17β-estradiol mitigates ovariectomy-induced defects in mitochondrial bioenergetics and redox balance after VML injury in female mice

Abstract

Volumetric muscle loss (VML) is characterized by an irrecoverable loss of skeletal muscle mass, persistent functional deficits, and metabolic dysfunction. A disrupted cellular redox homeostasis is one attribute of this metabolic dysfunction and can lead to excessive reactive oxygen species (ROS) emissions and chronic oxidative stress. The primary objective of this study was to define the role of ovarian hormones, specifically 17β-estradiol (17β-E2), in driving mitochondrial bioenergetic and redox balance after VML injury. Female C57BL/6J mice were randomized into experimental and control groups (VML-sham OVX, VML-OVX, and VML-OVX-E2). A time course of ROS emissions and antioxidant buffering capacity (AoxBC) for VML-injured muscles was established across the first 60 days post-injury (dpi) in ovary-intact females. Ovariectomy (OVX) was performed prior to injury to deplete ovarian hormones, and 17β-E2 was administered via continuous-release pellets to investigate the effects of hormone loss and repletion on ROS emissions and mitochondrial bioenergetics. The long-term effects of 17β-E2 were evaluated to determine whether restoring redox led to sustained redox balance long-term. Transcriptomic analyses were conducted to explore molecular mechanisms of 17β-E2 benefit after VML. In intact females, ROS emissions were greater during the first 14-dpi, but AoxBC recovered more rapidly than previously observed in males. OVX exacerbated VML-induced metabolic dysfunction, resulting in less AoxBC, greater ROS emissions, and an early suppression of mitochondrial gene networks. 17β-E2 repletion attenuated ROS emissions and improved AoxBC at 7-dpi, and led to greater mitochondrial respiratory capacity, conductance, and bioenergetic efficiency out to 60-dpi. Chronic 17β-E2 depletion resulted in impaired glucose tolerance and greater adiposity, which were mitigated by 17β-E2 treatment. Transcriptomic analyses suggest that 17β-E2 contributes to resolving inflammation and enforcing a temporal decoupling of cellular expansion and mitochondrial maturation after VML injury.

New and Noteworthy

Female mice exhibit accelerated recovery of mitochondrial redox balance after volumetric muscle loss (VML) compared to males. This study demonstrates that 17β-estradiol (17β-E2) drives this resilience. Following VML, ovariectomy induced an early transcriptional arrest and asynchronous repair signaling. 17β-E2 replacement restored regenerative coordination by temporally decoupling early cellular expansion from mitochondrial biogenesis. This precise transcriptional regulation translated to long-term functional resilience, restoring mitochondrial bioenergetic efficiency and resolving oxidative stress.

Graphical Abstract

INTRODUCTION

Volumetric muscle loss (VML) injury results in the irrecoverable loss of skeletal muscle mass and persistent functional deficits. Affecting both military personnel and the general population, VML often occurs secondary to severe traumatic events, including the estimated 150,000 open fractures, 30,000 gunshot wounds, 36,000 chainsaw accidents, and 13,000 soft-tissue sarcoma resections that occur annually in the United States (1, 2). The pathophysiology of VML extends far beyond the initial mechanical tissue loss, creating a cellular microenvironment characterized by fibrotic infiltration (3), chronic inflammation (4, 5), and metabolic dysfunction (6, 7). Importantly, this disrupted metabolic state impairs mitochondrial bioenergetics, creating an environment highly susceptible to the persistent overproduction of reactive oxygen species (ROS) and chronic oxidative stress. Despite the critical role of mitochondrial and redox homeostasis in determining tissue viability and recovery, research directed toward understanding VML pathophysiology has overwhelmingly focused on male subjects. Consequently, while the deleterious metabolic and oxidative consequences of VML are becoming clearer in males, a critical knowledge gap remains regarding female-specific metabolic responses to injury and the endogenous mechanisms that may govern redox balance in the female microenvironment.

Cellular redox homeostasis is dictated by the thermodynamic balance between the production of ROS and the tissue’s antioxidant buffering capacity (AoxBC) to reduce ROS to water. Disruption of this balance leads to chronic oxidative stress, a hallmark of the VML microenvironment. We previously reported metabolic and biochemical sex differences following VML, demonstrating that female mice are intrinsically less susceptible to VML-induced metabolic dysfunction compared to their male counterparts (8). Building on these findings, our recent investigations reveal a striking sex-dependent temporal divergence in post-injury redox homeostasis (Figure 1). As illustrated in this foundational summary comparison, while VML elicits a robust increase in ROS production in both sexes, the tissue AoxBC remains compromised in males for at least 30 days post-injury (dpi) [Figure 1A,C; (9)]. In stark contrast, female mice demonstrate a significantly accelerated recovery, with AoxBC returning to uninjured baseline levels as early as 14-dpi (Figure 1B,C). This divergence at the 14-dpi mark implies that females possess an innate, temporally protective mechanism against VML-induced oxidative stress; however, the molecular drivers underlying this resilience remain unresolved.

Figure 1:

Foundational comparison of sex-specific antioxidant buffering capacity (AoxBC) kinetics following VML. To establish the premise of the current study, male and female recovery profiles are contrasted. (A,C) Male permeabilized muscle fiber AoxBC in response to carbohydrate or fat substrates through 60 days post-injury (dpi). Data are republished with CC-BY-NC open access from Heo et al. (9) to demonstrate the delayed recovery of AoxBC in males (statistically significant through 14–30 dpi). (B,D) Female permeabilized muscle fiber AoxBC in response to carbohydrate or fat substrates through 60-dpi. In contrast to males, females demonstrate a rapid restoration of AoxBC by 14 dpi. Female data are summarized here for conceptual comparison and reported in primary detail (including individual replicates) in Figure 3. Bars represent mean ± SD. Statistical significance (*) denotes p < 0.05 from Naïve (uninjured) baseline via one-way ANOVA.

Emerging evidence implicates ovarian hormones, particularly 17β-estradiol (17β-E2), as primary candidates for this female-specific protection, given their critical role in regulating skeletal muscle contractile function, metabolic homeostasis, and redox balance (10–12). In uninjured models, the depletion of ovarian hormones via ovariectomy (OVX) precipitates mitochondrial dysfunction, reduces antioxidant gene expression, and exacerbates ROS emission (11–13). Conversely, exogenous 17β-E2 repletion effectively rescues these deficits, highlighting its indispensable role in maintaining mitochondrial bioenergetics. Furthermore, ovarian hormones have been shown to facilitate functional recovery and preserve muscle strength following milder forms of muscle injury (14–16). Yet, the extent to which ovarian hormones govern these protective adaptations within the severely compromised, highly fibrotic, and metabolically dysfunctional microenvironment of a VML injury is unclear.

