Influence of gonadectomy on muscle health in micro- and partial-gravity environments in rats

Abstract

Gonadal hormones, such as testosterone and estradiol, modulate muscle size and strength in males and females. However, the influence of sex hormones on muscle strength in micro- and partial-gravity environments (e.g., the Moon or Mars) is not fully understood. The purpose of this study was to determine the influence of gonadectomy (castration/ovariectomy) on progression of muscle atrophy in both micro- and partial-gravity environments in male and female rats. Male and female Fischer rats (n = 120) underwent castration/ovariectomy (CAST/OVX) or sham surgery (SHAM) at 11 wk of age. After 2 wk of recovery, rats were exposed to hindlimb unloading (0 g), partial weight bearing at 40% of normal loading (0.4 g, Martian gravity), or normal loading (1.0 g) for 28 days. In males, CAST did not exacerbate body weight loss or other metrics of musculoskeletal health. In females, OVX animals tended to have greater body weight loss and greater gastrocnemius loss. Within 7 days of exposure to either microgravity or partial gravity, females had detectable changes to estrous cycle, with greater time spent in low-estradiol phases diestrus and metestrus (∼47% in 1 g vs. 58% in 0 g and 72% in 0.4 g animals, P = 0.005). We conclude that in males testosterone deficiency at the initiation of unloading has little effect on the trajectory of muscle loss. In females, initial low estradiol status may result in greater musculoskeletal losses.

NEW & NOTEWORTHY We find that removal of gonadal hormones does not exacerbate muscle loss in males or females during exposure to either simulated microgravity or partial-gravity environments. However, simulated micro- and partial gravity did affect females’ estrous cycles, with more time spent in low-estrogen phases. Our findings provide important data on the influence of gonadal hormones on the trajectory of muscle loss during unloading and will help inform NASA for future crewed missions to space and other planets.

INTRODUCTION

Muscle mass is a significant predictor of health and longevity across a variety of conditions (1). Certain environmental stressors such as microgravity, and partial gravity as would be found on the Moon and Mars, elicit muscle atrophy in human and rodent models (2–4). Spaceflight-induced muscle loss is associated with worse health outcomes in astronauts upon return to Earth (5, 6). Generally speaking, muscle mass is maintained by the balance of protein anabolism and catabolism, with catabolism exceeding anabolism during atrophic states (7). Yet, despite years of research on spaceflight-induced muscle loss, the complex mechanisms underlying muscle catabolism and anabolism are not fully understood. Correspondingly, interventions to mitigate muscle loss, such as resistance exercise, have limited effectiveness and require very high exercise volumes (8).

In atrophic pathologies, endocrine function has been explored as a moderator of muscle mass. Prior studies have found alterations to normal endocrine function during simulated spaceflight in rodents, specifically, reductions in serum estradiol and testosterone levels. Testosterone is an anabolic steroid hormone contributing to enhanced protein synthesis through multiple mechanisms (9, 10). Preclinical models suggest that spaceflight itself may alter circulating testosterone levels. For example, hindlimb suspension in male rodents is sufficient to reduce testis size and plasma testosterone concentrations (11–13). Moreover, clinical research using spaceflight and bed rest has suggested decreased serum concentrations of luteinizing hormone (LH), which is necessary to induce testosterone synthesis (14). Given testosterone’s anabolic roles, changes in circulating testosterone levels associated with spaceflight could have meaningful implications for astronaut health. Separately, previous work has suggested that removal of testosterone via castration results in reduced muscle size (15, 16). However, a recent study proposed that testosterone status and muscle function may be more nuanced than previously believed. In castrated male mice, there were no differences in muscle cross-sectional area (CSA) between castrated and control animals across multiple ages (17). Correspondingly, others have found that castration in rats did not influence gastrocnemius muscle protein synthesis (18). Regardless, there is physiological plausibility that spaceflight may alter testosterone synthesis and circulating testosterone levels, which could impact muscle loss in microgravity states.

