Journal Logo


Exercise Protects Skeletal Muscle during Chronic Doxorubicin Administration


Author Information
Medicine & Science in Sports & Exercise: December 2017 - Volume 49 - Issue 12 - p 2394-2403
doi: 10.1249/MSS.0000000000001395


Cancer survivors represent a rapidly growing segment of our population (6). Unfortunately, many cancer survivors continue to experience deficits in physical function that occur not only from the burden of the disease but also from the side effects of chemotherapy treatments (15). In particular, anthracyclines, such as doxorubicin (DOX), are very effective chemotherapy agents with a broad clinical application for treatment of many cancers (23). However, DOX is also well known to produce devastating off-target effects (13,20,29–31) that promote a rapid reduction in cardiorespiratory fitness (27), the development of physical dysfunction (7,27), and an increased risk for hospital admissions (1). Consequently, the clinical utility of DOX is limited (23). On the other hand, the identification of effective strategies to attenuate or even eliminate these off-target effects could reduce the incidence of physical dysfunction and morbidity after treatment and improve the clinical utility of DOX.

In addition to the well-described effect of DOX on cardiac tissue and function (20,31), emerging data indicate DOX administration may directly promote atrophy (2,3) and dysfunction of skeletal muscle (4,14,17). The precise cellular mechanisms through which DOX exerts its effect on skeletal muscle continue to be unraveled; however, DOX is known to facilitate the formation of reactive oxygen species at complex I of the electron transport chain (23). In addition to promoting increased proteolytic activity and apoptosis, oxidative stress is also associated with the upregulation of the protein regulated in development and DNA damage 1 (REDD1) (19). Indeed, DOX administration was recently shown to result in increased expression of REDD1 mRNA in skeletal muscle (3,24). Importantly, REDD1 is a negative regulator of muscle size and appears to exert its effect through the inhibition of the mammalian/mechanistic target of rapamycin (mTORC1) signaling (16). mTORC1 signaling has a primary role in the stimulation of muscle protein synthesis (8) as well as the inhibition of autophagy (12), which is a process through which dysfunctional proteins, macromolecules, and organelles are degraded (9). Further, recent evidence indicates that the deleterious effect of DOX on skeletal muscle could be facilitated through reductions in protein synthesis rather than increases in protein breakdown (24). Collectively, these data highlight a potential role for REDD1 and mTORC1 signaling as important therapeutic targets for minimizing DOX-induced toxicities in skeletal muscle.

Previous studies have eloquently demonstrated the effectiveness of prior exercise training for protecting skeletal muscle from the acute damaging effects of a large, single bolus administration of DOX (4,20,29,30). Although the bolus doses used in these prior studies represent relevant accumulated clinical human doses, DOX is not typically administered as a single bolus but rather as multiple smaller doses over 2- to 4-wk cycles (36). Consequently, both the mechanisms that contribute to DOX-induced toxicities of skeletal muscle during chronic DOX treatment and the ability for exercise training to serve as a protective therapy for skeletal muscle during chronic DOX administration remain to be determined. In particular, exercise is a known stimulator of mTORC1 signaling activity (37); however, to our knowledge, the interaction among DOX, exercise, and mTORC1 signaling in skeletal muscle has yet to be examined. Therefore, the purpose of this study was to determine the extent to which interval-based exercise performed before and during biweekly DOX treatment can attenuate adverse effects of DOX on skeletal muscle. We hypothesized that DOX treatment would be associated with increased REDD1, impaired mTORC1 signaling, and reduced myocellular size in the soleus muscle, and that exercise training before and during treatment would attenuate these responses.


Study protocol

Eight-week-old ovariectomized female Sprague–Dawley rats were purchased from Charles River Labs (Wilmington, MA). DOX is considered to be one of the most effective treatments for breast cancer, and therefore, ovariectomized female rats were specifically studied to provide a gonadal estrogen-deficient model to align with the hormonal environment present in most breast cancer patients in which the median age of diagnoses is 61 yr (NCI SEER statistics). After 1 wk of acclimatization to the laboratory and 1 wk of familiarization to treadmill exercise (Columbus Instruments 3/6 treadmill, Columbus, OH), rats (n = 42) were randomized to one of four experimental groups (in accordance with American Physiological Society protocols for use of exercise in animal research): 1) sedentary + vehicle (Sed-Veh), 2) sedentary + DOX (Sed-Dox), 3) exercise trained + vehicle (Ex-Veh), and 4) exercise trained + DOX (Ex-Dox). On the basis of preliminary data obtained using our drug dosing protocol (see next section), a larger number of animals were randomized to the DOX groups, particularly Sed-Dox, to account for a greater risk of mortality in animals administered DOX. Total n size for each group after weighted randomization was Sed-Veh = 10, Sed-Dox = 14, Ex-Veh = 8, and Ex-Dox = 10. Rats were pair-housed with an animal from the same group, allowed access to food and water ad libitum, and maintained on a 12-h light–12-h dark cycle. All methods used in this investigation were approved by the Midwestern University Institutional Animal Care and Use Committee.

