Castration alters protein balance after high-frequency muscle contraction

In the absence of androgens, markers of autophagy were elevated, and these could not be normalized by muscle contractions. In the fasted state, REDD1 was identified as a potential contributor to autophagy in noncontracted muscle, whereas phosphorylation of ULK1 may contribute to this process in the contracted muscle. In the refed state, markers of autophagy remain elevated in both noncontracted and contracted muscles, but the relationship with REDD1 and ULK1 (Ser757) no longer existed.

Abstract

Resistance exercise increases muscle mass by shifting protein balance in favor of protein accretion. Androgens independently alter protein balance, but it is unknown whether androgens alter this measure after resistance exercise. To answer this, male mice were subjected to sham or castration surgery 7–8 wk before undergoing a bout of unilateral, high-frequency, electrically induced muscle contractions in the fasted or refed state. Puromycin was injected 30 min before euthanasia to measure protein synthesis. The tibialis anterior was analyzed 4 h postcontraction. In fasted mice, neither basal nor stimulated rates of protein synthesis were affected by castration despite lower phosphorylation of mechanistic target of rapamycin in complex 1 (mTORC1) substrates [p70S6K1 (Thr389) and 4E-BP1 (Ser65)]. Markers of autophagy (LC3 II/I ratio and p62 protein content) were elevated by castration, and these measures remained elevated above sham values after contractions. Furthermore, in fasted mice, the protein content of Regulated in Development and DNA Damage 1 (REDD1) was correlated with LC3 II/I in noncontracted muscle, whereas phosphorylation of uncoordinated like kinase 1 (ULK1) (Ser757) was correlated with LC3 II/I in the contracted muscle. When mice were refed before contractions, protein synthesis and mTORC1 signaling were not affected by castration in either the noncontracted or contracted muscle. Conversely, markers of autophagy remained elevated in the muscles of refed, castrated mice even after contractions. These data suggest the castration-mediated elevation in baseline autophagy reduces the absolute positive shift in protein balance after muscle contractions in the refed or fasted states.

NEW & NOTEWORTHY In the absence of androgens, markers of autophagy were elevated, and these could not be normalized by muscle contractions. In the fasted state, REDD1 was identified as a potential contributor to autophagy in noncontracted muscle, whereas phosphorylation of ULK1 may contribute to this process in the contracted muscle. In the refed state, markers of autophagy remain elevated in both noncontracted and contracted muscles, but the relationship with REDD1 and ULK1 (Ser757) no longer existed.

the increase in muscle mass after resistance exercise/muscle overload is due in part to a shift in protein balance in favor of net protein accretion. This is the result of an increase in rates of synthesis and decrease in autophagy, which together outweigh the increase in ubiquitin proteasome activity (3, 11, 14). Although these events are well characterized, the shift in protein balance after resistance exercise is reduced in certain populations, and accordingly, muscle mass is not increased to the same extent. For example, rates of protein synthesis after resistance exercise were lower in aged individuals compared with young, healthy subjects (13), likely contributing to the lower absolute degree of hypertrophy observed in the aged population (31). Thus it is important to identify the factors that regulate the shift in protein balance toward net accretion in response to resistance exercise/muscle overload.

Although it is generally accepted that androgens significantly alter protein balance and thereby muscle mass, corresponding data suggest that long-term changes in these hormones may modulate the absolute magnitude of muscle hypertrophy achieved by resistance exercise/muscle overload. For instance, nandralone decanoate administration overcame the aging-induced anabolic resistance to overload and increased soleus muscle mass, whereas no change was seen in untreated rats (22). Similarly, administration of a gonadotropin-releasing hormone analog to deplete endogenous androgen production in humans reduced the overall change in lean leg mass and strength after a resistance exercise training program (21), further supporting an interaction between androgens and protein balance in response to resistance exercise. Contrasting data also exist, including work suggesting that androgens raise the baseline from which hypertrophy begins instead of affecting the magnitude of the change (5).