Therefore, the primary objective of this study was to elucidate the role of ovarian hormones, specifically 17β-E2, in driving female-specific mitochondrial bioenergetic and redox resilience following VML injury. We hypothesized that the loss of ovarian hormones via OVX would exacerbate VML-induced mitochondrial dysfunction and prolong redox imbalance (manifesting as increased ROS production and diminished AoxBC), and that exogenous E2 repletion would mitigate these deficits. Mechanistically, we hypothesized that 17β-E2-mediated protection is driven by acute transcriptomic programming that helps to preserve mitochondrial integrity and antioxidant networks. To test this, we employed a comprehensive, multi-tiered approach. First, we characterized the natural temporal progression (7-, 14-, 30-, and 60-dpi) of mitochondrial and redox recovery in ovary-intact females. Second, we evaluated intact, OVX, and OVX+E2 mice during the acute injury phase (3- and 7-dpi), integrating physiological bioenergetic assays with robust transcriptomic analyses to identify early molecular regulators of metabolic resilience. Finally, we assessed the chronic (60-dpi) biochemical consequences of ovarian hormone-depletion and rescue on long-term bioenergetic and redox homeostasis. Collectively, this study provides critical mechanistic insights into the sex-specific pathophysiology of VML and highlights the therapeutic potential of estrogenic pathways for improving muscle recovery in estrogen-deficient populations, such as those experiencing menopause, ovarian dysfunction, or low energy availability-induced hormone decline.

MATERIALS AND METHODS

Animals and ethical approval

Adult female C57BL/6J mice (N=80, aged 11-weeks) were purchased from Jackson Laboratories (Stock #000664, Bar Harbor, ME, USA; RRID:IMSR_JAX:000664) and were group-housed with food and water provided ad libitum. Housing temperature was controlled at 20–23°C on a 12-h light-dark cycle. Upon arrival, mice were acclimated to the facility for 1-week before being randomly assigned to groups. Humane euthanasia was performed using carbon dioxide and cervical dislocation. All procedures and animal care guidelines were approved and conducted in accordance with the guidelines and regulations of the Institutional Animal Care and Use Committee at the University of Georgia. Procedures were conducted in compliance with the Animal Welfare Act and the Implementing Animal Welfare Regulations, in compliance with the principles of the Guide for the Care and Use of Laboratory Animals.

Experimental designs

Study 1: Time course of ROS emission and association with mitochondrial dysfunction after VML injury in intact females

Permeabilized gastrocnemius muscle fiber bundles from VML-injured mice were assessed at 7-, 14-, 30-, and 60-day post-injury for ROS emission, ROS production, AoxBC, and mitochondrial function (n=4 per time point). Results were compared to gastrocnemius fiber bundles from injury naïve (uninjured) mice (n=4) (Figure 2).

Figure 2:

Schematic of experimental designs for studies 1-3 including primary dependent variables. Day post-injury, dpi; volumetric muscle loss, VML; ovariectomy, OVX; 17β-estradiol pellet, E2; mitochondrial respiration, JO₂; mitochondrial membrane potential, Δψm; reactive oxygen species and antioxidant buffering capacity, JH₂O₂. This figure was created with BioRender.com.

Study 2: The effect of ovarian hormone loss and 17β-E2 rescue on acute redox balance and transcriptomic regulation after VML

Based on the results from study 1, the first week post-injury is a critical time to evaluate the physiology and transcriptional patterns of mitochondrial function and redox balance in the remaining muscle of female mice. Importantly, transcriptional regulation plays an essential role in muscle adaptation and function, as transcripts are available for translation well before functional changes are evident (17). Therefore, 3- and 7-dpi timepoints were selected for this study (Figure 2). To examine the acute effect of 17β-E2 on mitochondrial function, redox balance, and transcriptomic regulation, mice were randomly divided into three groups: VML-sham OVX, VML-OVX given placebo pellets (VML-OVX), or VML-OVX given 17β-E2 pellets (VML-OVX-E2) (n=5 per group per time point). Sham or OVX surgery was conducted at 12 weeks of age. 17β-E2, designed to release 0.18 mg of 17β-E2 for 60 days or placebo pellets (Innovative Research of America, Inc.), was inserted at the time of surgery (13, 18). Two weeks later, at 14 weeks of age, mice underwent a unilateral VML surgery at the center of the plantar flexor muscles (gastrocnemius/plantaris/soleus muscles). Terminally, 3- and 7-dpi, uterine mass was recorded as a secondary marker of circulating estrogen levels (19). All mice were fed phytoestrogen-free food (Harlan-Teklad #2019) to completely exclude exogenous estrogen intake. Mitochondrial ROS and AoxBC were measured using permeabilized gastrocnemius muscle fibers at 3- and 7-dpi. The ~20 mg muscle tissue was snap-frozen in liquid nitrogen and stored at −80 °C for transcriptomic analysis. We excluded two animals, one from each VML-OVX and VML-OVX-E2 groups at 7-dpi cohort based on the results of the uterine mass.

Study 3: The effect of ovarian hormone loss and 17β-E2 rescue on chronic bioenergetics and redox balance post-VML.

In order to determine whether the long-term effects of 17β-E2 alter mitochondrial function and AoxBC following VML in OVX mice, mice were randomly separated into three groups as study 2: VML-sham OVX, VML-OVX, and VML-OVX-E2 (n=10 per group). Mitochondrial bioenergetic profiles, ROS emission/production, and AoxBC were measured using permeabilized gastrocnemius fiber bundles.

OVX surgery

Female mice randomized to undergo bilateral OVX procedure were administered a pre-procedural dose of Carprofen (20mg/kg, s.q) ~30-min prior to the start of the procedure for analgesia control. Mice were anesthetized by isoflurane inhalation (1-2%). Immediately prior to initial incision, Bupivacaine (1-2mg/kg, s.q.) was administered for additional local analgesia at the surgical site. Briefly, two small dorsal incisions were made through the skin and muscular layer between the iliac crest and the lower ribs. Ovarian tissue was visualized and excised. The muscles and skin were closed in layers. Animals were monitored throughout acute recovery and twice daily for 72-hr following the procedure, an additional dose of Carprofen was administered daily over this time. At the time of OVX surgery, mice were subcutaneously implanted with either 17β-estradiol, designed to release 0.18 mg of E2 for 60 days, or placebo pellets (Innovative Research of America, Inc.) (13, 18).