The experimental design is presented in Figure 1. Animals were injected (ip) with DOX (hydrochloride >99%; LC Laboratories, Woburn, MA) dissolved in 0.9% normal saline or vehicle (0.9% normal saline) in 2-wk intervals. During each injection, DOX was administered in doses of 4 mg·kg−1, totaling 12 mg·kg−1 received by each animal by the end the experimental protocol. Importantly, this dosing strategy and cumulative dose closely mimics doses and dosing strategies used in the clinical treatment of human cancer patients (36) (consultation from Mayo Clinic Arizona Breast Clinic). Animals assigned to exercise groups underwent interval style exercise 5 d·wk−1, beginning 1 wk before the first injection and sustained throughout the study duration. Specifically, exercise animals completed four 4-min intervals separated by 2 min of active rest (10 m·min−1). During the first week, exercise intervals were completed at 20 m·min−1 at 0° incline, which progressed to 25 m·min−1 at 10° incline by the last week of training. Each exercise session began and ended with a 5-min warm-up and cooldown at 10 m·min−1. Sedentary animals were not provided any access to exercise.

General schematic of the study design and timeline. After acclimatization, animals were randomized to one of four experimental groups, sedentary + vehicle (Sed-Veh), sedentary + DOX (Sed-Dox), exercise trained + vehicle (Ex-Veh), and exercise trained + DOX (Ex-Dox). Exercise groups began exercise 1 wk before first drug treatment and continued to exercise throughout study duration. Sed animals were not provided any access to running wheel. Veh animals were administered normal saline. DOX, doxorubicin.

Five days after the final injection, animals were anesthetized with isoflurane and then decapitated. This timeframe was chosen as DOX has been shown to accumulate in tissues and facilitate dysfunction for up to 5 d after administration (17). Exercise animals completed their last exercise session 24 h before sacrifice. Upon euthanasia, the soleus muscle of each leg was removed. The soleus muscle from one leg was immediately frozen in liquid nitrogen for immunoblot and mRNA analyses. The soleus muscle from the opposite leg was carefully embedded on a cork in Tissue Tek optimal cutting temperature (Thermo Fisher Scientific, Rockford, IL) and frozen in liquid nitrogen-cooled isopentane for immunohistochemical analysis. All samples were stored at −80°C.

Immunoblot analyses

Frozen soleus tissue was weighed (mean ± SD = 30.7 ± 11.1 mg) and homogenized (1:9 wet weight/volume) in an ice-cold buffer containing 50 mM Tris–HCl, 250 mM mannitol, 50 mM NaF, 5 mM sodium pyrophosphate, 1 mM ethylenediaminetetraacetic acid, 1 mM ethyleneglycotetraacetic acid, 1% Triton X-100, pH 7.4, 1 mM DTT, 1 mM benzamidine, 0.1 mM PMSF, and 5 μg·mL−1 soybean trypsin inhibitor. Samples were then centrifuged at 3400g for 10 min at 4°C, and the supernatant was collected. Total protein concentrations were determined using the Bradford assay (H1 Synergy, Biotek, Winooski, VT). The supernatant was diluted (1:1) in a 2× sample buffer mixture containing 125 mM Tris, pH 6.8, 25% glycerol, 2.5% SDS, 2.5% β-mercaptoethanol, and 0.002% bromophenol blue and then boiled for 3 min at 100°C. Equal amounts of total protein (50 μg) were loaded into each lane, and the samples were separated by electrophoresis (150 V) on a 7.5% or 12% polyacrylamide gel as determined by the size of the target protein (Mini-PROTEAN; BioRad, Hercules, CA). Each gel contained an internal control sample and molecular weight ladder (Precision Plus, BioRad).

After electrophoresis, protein was transferred to a polyvinylidene difluoride membrane (TransBlot Turbo, BioRad) using manufacturer-designed protocols. Blots were then blocked for 1 h in 5% nonfat dry milk and incubated in primary antibody overnight at 4°C (see Antibodies section). The next morning, blots were incubated in secondary antibody for 1 h at room temperature. Blots were then incubated in a chemiluminescent solution (Pierce ECL Western Blotting Substrate; Thermofisher, Waltham, MA) and optical density measurements were obtained with a phosphoimager (ChemiDoc MP, BioRad), and densitometric analysis was performed using Image Lab software (version 5.2.1, BioRad). Membranes containing phosphodetected proteins were stripped of primary and secondary antibodies (25 mM glycine, pH 2.0 and 1% SDS) and were subsequently reprobed for total protein with the specific antibody of interest. Phosphorylation and total density values were normalized to the internal control, and the phospho/total protein ratios were determined. Immunoblot data are expressed as phosphorylation divided by total protein or total protein relative to α-tubulin and adjusted to represent fold difference from Sed-Veh. All gels were loaded and analyzed by a blinded investigator.


Phosphospecific antibodies were mTORSer2448 (no. 2971), 4E-BP1Thr37/46 (no. 9459), and eEF2Thr56 (no. 2331) from Cell Signaling (Danvars, MA). Total antibodies included mTOR (no. 2972), 4E-BP1 (no. 9452), eEF2 (no. 2332), LC3B (no. 2775), p62, (no. 5114), and α-Tubulin (no. 2144) from Cell Signaling and REDD1 (no. 10638-1) from Proteintech Group, Inc. (Rosemont, IL). Antirabbit IgG horseradish peroxidase-conjugated secondary antibody was purchased from Santa Cruz Biotechnology (sc-2313).

