Limitations in metabolic plasticity after traumatic injury are only moderately exacerbated by physical activity restriction

Ethical approval and experimental design

All protocols were approved by the Institutional Animal Care and Use Committee at the University of Minnesota (#2110-39493A), in compliance with the Animal Welfare Act, the Implementing Animal Welfare Regulations and in accordance with the principles of the Guide for the Care and Use of Laboratory Animals. Adult male C57Bl/6 J mice (n = 40; 8–12 per group) were purchased from Jackson Laboratories (Stock #000664; Bar Harbor, ME). Mice were given at least a one-week acclimation period prior to study initiation. Mice were housed on a 12-h light-dark cycle (light phase begins at 06:00) with ad libitum access to chow (LabDiet #5053, Land O’ Lakes, Inc.) and water.

At ~13 weeks of age, mice were randomized to VML injury or served as age-matched injury naïve controls. Mice were further randomized to standard or restricted cages^16,17. As we have previously used, restricted cages (12.5 × 8.5 × 6.3 cm) reduce physical activity by ~50% and alter whole-body metabolism within the first week following restriction, compared to standard cages (28 x 18 x 12.5 cm)^16,17. At 6 weeks following VML, whole-body metabolic and physical activity measurements were evaluated for all mice using the Comprehensive Lab Animal Monitoring System (CLAMS; Columbus Instruments). Glucose tolerance was evaluated at 7 weeks following VML or study start. Terminally 8 weeks post-VML ( ~ 21 weeks of age) in vivo muscle function was assessed, and skeletal muscles, liver, and serum were subsequently extracted. Gastrocnemius muscles were weighed, cut into thirds (proximal, mid, distal sections), frozen in liquid nitrogen, and stored at −80^oC for later analyses. A subset of the mid-portion of the gastrocnemius muscle, encompassing the VML defect, was saved in OCT, frozen in isopentane cooled with liquid nitrogen, and stored at −80^oC for histological evaluation. A portion of liver tissue was weighed, frozen in liquid nitrogen, and stored at −80^oC. Blood was clotted and centrifuged at 2000 rpm for 15 min at 4°C, serum was collected and stored at −20^oC. Mice were euthanized with pentobarbital ( > 100 mg/kg; s.q.).

Volumetric muscle loss (VML) surgical procedure

As described previously^19,47, a full thickness VML injury was surgically created to the muscles of the posterior hindlimb compartment (gastrocnemius, soleus, plantaris muscles). Mice received buprenorphine SR (1.2 mg/kg; s.q.) approximately 2 h prior to surgery for pain management. Mice were anesthetized by isoflurane inhalation ( ~ 2.0%) under aseptic surgical conditions. Briefly, a posterior-lateral incision was created through the skin to reveal the gastrocnemius muscle. Blunt dissection isolated the posterior muscle compartment, and a metal plate was inserted between the tibia and the deep aspect of the soleus. A 4-mm punch biopsy (19.1 ± 1.6 mg, ~15% volume loss of muscle) was performed on the middle third of the muscle compartment. Any bleeding was stopped with light pressure. Skin incisions were closed with 6–0 PGA suture and animals were monitored through recovery.

Evaluation of physical activity and whole-body metabolism

Physical activity and whole-body metabolic assessments were conducted as previously described^17,18 using the CLAMS system (Columbus Instruments) and data examination tool (Clax, v2.2.15; Columbus Instruments, Columbus, OH, USA). Physical activity data were collected over 10-sec increments and metabolic data were collected in 10-min intervals over 24 h and processed. MATLAB (version R2020a, MathWorks, Natick, MA, USA) was used to calculate metabolic rate, RER, and carbohydrate and lipid oxidation rate averages over 24-h and 12-h active and inactive periods, and to evaluate 24-h moving averages for RER. Area under the curve (AUC) for RER was calculated over 24-h and 12-h active and inactive periods. Delta RER (Δ RER) was determined as the difference between average RER across the 12-h active and 12-h inactive periods. Mice were given a 24-h acclimation period prior to the data collection period.

Glucose tolerance test

As previously described¹⁷, at 7 weeks following VML (one week prior to harvest), glucose tolerance testing was performed after a 6-h fast. Baseline blood glucose was obtained from the lateral tail vein, nicked with a 20-G needle, with a glucometer (Freestyle Lite, Abbott). Following injection of D-glucose saline solution (Sigma #G7021; 2 mg/g, i.p.), glucose measurements were obtained at 15−, 30−, 45−, 60−, 90−, and 120-min following injection. The glucometer measured a range of glucose values up to 500 mg/dl; readings above this range were recorded, plotted, and analyzed as 500 mg/dl. Mice were continuously monitored, and any additional bleeding was stopped with light pressure. Following testing, mice were returned to home cages with ad libitum access to food and water.