Surgical creation of VML injury

The VML injury was performed on the posterior compartment of anesthetized (isoflurane inhalation 1.5-3.0%) mice under aseptic surgical conditions as described previously (8, 20, 21). Mice received buprenorphine (Patterson Veterinary Supply, Inc.; 1.2mg/kg; s. q.) prior to surgery. Buprenorphine was given at 12 and 24 h, and meloxicam (2.0 mg/kg) was given at 24, 48, and 72 h post-procedure for pain management. A single incision was made in the mid-gastrocnemius to expose the posterior compartment muscles. A 4-mm biopsy punch was employed to induce VML (20.17±3.94 mg) from the center of the plantar flexor muscles (gastrocnemius/plantaris/soleus) complex. The skin incision was closed with 6.0 silk suture (ETHICON, 668G S32), and the mice were monitored throughout recovery.

Intraperitoneal glucose tolerance test (ipGTT)

One week before harvest, an intraperitoneal glucose tolerance test (ipGTT) was conducted following glucose injection (2.0g/kg body weight, i.p.) during the light-cycle in the animal facility in Study 3 (Figure 2). After 6 hours of fasting, blood glucose was evaluated before injection and again at 30, 60, and 120 min after injection.

Mitochondrial respiration

High-resolution oxygen respiratory measurements were conducted on permeabilized gastrocnemius fiber bundles using an Oroboros Oxygraph-2K (Oroboros Instruments, Innsbruck, Austria) with a modified creatine kinase (CK) energetic clamp technique (9, 22–24). This method precisely controls the extramitochondrial ATP:ADP ratio and ΔG_ATP by utilizing excess CK along with defined concentrations of creatine, phosphocreatine (PCr), and adenylates to simulate physiological energy demand shifts. To assess mitochondrial respiratory capacity under near-exercise conditions, fiber bundles were energized with either carbohydrate (5 mM pyruvate, 2 mM malate) or fat (40 μM palmitoyl-carnitine, 2 mM malate) substrates in the presence of CK (20 U/mL), PCr (1 mM), and ATP (5 mM). Cytochrome c (10 μM) was added to verify mitochondrial integrity, followed by sequential PCr titrations (6, 12, 15 mM) to progressively lower ΔG_ATP to resting conditions. This approach enables measurement of the linear relationship between ATP:ADP (ΔG_ATP) and oxygen flux (JO₂), allowing estimation of respiratory conductance, where a steeper slope indicates greater sensitivity and improved kinetics. The CK clamp models energetic demands and thermodynamic constraints comparable to in vivo conditions. To normalize results, oxygen flux rates were adjusted for tissue wet weight and citrate synthase (CS) activity to account for differences in mitochondrial content (25).

Mitochondrial membrane potential

The Δψm was measured fluorometrically in buffer Z containing 5 mM creatine, using a spectrofluorometer (FluoroMax Plus-C; Horiba Instruments Inc., Irvine, CA, USA) (9, 20, 22). The membrane potential was assessed with tetramethylrhodamine methyl ester (TMRM) at 30°C, in accordance with the CK clamp assay protocol and with constant stirring. TMRM excitation/emission [(572/590 nm)/(551/590 nm)] fluorescence is quenched, meaning the 572/551 ratio increases with greater mitochondrial membrane polarization. To our knowledge, the 572/551 ratio is represented here, as there has been no report validating the conversion of the 572/551 ratio to millivolts in permeabilized muscle fiber.

ROS emission, ROS production, Antioxidants buffering capacity, and site-specific ROS

The ROS emission and production were measured in buffer Z supplemented with Amplex Ultrared (5 μM), Cu-Zn superoxide dismutase (25 units/ml), and horseradish peroxidase (1 U/mL) detection system of H₂O₂ (Ex:Em 565:600), as previously described (9, 26, 27) with a minor modification. After recording the basal rate, ROS emission was assessed by adding 5 mM pyruvate and 2 mM malate (carbohydrate substrates), or 40 μM palmitoyl-carnitine and 5 mM malate (fat substrates). ROS production was assessed by the addition of 1 μM auranofin, an inhibitor of thioredoxin, and 100 μM carmustine (BCNU), an inhibitor of glutathione reductase. The AoxBC (percentage of ROS buffered by antioxidant enzymes thioredoxin and glutathione reductase) reflects the percentage of ROS produced but not emitted (i.e., (ROS production – ROS emission) / ROS production) x 100) (9, 26).

In vivo plantarflexion muscle function

Peak-isometric torque of hindlimb plantar flexor muscles was measured in vivo as previously described (8, 28–30) using electrophysiological stimulation of the sciatic nerve and the Model 300C muscle lever system (Aurora Scientific, Aurora, Ontario, Canada). Peak-isometric torque was normalized to body mass. Mice were then euthanized immediately, and tissue was harvested for metabolic assays.

RNA Sequencing

Total RNA was extracted from mouse gastrocnemius muscle using the RNeasy Fibrous Tissue Mini Kit (Qiagen) from VML-sham OVX, VML-OVX, and VML-OVX-E2 mice at 3- and 7-days post-VML injury. Sample quantity was determined to be ≥ 400 ng/uL using the Qubit RNA HS assay (ThermoFisher) and fragment size distribution (DV200) was assessed by Agilent TapeStation where DV200 ≥ 82.5%. Libraries were prepared using ribosomal RNA depletion (KAPA Hyper RNA with Riboerase HMR) and sequenced using the NovaSeq X plus (Illumina) in a single flow lane with 150-bp pair-end reads. Sequencing quality was high with ≥ 94% of bases at Q30 and ≥ 60 million total pair-end reads. Read quality was assessed using FastQC (v0.12.1) (31). GC content, read length, adapter contamination and low-quality bases were evaluated and all samples were of sufficient quality. Sequencing reads were pseudo aligned to GRCm39 mouse reference cDNA (NCBI RefSeq assembly GCF_000001635.27) and count estimates were generated using Kallisto (v0.46.1) with default settings and aggregated using tximport (32, 33). Pairwise differential expression analysis was performed using DESeq2 where statistical significance was defined as |log2Fold-Change| > 1 and Benjamini-Hochberg false discovery rate (FDR) < 0.05 (34, 35). Gene set enrichment analysis (GSEA) was performed using fgseaMultilevel on genes ranked by the Wald statistic (min size = 3). GSEA was conducted using the Hallmark gene sets from Molecular Signatures Database (MSigDB) and custom curated mitochondrial gene sets based on the Broad Institute’s MitoCarta3.0 pathways (36–38). For GSEA significance was defined as |Normalized Enrichment Score (NES)| > 1 and FDR < 0.05. Transcription factor activity was investigated by virtual inference of protein-activity by enriched regulon (VIPER) analysis using the packages Dorothea and viper (regulon size ≥ 20, confidence A-D, including Tfam) (39–42). Statistical significance was defined as |NES| > 1 and FDR < 0.05. All analyses were performed in R (v4.6.0); visualizations were generated using ggplot2 and gridExtra (43).