The resistance exercise-induced shift in protein balance is regulated in part by signaling through the mechanistic target of rapamycin in complex 1 (mTORC1) (11, 32). Signaling through mTORC1 regulates protein synthesis via phosphorylation of at least two known downstream substrates termed the 70-kDa ribosomal protein S6 kinase 1 (p70S6K1) and the eIF4E binding protein 1 (4E-BP1) (7–9). It also regulates autophagy initiation via its inhibitory phosphorylation of proteins such as uncoordinated like kinase 1 (ULK1) on Ser757 (20). Inhibiting mTORC1 suppresses the anabolic response to resistance exercise/muscle overload. For example, expression of the mTORC1 repressor, Regulated in Development and DNA Damage 1 (REDD1), limited the peak phosphorylation of p70S6K1 (Thr389) and 4E-BP1 (Ser65) as well as protein synthesis after a single bout of muscle contractions by altering the baseline from which the increase occurred (17). REDD1 expression also limited the muscle overload-induced suppression of autophagy by altering the baseline from which this response occurred (16). Furthermore, other mTORC1 repressors such as the 5′-AMP-activated protein kinase (AMPK) also blunt the peak phosphorylation of mTORC1 substrates after muscle contractions (28). Because androgens are known to modify at least mTORC1 signaling and REDD1 expression (33, 34), it is plausible that the shift in protein balance after resistance exercise would be lower in the absence of these hormones through a change in baseline levels. Thus the current objective was to determine whether androgens alter the different components of protein balance in response to a single bout of simulated resistance exercise.

METHODS

Animals.

Male, C57BL/6NHsd mice (aged 9 wk) were obtained from Envigo (Indianapolis, IN). Upon arrival, all mice were housed individually in ventilated cages during a 7-day acclimation in a temperature (25°C)- and light (12:12-h light/dark)-controlled environment on corn cob bedding within the barrier containment vivarium at the Burnett School of Biomedical Sciences. Mice were provided irradiated PicoLab 5053 rodent chow (LabDiet, St. Louis, MO) and water ad libitum. The Institutional Animal Care and Use Committee of The University of Central Florida approved the animal facilities and all experimental protocols.

Experimental design and castration surgery.

After a 7-day acclimation, mice were randomized into two groups of equal body weight. One group underwent castration surgery to decrease the concentration of circulating androgens, whereas the other group was subjected to a sham surgery. Mice were deeply anesthetized with isoflurane anesthesia (~3%), and by using sterile surgical technique, the lower abdomen was shaved, cleaned, and disinfected with iodine solution. A small longitudinal incision (~1 cm) was made through the abdominal wall. For those mice randomized to the castration surgery, the testes were located and removed, leaving the epididymal fat pad and seminal vesicles in place. An identical procedure was performed on the sham-castrated group except that the testes were left intact. The abdominal wall was closed using 2.0 absorbable suture, and the skin was closed using wound clips. A subcutaneous injection of buprenorphine (0.05 mg/kg in 500 µl of sterile saline) was administered immediately after closure of the wound, and the mice were allowed to recover in their cages. A second subcutaneous injection was administered 5 h later. Mice were allowed to recover for 7–8 wk, which is sufficient to observe a castration-mediated reduction in muscle mass (33). No adverse side effects were observed in either group throughout the recovery. The night before the muscle contractions protocol, food was removed from all mice beginning at 1600. The next morning (~0800–1000), mice were randomized into one of two groups that either remained fasted or were given access to food pellets for 30 min immediately before the muscle contractions.

Muscle contractions protocol.

A single bout of resistance exercise was simulated via unilateral, high-frequency, electrically induced, muscle contractions as previously described (17). In brief, mice were deeply anesthetized with 3% isoflurane. The left thigh was shaved, and a small incision (~1 cm) was made to expose the sciatic nerve. Two bipolar electrodes were used to stimulate the nerve, causing all muscles of the lower limb to contract. The nerve was stimulated with a 1-mAmp constant current pulse at a frequency of 100 Hz using a constant current stimulator (Aurora Scientific, Ontario, Canada). Each stimulus consisted of 300 pulses, each 1 ms in duration. The entire protocol lasted ~22 min and consisted of 10 sets of six stimuli. Each stimulus within a set was separated by a 10-s rest period, and between each of the 10 sets was a 60-s rest period. After contractions, the skin was closed using wound clips, and all mice received a subcutaneous injection of warm saline (500 µl) before returning to their cages, where they had free access to water but not food. Mice were allowed to recover for 4 h before being anesthetized with isoflurane for removal of the tibialis anterior muscles, which were snap frozen in liquid nitrogen, and stored at −80°C until further analysis. This time point was chosen based on our previous work and that of others showing anabolic signaling is maximal at this time point and corresponds to the contraction-induced increase in the rates of muscle protein synthesis (17, 25).

Protein synthesis.