Estradiol, the most abundant circulating estrogen, is a key ovarian hormone in the female estrous (rodent) and menstrual (human) cycle, with low estrogenic states, such as in menopause, being associated with reduced muscle strength (19). Moreover, ovariectomy in female mice results in lower muscle size and strength (20) [a comprehensive review of the role of estrogen in muscle quality has been published elsewhere (21)]. Specific to spaceflight, changes in circulating estrogen levels and effects on corresponding musculoskeletal health are more challenging to quantify. One manuscript reporting studies of space-flown mice concluded that spaceflight did not induce estrous cycle changes, based on histological examination of the uterus (22). However, given the cyclical nature of estradiol secretion, it is unlikely that a single histological examination of the uterus would appropriately reflect estradiol levels due to the estrous cycle. We recently reported alterations in estrous cyclicity in female rats after exposure to simulated microgravity (23). Importantly, these alterations correlated with changes in muscle strength (23). As for testosterone, it is not known whether the presence of low circulating estradiol levels before the initiation of atrophic stimuli affects further development or progression of muscle loss.

NASA’s future crewed missions will have equal representation of males and females. As such, it is critical to understand how initial content of serum sex hormones influences the trajectory of musculoskeletal outcomes during spaceflight. Therefore, the purpose of this study was to determine the influence of initial hormonal status on progression of muscle atrophy in both analog micro- and partial-gravity environments in male and female rats. We hypothesized that the absence of gonads and corresponding sex steroid hormones would exacerbate musculoskeletal responses to decreased mechanical loading.

METHODS

Animal Experiments

Overall design.

All experiments were approved by the Beth Israel Deaconess Medical Center (BIDMC) Institutional Animal Care and Use Committee (IACUC, no. 081-2020). One hundred twenty (60 males, 60 females) Fischer rats were purchased from Charles River Laboratory (Wilmington, MA; stock no. 002). At 11 wk of age, animals underwent castration (CAST) or ovariectomy (OVX) (for males and females, respectively) or sham (SHAM) surgeries at Charles River Laboratory. Surgeries and postoperative care were approved by Charles River Laboratory’s IACUC. Animals were allowed to recover for ∼1 wk and then shipped to BIDMC. Rats arrived at BIDMC at ∼12 wk of age. After 48 h of acclimation to BIDMC animal facilities, estrous cycle monitoring of females was initiated (described below). At ∼14 wk of age, rats underwent baseline testing for longitudinal variables (described below). Animals were then divided into three different loading groups: 0% partial weight bearing (hindlimb unloading, 0 g), 40% partial weight bearing (simulating Martian gravity, 0.4 g), and 100% partial weight bearing (fully weight bearing controls, 1 g; further described below) and underwent interventions for 28 days. Approximately 48 h before baseline measurements, animals were acclimated to specialized cages suitable for unloading, as described previously (4, 24). After longitudinal data collection, animals were euthanized with 30–70% carbon dioxide of chamber volume per minute; death was confirmed by cardiac puncture per the 2020 American Veterinary Medical Association Guidelines on Euthanasia. Muscle tissues were then collected and weighed on a precision scale. Right leg muscles [gastrocnemius, soleus, tibialis anterior (TA), and extensor digitorum longus (EDL)] were flash frozen in liquid nitrogen. Left leg soleus was fixed in 10% neutral buffered formalin for 48 h, rinsed with phosphate-buffered solution (PBS), and used for histological analysis.

Hindlimb unloading to simulate microgravity (0 g) and simulated partial gravity (0.4 g) interventions.

0 g was accomplished with a pelvic harness as our group has described previously (24, 25). Simulated partial gravity at 0.4 g was accomplished with our specialized harness system and an animal-grade scale as our group has described previously (24, 25). Images of cages and harness system are depicted in prior reports (4, 24). A separate group of rats served as full weight-bearing controls (1 g) without any harness system, as we previously established that the presence or absence of the harness in control animals does not impact musculoskeletal health (25). Animals were checked at least weekly to ensure continuous appropriate loading at 0.4 g with body weight changes. Animals were randomly assigned to groups; within OVX/CAST/SHAM, there were no differences between baseline body weights within hormone status. For all in vivo animal assessments [grip strength, maximally stimulated muscle force, peripheral quantitative computed tomography (pQCT), and estrous cycle], experimenters were blinded to animal groups.

Animal housing and monitoring.

During inventions all animals were singly housed in a temperature-controlled (∼23°C) facility with a 12:12-h light-dark cycle. Animals were fed standard LabDiet Formulab Diet rodent chow (catalog no. 5008, irradiated). This product contains a caloric breakdown of 26.6% protein, 16.5% fat, and 56.8% carbohydrate. Food intake was recorded daily. Daily checks were performed to ensure that animals remained in partial-gravity apparatuses and to monitor rodent health. Additionally, animals were weighed weekly. If an animal exhibited significant signs of distress or health problems (>20% body weight loss, anorexia, nongrooming behaviors, etc.) it was removed from the study and euthanized. One animal (M-CAST-0.4g) was removed from the study because of excessive body weight loss.