Statistical analysis

All statistical analyses were performed using JMP Pro statistical software (version 16.0.0; SAS Institute, Cary, NC), and figures were generated using GraphPad Prism (version 9.4.1; GraphPad Software, San Diego, CA). All data are presented as mean ± standard deviation (SD), and statistical significance was defined a priori at an alpha level of α≤0.05. To account for technical variability during permeabilized fiber assays, specifically for mitochondrial respiration, membrane potential (Δψm), and ROS emissions and production, a mixed linear model was employed. This model nested 2–3 technical replicates (fiber bundles) within each biological replicate. Across both Study 1 and Study 3, mitochondrial membrane potential as a function of energy demand (Gibbs free energy states), was evaluated using a two-way analysis of variance (ANOVA) with repeated measures. Study 1: Changes in bioenergetic and redox parameters across the VML injury time course were assessed using a one-way ANOVA. When significant main effects were observed, a Dunnett’s post-hoc test was utilized to compare each post-injury timepoint directly to the uninjured Naive control group. Studies 2 and 3: Differences in morphological, bioenergetic, and redox features among the three experimental cohorts (VML-sham OVX, VML-OVX, and VML-OVX-E2) during the acute phases of injury (3- and 7-dpi; Study 2) and the chronic phase (60-dpi; Study 3) were analyzed using a one-way ANOVA. A Tukey’s Honestly Significant Difference (HSD) post-hoc test was employed to examine pairwise differences among groups following a statistically significant main effect.

RESULTS

Study 1: Time course of ROS emission and association with mitochondrial dysfunction after VML injury in intact females

Consistent with prior reports utilizing the CK clamp to investigate mitochondrial function after VML injury (9, 20), there is robust evidence supporting an immediate and lasting change in carbohydrate-mediated JO₂ and respiratory conductance in permeabilized fiber bundles from 7- to 60-dpi (Figure 3A–C; p<0.001). Fat-mediated JO₂ and respiratory conductance were lower in VML-injured muscle through 30-dpi (Figure 3E–G, p<0.001), but not statistically different from Naive mice at 60-dpi. This early recovery of fat-mediated mitochondrial function contrasts with male mice (9). The time course of mitochondrial membrane potential also differs from that reported in males. Herein, the female carbohydrate-mediated analyses showed only a statistically significant depolarization at 60-dpi (Figure 3D, p<0.001) and for fat-mediated analysis a statistically significant hyper-polarization at 7-dpi and then depolarization at 60-dpi for females (Figure 3H, p≤0.003); whereas male mice demonstrated a consistent hyper-polarized membrane across the entire time course (9).

Figure 3.

Time course of permeabilized muscle fiber bundle mitochondrial bioenergetics and reactive oxygen species after volumetric muscle loss injury. A,E: The relationship between ATP re-synthesis demand (ΔGATP) and mitochondrial oxygen consumption (JO2) normalized to citrate synthase (CS) activity for carbohydrate and fat substrates, respectively. B,F: Maximal JO2 for carbohydrate and fat substrates, respectively. Statistical significance was determined by one-way ANOVA. C,G: Respiratory conductance for carbohydrate and fat substrates, respectively. Statistical significance was determined by one-way ANOVA. D,H: The relationship between ΔGATP and mitochondrial membrane potential for carbohydrate and fat substrates, respectively. Statistical significance was determined by two-way ANOVA. I,K: Reactive oxygen species emission, production, and J,L: antioxidant buffer capacity for carbohydrate and fat substrates, respectively. Statistical significance was determined by one-way ANOVA. Data are expressed as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. dpi, days post-injury.

To determine the extent to which the impaired bioenergetic efficiency after VML injury was associated with ROS, the rate of H₂O₂ (JH₂O₂) was assessed in permeabilized fiber bundles using carbohydrate- or fat-substrates in the presence of endogenous antioxidant buffering systems (ROS emission) and when endogenous antioxidants were inhibited (ROS production). Compared to Naive, carbohydrate-mediated ROS emission (Figure 3I, p≤0.019) and production (p=0.026) were significantly higher in VML-injured muscle until 14-dpi. Similarly, fat-derived H₂O₂ emission and production were significantly greater by 14-dpi (Figure 3K, emission; p≤0.008 and production; p≤0.022). Carbohydrate-mediated AoxBC was 35% less at 7-dpi (Figure 3J, p=0.005); however, AoxBC was not statistically different from Naive from 14- to 60-dpi (Figure 3L, p≥0.98). Intriguingly, there were no changes in fat-mediated AoxBC (Figure 3L, p=0.131). Taken together, these data suggest that 7-dpi is a critical transition point for mitochondrial redox balance in female mice.

Study 2 Physiology: The effect of ovarian hormone loss and 17β-E2 rescue on acute redox balance after VML

To determine the extent to which ovarian hormones are involved in the early redox balance recovery in female mice, VML-injured skeletal muscles from VML-sham OVX and VML-OVX mice were assessed at 3- and 7-dpi for AoxBC. The uterine mass was 82% and 78% lower in the VML-OVX group compared to the VML-sham OVX group after 3- and 7-dpi (Figure 4A, E; p ≤ 0.001); however, the uterine mass was significantly greater in the VML-OVX-E2 group compared to the VML-OVX group (Figure 4A, E; p ≤ 0.001). The body weights at 3- and 7-dpi were 16% and 13% greater than in the VML-sham OVX group (Figure 4B, F; p ≤ 0.005). There was a trend in body mass for the VML-OVX-E2 group compared to the VML-OVX group at 3-dpi (Figure 4B, F; p = 0.050). Absolute gastrocnemius muscle mass was not significantly altered at either 3- or 7-dpi (Figure 4C, G; p ≥ 0.981). Notably, normalized muscle mass by body mass at 7-dpi was 14% lower in the VML-OVX group compared to the VML-sham OVX group (Figure 4H, p = 0.001), which rebounded with E2 repletion in VML-OVX-E2 mice at 7-dpi (Figure 4H, p = 0.003).

Figure 4:

Effect of E2 on morphology and redox balance at 3- and 7-days post-injury (dpi). Effect of ovarian hormone loss and E2 rescue on uterine mass, body mass (BM), gastrocnemius muscle mass, and body mass-normalized gastrocnemius muscle mass at 3-dpi (A-D) and at 7-dpi (E-H). Statistical significance was determined by one-way ANOVA with Tukey’s HSD post-hoc test. Data are expressed as mean ± SD. Dots equal individual mice.

