Figure 3b).

3.4 Exercise and metformin effects on CSA in skeletal muscle

Exercise and metformin did not recover the loss of CSA in gastrocnemius muscle (Figure 4a–c). As shown in Figure 4a–c, chemotherapy alone causes severe reduction of fiber area.

3.5 Effects on proteins involved in muscle metabolism

Although PGC1 alpha levels are unaffected by DOX treatment alone, a combination of DOX treatment with either metformin or exercise training significantly reduces PGC-1 alpha levels (Figure 5a). However, another important protein involved in skeletal metabolism was altered. Physical exercise and metformin are both AMPk activators. On the basis of this, we investigated the effect of the interventions. The AMPKα phT172/total AMPkα expression was increased in the DOX+T group (Figure 5b). There was a tendency to increase this protein in the DOX+MET group (p = 0.069). Adiponectin is one of the possible activators of AMPk. In this sense, we evaluated the receptor expression of this protein in the skeletal muscle. There was a reduction of gene expression of ADIPOR1 in the DOX+T group compared with the CT group (Figure 5c) without alterations in ADIPOR2 gene expression (Figure 5d).

3.6 Effects on protein synthesis and its signaling

On the basis of our results demonstrating reduction of muscle mass, we investigated whether protein synthesis could be involved in this process. We observed that protein synthesis is decreased with DOX treatments (DOX group and DOX+MET), and aerobic exercise rescue this effect (Figure 6a). Then, we investigated possible targets in the protein synthesis signaling pathway that could be involved in this process. We did not observe change in pS6k (Thr389) or p4EBP1(Thr37/46)/4EBP1 ratio (Figure 6b,c).

3.7 Effects on protein degradation pathways

Autophagy and proteasomal degradation could be other possible mechanisms that could be involved in muscle wasting previously observed. Thus, we investigated how the treatments could change the proteasomal activity. Treatment with DOX alone does not alter the proteasome activity. However, the DOX+MET group showed increase in the 20S β5 activity in comparison with all other groups (Figure 7a). There was no change in 26S β5 activity (Figure 7b). MAFbx/atrogin-1, an important protein expressed during muscular atrophy, was not changed either (Figure 7c).

With regard to autophagy, we observed that aerobic exercise and metformin reduces the expression of proteins involved in this process. The expression of autophagic proteins Atg7 and Beclin was reduced by both exercise and metformin (Figure 8b,c), while the protein expression of LC3β was reduced just for the DOX+T group (Figure 8a). These results are caused by the treatments themselves since there was no statistical difference between the CT and DOX groups.