Protein synthesis was assessed using the SUnSET method with antibodies against puromycin (Kerafast, Boston, MA) as previously described (27). Thirty minutes before death, mice were administered an intraperitoneal injection of 0.04 µmol/g body weight of puromycin (AG Scientific; San Diego, CA) dissolved in phosphate-buffered saline. Western blotting procedures were performed to visualize puromycin incorporation into protein of the tibialis anterior. Although we are aware that our anti-mouse IgG secondary antibody (Bethyl Laboratories, Montgomery, TX; cat. #A90–116P) reacts nonspecifically with bands corresponding to ~25 kDa, and to a much lesser extent, bands corresponding to ~50 kDa, control experiments revealed that these interactions were not altered by castration, and thus all immunopositive bands were used for analysis.

Western blot analysis.

Western blotting was conducted as previously described (15, 18). Briefly, whole muscle protein was extracted by glass on glass homogenization in 10 volumes of buffer (10 µl/mg) consisting of 50 mM HEPES (pH 7.4), 0.1% Triton-X 100, 4 mM EGTA, 10 mM EDTA, 15 mM Na₄P₂O₇, 100 mM β-glycerophosphate, 25 mM NaF, 5 mM Na₃VO₄, and 10 µl/ml protease inhibitor cocktail (Sigma Aldrich #P8340). The muscle extract was centrifuged at 10,000 g for 10 min, and the protein content of the supernatant fraction was quantified by the Bradford method. For analysis, proteins in the supernatant were fractionated on 4–20% Bio-Rad (Hercules, CA) Criterion precast gels and transferred to PVDF membranes. Membranes were stained with Ponceau-S to confirm effective transfer and equal protein loading. Membranes were incubated with appropriate antibodies overnight at 4°C. Antibodies against phospho-p70S6K1 (Thr389) (cat. #9205), phospho-4E-BP1 (Ser65) (cat. #9451), LC3B (cat. #2775), p62 (cat. #5114), phospho-ULK1 (Ser757) (cat. #6888), phospho-ULK1 (Ser317) (cat. #12753), total ULK1 (cat. #8054), phospho-Akt (Thr308) (cat. #9275), total Akt (cat. #9272), phospho-FoxO3a (Ser253) (cat. #9466), total FoxO3a (cat. #2497), AMPKα (Thr172) (cat. #2531), total AMPKα (cat. #2532), Ubiquitin (cat. #3933), and BNIP3 (cat. #3769) were obtained from Cell Signaling Technology (Danvers, MA). Antibodies against total p70S6K1 and total 4E-BP1 were custom made by Bethyl Laboratories (Montgomery, TX) and kindly provided by Dr. Scot R. Kimball (The Pennsylvania State College of Medicine). Antibodies against REDD1 (cat. #10638–1-AP) were obtained from ProteinTech (Chicago, IL). Monoclonal antibodies against puromycin were obtained through Kerafast (3RH11) (Boston, MA). After incubation with appropriate secondary antibodies (Bethyl Laboratories) (cat. #A120–101P, A90–116P), the antigen-antibody complex was visualized by enhanced chemiluminescence using Clarity reagent (Bio-Rad) and a ChemiDoc Touch imaging system (Bio-Rad). The pixel density from all blots were quantified using Image J software (NIH, Bethesda, MD) or ImageLab software (Bio-Rad).

Plasma testosterone.

Whole blood was extracted in EDTA-treated syringes, and the plasma fraction was isolated after centrifugation at 2,000 g for 10 min at 4°C. Total testosterone was determined from the plasma fraction by ELISA according to manufacturer’s instructions (CALBIOTECH Inc., Spring Valley, CA; cat. #TE187S-100). The data are expressed as nanograms per milliliter.

RNA Extraction, cDNA synthesis, and quantitative RT-PCR.