We sought to investigate the role of E2 on ROS emission, ROS production, and AoxBC using carbohydrate and fat substrates at 3- and 7-dpi. A 3-dpi, both ROS emissions and production from permeabilized fiber bundles were not altered across all groups from carbohydrate substrates (Figure 5A, B; p ≥ 0.991). Fat-mediated AoxBC was 26% less in the VML-OVX group compared to the VML-sham OVX (Figure 5D, p = 0.015); however, the AoxBC was 35% greater in the VML-OVX-E2 compared to the VML-OVX group (Figure 5D, p = 0.010). At 7-dpi, there were no changes in carbohydrate-fueled ROS emission and production across the groups (Figure 5E, p = 0.998). Fat-derived ROS emission was 64% greater in the VML-OVX than the VML-sham OVX group (Figure 5G, p = 0.026); however, ROS emission was 51% lower after E2 repletion in the VML-OVX-E2 groups compared to the VML-OVX group (Figure 5G, p = 0.004). Carbohydrate-mediated AoxBC was 25% less in the VML-OVX compared to VML-sham OVX (Figure 5F, p = 0. 040); however, E2 repletion significantly restored carbohydrate-mediated AoxBC in VML-OVX-E2 compared to VML-OVX (Figure 5F, p = 0.049). Fat-derived AoxBC was not significantly reduced in the VML-OVX (Figure 5H, p = 0.113). The VML-OVX-E2 group has 30% greater fat-induced AoxBC compared to the VML-OVX (Figure 5H, p = 0.012).

Figure 5:

Effect of E2 on permeabilized muscle fiber bundle reactive oxygen species after VML injury. A-D: Reactive oxygen species emission, production, and antioxidant buffer capacity for carbohydrate and fat substrates, respectively, at 3-days post-injury (dpi). E-H: Reactive oxygen species emission, production, and antioxidant buffer capacity for carbohydrate and fat substrates, respectively, at 7-dpi. Statistical significance was determined by one-way ANOVA with Tukey’s HSD post-hoc test. Data are expressed as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. AoxBC, antioxidant buffering capacity. ROS, reactive oxygen species.

Study 2 Transcriptomics: The effect of ovarian hormone loss and 17β-E2 rescue on acute transcriptomic regulation after VML

To investigate the molecular mechanisms underlying the observed redox imbalance and impaired bioenergetic recovery, we performed bulk RNA-sequencing on VML-injured muscle at 3- and 7-dpi. Given the distinct physiological phenotypes observed, we employed a two-stage analytical approach: first, comparing VML-OVX and VML-sham OVX mice to define the fundamental pathology of ovarian hormone loss, and second, comparing VML-OVX-E2 and VML-OVX mice to isolate the specific pathways through which 17β-E2 restores muscle repair.

We first characterized the broad transcriptional impact of ovarian hormone loss by identifying differentially expressed genes (DEGs) in the VML-OVX versus VML-sham OVX comparison. At 3-dpi, VML-OVX mice exhibited 90 upregulated and 73 downregulated genes compared to VML-sham OVX, a signature that contracted to only 11 total DEGs by 7-dpi. While these low DEG counts at 7-dpi might suggest a return to baseline, we hypothesized that they instead reflected a state of asynchronous signaling. To visualize the gene-level transcriptional response, we generated a heatmap of the top 50 genes from the VML-OVX vs. VML-sham OVX comparison at 3-dpi, and examined the expression of these same genes in VML-OVX-E2 mice (Supplementary Figure 1). This analysis revealed that 17β-E2 repletion substantially reversed the OVX-induced transcriptional signature, providing gene-level corroboration of the pathway enrichment findings. To move beyond individual gene changes and identify the coordinated biological programs disrupted by hormone loss, we employed HM-GSEA, Mito-GSEA, and VIPER analysis.

At 3-dpi, the transcriptomic profile of VML-OVX mice revealed a blunted induction of regenerative and metabolic programs compared to VML-sham OVX. Specifically, HM-GSEA identified a significant suppression of essential cell-cycle pathways, including E2F_Targets, Myc_Targets, and the G2M_Checkpoint (Figure 6A), which was corroborated by a marked reduction in the inferred activity of master regulators Myc and E2f4 via VIPER analysis (Figure 6E). Most strikingly, the loss of ovarian hormones resulted in a mitochondrial gene regulation stall. All significant Mito-GSEA pathways were downregulated in VML-OVX mice at 3-dpi (Figure 6B). This suppression included critical bioenergetic nodes such as oxidative phosphorylation (OxPhos_All), the TCA Cycle (Met_TCA), and mtDNA Maintenance (mt_DNA). Instead of initiating repair, the VML-OVX muscle environment shifted toward a pro-fibrotic and myeloid-dominant signature, evidenced by the upregulation of Epithelial_Mesenchymal_Transition (EMT; Figure 6A) and an increased inferred activity of the macrophage-associated transcription factor Spi1 (Figure 6E).

Figure 6.

Impact of ovarian hormone depletion on the transcriptomic landscape and mitochondrial gene networks following volumetric muscle loss (VML). All panels represent the differential enrichment and transcription factor activity of injured, ovariectomized mice compared to injured, ovary-intact sham controls (VML-OVX vs. VML-sham OVX) at 3- and 7-days post-injury (dpi). A,C: Gene set enrichment analysis (GSEA) of Molecular Signatures Database (MSigDB) Hallmark gene sets at 3-dpi and 7-dpi. B,D: GSEA of mitochondrial-specific pathways curated from the MitoCarta3.0 database at 3-dpi and 7-dpi. E,F: Virtual inference of protein-activity by enriched regulon (VIPER) analysis identifying differentially active transcription factors at 3-dpi and 7-dpi. For all analyses, data are presented as Normalized Enrichment Scores (NES). Positive values indicate pathways or transcription factors significantly upregulated following the loss of ovarian hormones; negative values indicate pathways or transcription factors downregulated by hormone depletion. Statistical significance for all panels was defined as |NES| > 1 and Benjamini-Hochberg false discovery rate (FDR) < 0.05. OVX, ovariectomy; TF, transcription factor.

By 7-dpi, the transcriptional landscape in VML-OVX mice transitioned from a paused regenerative state to one of maladaptive, asynchronous repair signaling. VML-OVX mice showed a delayed and uncoordinated increase across multiple pathways and transcription factors. In sharp contrast to 3-dpi, every significant Mito-GSEA pathway was upregulated at 7-dpi in the VML-OVX mice, representing a large mitochondrial rebound (Figure 6D). This change was driven by a sudden induction of Tfam inferred activity (Figure 6F) and was accompanied by the simultaneous upregulation of Myogenesis, Adipogenesis, and Fatty acid metabolism (FA Metabolism) (Figure 6C). However, this metabolic induction appeared to be temporally uncoupled from functional recovery. The concurrent upregulation of muscle-specific repair pathways (Myogensis) alongside markers of fatty degeneration (Adipogenesis) suggests that ovarian hormone loss results in a failure of sequential coordination, leading to a discordant and ultimately less effective regenerative process.