Estrous Cycle Monitoring

The estrous cycle was monitored for ∼2 wk before the start of interventions to establish a baseline for female rats. Estrous cycle monitoring was performed as we (23) and others (26) have previously described. Briefly, a small amount of sterile dH₂O (∼100 µL) was inserted at the vaginal canal into the vagina with a standard pipette and pipette tip. The vagina was flushed two or three times with the dH₂O to collect cells from the vaginal wall. The dH₂O containing the vaginal cells was then pipetted onto a microscope slide and allowed to dry (∼2–3 h minimum). Microscope slides were then stained with 0.1% crystal violet solution in dH₂O (Sigma Aldrich, catalog no. C0775-025G). After staining, vaginal cells were visualized with a standard white light microscope using a ×40 objective. Based on the relative numbers of nucleated epithelial cells, cornified epithelial cells, and leukocytes, estrous cycle phase was determined to be proestrus, estrus, metestrus, or diestrus per previous protocols (23, 26). All sample collections were completed by trained research staff between 8:00 and 10:00 AM. All visualization of vaginal staining was completed by the same researcher (M.E.R.-C.). Our prior studies have found excellent retest reliability for these measurements (23).

Grip Strength

Rear paw grip strength was measured as previously described (11, 24). Rats were gently held and restrained by a trained researcher. Then, rodents’ rear paws were placed on a grip bar attached to a 50-N-capacity force transducer (Chatillon, Agawam, MA). Animals were gently pulled from the bar until they voluntarily released their grip. The corresponding force was recorded. After a brief (∼30–60 s) recovery, the measurement was repeated twice. The maximal force across the three trials was used for analysis. Grip strength was measured before and after interventions.

Maximally Stimulated Muscle Force

Rats were anesthetized with ∼2% isoflurane and a standard vaporizing system. Under anesthesia, animals were placed on a foot plate system (Dual Mode Muscle Lever System; Aurora Scientific) as previously reported (11). The rodents’ left foot was taped to the foot plate with medical-grade tape (3M Transpore Surgical Tape, catalog no. 1527). To ensure consistency of data collection, the foot plate position was adjusted to 90° of dorsiflexion. The left peroneal and tibial nerves were stimulated to produce dorsiflexion and plantarflexion, respectively. To ensure appropriate needle placement, a small (10 Hz) twitch was applied and recorded. Once needle placement was finalized, nerves were stimulated at 200 Hz for 200 ms to achieve a maximal tetanus. Visual inspection of the contraction was performed to ensure that tetanus was reached. Animals were anesthetized for ∼5 min for this procedure; therefore, we do not anticipate that animals had significant drop in force production due to temperature loss.

Peripheral Quantitative Computed Tomography

Muscle cross-sectional area was quantified in anesthetized animals by peripheral quantitative computed tomography (pQCT) (Stratec; XCT Research SA+, Pforzheim, Germany). Measurements were taken 40 mm distally from the tibial plateau with the following acquisition parameters: voxel size: 0.10 mm, Ct speed: 10 mm/s. Images were then analyzed for density by parsing out different tissues (muscle, fat, bone) through different density thresholds and filters as provided by the manufacturer. Parameters for this analysis have previously been used in our prior experiments (27).

Histology

Histological preparation and analysis was performed as previously described (11). Soleus muscles were fixed in 10% formalin for 48 h. Afterwards, tissues were embedded in paraffin and immunofluorescence staining was used to detect muscle fiber type. Immunofluorescence tagging was performed with antibodies specific to myosin heavy chain (MHC) I (ab11083; Abcam) and MHC II (ab91506; Abcam). Wheat germ agglutinin antibodies were used to detect myofiber cell borders (W6748; ThermoFisher Scientific). Images were collected with an epifluorescent microscope at ×20 magnification (Zeiss Axio Imager M1). Muscle fiber cross-sectional area was quantified with Fiji (ImageJ, NIH) and muscle morphometry plug‐in (by Anthony Sinadinos using Eclipse IDE). All analysis was completed by the same investigator (D.-M.S.), who was blinded to experimental conditions. The average myofiber CSA was used for analysis.