Muscle samples were homogenized in 600 μl of Zymo Tri Reagent (Irvine, CA), and RNA was isolated using a Zymo RNA Miniprep kit with on column DNase treatment (Irvine, CA). RNA quantity and purity was determined spectrophotometrically by 260- to 280-nm ratio, and cDNA was synthesized using a Superscript VILO cDNA synthesis kit (Invitrogen) from 2 μg RNA. Quantitative (q) RT-PCR was performed using PowerUp Sybr Green Master Mix (Thermo Fisher Scientific, Waltham, MA). Each sample was analyzed in triplicate. The conditions for the qRT-PCR included an initial 2 min at 50°C and 2 min at 95°C, followed by 40 cycles that included a 15-s denature step at 95°C, a 15-s annealing step at 55°C, and a 1-min extension step at 72°C within each cycle. A melting curve analysis was performed for each primer pair to ensure that a single product was efficiently amplified. Relative expression levels of Muscle RING-Finger Protein-1 (MuRF-1) and Muscle Atrophy F-Box (MAFbx) were normalized by using the delta delta Ct method with GAPDH as the control gene. Primer sequences for MuRF-1 were as follows: forward, 5′-AAG CAG GAG TGC TCC AGT CG-3′; reverse, 5′-ACC AGC ATG GAG ATG CAG TTA C-3′. Primer sequences for MAFbx were as follows: forward, 5′-GTC GCA GCC AAG AAG AGA AAG-3′; reverse, 5′-ACT CAG GGA TGT GAG CTG TGA-3′. Primer sequences for GAPDH were as follows: forward, 5′-GTT GTC TCC TGC GAC TTC A-3′; reverse, 5′-TGC TGT AGC CGT ATT CAT TG-3′.

Statistical analysis.

All data are presented as means ± SE. Two-way ANOVA with repeated measures [treatment (castration vs. sham) × time (initial vs. final)] was used to evaluate body weight. Separate two-way ANOVA with repeated measures [treatment (castration vs. sham) × contraction (contracted vs. noncontracted)] were used to evaluate all other dependent variables in the fasted and refed conditions. Where appropriate, Student’s t-test was used to compare the means of two variables. Analysis of select relationships was performed using Pearson product moment correlation. All analysis was performed using GraphPad Prism Software (La Jolla, CA). Significance was set at P < 0.05 for all analyses.

RESULTS

In the fasted state, castration did not affect protein synthesis despite suppressed mTORC1 signaling.

Consistent with others (2, 19), final body weight was lower in castrated mice relative to sham-castrated mice, and this coincided with a lower tibialis anterior muscle mass but no difference in tibia length (Table 1). Plasma testosterone concentrations were also reduced by castration as was the mass of the androgen-sensitive seminal vesicle (Table 1).

Table 1.

Descriptive characteristics of sham-castrated and castrated mice

	Initial Body Wt, g	Final Body Wt, g	TA Mass, g	Tibia Length, mm	Seminal Vesicle Mass, mg	Plasma Testosterone, ng/ml
Sham	23.4 ± 0.3	27.8 ± 0.5*	46.0 ± 1.0	17.4 ± 0.1	286 ± 11.0	0.77 ± 0.03
Castrated	23.1 ± 0.3	25.7 ± 0.5*#	39.1 ± 1.2#	17.5 ± 0.1	19.3 ± 1.2#	0.23 ± 0.02#

Data are means ± SE. N = 6–18/group from 3 to 4 independent experiments. Significance set at P < 0.05. BW = body weight; TA = tibialis anterior.

^*Significantly different compared with initial BW;

^#significantly different than sham within the measure.

In the fasted state, there was a main effect of contractions to increase rates of protein synthesis as assessed by puromycin incorporation (P < 0.05), but there was no effect of castration (Fig. 1, A and D). Contractions increased the phosphorylation of p70S6K1 (Thr389) and 4E-BP1 (Ser65) (Fig. 1, B–D), whereas castration reduced these measures. Thus, despite the castration-mediated reduction in mTORC1 signaling, no difference in either basal or contraction-induced rates of muscle protein synthesis was observed after castration.

Fig. 1.

Rates of muscle protein synthesis and mTORC1 signaling in the fasted metabolic state. A: muscle protein synthesis was determined by the SUnSET method. Phosphorylated to total protein ratio of p70S6K1 (Thr389) (B) and 4E-BP1 (Ser65) (C) was determined by Western blot analysis. D: representative Western blots. Significance set at P < 0.05. ME, main effect; P, phosphorylated protein; T, total protein. N = 7 or 8/group from 3 independent experiments.

In the fasted state, castration increased basal markers of autophagy, which remained elevated after muscle contractions.

In agreement with recent work (14, 16), contractions reduced the LC3 II/I ratio in both sham-castrated and castrated mice (Fig. 2, A and F). However, castration increased this measure in both the noncontracted and contracted muscle (Fig. 2, A and F). Measurement of p62 protein content is a complimentary marker of autophagy activation (16), and accordingly, there was a main effect of castration to reduce the protein content of this marker despite no change with contractions (Fig. 2, B and F). These results suggest that in the fasted state, castration alone increased markers of autophagy, and muscle contractions were unable to remediate this effect.