Determination of fiber type–specific cross-sectional area

Transverse 7-μm-thick sections were cut from tissue previously frozen in isopentane using a cryostat (CM1950; Leica, Wetzlar, Germany). Sections were cut at −25°C and mounted on uncoated glass slides. Slides were allowed to air dry for 30 min at room temperature and then rehydrated in phosphate-buffered saline (PBS) for 3 min. After blocking for 1 h in 10% normal goat serum at room temperature, sections were incubated at room temperature for 90 min in antilaminin (2E8; IgG2a; supernatant; Developmental Studies Hybridoma Bank [DSHB], Iowa City, IA), anti-MHC I (BA.D5; IgG2b supernatant; DSHB), and anti-MHC IIa (SC.71; IgG1; supernatant; DSHB) antibodies. After a series of four 5-min washes in PBS, slides were incubated at room temperature for 60 min in goat antimouse IgG2a AF633 (no. A21136; Invitrogen, Grand Island, NY), goat antimouse IgG2b AF488 (no. A21141, Invitrogen), and goat antimouse IgG1 AF546 (no. A21123, Invitrogen). Finally, slides were washed in PBS and mounted with a fluorescent mounting media (Fluormount G, no. 0100-01; Southern Biotech, Birmingham, AL). Images were captured at room temperature with a Zeiss upright fluorescent microscope (AxioImager M2; Zeiss, Oberkochen, Germany). Individual muscle fiber cross-sectional area (CSA) and minimum Feret diameter (MFD) for MHC I and MHC IIa fibers were analyzed from 10× magnification images using a semiautomated threshold analysis (ImageJ version 1.50b; NIH, Bethesda, MD). Fiber type–specific fiber size analyses was performed using n = 4 for Sed-Dox, n = 4 for Sed-Veh, n = 3 for Ex-Dox, and n = 4 for Ex-Veh. Total MHC I fibers analyzed for each group were 799 for Sed-Dox, 534 for Sed-Veh, 543 for Ex-Dox, and 734 for Ex-Veh, whereas total MHC IIa fibers analyzed for each group were 38 for Sed-Dox, 33 for Sed-Veh, 61 for Ex-Dox, and 11 for Ex-Veh.

RNA extraction and semiquantitative real-time PCR

Frozen soleus tissue was weighed (mean ± SD = 24.9 ± 8.2 mg) at 4°C and homogenized with a handheld homogenizing dispenser (Bio-Gen PRO200; Pro Scientific, Oxford, CT) in 1 mL of Tri reagent (Molecular Research Center, Cincinnati, OH). The RNA was separated into an aqueous phase using 0.20 mL of chloroform and precipitated using 0.50 mL of isopropanol. The RNA pellet was then washed with 1.0 mL of 75% ethanol, dried, and dissolved in a predetermined amount (1.5 μL·mg−1 tissue) of nuclease free water. RNA concentration was determined using a Take3 plate (Biotek) and Biotek H1 Synergy H1 System. RNA (5 μg) was DNase-treated using a commercially available kit (DNA-free; Ambion, Austin, TX). A total of 1 μg of RNA was reverse transcribed into cDNA according to the directions provided by the manufacturer (iScript, BioRad). Real-time qPCR was conducted with a CFX Connect Real-Time PCR Detection System (BioRad). cDNA was analyzed with SYBR green fluorescence (iTaq Universal SYBR green supermix, BioRad). Primer sequences were designed using the National Center for Biotechnology Information database and carefully optimized (see Table, Supplemental Digital Content 1, Primer sequences used for PCR analyses, β2-Microglobulin was used as a normalization/housekeeping gene. Relative fold changes from Sed-Veh were determined from the Cq values using the 2ΔΔCt method (22).

Statistical analyses

All data were tested for normality through skewness and kurtosis analyses and visual inspection of the normality plots using SPSS version 24 (IBM). For nonnormally distributed data, a natural log (ln) transformation was performed before statistical analyses. Body mass, soleus mass, signaling, and mRNA expression data were analyzed using a 2 × 2 ANOVA to test for a drug effect, exercise effect, and interaction. If a significant interaction or effect was observed, Tukey’s post hoc testing was used to determine specific differences within the interaction or individual effect. To account for multiple measures obtained from the same animal, all MHC I and MHC IIa fiber CSA and MFD data from an individual animal were nested, and data were analyzed using a multilevel mixed model (nested ANOVA). This statistical model allows for the inclusion of the multiple measures (i.e., all fibers analyzed), which is the correct approach because it avoids treating each fiber as an independent observation. This model also takes into account variability for measures obtained both within and between animals. To determine differences within the multilevel mixed model, post hoc analyses with Bonferroni corrections for the following a priori comparisons were conducted: drug effect within sedentary, drug effect within exercise, and exercise effect within DOX. Data analyses were conducted using SigmaStat version 12.0 (Systat Software) and SPSS version 24 (IBM). Significance was set a priori at P ≤ 0.05. Data are presented as mean ± 95% confidence interval (CI).


Experimental group characteristics

All animals in the Sed-Veh and Ex-Veh groups were able to complete their respective protocols. In the Sed-Dox group, 6 of the 14 animals were unable to complete the Sed-Dox protocol, whereas 1 of the 10 animals in the Ex-Dox group was unable to complete the Ex-Dox protocol. Thus, data are presented as Sed-Veh, n = 10; Sed-Dox, n = 8; Ex-Veh, n = 8; and Ex-Dox, n = 9.