Statistical Analysis

Longitudinal data were analyzed for change between 0 and 28 days of intervention with a covariate of food consumption throughout the intervention. Food consumption was used as a covariate because of slightly different food consumption between CAST/OVX and SHAM animals throughout the intervention (Supplemental Fig. S1; all Supplemental Materials are available at http://doi.org/10.17605/OSF.IO/ND7XS). Cross-sectional data (e.g., muscle fiber cross-sectional area and tissue masses) were analyzed by determining the average difference from 1 g animals within each interventional group (SHAM or CAST/OVX). Group means were calculated as follows:

$$\text{group mean}=\frac{\sum \left[\text{average }\mathrm{1}g\left(\text{SHAM},\text{ CAST},\text{ or OVX}\right)-\text{individual data point}\right]}{\text{group }n}$$

The specific research question was to investigate how the presence or absence of sex steroid hormones influenced trajectory of muscle atrophy in simulated microgravity and partial gravity in males and females. Therefore, data were analyzed with a customized statistical code in SAS using the LSMESTIMATE statement within PROC MIXED. Primary comparisons of interest included

• CAST vs. SHAM at 0 g in males

• CAST vs. SHAM at 0.4 g in males

• CAST vs. SHAM at 1 g in males

• OVX vs. SHAM at 0 g in females

• OVX vs. SHAM at 0.4 g in females

• OVX vs. SHAM at 1 g in females

These four primary comparisons were condensed into an a priori contrast and globally were tested with a joint F test. If the joint F test was significant (P < 0.05), then individual contrasts were inspected. Significant pairwise differences were determined at a Holm’s adjusted P < 0.05. Significant differences in figures are denoted by *. We also completed a series of equivalence tests as previously described (28). These statistical tests are noted in the Supplemental Files.

Additionally, as a model validation, within each sex and hormonal status (SHAM or CAST/OVX) a one-way ANOVA with a factor of loading condition (0 g, 0.4 g, 1 g) was analyzed, with global significance denoted at P < 0.05 and pairwise comparisons denoted at a Tukey-adjusted P < 0.05. Graphical representation of these analyses can be found in Supplemental Figs. S2–S17.

All data were analyzed with SAS statistical software (SAS Institute, Cary, NC). All SAS coding is available in the Supplemental Files.

RESULTS

Castrated Males Had Similar Body Weight and Grip Strength Losses Compared to Sham, Whereas Females Tended to Have Greater Body Weight Losses

All statistical comparisons with exact P values can be found in the Supplemental Files.

After 28 days of intervention, there were no differences between males at either 1 g or 0.4 g. At 0 g, male CAST-0g rats lost ∼10 g less body weight compared with male SHAM-0g rats (Fig. 1A). For females, there were no differences at 1 g. At 0 g and 0.4 g, OVX animals tended to lose more body weight (∼7.57 g and ∼6.75 g, respectively), although these differences were not statistically significant (Fig. 1B).

Figure 1.

Body weight changes after interventions. A: change in body weight between sham-operated (SHAM) and castrated (CAST) males at 1 g, 0.4 g, and 0 g. B: change in body weight between SHAM and ovariectomized (OVX) females at 1 g, 0.4 g, and 0 g. Individual data points are the difference between preintervention and postintervention. Data are plotted as means ± SE (adjusted for food consumption as a covariate); n = 8–10 rats/group.

Grip Strength and Muscle Area Were Not Affected by Gonadectomy Status

In males there were no differences between SHAM and CAST in grip strength changes across 1 g, 0.4 g, or 0 g (Fig. 2A). Similarly, females also had no differences in grip strength between SHAM and OVX at 1 g, 0.4 g, or 0 g (Fig. 2B).

Figure 2.

Change in grip strength and muscle area after interventions. A: change in grip strength between sham-operated (SHAM) and castrated (CAST) males at 1 g, 0.4 g, and 0 g. B: change in grip strength between SHAM and ovariectomized (OVX) females at 1 g, 0.4 g, and 0 g. C: change in muscle area between SHAM and CAST males at 1 g, 0.4 g, and 0 g. D: change in muscle area between SHAM and OVX females at 1 g, 0.4 g, and 0 g. Individual data points are the difference between preintervention and postintervention. Data are plotted as means ± SE (adjusted for food consumption as a covariate); n = 8–10 rats/group. Muscle area was quantified in the lower leg, ∼40 mm from the tibial plateau.