Fig. 2.

Markers of protein degradation in the fasted metabolic state. LC3 II/I ratio (A), p62 protein content (B), and the content of ubiquitylated proteins (C) were determined by Western blot analysis. Relative mRNA abundance of MAFbx (D) and MuRF-1 (E) were determined by qRT-PCR using GAPDH as an internal control. F: representative Western blots. Significance set at P < 0.05. ME, main effect. N = 6–8/group from 3 independent experiments. *Significance between noncontracted and contracted muscle within a treatment group. #Significance between sham-castrated and castrated groups for specified condition.

We then measured the content of ubiquitylated proteins as a marker of changes in the ubiquitin proteasome system (UPS). In sham-castrated mice, contractions decreased the content of protein ubiquitylation by 31% (P < 0.05) (Fig. 2, C and F). Castration alone significantly decreased ubiquitylation in the noncontracted muscle, and although muscle contractions resulted in a small increase in this measure (Fig. 2, C and F), it was not different from the contracted muscles of sham-contracted mice. Independent of these changes in protein ubiquitylation, there was a main effect of contractions to decrease the relative mRNA abundance of MAFbx, whereas MuRF-1 mRNA abundance was not significantly changed (Fig. 2, D and E). In all, although the content of protein ubiquitylation was differentially altered by castration and muscle contractions, the value in the contracted muscle was not different between sham-castrated and castrated mice.

In the fasted state, markers of autophagy were associated with changes in REDD1 protein content and phosphorylation of ULK1.

After the observation that castration enhanced autophagy and that it remained elevated despite muscle contractions, we analyzed upstream regulatory factors to determine what contributed to this effect (20, 23, 26, 30, 35). There was no effect of castration on the phosphorylation of ULK1 (Ser317), Akt (Thr308), FoxO3a (Ser253), or the protein content of BNIP3 (Fig. 3, A, C, D, F, and H). Conversely, castration increased the phosphorylation of AMPKα (Thr172) and REDD1 protein content, whereas phosphorylation of ULK1 (Ser757) was decreased (Fig. 3, B, E, G, and H). Further analysis showed a significant positive relationship between REDD1 protein content and the LC3 II/I ratio only in the noncontracted muscle (r = 0.69; P = 0.0029), whereas a significant negative relationship between ULK1 (Ser757) and the LC3 II/I ratio was detected in only the contracted muscle (r = −0.58; P = 0.0012) of both sham-castrated and castrated mice. Despite reaching significance, the R² values for these relationships were only R² = 0.48 (REDD1 with LC3 II/I) and R² = 0.30 (ULK1 Ser757 with LC3 II/I), suggesting that other factors likely contribute to autophagy regulation under these conditions. Regardless, it appears that the castration-mediated increase in the basal LC3 II/I ratio was due in part to increased REDD1 protein content, whereas the sustained elevation of this measure in the contracted muscle of castrated mice was due in part to lower phosphorylation of ULK1 (Ser757).

Fig. 3.

Analysis of autophagy regulatory factors in the fasted metabolic state. Phosphorylated to total protein ratio of ULK1 (Ser317) (A), ULK1 (Ser757) (B), Akt (Thr308) (C), FoxO3a (Ser253) (D), and AMPK (Thr172) (E) were determined by Western blot analysis. Total protein content of BNIP3 (F) and REDD1 (G) protein content were determined by Western blot analysis. H: representative Western blots. Significance set at P < 0.05. ME, main effect; P, phosphorylated protein; T, total protein. N = 8/group from 3 independent experiments.

Nutrient consumption before muscle contractions normalized the castration-induced changes in mTORC1 signaling and content of ubiquitylated proteins, but not markers of autophagy.