A drug effect was observed for body mass at time of euthanasia (P < 0.001). Body mass was lower in Sed-Dox (262 ± 31 g) compared with Sed-Veh (309 ± 17 g) (P < 0.001) and in Ex-Dox (267 ± 26 g) compared with Ex-Veh (302 ± 22 g) (P = 0.006). No interaction or effects were observed (P > 0.05) at time of euthanasia for absolute soleus mass (Sed-Veh, 0.127 ± 0.016 g; Sed-Dox, 0.118 ± 0.014 g; Ex-Veh, 0.130 ± 0.021 g; Ex-Dox, 0.120 ± 0.015 g) or soleus mass relative to body mass or tibia length (data not shown).

REDD1 and mTOR signaling

No interactions or effects were observed for total proteins or α-tubulin (P > 0.05). An interaction (P = 0.023) was observed for REDD1 mRNA. Post hoc analyses revealed REDD1 mRNA was higher with DOX administration in sedentary animals (P = 0.006 Sed-Dox vs Sed-Veh), whereas REDD1 mRNA was unaffected by DOX administration in exercising animals (P = 0.763 Ex-Dox vs Ex-Veh) (Fig. 2A). A drug effect was observed for REDD1 protein (P = 0.011) with post hoc analyses revealing higher REDD1 protein with DOX administration only in sedentary (P = 0.015 Sed-Dox vs Sed-Veh) but not in exercising animals (P = 0.235 Ex-Dox vs Ex-Veh) (Fig. 2B). An interaction (P = 0.003) and a drug effect (P = 0.011) were observed for mTOR phosphorylation. Post hoc analyses revealed mTOR phosphorylation was lower with DOX administration in sedentary animals (P < 0.001 Sed-Dox vs Sed-Veh), whereas mTOR phosphorylation was unaffected by DOX administration in exercising animals (P = 0.702 Ex-Dox vs Ex-Veh) (Fig. 3A). Further, mTOR phosphorylation was higher in Ex-Dox compared with Sed-Dox (P = 0.012) (Fig. 3A). A drug effect was also observed for 4E-BP1 phosphorylation (P = 0.031), a downstream target of mTORC1, in which post hoc analyses revealed that 4E-BP1 phosphorylation was lower with DOX administration in sedentary animals (P = 0.039 Sed-Dox vs Sed-Veh) but not exercising animals (P = 0.310 Ex-Dox vs Ex-Veh) (Fig. 3B). By contrast, no interaction or effects (P > 0.05) were observed for eEF2 phosphorylation (Fig. 3C).

The effect of chronic exercise and doxorubicin administration on regulated in development and DNA damage 1 (REDD1) mRNA and protein levels in the soleus muscle. mRNA values represent relative fold change from Sed-Veh as determined from the Cq values using the 2–ΔΔCt method (22). Data for REDD1 protein are relative to α-tubulin and normalized to Sed-Veh (fold difference). Data are presented as mean ± 95% CI. (A) mRNA levels of REDD1 (interaction effect, P = 0.006); (B) protein levels of REDD1 (drug effect, P = 0.011). Sed, sedentary; Ex, exercise; Veh, vehicle; Dox, doxorubicin; T, total protein, α-tub, α-tubulin. Sed-Veh, n = 10; Sed-Dox, n = 8; Ex-Veh, n = 8; Ex-Dox, n = 9. Post hoc analyses: *P < 0.05 Sed-Dox vs Sed-Veh. All blots were loaded and analyzed in a blinded fashion. Representative images have been rearranged to mimic order of presentation in the figure.
The effect of chronic exercise and doxorubicin administration on mTORC1 signaling in the soleus muscle. Data for phosphorylated proteins are relative to respective total protein. All data are normalized to Sed-Veh (fold difference) and are presented as mean ± 95% CI. (A) Phosphorylation of mTOR, mammalian/mechanistic target of rapamycin (interaction effect, P = 0.003); (B) phosphorylation of 4E-BP1, 4E binding protein 1 (drug effect, P = 0.031); and (C) phosphorylation of eEF2, eukaryotic elongation factor 2. Sed, sedentary; Ex, exercise; Veh, vehicle; Dox, doxorubicin; P, phosphorylated protein; T, total protein. Sed-Veh, n = 10; Sed-Dox, n = 8; Ex-Veh, n = 8; Ex-Dox, n = 9. Post hoc analyses: *P < 0.05 Sed-Dox vs Sed-Veh. ‡P < 0.05 Ex-Dox vs Sed-Dox. All blots were loaded and analyzed in a blinded fashion. Representative images have been rearranged to mimic order of presentation in the figure.

Markers of autophagy, ubiquitin proteasome system, and PGC-1α

A drug effect was observed for LC3BI (P = 0.006) with post hoc analyses revealing LC3BI protein levels were higher with DOX administration in sedentary animals (P = 0.007 Sed-Dox vs Sed-Veh) but were unaffected by DOX administration in exercising animals (P = 0.238 Ex-Dox vs Ex-Veh) (Fig. 4A). No interaction or effects were observed for LC3BII protein levels (P > 0.05) (Fig. 4B); however, there was a drug effect for the LC3BII/I ratio (P = 0.049), with post hoc analyses revealing the LC3BII/I ratio was lower with DOX administration only in sedentary animals (P = 0.017 Sed-Dox vs Sed-Veh) but not in exercising animals (P = 0.796 Ex-Dox vs Ex-Veh) (Fig. 4C). No interaction or effects were observed for p62 protein levels (P > 0.05) (Fig. 4D).