In males, CAST animals had ∼19.8 mm² greater muscle area than SHAM animals at 1 g (Fig. 2C). At 0.4 g, CAST animals lost less muscle area (∼15.6 mm²) than SHAM animals (Fig. 2C). At 0 g, CAST rats lost less muscle area (∼17.3 mm²) than SHAM rats (Fig. 2C). In females there were no differences between SHAM and OVX animals in muscle area changes at 1 g, 0.4 g, or 0 g (Fig. 2D).

Gonadectomy Status Did Not Affect Changes in Maximal Dorsiflexion or Plantar Flexion

In males, there were no differences between SHAM and CAST at 1 g, 0.4 g, or 0 g in maximal dorsiflexion change (Fig. 3A). Similarly, in females there were no differences between SHAM and OVX at 1 g, 0.4 g, or 0 g in maximal dorsiflexion change (Fig. 3B). In males, there were no differences between SHAM and CAST at either 1 g or 0.4 g in maximal plantar flexion (Fig. 3C). At 0 g, CAST animals tended to have more protection of plantar flexion muscle strength (∼10 N·mm); however, these differences did not reach statistical significance (Fig. 3C). In females, there were no differences in maximal plantar flexion strength between SHAM and OVX at 1 g, 0.4 g, or 0 g (Fig. 3D).

Figure 3.

Change in maximal dorsiflexion and plantar flexion force production after interventions. A: change in maximal dorsiflexion force production between sham-operated (SHAM) and castrated (CAST) males at 1 g, 0.4 g, and 0 g. B: change in maximal dorsiflexion force production between SHAM and ovariectomized (OVX) females at 1 g, 0.4 g, and 0 g. C: change in maximal plantar flexion force production between SHAM and CAST males at 1 g, 0.4 g, and 0 g. D: change in maximal plantar flexion force production between SHAM and OVX females at 1 g, 0.4 g, and 0 g. Individual data points are the difference between preintervention and postintervention. Data are plotted as means ± SE (adjusted for food consumption as a covariate); n = 8–10 rats/group.

Muscle CSA Changes Were Similar in Sham-Operated and Gonadectomized Animals at All Simulated Gravitational Loads

Changes in muscle CSA were quantified by determining the average difference from 1 g animals within each interventional group (SHAM or CAST/OVX). In males, there were no differences between SHAM and CAST in overall muscle fiber CSA at 1 g, 0.4 g, or 0 g (Fig. 4, A and I). In females, there were also no differences between SHAM and OVX in overall muscle fiber CSA at 1 g, 0.4 g, or 0 g (Fig. 4, B and J). Across different MHC isoforms (type I, type II, type I/II hybrid), there were no differences in fiber CSA between SHAM and CAST at any simulated gravitational loads in males (Fig. 4, C, E, G, and I). Similarly, in females there were no differences in fiber CSA across any isoforms between SHAM and OVX at any simulated gravitational loads (Fig. 4, D, F, H, and J).

Figure 4.

Differences in muscle fiber cross-sectional area (CSA) after interventions. A: differences in overall CSA between sham-operated (SHAM) and castrated (CAST) males at 1 g, 0.4 g, and 0 g. B: differences in overall CSA between SHAM and ovariectomized (OVX) females at 1 g, 0.4 g, and 0 g. C: differences in type I muscle fiber CSA between SHAM and CAST males at 1 g, 0.4 g, and 0 g. D: differences in type I muscle fiber between SHAM and OVX females at 1 g, 0.4 g, and 0 g. E: differences in type II muscle fiber CSA between SHAM and CAST males at 1 g, 0.4 g, and 0 g. F: differences in type II muscle fiber CSA between SHAM and OVX females at 1 g, 0.4 g, and 0 g. G: differences in hybrid type I/type II muscle fiber CSA between SHAM and CAST males at 1 g, 0.4 g, and 0 g. H: differences in hybrid type I/type II muscle fiber between SHAM and OVX females at 1 g, 0.4 g, and 0 g. I: representative CSA images in males. J: representative CSA images in females. Pink, type I; green, type II. Individual data points are the difference from the average CSA from within-sex 1 g animals within SHAM or CAST/OVX. All images were collected at ×40 magnification. Data are plotted as means ± SE (adjusted for food consumption as a covariate); n = 8–10 rats/group.