To determine if exogenous 17β-E2 could rescue the disrupted regenerative timeline observed in VML-OVX mice, we administered 17β-E2 and analyzed the transcriptional response at 3- and 7-dpi compared to VML-OVX. The results indicate that 17β-E2 restores the sequential time course of repair by separating the proliferative and metabolic phases. At 3-dpi, 17β-E2 administration effectively reversed the proliferative arrest seen in VML-OVX mice. HM-GSEA and VIPER analysis revealed an upregulation of E2F_Targets, Myc_Targets, and the G2M_Checkpoint (G2/M) (Figure 7A), driven by the increased inferred activity of master regulators Myc, E2f1, and E2f4 (Figure 7E).

Figure 7:

Transcriptomic regulation of cellular signaling and mitochondrial networks by 17β-estradiol following volumetric muscle loss (VML). All panels represent the differential enrichment and transcription factor activity of injured, ovariectomized mice with estradiol repletion compared to injured, ovariectomized mice without repletion (VML-OVX-E2 vs. VML-OVX) at 3- and 7-days post-injury (dpi). A,C: Gene set enrichment analysis (GSEA) of Molecular Signatures Database (MSigDB) Hallmark gene sets at 3-dpi and 7-dpi. B,D: GSEA of mitochondrial-specific pathways curated from the MitoCarta3.0 database at 3-dpi and 7-dpi. E,F: Virtual inference of protein-activity by enriched regulon (VIPER) analysis identifying differentially active transcription factors at 3-dpi and 7-dpi. For all analyses, data are presented as Normalized Enrichment Scores (NES). Positive values indicate pathways or transcription factors significantly upregulated by 17β-E2 repletion; negative values indicate pathways or transcription factors downregulated. Statistical significance for all panels was defined as |NES| > 1 and Benjamini-Hochberg false discovery rate (FDR) < 0.05. OVX, ovariectomy; TF, transcription factor.

Critically, while 17β-E2 stimulated this mitogenic rescue, it produced only a modest change in the mitochondrial gene regulation as most statistically significant Mito-GSEA pathways were downregulated in VML-OVX-E2 muscle at 3-dpi (Figure 7B). This metabolic suppression relative to VML-OVX was further reflected in the downregulation of Tfam inferred activity (Figure 7E) and Oxidative Phosphorylation (OxPhos and OxPhos_All) (Figure 7A and 7B). This suggests that 17β-E2 promotes an early regenerative strategy, prioritizing the restoration of the progenitor pool while delaying the high energy expenditure associated with mitochondrial biogenesis.

By 7-dpi, the VML-OVX-E2 tissue transitioned into a high-energy regenerative state, successfully avoiding the uncoordinated signaling observed in VML-OVX mice. 17β-E2 promoted a synergistic resolution of inflammation, evidenced by the downregulation of TNFa_Signaling_via_NFkB, IL6_Jak_Stat3_Signaling, and the Inflammatory_Response (Figure 7C). VIPER analysis confirmed this resolution through the marked inferred suppression of myeloid and pro-inflammatory transcription factors, including Spi1, Stat1, and Batf (Figure 7F).

Concurrently, 17β-E2 flipped the metabolic switch to fuel the repair state. Following the delay at 3-dpi, every significant Mito-GSEA pathway was upregulated at 7-dpi (Figure 7D). This late-phase bioenergetic activation was coordinated by a significant increase in Tfam inferred activity (Figure 7F) absent the competing adipogenic (Adipogenesis) or fibrotic (EMT) signals seen in the VML-OVX state (Figure 6C). Together, these data demonstrate that 17β-E2 rescues VML-injured muscle by enforcing a strict temporal decoupling of cell-cycle entry and mitochondrial maturation, ensuring a streamlined transition from inflammation to functional tissue reconstruction.

Study 3: The effect of ovarian hormone loss and 17β-E2 rescue on chronic bioenergetics and redox balance post-VML

While the transcriptomic data at 3- and 7-dpi established that 17β-E2 restores the temporal coordination of the early repair environment, it remained unknown whether this early organizational advantage translates into long-term functional and metabolic resilience. We therefore extended our investigation to 60-dpi to assess the chronic impact of ovarian hormone loss and 17β-E2 replacement on muscle mass and mitochondrial bioenergetics following VML. Consistent with the early failure of the regenerative niche observed in Study 2, VML-OVX mice at 60-dpi exhibited a state of chronic metabolic dysfunction. At the study endpoint, body weight was 21% greater in the VML-OVX mice compared to VML-sham OVX (Figure 8B, p = 0.001), which is 13% less in the VML-OVX-E2 group compared to the VML-OVX (Figure 8B, p = 0.002). Absolute gastrocnemius muscle mass was 15% and 10% greater in the VML-OVX-E2 mice compared to the VML-sham OVX and VML-OVX, respectively (Figure 8C, p ≤ 0.026). Body mass-normalized muscle mass was 15% lower in the VML-OVX relative to the VML-sham OVX (Figure 8D, p = 0.040); however, it was 26% higher in the VML-OVX-E2 mice (Figure 8D, p = 0.001). Uterus mass was markedly reduced in both VML-OVX groups compared to the VML-sham OVX (Figure 8A, p ≤ 0.001), which is rebounded by 17β-E2 replacement (Figure 8A, p = 0.041). Gonadal adipose tissue mass was 4.9-fold and 2.6-fold greater in both VML-OVX groups compared to the VML-sham OVX (Figure 8E, p ≤ 0.014). Between the VML-OVX groups, the VML-OVX-E2 mice have 46% lower gonadal adipose tissue mass compared to the VML-OVX mice (Figure 8E, p = 0.001). The area under the curve (AUC) from the ipGTT was greater in the VML-OVX group compared to the VML-sham OVX (Figure 8G, p = 0.006), which was reduced by 17β-E2 replacement (Figure 8G, p = 0.002). The results suggest that VML-OVX impairs whole-body metabolism after VML injury, which is reversed by 17β-E2 treatment. Body normalized peak-isometric torque of the VML-injured plantar flexors was ~20% less in VML-OVX compared to both VML-sham OVX and VML-OVX-E2 (Figure 8H, p = 0.001).