The effect of chronic exercise and doxorubicin administration on protein levels of autophagy markers in the soleus muscle. Data are relative to α-tubulin and normalized to Sed-Veh (fold difference) and are presented as mean ± 95% CI. (A) LC3BI, microtubule-associated protein 1 light chain 3 BI (drug effect, P = 0.006); (B) LC3BII; (C) the ratio of LC3BII/LC3BI; (drug effect, P = 0.049); and (D) p62, sequestosome-1. Sed, sedentary; Ex, exercise; Veh, vehicle; Dox, doxorubicin; α-tub, α-tubulin. Sed-Veh, n = 8; Sed-Dox, n = 6; Ex-Veh, n = 7; Ex-Dox, n = 9. Post hoc analyses: *P < 0.05 Sed-Dox vs Sed-Veh; All blots were loaded and analyzed in a blinded fashion. Representative images have been rearranged to mimic order of presentation in the figure.

The mRNA expression of several genes associated with autophagy was unaffected by DOX administration or exercise (Table 1). There was a drug effect for MuRF1 mRNA expression (P = 0.016). Post hoc analyses revealed that MuRF1 mRNA was unaffected by DOX administration in sedentary animals (P = 0.158 Sed-Dox vs Sed-Veh) but was reduced with DOX administration in exercising animals (P = 0.037 Ex-Dox vs Ex-Veh). Further, there was a trend for a drug effect for PGC-1α mRNA expression (P = 0.076); however, post hoc analyses did not reveal any differences within the sedentary or exercising groups (P > 0.05).

Soleus muscle mRNA analyses for markers of autophagy, ubiquitin proteasome system, and mitochondrial biogenesis.

Fiber type–specific CSA

To assess the effect of DOX treatment and exercise on muscle size, we examined soleus MHC I and MHC IIa fiber size, using measures of both CSA and MFD (see Figure, Supplemental Digital Content 2, Representative images for immunohistochemical analyses, The multilevel mixed model revealed MHC I fiber CSA (Fig. 5A) and MFD (Fig. 5B) was smaller with DOX administration in sedentary animals (P = 0.033 and P = 0.011, respectively, Sed-Dox vs Sed-Veh), whereas these measures of MHC I fiber size were unaffected by DOX administration in exercising animals (P = 1.00 Ex-Dox vs Ex-Veh). MHC IIa fiber CSA was smaller with DOX administration in both sedentary (P < 0.001 Sed-Dox vs Sed-Veh) and exercising animals (P = 0.001 Ex-Dox vs Ex-Veh) (Fig. 5A). Similarly, MHC IIa fiber MFD was also smaller with DOX administration in sedentary animals (P = 0.005 Sed-Dox vs Sed-Veh) and exercising animals (P = 0.009 Ex-Dox vs Ex-Veh) (Fig. 5B). However, MHC IIa fiber CSA was larger in Ex-Dox compared with Sed-Dox (P = 0.049 Ex-Dox vs Sed-Dox) (Fig. 5A), whereas there was a trend for larger MHC IIa fiber MFD in Ex-Dox compared with Sed-Dox (P = 0.069 Ex-Dox vs Sed-Dox) (Fig. 5B).



The findings from this study show that chronic DOX administration in sedentary animals is associated with increased REDD1 mRNA and protein, reduced mTORC1 signaling activity, and smaller MHC I and MHC IIa fibers in the soleus muscle. On the other hand, REDD1 levels and mTORC1 signaling were preserved in the soleus of animals that underwent exercise training before and during DOX administration, in conjunction with better preservation of MHC I and MHC IIa fiber size. These data indicate that 1) DOX may impair the regulation of muscle fiber size, in part, through altered mTORC1 signaling that may be related to elevated REDD1 and 2) interval exercise initiated before and maintained throughout chronic DOX administration could provide an important strategy to attenuate the negative effects of DOX on skeletal muscle.

REDD1 has been identified as an important negative regulator of skeletal muscle size. In particular, upregulation of REDD1 is observed in several pathological states associated with muscle wasting, including cancer cachexia (28), sepsis (32), and diabetes (38). Similarly, the overexpression of REDD1 in skeletal muscle also reduces fiber CSA (10). Recent studies have shown that acute DOX administration increases the expression of REDD1 mRNA in skeletal muscle (3,24). In agreement with these studies, we observed increased levels of REDD1 mRNA and protein in the soleus muscle of sedentary animals chronically administered DOX. Further, in conjunction with elevated REDD1 levels, Sed-Dox animals also showed reduced MHC I and MHC IIa fiber size. By contrast, exercise training before and during DOX administration blocked this DOX-induced increase in skeletal muscle REDD1 mRNA and protein and preserved skeletal muscle fiber size. Given that exercise training in the absence of DOX treatment (Ex-Veh) was not associated with a difference in REDD1 levels or fiber size in the soleus, the ability for exercise training to block the DOX-induced increase in REDD1 is not likely facilitated through a direct action of exercise on REDD1 expression. Instead, the ability for exercise training to blunt the increase in REDD1 is likely mediated through alterations in upstream regulation of REDD1 in response to DOX. Given the well-described role of REDD1 as a negative regulator of muscle fiber size, these data highlight REDD1 as a potential therapeutic target for interventions aimed to preserve muscle size during DOX treatment. However, further research is necessary to more precisely uncover the upstream link(s) between interval exercise, REDD1 and DOX.