Castrated Males Lost Comparable Muscle Weight Compared to Sham, Whereas Ovariectomized Females Tended to Lose More Muscle Weight Compared to Sham

In males, there were no differences in gastrocnemius mass change between SHAM and CAST at 1 g or 0.4 g (Fig. 5A). At 0 g, SHAM males had greater gastrocnemius mass loss compared with SHAM (∼0.105 g) (Fig. 5A). In females there were no differences between SHAM and OVX in gastrocnemius mass change at 1 g (Fig. 5B). At both 0.4 g and 0 g, OVX females tended to have greater gastrocnemius mass loss compared with SHAM (∼0.066 g and ∼0.081 g), although these differences did not reach statistical significance (Fig. 5B).

Figure 5.

Difference in hindlimb tissue masses after interventions. A: differences in gastrocnemius mass between sham-operated (SHAM) and castrated (CAST) males at 1 g, 0.4 g, and 0 g. B: differences in gastrocnemius mass between SHAM and ovariectomized (OVX) females at 1 g, 0.4 g, and 0 g. C: differences in soleus mass between SHAM and CAST males at 1 g, 0.4 g, and 0 g. D: differences in soleus mass between SHAM and OVX females at 1 g, 0.4 g, and 0 g. E: differences in tibialis anterior mass between SHAM and CAST males at 1 g, 0.4 g, and 0 g. F: differences in tibialis anterior mass between SHAM and OVX females at 1 g, 0.4 g, and 0 g. G: differences in extensor digitorum longus (EDL) mass between SHAM and CAST males at 1 g, 0.4 g, and 0 g. H: differences in EDL mass between SHAM and OVX females at 1 g, 0.4 g, and 0 g. Individual data points are the difference from the average tissue mass from within-sex 1 g animals within SHAM or CAST/OVX. Individual data points are raw values for individual animals. Data are plotted as means ± SE (adjusted for food consumption as a covariate); n = 8–10 rats/group.

In males, there were no differences in soleus mass changes between SHAM and CAST at 1 g, 0.4 g, or 0 g (Fig. 5C). In females, there were also no differences between SHAM and OVX in soleus mass changes at 1 g, 0.4 g, or 0 g (Fig. 5D).

Within the tibialis anterior (TA) muscle, there were no differences in mass change between SHAM and CAST males at 1 g or 0 g (Fig. 5E). At 0.4 g, CAST males tended to have greater TA loss compared with SHAM (∼0.026 g), although this difference did not reach statistical significance (Fig. 5E). In females, there were no differences in TA mass change at 1 g; at 0.4 g and 0 g, OVX animals tended to have greater TA loss compared with SHAM (∼0.028 g and ∼0.020 g, respectively), although these differences did not reach significance (Fig. 5F).

Within the extensor digitorum longus (EDL) muscle there were no differences in mass change between SHAM and CAST males at any simulated gravitational loads (Fig. 5G). Within females, there were no differences between SHAM and OVX at either 1 g or 0.4 g; at 0 g OVX females tended to have greater EDL loss (∼0.006 g) compared with SHAM, although this difference did not reach significance (Fig. 5F).

Estrous Cycle and Gonadal Assessment

Testicular weight was ∼0.92 g lower in 0 g animals compared to 1 g and 0.4 g (Fig. 6A). In females, ovarian weight was ∼0.11 g smaller in 0.4 g compared to 1 g (Fig. 6B), with no difference in ovarian mass between 0 g and 1 g or 0.4 g animals (Fig. 6B). As expected, OVX females spent essentially 100% of time in metestrus (M) or diestrus (D) (Fig. 6C), confirming ovariectomy efficacy. Within SHAM females, 0.4 g animals spent ∼29% more time in M/D compared with 1 g (Fig. 6C). 0 g animals spent ∼16% more time in M/D compared to 1 g (Fig. 6C). Additionally, 0.4 g animals spent ∼13% more time in M/D compared to 0 g (Fig. 6C). To temporally determine how quickly estrous cycle alterations occur, we analyzed cumulative time in M/D over the 4-wk intervention. Within 1 wk, there were statistical differences in time spent in M/D between groups. These differences remained throughout the 28-day intervention (Fig. 6D).

Figure 6.

Gonadal mass differences in intact animals after micro- and partial-gravity interventions. A: testicle mass in intact [sham operated (SHAM)] male animals after interventions. B: ovary mass in intact female animals after interventions. C: overall time spent in either metestrus (M) or diestrus (D) in female animals over the 28-day intervention. OVX, ovariectomized. D: cumulative time spent in M/D over the 28-day intervention. Individual data points are raw values for individual animals. Data are plotted as means ± SE (adjusted for food consumption as a covariate); n = 8–10 rats/group.