Previous work showed that nutrient consumption coupled with resistance exercise leads to a greater positive shift in protein balance than resistance exercise alone (10, 29). To determine if nutrients altered our measures, overnight fasted sham-castrated and castrated mice were provided access to food pellets for 30 min before muscle contractions. Both groups consumed similar quantities of food in this refeeding period (455 ± 24 mg of food and 460 ± 52 mg of food, respectively) with no significant difference observed between groups (P = 0.54). Importantly, this quantity of food has been shown to induce a significant anabolic response in mouse skeletal muscle (18). Similar to the fasted condition, there was a main effect of contractions to increase rates of protein synthesis with no effect of castration detected (Fig. 4, A and D). Contractions also increased measures of mTORC1 signaling [i.e., phosphorylation of p70S6K1 (Thr389) and 4E-BP1 (Ser65)] (Fig. 4, B–D), but unlike the fasted state, nutrient consumption negated the castration-mediated reduction in these measures. Despite this normalization of mTORC1 signaling, the main effect of castration to increase the LC3 II/I ratio remained even after refeeding (Fig. 5, A and F). Likewise, a main effect of castration to reduce p62 protein content was also evident (Fig. 5, B and F), suggesting that the baseline of autophagy remained elevated after refeeding and that muscle contractions could not normalize this process. Furthermore, nutrient consumption before muscle contractions negated the changes in the content of protein ubiquitylation observed in the fasted state (Fig. 5, C and F).

Fig. 4.

Protein synthesis and mTORC1 signaling in the refed metabolic state. A: muscle protein synthesis was determined by the SUnSET method. Phosphorylated to total protein ratio of p70S6K1 (Thr389) (B) and 4E-BP1 (Ser65) (C) were determined by Western blot analysis. D: representative Western blots. Significance set at P < 0.05. ME, main effect; P, phosphorylated protein; T, total protein. N = 7–10/group from 4 independent experiments.

Fig. 5.

Markers of protein degradation in the refed metabolic state. LC3 II/I ratio (A), p62 protein content (B), and content of ubiquitylated proteins (C) were determined by Western blot analysis. Phosphorylated to total protein ratio of ULK1 (Ser757) (D) and the protein content of REDD1 (E) were determined by Western blot analysis. F: representative Western blots. ME, main effect; P, phosphorylated protein; T, total protein. N = 8–10/group from 4 independent experiments.

Because REDD1 protein content and ULK1 (Ser757) phosphorylation were related to the LC3 II/I ratio in the fasted state and both measures are sensitive to nutrient consumption (18, 20, 24), we sought to determine whether these relationships remained after refeeding. Interestingly, and unlike those data observed in the fasted condition, nutrient consumption before muscle contractions negated the reduction in the phosphorylation of ULK1 (Ser757) and the increase in REDD1 protein content by castration (Fig. 5, D–F). Additionally, the upstream factors previously analyzed [ULK1 (Ser317), Akt (Thr308), FoxO3a (Ser253), and BNIP3] failed to provide a potential explanation of the maintained elevation in autophagy markers (data not shown). Therefore, castration increases markers of autophagy after refeeding and muscle contractions, and this appears to be independent of REDD1 protein content and the phosphorylation of ULK1 (Ser757).

DISCUSSION

Although it is well accepted that androgens significantly affect muscle mass through changes in protein balance, the factors that contribute to these changes in the basal state and after an anabolic stimulus remain undefined. Signaling through mTORC1 has been proposed to be a contributing mechanism to the androgen-induced change in baseline protein balance by influencing rates of protein synthesis. For instance, treatment of muscle cells in culture with supraphysiological concentrations of testosterone (i.e., 100 to 500 nM) increased mTORC1 signaling and cell size in a rapamycin sensitive manner (1, 4, 33). Furthermore, castration decreased phosphorylation of mTORC1 substrates [i.e., p70S6K1 (Thr389)] and protein synthesis in the gastrocnemius of mice, whereas supraphysiological androgen administration restored or enhanced these measures (33). However, these previous mechanistic studies failed to control for feeding/nutrient status, and supraphysiological concentrations of androgens were administered to artificially enhance hormone levels. In the present study, when feeding status was controlled and a model of consistent androgen deprivation was used, we also observed a reduction in the phosphorylation of mTORC1 substrates [i.e., p70S6K1 (Thr389) and 4E-BP1 (Ser65)] by castration in the noncontracted and contracted muscles, but only in the fasted state. Furthermore, this reduction in mTORC1 signaling did not coincide with a measurable difference in global rates of protein synthesis in either the noncontracted or contracted muscle. This suggests that the reduction in mTORC1 signaling was not sufficient to alter global rates of protein synthesis, and under these deprivation circumstances, muscle mass may be regulated independent of protein synthesis. This is consistent with a report in which rates of protein synthesis in either the basal or stimulated conditions (i.e., after amino acid infusion) were unaffected 6 mo after androgen restoration to the physiological range in hypogonadal males despite an increase in lean leg mass (12).