The downstream mechanism through which REDD1 facilitates its negative regulation of muscle fiber size appears to occur, in part, through inhibition of mTORC1 signaling (16). Thus, to obtain insight into the potential downstream mechanisms, we examined the phosphorylation of mTOR and downstream targets in the signaling pathway. Consistent with the enhanced expression of REDD1 and reduced fiber size specific to Sed-Dox, mTOR and 4E-BP1 phosphorylation in skeletal muscle were also reduced by DOX in sedentary animals. On the other hand, eEF2 phosphorylation was not altered by DOX, indicating that the effect of DOX may be more directed to translation initiation rather than elongation. To the best of our knowledge, mTORC1 activity has not been examined extensively in skeletal muscle after DOX administration; however, previous work has shown mTOR phosphorylation to be reduced (39) or unchanged (33) in cardiomyocytes after DOX administration. Discrepancies between studies may be the result of different DOX doses as this study and that of Zhu et al. (39) used higher accumulative doses (20 and 12 mg·kg−1, respectively), compared with that of Sturgeon et al. (4 mg·kg−1) (33). Nonetheless, our observation of reduced mTORC1 activity in the skeletal muscle of Sed-Dox is aligned with recent observations, indicating that DOX may promote atrophy of skeletal muscle through reducing protein synthesis rather than enhancing breakdown (24).

Consistent with the preservation of REDD1 mRNA and protein levels and skeletal muscle fiber size, exercise training before and during DOX administration preserved mTORC1 signaling. The exact mechanism through which exercise preserved mTORC1 signaling is not readily available; however, it could be related to the lack of increase in REDD1 protein or to the ability for exercise to repeatedly stimulate mTORC1 signaling (37). For instance, a mouse model of cardiac-restricted, constitutively active mTOR was shown to protect cardiomyocytes from DOX-induced cardiac dysfunction (39). Further, we sacrificed the animals 24 h after the last exercise bout. Given that exercise alone (Ex-Veh) was not associated with elevated mTORC1 activity, the preservation of mTORC1 signaling in Ex-Dox was not likely the result of enhanced mTORC1 activity from the previous exercise bout but rather a preservation of basal activation. Future research is necessary to determine whether exercise training elicits its protective effects on skeletal muscle fiber size through the preservation of basal mTORC1 activity or the preservation of the mTORC1 signaling response to a stimulus (e.g., nutrients and exercise).

In contrast to a previous study that sacrificed animals 24 h after a large bolus injection of DOX (29), we did not observe an increase in genes associated with autophagy. However, we did observe an increase in LC3BI protein specific to Sed-Dox, which is consistent with previous reports (5) and the increase in skeletal muscle LC3 mRNA expression observed by Smuder et al. (29). Further, the ratio of LC3BII/I was reduced only in Sed-Dox. A change in the LC3BII/LC3BI ratio alone can be difficult to interpret (21); however, in conjunction with a lack of change in p62, which is degraded during autophagy flux (25), it appears that chronic DOX administration did not increase basal levels of autophagy flux. Instead, the increase in LC3BI coupled with a lower conversion of available LC3BI to LC3BII in Sed-Dox may indicate an attenuated progression through autophagy. Collectively, it appears that DOX administration may reduce autophagy flux, perhaps leading to the retention of damaged proteins/organelles that may be related to reduced muscle function in response to DOX (4,14,17).

Recent reports indicate that knockout of PGC-1α is associated with reduced mitochondrial turnover, both through reductions in mitochondrial biogenesis and as well as slowed autophagy/mitophagy (35). Although PGC-1α expression was not statistically reduced in our study, there was a tendency for DOX administration to downregulate PGC-1α. Importantly, the slower autophagy flux observed in the previous article was associated with reduced lipidation of LC3B protein (35) (conversion of LC3BI to LC3BII), which is consistent with the reduced ratio of LC3BII/LC3BI observed in Sed-Dox. Interestingly, the ratio of LC3BII/I was not reduced by DOX in exercising animals, indicating that exercise training may have preserved the rate of LC3 lipidation, at least at the time point studied, which may have preserved the breakdown of any damaged mitochondria, at least through autophagy. Importantly, slower mitochondrial turnover (reduced autophagy/mitophagy and/or biogenesis) could promote retention of compromised mitochondria, which may stimulate the formation of ROS (34) that is known to occur in the skeletal muscle of DOX-treated animals (23). Nonetheless, the interaction among DOX, autophagy/mitophagy, and PGC-1α requires further study.

The protective effects of exercise before DOX administration on an array of DOX-induced cellular alterations in skeletal muscle have been exquisitely demonstrated in prior studies using large, single bolus injections of DOX (4,20,29,30). It is also important to note that additional mechanisms through which DOX negatively affects skeletal muscle have been demonstrated using bolus injections (13). It is important to consider that DOX administration in human clinical populations is often provided over multiple cycles (36), and therefore, our approach provided a means to determine whether exercise can serve as a strategy to protect skeletal muscle from multiple DOX treatments. Importantly, the accumulated dose of DOX provided in this study is similarly scaled to human doses used clinically (11,36). In addition, these prior studies used male rats (4,20,29,30). By contrast, we used a model of female, estrogen-deficient animals to simulate the hormonal physiological state present in many breast cancer patients, in which the median age for diagnosis (61 yr) occurs after menopause (NCI SEER statistics). Further, we examined the soleus muscle based on previous data demonstrating a relatively high degree of impairment in soleus muscle function (vs extensor digitorum longus) (18). Thus, to what extent our findings extrapolate to other skeletal muscles with different characteristics requires further investigation. Lastly, our study was not performed in a cancer model, and therefore to what extent our exercise approach is able to protect skeletal muscle during chronic DOX administration in the presence of cancer remains to be determined. However, it is important to note that previous research has demonstrated that exercise does not impair the antitumor effects of DOX (26).