DISCUSSION

Sex hormones have long been lauded as key regulators of muscle size. Yet, to date, potential additive effects of initial hormonal status and trajectory of muscle loss during simulated micro- and partial-gravity environments have not been fully investigated. We found that castration in adult male rats did not exacerbate muscle loss compared with SHAM during exposure to simulated micro- or partial-gravity environments. In females, OVX does appear to mildly exacerbate muscle loss during exposure to microgravity and partial gravity. However, given that few of these differences reached statistical significance, even if additive effects exist they likely have overall modest effects compared with the impact of mechanical unloading.

In contrast to our original hypothesis, CAST males did not have greater muscle loss compared with SHAM animals across multiple outcome variables. In fact, there were several instances where castrated animals had attenuated musculoskeletal losses compared with SHAM animals. For example, CAST males had mitigated body weight and muscle area loss (Figs. 1 and 2). Given that previous research in humans and rodents has detailed the anabolic properties of testosterone in relation to muscle size across other muscle pathologies (9, 10, 29–32), the results of the present study may appear paradoxical. However, our data replicate prior work in rodents using similar methods (29). There are a few plausible interpretations of these data. First, it is possible that animals (or humans) that have normal androgens and undergo exposure to micro- or partial-gravity environments undergo essentially a “double hit” of androgen depletion and unloading. These two stimuli combined together may then have an additive effect on muscle loss. Comparatively, CAST animals only have one atrophic stimuli occurring during microgravity or partial gravity. This interpretation would imply that the active alterations to androgen secretion during disuse may be an important moderator of muscle mass during exposure to micro- or partial gravity. We speculate that if the double hit of androgens and disuse was driving our findings, we would see differences between SHAM and CAST animals at 1 g. We do not see differences between SHAM and CAST at 1 g across multiple variables, and raw muscle weights of these animals also suggest no differences between our SHAM and CAST animals (data not shown, available in Supplemental Files). Alternatively, it is possible that the stimuli of microgravity or partial gravity are sufficiently aggressive to result in dramatic muscle loss, regardless of initial testosterone status, and alterations in serum testosterone noted during disuse (11–13) are concurrent with muscle loss but not causative. In fact, supporting this view, prior work has clearly shown that testosterone treatment is insufficient to protect muscle mass during hindlimb unloading (33, 34). Finally, it could be speculated that the lack of differences between CAST and SHAM rats noted in the present study could be attributed to differences in muscle size and strength between SHAM and CAST males before the start of interventions. Our preintervention data largely did not show any differences between SHAM and CAST animals (data not shown, see Supplemental Files). Overall, the present data and the scientific literature imply that we may need to look beyond simple androgen supplementation to mitigate muscle loss in micro- and partial-gravity environments.

Whereas CAST overall had little effect on the trajectory of muscle loss, OVX appeared to have modest effects to accelerate muscle loss in micro- and partial gravitational environments. OVX females had greater losses in body weight, gastrocnemius, tibialis anterior, and EDL compared with SHAM (Figs. 1, 2, and 5). Our data complement previous work in ovariectomized mice that demonstrated reduced force production compared with sham-operated mice after 14 days of hindlimb unloading (35). Although the differences between SHAM and OVX animals rarely reached statistical significance, it is important to note the overall pattern that OVX females tended to have worse outcomes compared with SHAM. Indeed, an additional supplemental analysis of equivalence demonstrated that ovariectomized females were not generally statically equivalent compared with sham-operated females (see Supplemental Files). Comparatively, CAST males had largely equivalent (or attenuated) muscle loss in both micro- and partial-gravity environments compared with SHAM males. Estradiol has recently been recognized as a key modulator of muscle size in females (21); therefore, lack of estradiol may contribute to exacerbated muscle loss during exposure to micro- or partial-gravity environments.