Figure 8.

Effect 17β-estradiol on morphologic adaptations 60-days post-VML injury. A-D: Uterus mass, Body mass (BM), VML-injured gastrocnemius muscle mass, and body mass-normalized muscle mass at 8-weeks post-injury. E: Gonadal fat mass. F-G: Two-hour blood glucose response to a single, fasted glucose challenge and calculated area under the curve at 8-weeks post-injury. H: Body mass normalized peak-isometric torque of the VML-injured hindlimb plantar flexors. Statistical significance was determined by one-way ANOVA with Tukey’s HSD post-hoc test. Data are expressed as mean ± SD. *p < 0.05.

Both carbohydrate- and fat-mediated maximal JO₂ in permeabilized fiber bundles were 30% and 44% lower in the VML-OVX group (Figure 9A,B,E,F, G, p ≤ 0.008); and, carbohydrate-mediated JO₂ was 49% greater in VML-OVX-E2 muscle compared to the VML-OVX muscle (Figure 9C, p < 0.001). Carbohydrate-mediated respiratory conductance was >30% greater in the VML-sham OVX and VML-OVX-E2 compared to the VML-OVX (Figure 9C, p = 0.023). For fat substrates, respiratory conductance was 41% less in the VML-OVX compared to the VML-sham OVX (Figure 9G, p = 0.023); however, the conductance was not significantly different between VML-OVX-E2 and VML-OVX (Figure 9G, p = 0.208). For Δψm, the results suggest that both VML-OVX and VML-OVX-E2 groups exhibited a less polarized mitochondrial membrane potential compared to VML-sham OVX muscle when utilizing both carbohydrate and fat substrates (Figure 9D, H, p<0.001). These results suggest that the VML-OVX-E2 group is more efficient at utilizing the proton motive force for ATP synthesis compared to VML-OVX mice and that electrons are appropriately used in complex-IV to reduce O₂ to H₂O.

Figure 9.

Effect of 17β-estradiol metabolic adaptations after VML injury. A,F: The relationship between ATP re-synthesis demand (ΔG_ATP) and mitochondrial oxygen consumption (JO₂) normalized to citrate synthase (CS) activity for carbohydrate and fat substrates, respectively, at 8-weeks post-injury. B,G: Maximal JO₂ for carbohydrate and fat substrates, respectively, at 8-weeks post-injury. C,H: Respiratory conductance for carbohydrate and fat substrates, respectively, at 8-weeks post-injury. D,I: The relationship between ΔG_ATP and mitochondrial membrane potential for carbohydrate and fat substrates, respectively, at 8-weeks post-injury. E,J: The relationship between JO₂ and mitochondrial membrane potential (i.e., bioenergetic efficiency) for carbohydrate and fat substrates, respectively, at 8-weeks post-injury. K-N: Reactive oxygen species emission, production, and antioxidant buffer capacity for carbohydrate and fat substrates, respectively, at 8-weeks post-injury. Statistical significance was determined by one-way ANOVA with Tukey’s HSD post-hoc test. Data are expressed as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. AoxBC, antioxidant buffering capacity. ROS, reactive oxygen species.

ROS emission from both carbohydrate and fat substrates was 65% and 95% greater, respectively, in the VML-OVX compared to the VML-sham OVX mice (Figure 9I,K, p ≤ 0.006). There was no statistical difference in ROS production for both substrates across group. Both carbohydrate- and fat-mediated AoxBC were 27% and 40% less, respectively, in the VML-OVX compared to the VML-sham OVX mice (Figure 9J,L, p ≤ 0.001); however, they were 40% and 61% greater in the VML-OVX-E2 relative to the VML-OVX (Figure 9J,L, p ≤ 0.001).

Discussion

VML induces a severe, irrecoverable loss of muscle mass compounded by persistent metabolic and mitochondrial dysfunction. The primary objective of this study was to define the role of 17β-E2 in modulating mitochondrial bioenergetics and redox balance following VML injury. The most significant finding of the current study is that 17β-E2 acts as a critical temporal regulator of the transcriptomic response to muscle trauma, VML specifically. 17β-E2 drives a temporal decoupling of acute cellular expansion (i.e., cell cycle) from subsequent mitochondrial biogenesis. In the absence of ovarian hormones, VML-injured muscle exhibits an early transcriptional arrest followed by an asynchronous induction of metabolic and adipo-fibrogenic gene pathways that agrees with established maladaptive pathologies of VML (4, 5, 37). 17β-E2 replacement effectively resolves this signaling, prioritizing inflammatory clearance before sequentially initiating mitochondrial gene response. This transcriptional coordination ultimately supports chronic functional resilience, restoring bioenergetic efficiency, mitigating ROS emissions, and preventing the systemic metabolic dysfunction observed in chronic VML pathology found in male mice (see (9)).

Our acute and chronic time point functional data highlight a distinct, 17β-E2-dependent physiological resilience to VML-induced oxidative stress. Prior studies establish that male mice suffer from prolonged mitochondrial suppression and elevated ROS emissions for up to 1-month. In contrast, our ovary-intact female cohort recovered endogenous AoxBC as early as 2-weeks post-injury. This resilience appears to be rooted in sex-specific management of the proton motive force. In previous studies, VML-injured male mice exhibited a persistently hyperpolarized mitochondrial membrane potential coupled with diminished respiratory capacity, a state that exacerbates electron leak and superoxide generation (9). In contrast, our bioenergetic data indicate that female mice adopt a less hyperpolarized, and eventually depolarized, mitochondrial membrane potential trajectory following injury. By limiting hyperpolarization, female mitochondria intrinsically restrict excessive electron backflow at complexes I, II, and III. 17β-E2 further capitalizes on this thermodynamic advantage by rapidly upregulating antioxidant defenses; early 17β-E2 replacement primarily bolstered AoxBC prior to directly reducing raw ROS emissions, a finding supported by literature demonstrating 17β-E2 -mediated upregulation of Gpx3 and superoxide dismutase (13, 44).

Our data suggest this resilience is tightly governed by 17β-E2 rather than representing a generalized sex-dimorphic trait. 17β-E2 is well-documented to influence skeletal muscle regeneration, satellite cell function, and structural repair following injury (45–48). By pacing the regenerative timeline, 17β-E2 lessens the trajectory of redox imbalance characteristic of male VML models. In the absence of 17β-E2, the female response to VML defaults to a "male-like" phenotype characterized by longer ROS emission and suppressed AoxBC timeline. Interestingly, recent studies utilizing 17α-estradiol, a non-feminizing enantiomer, in male mice have demonstrated profound metabolic benefits, mitigating oxidative stress and improving systemic metabolism (49, 50). This suggests that mechanistically estrogen receptor signaling operates as a potent regulator of redox recovery, dictating the specific timing of metabolic repair following severe trauma. The 17β-E2 delivery system utilized in this study (continuous-release pellet, 0.18 mg over 60 days) produces circulating estrogen levels comparable to those of ovary-intact female mice, as previously validated by LC-MS/MS (51). This suggests that the benefits observed herein reflect physiological estrogenic signaling rather than supraphysiological E2 concentration. Nevertheless, the current study was not designed to resolve a full dose-response relationship, and future investigations systematically comparing various E2 concentrations will be necessary to determine whether a U-shaped relationship exists in the context of VML recovery.