In conclusion, our data support a model in which chronic DOX administration promotes reduced skeletal muscle fiber size through the upregulation of REDD1 and subsequent attenuation of mTORC1 signaling. Further, smaller muscle fiber size with chronic DOX administration did not appear to be the result of enhanced mRNA expression of markers in the ubiquitin proteasome system or the stimulation of autophagy, supporting a hypothesis in which chronic DOX administration promotes reduced muscle fiber size through downregulation of muscle growth processes. By contrast, exercise initiated before and continued throughout chronic DOX administration protected skeletal muscle fiber size and prevented the increase in REDD1 and associated reduction in mTORC1 signaling. More research involving human clinical trials is certainly warranted, and particular attention needs to be given to the timing of exercise relative to DOX administration as well as the role of exercise mode and exercise intensity.

The authors thank the animal care staff at Midwestern University and Matthew P. Buman, Ph.D., for his statistical guidance/support. J. M. D., F. M., R. J. G., T. M. H., C. C. C., and S. S. A. designed the research; all authors conducted the research; J. M. D. and A. C. D. analyzed data and performed statistical analyses; J. M. D. had primary responsibility for final product. All authors read and approved the final manuscript.

This research was supported by intramural Funds from Arizona State University and Midwestern University.

The authors have no conflicts of interests to disclose. The results of the present study do not constitute endorsement by the American College of Sports Medicine. The results of this study are presented clearly, honestly, and without fabrications, falsification, or inappropriate data manipulation.

Chad C. Carroll and Siddhartha S. Angadi denote equal contributions.