It is possible, if not likely, that these differences between OVX and SHAM females would have been greater and reached statistical significance if SHAM females maintained their own estradiol levels. As demonstrated in Fig. 6, females at microgravity and partial gravity had early and significant estrous cycle alterations, which would be expected to be associated with reduced estradiol and progesterone levels. Specific interactions between estradiol and progesterone have been published elsewhere (9, 21), though there is reason to believe that both hormones may have protective effects on skeletal muscle. Although it is a clear limitation that SHAM animals had estrous cycle changes that may have attenuated our ability to detect differences, a secondary goal of the study was to also evaluate how these interventions altered estrous cycle in normally cycling animals. Recently, there has been scientific debate on the influence of reduced mechanical loading on the estrous cycle, with some studies of mice sent to the International Space Station finding no estrous cycle alterations following spaceflight (22) whereas our laboratory has found that hindlimb unloading is sufficient to induce estrous cycle alterations and musculoskeletal deficits (23). Based on our data, there appears to be evidence to implicate altered gravitational environments in the modification of the female estrous cycle, resulting in reduced sex hormone levels. Although not fully conclusive, these alterations are likely conserved across species, and we would also expect micro- and partial-gravity environments to alter the human female menstrual cycle. These data have interesting scientific and clinical implications; however, they may have limited our ability to detect differences between SHAM and OVX females. Plausibly, this same mechanism could also be at play in males. However, we did not have a method to reliably estimate testosterone status longitudinally in male animals. Regardless, we must acknowledge that it is possible that our interventions altered hormonal status sufficiently in both males and females that SHAM animals had comparable hormonal profiles after interventions compared with CAST/OVX.

It should be noted that our study only evaluated the effects of hypogonadism on muscle health during atrophic stimuli and did not evaluate how hypogonadism may alter or attenuate muscle recovery after exposure to deconditioning. Previous work has demonstrated that both testosterone status (29) and estradiol status (36, 37) moderate musculoskeletal recovery after cessation of microgravity environments. Mechanistically, androgens and estrogens are regulators of muscle satellite cell activation and proliferation (20, 38, 39). Attenuation of satellite cell function could in turn diminish skeletal muscle regeneration after exposure to atrophic stimuli and correspondingly reduced muscle accrual. Of note, given that partial gravity induces relatively less muscle mass and strength losses compared with microgravity (24, 25), it is unknown how or if hypogonadism would have any effect on recovery after exposure to only partial deconditioning. Presumably, altered hormonal statuses are equilibrated (or at least return to a level sufficient for muscle mass gain) upon return to normal loading, which allows for the enhanced muscle regeneration compared with gonadectomized animals. Given our data and the current scientific literature, it is plausible that individuals with lower baseline sex hormones (due to aging or other endocrine alterations) may have similar muscle loss compared to individuals with normal levels of sex hormones but may require extended recovery periods to regain muscle mass.

Taken together, our data demonstrate that undergoing disuse atrophy during an androgen/estrogens-deprived state does not exacerbate muscle atrophy. In female rats, if effects exist, these are modest compared with the atrophic stimuli itself. These data imply that hormonal status may not itself be a key driver for muscle loss in atrophic environments in males and may therefore not be an optimal therapeutic target for future missions to the Moon and beyond. Just as it is important to understand mechanisms directly relating to muscle loss during exposure to atrophic stimuli, it is also just as important to understand mechanisms not directly contributing to muscle loss. Only by understanding both can we appropriately develop effective therapeutics for astronauts to maximize their health and safety in future exploration of our solar system.

DATA AVAILABILITY

Raw data and statistical coding can be found in the Open Science Framework (OSF) page for this project at http://doi.org/10.17605/OSF.IO/ND7XS.

SUPPLEMENTAL MATERIALS

Supplemental Files and Supplemental Figs. S1–S17: http://doi.org/10.17605/OSF.IO/ND7XS.

GRANTS

This study was funded by NASA Awards 80NSSC21K0311 (M.E.R.-C./S.B.R.) and 80NSSC19K9518 (S.B.R.) as well as NIH Award R37HD019938 (U.B.K.) and T32GM144273 (A.W.) from the National Institute of General Medical Sciences.

DISCLAIMERS

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of General Medical Sciences or the National Institutes of Health.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

M.E.R.-C., U.B.K., M.L.B., and S.B.R. conceived and designed research; M.E.R.-C., M.M., A.W., and D.-M.S. performed experiments; M.E.R.-C. and D.-M.S. analyzed data; M.E.R.-C., M.M., A.W., U.B.K., M.L.B., and S.B.R. interpreted results of experiments; M.E.R.-C. prepared figures; M.E.R.-C. drafted manuscript; M.E.R.-C., M.M., A.W., U.B.K., D.-M.S., M.L.B., and S.B.R. edited and revised manuscript; M.E.R.-C., M.M., A.W., U.B.K., D.-M.S., M.L.B., and S.B.R. approved final version of manuscript.