Although the importance of mTORC1 in the regulation of protein balance after changes in androgen levels remains undefined, it is likely to have a role in the regulation of autophagy after muscle contractions. Indeed, we observed a reduction in the phosphorylation of ULK1 (Ser757), a putative mTORC1 site (20), in the contracted muscle of fasted castrated mice, and this measure was related to the LC3 II/I ratio. The reduced phosphorylation of ULK1 (Ser757) in castrated mice may have been due to the increase in the expression/activation of the mTORC1 repressors, REDD1 and AMPK. Although neither alone was directly related to the LC3 II/I ratio in the contracted muscle, it is likely their combined elevation contributed to the suppression of mTORC1-mediated phosphorylation of ULK1 (Ser757) and thus indirectly contributed to the regulation of autophagy (6, 17). Further evidence for an indirect effect of AMPK in the regulation of autophagy through mTORC1 is garnered by the inconsistent phosphorylation pattern between ULK1 (Ser317), a putative AMPK site (20), and the phosphorylation of AMPK (Thr172) (Fig. 3, A, E, and H).

The finding that REDD1 protein content was increased by androgen depletion is consistent with previous data showing that REDD1 expression is regulated by testosterone. For example, administration of testosterone to male rats was sufficient to block the glucocorticoid-mediated increase in REDD1 expression in the gastrocnemius, a response that required a functional androgen receptor (34). Because the fasting-induced increase in glucocorticoids contributes to increased REDD1 expression in skeletal muscle (24), it is likely that the depletion of testosterone enabled a greater glucocorticoid-mediated induction of REDD1 in the fasted state. Although underpowered, further support for this idea is garnered by a trend (P = 0.08) for an inverse relationship between circulating testosterone concentrations and REDD1 protein expression in the noncontracted tibialis anterior of fasted sham-castrated mice (r = −0.55). Thus it is likely that the lack of androgens increases autophagy during fasting in part by enabling a greater glucocorticoid-mediated increase in REDD1 expression. However, refeeding negated the increase in REDD1 expression despite a sustained elevation in markers of autophagy in castrated mice. Hence, identifying the factors that contribute to the sustained elevation in markers of autophagy after refeeding requires further investigation.

An interesting outcome from this study was the interaction between castration and muscle contractions in relation to the ubiquitylated protein content. The decrease in ubiquitylation after muscle contractions in sham-castrated animals is consistent with data showing that the activity of the UPS was increased after muscle overload in a manner that is independent of atrogene mRNA expression (i.e., MuRF-1 and MAFbx) (3). The content of ubiquitylated proteins was already decreased in the noncontracted muscle of castrated mice, and this value was not further decreased, rather slightly increased, by contractions. The lower content of ubiquitylated proteins in the noncontracted muscle of castrated mice may be due to an already elevated level of UPS activity during fasting, which is consistent with previous reports in humans and rodents (12, 19). Although the reason for the small increase in ubiquitylated proteins after contractions in castrated mice is unknown, it may be due to an increase in ubiquitin ligase activity without a corresponding changes in UPS activity. Overall, this interaction requires further investigation.

In conclusion, the results of this investigation are consistent with the idea that depletion of androgens reduces the positive shift in protein balance after muscle contractions due at least in part to the sustained elevation in markers of autophagy. Additionally, these data highlight potential roles for REDD1, AMPK, and ULK1 in the regulation of autophagy and mTORC1 signaling in the fasted state after androgen depletion. In all, these results provide a molecular basis for understanding how androgens alter muscle protein balance after muscle contractions.

GRANTS

This project was supported by National Institute on Alcohol Abuse and Alcoholism Grant F32 AA023422 (to J.L.S.).

AUTHOR CONTRIBUTIONS

J.L.S. and B.S.G. conceived and designed study; J.L.S., M.L.R., and B.S.G. performed experiments; J.L.S., D.H.F., M.L.R., and B.S.G. analyzed data; J.L.S., M.L.R., J.R.H., and B.S.G. interpreted results of experiments; J.L.S. and B.S.G. drafted manuscript; J.L.S., D.H.F., J.R.H., and B.S.G. edited and revised manuscript; J.L.S., D.H.F., M.L.R., J.R.H., and B.S.G. approved final version of manuscript; B.S.G. prepared figures.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).