This 17β-E2-mediated regulation of the regenerative and metabolic transcriptome complements and may directly impact the functional efficiency of the recovering mitochondrial network. A defining feature of VML is the immediate and severe disruption of the bioenergetic grid. Previous work demonstrated a temporal breakdown of the mitochondrial network following VML (7), an observation externally validated by retrospective analyses of independent RNA-sequencing datasets revealing widespread acute suppression of mitochondrial gene networks in the injured muscle (37). Our transcriptomic data aligns with this acute bioenergetic stall at 3-dpi; however, our findings demonstrate that simply re-initiating these networks does not equate to physiological recovery. In mice with ovarian hormone loss, an uncoordinated increase in mitochondrial gene expression at 7-dpi fails to restore bioenergetic efficiency, resulting instead in greater ROS emissions and sustained hyperpolarization. By delaying mitochondrial biogenesis until the early proliferative phase (Myc, E2f4) is established, 17β-E2 may help ensure that the mitochondria constructed during the later stages of repair are highly coupled and efficiently utilize the proton motive force for ATP synthesis. This transcriptomic pacing provides the molecular mechanism for the findings of Bruzina et al. (Manuscript Accepted at AJP-Cell C-00251-2026R1), who demonstrated that while total mitochondrial content (citrate synthase activity) remains unchanged across intact and ovarian hormone-deficient groups at 12-weeks post-VML, it is the functional capacity of the respiratory chain (complex I and pyruvate dehydrogenase) that is critically protected by ovarian hormones.

Furthermore, this strict temporal coordination alters the cellular trajectory within the VML-injured wound environment. Successful skeletal muscle regeneration relies on the highly coordinated activity of stem cell populations, particularly satellite cells and fibro-adipogenic progenitors (FAPs). In severe trauma models like VML, the regenerative niche is often overwhelmed, driving FAPs toward maladaptive adipogenic and fibrotic differentiation (52, 53). Although bulk transcriptomics do not isolate specific cell populations, the coordinated upregulation of EMT and adipogenesis signatures in ovarian hormone-deficient muscle serves as a bulk-level proxy for a cellular environment responding to disparate signaling strategies. In this context, 17β-E2 acts as a critical fate-diverting signal. By maintaining progenitor populations in an active, proliferative state long enough for the inflammatory response to resolve, 17β-E2 subsequently orchestrates a coordinated transition to myogenesis without competing fibrotic or adipogenic signals. This 17β-E2-mediated diversion of progenitor fate provides a direct cellular explanation for the parallel findings of Bruzina et al. (Manuscript Accepted at AJP-Cell C-00251-2026R1), who reported that intact females are protected from the ectopic lipid accumulation and elevated perilipin expression that physically impairs contractile torque in ovarian hormone-deficient mice at chronic time points.

Ultimately, the failure to resolve this localized tissue trauma precipitates profound systemic metabolic inflexibility. VML is known to provoke an exacerbated and persistent inflammatory response (5) and acute and chronic local lipid dysregulation (21, 53). We observed this pathology natively at 7-dpi, where ovarian hormone deficient tissue exhibited an asynchronous upregulation of fatty acid metabolism concurrent with persistent inflammation. Without the regulatory pacing of 17β-E2, the local microenvironment defaults to a chronic lipid-sequestering and inflammatory state. We propose that this acute local metabolic failure acts as the primary catalyst for broader systemic decline. By 60-dpi, this unresolved local signaling translated into significant whole-body metabolic dysfunction, including impaired glucose tolerance and elevated gonadal adiposity, a finding corroborated by Bruzina et al. (Manuscript Accepted at AJP-Cell C-00251-2026R1) at 3-months post-injury. Re-introducing 17β-E2 extinguished this chaotic local signaling, restoring local redox balance and subsequently protecting against systemic metabolic disease.

Beyond the acute transcriptomic pacing, 17β-E2 orchestrates long-term, substrate-specific bioenergetic adaptations. Consistent with prior work, we observed that 17β-E2 improves overall respiratory capacity following OVX (11, 55). However, our data uniquely reveal that 17β-E2 influences respiratory conductance in a highly substrate-dependent manner, profoundly rescuing fat-fueled, but not carbohydrate-fueled, conductance at 60-dpi. This finding contrasts with earlier reports by Torres et al., which demonstrated no effect of 17β-E2 on fatty acid-derived mitochondrial respiration in permeabilized fibers from uninjured ovarian hormone-deficient animals (11). We posit that this discrepancy highlights a trauma-specific adaptation. Severe trauma like VML drastically alters the local lipid microenvironment. Our transcriptomic data indicate that 17β-E2-repleted mice exhibit a highly coordinated upregulation of fatty acid metabolism genes compared to the uncoordinated signaling in ovarian hormone-deficient mice. Over 2-months, this 17β-E2-directed transcriptional programming likely optimizes the mitochondrial machinery to efficiently oxidize the ectopic lipids accumulating in the wound bed, thereby restoring fat-fueled respiratory conductance and preventing localized lipotoxicity.

In conclusion, this study establishes 17β-E2 as an essential regulator of mitochondrial function, redox homeostasis, and systemic metabolic resilience following VML injury. By strictly coordinating the temporal phases of cellular expansion and mitochondrial maturation, 17β-E2 mitigates maladaptive repair signaling and preserves bioenergetic efficiency. Furthermore, by characterizing the female-specific response to severe muscle trauma, this research directly addresses the critical need to integrate Sex as a Biological Variable (SABV) into the VML landscape, a field that has historically favored single sex-focused models. These findings underscore the critical role of endocrine and sex hormone signaling in the resolution of severe musculoskeletal trauma and highlight estrogen receptor pathways as a potential therapeutic target for mitigating the chronic functional and metabolic deficits associated with VML.

Supplementary Material

Supplemental Figs. S1: https://doi.org/10.6084/m9.figshare.32537550

Data availability:

The datasets used and/or analyzed during the current study are primarily presented in the current manuscript and are available from the corresponding author on request.