In vivo muscle function

Muscle function of the posterior compartment was evaluated 8 weeks following VML or study start (naïve) as previously described^19,47. Briefly, mice were anesthetized using inhaled isoflurane (1.5–2.0%) and body temperature was maintained at 37^oC. Mice were subsequently positioned on the right side with the left foot attached to the footplate of the dual‐mode muscle lever system (300C-LR; Aurora Scientific, Aurora, Ontario, Canada). The knee and hip were stabilized at 90^o. First, the common peroneal nerve was severed to isolate stimulation to the posterior compartment. Maximal isometric torque was evaluated by stimulating the sciatic nerve using Platinum-Iridium percutaneous needle electrodes. Torque frequency was conducted (5, 10, 20, 40, 60, 80, 100, 150, and 200 Hz) and expressed as mN·m per kg body weight.

Histological analyses

Histological evaluation was performed in the mid-belly of the gastrocnemius muscle. Ten-µm serial cross-sections were obtained using a Leica cryostat and microtome, and subsequently stained for hematoxylin and eosin to evaluate myofiber number and morphology and NADH-tetrazolium reductase to assess percentage of oxidative myofibers.

The NADH-TR staining was performed as previously described by incubating tissues at 37°C for 20 min in a solution containing 0.2 M Tris, 1.5 mM NADH, and 1.5 mM nitro blue tetrazolium, as previously described¹⁸. Sections were washed, dehydrated, and cleared in xylenes. Brightfield images were acquired using the TissueScope LE slide scanner (Huron Digital Pathology, St. Jacobs, ON, Canada) using a 20X objective (0.75 NA, 0.5 µm/pixel resolution). Following imaging, HuronViewer (Huron Digital Pathology) was used to export three standardized non-overlapping regions of interest (ROI) for each muscle. All ROIs were standardized to 1000 × 1000 µm. Regions encompassed the VML defect, the border of the defect which included the defect and remaining muscle, and the remaining muscle tissue. The corresponding three ROI areas were also obtained in muscles of the injury naïve groups. All ROIs were systematically evaluated in NADH-stained muscles. Data from these ROIs were averaged across the gastrocnemius muscle and compared across groups.

Analyses of muscle sections were conducted using Fiji⁴⁸. The multipoint tool was used to manually count myofiber numbers and types, in addition to the number of darkly and lightly stained muscle fibers for NADH-stained muscle. Darkly stained fibers were classified as positive for NADH, representing high metabolic activity, while lightly stained fibers were classified as negative for NADH, representing low metabolic activity. Investigators were blinded during all imaging and post-imaging analyses.

Biochemical analyses

The proximal gastrocnemius muscle, liver, and serum were used in biochemical analyses. Gastrocnemius muscles and liver were homogenized in 10 mM phosphate buffer at a ratio of 1:10 (mg/µl) with the addition of 1:100 protease and phosphatase inhibitors. Total protein content was analyzed using the Protein A280 setting on a NanoDrop One spectrophotometer (Thermo Scientific) in triplicate and averaged.

Muscle adiponectin concentrations were assessed in duplicate (Milliplex #MADPNMAG-70K-01); plates were read using a Bio-Plex 200 system (Bio-Rad Laboratories; Hercules, CA, USA). Prepared standards were validated within an acceptable recovery range of 70–130% (observed concentration/expected concentration). Serum leptin concentrations were evaluated by ELISA (R&D Systems; #MOB00B), according to the manufacturer’s protocol. Values were excluded if the observed concentration was below the minimum detectable concentration of 2.58 pg/ml.