1. Barcenas CH, Niu J, Zhang N, et al. Risk of hospitalization according to chemotherapy regimen in early-stage breast cancer. J Clin Oncol. 2014;32(19):2010–7.
2. Bonifati DM, Ori C, Rossi CR, Caira S, Fanin M, Angelini C. Neuromuscular damage after hyperthermic isolated limb perfusion in patients with melanoma or sarcoma treated with chemotherapeutic agents. Cancer Chemother Pharmacol. 2000;46(6):517–22.
3. Braun TP, Szumowski M, Levasseur PR, et al. Muscle atrophy in response to cytotoxic chemotherapy is dependent on intact glucocorticoid signaling in skeletal muscle. PLoS One. 2014;9(9):e106489.
4. Bredahl EC, Pfannenstiel KB, Quinn CJ, Hayward R, Hydock DS. Effects of exercise on doxorubicin-induced skeletal muscle dysfunction. Med Sci Sports Exerc. 2016;48(8):1468–73.
5. Campbell TL, Quadrilatero J. Data on skeletal muscle apoptosis, autophagy, and morphology in mice treated with doxorubicin. Data Brief. 2016;7:786–93.
6. de Moor JS, Mariotto AB, Parry C, et al. Cancer survivors in the United States: prevalence across the survivorship trajectory and implications for care. Cancer Epidemiol Biomarkers Prev. 2013;22(4):561–70.
7. Der-Torossian H, Couch ME, Dittus K, Toth MJ. Skeletal muscle adaptations to cancer and its treatment: their fundamental basis and contribution to functional disability. Crit Rev Eukaryot Gene Expr. 2013;23(4):283–97.
8. Drummond MJ, Fry CS, Glynn EL, et al. Rapamycin administration in humans blocks the contraction-induced increase in skeletal muscle protein synthesis. J Physiol. 2009;587(Pt 7):1535–46.
9. Dunn WA Jr. Autophagy and related mechanisms of lysosome-mediated protein degradation. Trends Cell Biol. 1994;4(4):139–43.
10. Favier FB, Costes F, Defour A, et al. Downregulation of Akt/mammalian target of rapamycin pathway in skeletal muscle is associated with increased REDD1 expression in response to chronic hypoxia. Am J Physiol Regul Integr Comp Physiol. 2010;298(6): R1659–66.
11. Freireich EJ, Gehan EA, Rall DP, Schmidt LH, Skipper HE. Quantitative comparison of toxicity of anticancer agents in mouse, rat, hamster, dog, monkey, and man. Cancer Chemother Rep. 1966;50(4):219–44.
12. Gallagher LE, Williamson LE, Chan EY. Advances in autophagy regulatory mechanisms. Cells. 2016;5(2):24.
13. Gilliam LA, Ferreira LF, Bruton JD, et al. Doxorubicin acts through tumor necrosis factor receptor subtype 1 to cause dysfunction of murine skeletal muscle. J Appl Physiol (1985). 2009;107(6): 1935–42.
14. Gilliam LA, Lark DS, Reese LR, et al. Targeted overexpression of mitochondrial catalase protects against cancer chemotherapy-induced skeletal muscle dysfunction. Am J Physiol Endocrinol Metab. 2016;311(2):E293–301.
15. Gilliam LA, St Clair DK. Chemotherapy-induced weakness and fatigue in skeletal muscle: the role of oxidative stress. Antioxid Redox Signal. 2011;15(9):2543–63.
16. Gordon BS, Steiner JL, Williamson DL, Lang CH, Kimball SR. Emerging role for regulated in development and DNA damage 1 (REDD1) in the regulation of skeletal muscle metabolism. Am J Physiol Endocrinol Metab. 2016;311(1):E157–74.
17. Hayward R, Hydock D, Gibson N, Greufe S, Bredahl E, Parry T. Tissue retention of doxorubicin and its effects on cardiac, smooth, and skeletal muscle function. J Physiol Biochem. 2013;69(2):177–87.
18. Hydock DS, Lien CY, Jensen BT, Schneider CM, Hayward R. Characterization of the effect of in vivo doxorubicin treatment on skeletal muscle function in the rat. Anticancer Res. 2011;31(6): 2023–8.
19. Jin HO, Seo SK, Woo SH, et al. Activating transcription factor 4 and CCAAT/enhancer-binding protein-beta negatively regulate the mammalian target of rapamycin via Redd1 expression in response to oxidative and endoplasmic reticulum stress. Free Radic Biol Med. 2009;46(8):1158–67.
20. Kavazis AN, Smuder AJ, Powers SK. Effects of short-term endurance exercise training on acute doxorubicin-induced FoxO transcription in cardiac and skeletal muscle. J Appl Physiol (1985). 2014;117(3):223–30.
21. Klionsky DJ, Abeliovich H, Agostinis P, et al. Guidelines for the use and interpretation of assays for monitoring autophagy in higher eukaryotes. Autophagy. 2008;4(2):151–75.
22. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001;25(4):402–8.
23. Minotti G, Menna P, Salvatorelli E, Cairo G, Gianni L. Anthracyclines: molecular advances and pharmacologic developments in antitumor activity and cardiotoxicity. Pharmacol Rev. 2004;56(2):185–229.
24. Nissinen TA, Degerman J, Rasanen M, et al. Systemic blockade of ACVR2B ligands prevents chemotherapy-induced muscle wasting by restoring muscle protein synthesis without affecting oxidative capacity or atrogenes. Sci Rep. 2016;6:32695.
25. Pankiv S, Clausen TH, Lamark T, et al. p62/SQSTM1 binds directly to Atg8/LC3 to facilitate degradation of ubiquitinated protein aggregates by autophagy. J Biol Chem. 2007;282(33):24131–45.
26. Parry TL, Hayward R. Exercise training does not affect anthracycline antitumor efficacy while attenuating cardiac dysfunction. Am J Physiol Regul Integr Comp Physiol. 2015;309(6):R675–83.
27. Peel AB, Thomas SM, Dittus K, Jones LW, Lakoski SG. Cardiorespiratory fitness in breast cancer patients: a call for normative values. J Am Heart Assoc. 2014;3(1):e000432.
28. Puppa MJ, Gao S, Narsale AA, Carson JA. Skeletal muscle glycoprotein 130’s role in Lewis lung carcinoma-induced cachexia. FASEB J. 2014;28(2):998–1009.
29. Smuder AJ, Kavazis AN, Min K, Powers SK. Exercise protects against doxorubicin-induced markers of autophagy signaling in skeletal muscle. J Appl Physiol (1985). 2011;111(4):1190–8.
30. Smuder AJ, Kavazis AN, Min K, Powers SK. Exercise protects against doxorubicin-induced oxidative stress and proteolysis in skeletal muscle. J Appl Physiol (1985). 2011;110(4):935–42.
31. Smuder AJ, Kavazis AN, Min K, Powers SK. Doxorubicin-induced markers of myocardial autophagic signaling in sedentary and exercise trained animals. J Appl Physiol (1985). 2013;115(2):176–85.
32. Steiner JL, Crowell KT, Kimball SR, Lang CH. Disruption of REDD1 gene ameliorates sepsis-induced decrease in mTORC1 signaling but has divergent effects on proteolytic signaling in skeletal muscle. Am J Physiol Endocrinol Metab. 2015;309(12):E981–94.
33. Sturgeon K, Schadler K, Muthukumaran G, et al. Concomitant low-dose doxorubicin treatment and exercise. Am J Physiol Regul Integr Comp Physiol. 2014;307(6):R685–92.
34. Terman A, Kurz T, Navratil M, Arriaga EA, Brunk UT. Mitochondrial turnover and aging of long-lived postmitotic cells: the mitochondrial-lysosomal axis theory of aging. Antioxid Redox Signal. 2010;12(4):503–35.
35. Vainshtein A, Desjardins EM, Armani A, Sandri M, Hood DA. PGC-1α modulates denervation-induced mitophagy in skeletal muscle. Skelet Muscle. 2015;5:9.
36. van Dalen EC, van der Pal HJ, Kremer LC. Different dosage schedules for reducing cardiotoxicity in people with cancer receiving anthracycline chemotherapy. Cochrane Database Syst Rev. 2016;3:CD005008.
37. Walker DK, Dickinson JM, Timmerman KL, et al. Exercise, amino acids, and aging in the control of human muscle protein synthesis. Med Sci Sports Exerc. 2011;43(12):2249–58.
38. Williamson DL, Dungan CM, Mahmoud AM, Mey JT, Blackburn BK, Haus JM. Aberrant REDD1-mTORC1 responses to insulin in skeletal muscle from type 2 diabetics. Am J Physiol Regul Integr Comp Physiol. 2015;309(8):R855–63.
39. Zhu W, Soonpaa MH, Chen H, et al. Acute doxorubicin cardiotoxicity is associated with p53-induced inhibition of the mammalian target of rapamycin pathway. Circulation. 2009;119(1):99–106.


Supplemental Digital Content

© 2017 American College of Sports Medicine