Immunoblot analyses of muscle and liver were performed by separating 20–50 µg of protein by 4–15% Criterion TGX Stain-Free Gel (Bio-Rad), transferring protein onto a PVDF membrane, and immunoblotting. Primary antibodies were used to probe for acetyl-CoA carboxylase (Cell Signaling #3676; Lot #12; RRID: AB2219397; 1:000), phospho-acetyl-CoA carboxylase (Cell Signaling #3661; Lot #10; RRID:AB330337; 1:1000), fatty acid synthase (Cell Signaling #C20G5, Lot #7; RRID:AB_2100796; 1:1000), FABP4 (Abcam #ab92501; Lot#1006255-21; RRID:AB_10562486; 1:1000) and GLUT4 (Millipore Sigma #07-1404; Lot #3829399; RRID:AB1587080; 1:2000). Antibodies were detected using a corresponding host- and isotype-specific horseradish peroxidase conjugated secondary antibody (Cell Signaling #7074, 1:1000). Immunoblots were blocked with 5% nonfat milk, and primary and secondary antibodies were diluted in 5% nonfat milk, followed by incubation in Clarity Max ECL Western Blotting Substrate for protein detection (Bio-Rad). Immunoblots were visualized with stain-free and chemiluminescent imaging on a ChemiDoc System (Bio-Rad) for total lane protein and band intensity quantification, respectively⁴⁹. The intensity of each band was normalized to total protein in each respective lane using Bio-Rad Laboratories Image Lab software (Hercules, CA). The gastrocnemius muscle of injury naïve mice was used for relative comparison across experimental groups on all immunoblots, full images of all Western Blots is included in Supplementary Fig. 1.

The muscle was used to evaluate mitochondrial content^15,19 and β-hydroxyacyl-CoA dehydrogenase (β-HAD)^21,50, as previously described. Briefly, the muscle homogenate was further diluted with10 mM phosphate buffer to a final ratio of 1:40 (mg/µl). Mitochondrial content was determined by citrate synthase (CS) activity by evaluating the reduction of 5,5’-dithio-bis(2-nitrobenzoic) acid (DTNB) over time, at 412 nm evaluated using a spectrophotometer. The activity of β-HAD was determined by incubating muscle homogenate in a buffer containing 100 mM triethanolamine, 451 µM β-nicotinamide adenine dinucleotide and 5 mM ethylenediaminetetraacetic acid (EDTA). The enzyme activity of β-HAD was normalized to mitochondrial content (i.e., CS enzyme activity).

Measurement of tissue metabolites

Metabolomics measurements were obtained using the MxP Quant 500 kit (Biocrates Life Sciences AG, Innsbruck, Austria) following the manufacturer’s protocol for tissue samples¹⁷. Distal gastrocnemius muscles ( ~ 50 mg) were placed in 2 mL Precellys CKMix lysis tubes (Bertin Corp., Rockville, MD, USA) and diluted threefold with cold isopropanol. Samples were homogenized using a Precellys bead beater set to 4°C three times for 30 s at 5800 rpm, 30 s pause between pulses. Samples were centrifuged for five min at 10,000 x g and the supernatant was transferred to a new vial. Samples were analyzed in two batches using a 56-well and 96-well based sample preparation plate, respectively. The 56-well plate was analyzed using a Sciex QTRAP 5500 mass spectrometer (Sciex, Framingham, MA, USA) coupled to a Shimadzu LC-20AD XR (Shimadzu USA Manufacturing Inc., Canby, OR, USA) liquid chromatography platform. The 96-well plate was analyzed using an Agilent 6495 C triple quadrupole platform coupled with an Agilent Infinity II liquid chromatography system (Agilent, Santa Clara, CA, USA). Small molecule metabolites were measured by liquid chromatography-tandem mass spectrometry (LC-MS/MS). Lipids and hexoses were measured by flow injection analysis-tandem mass spectrometry (FIA-MS/MS). The metabolomics measurement technique is described in detail by patents EP1875401B1⁵¹ and EP1897014B1⁵². Data was processed using the WebIDQ online software (Biocrates Life Sciences AG, Innsbruck, Austria). Metabolite concentrations were adjusted using a tissue correction factor and between-plate concentrations were normalized to an internal quality control sample.

Statistical analyses

All statistical analyses were conducted using JMP (version 16.0.0, SAS Institute, Inc.) and R programming language⁵³. Two-way ANOVA with Tukey’s honest significant (HSD) difference post hoc test was used to evaluate differences across injury and physical activity status groups for whole-body metabolism (24- and 12-h metabolic rate, RER, lipid oxidation, carbohydrate oxidation) and ambulatory distance, in vivo twitch and maximal isometric torque and corresponding contractile curve analyses (e.g., average rates of contraction and relaxation, half relaxation time, time to peak), terminal body and muscle masses, protein expression, enzyme activity and signaling markers in muscle, serum, and/or liver. The RER AUC was compared by one-way ANOVA with Tukey’s HSD post hoc test. Three-way ANOVA with Tukey’s HSD post hoc test evaluated differences across injury status, physical activity status, and time during the glucose tolerance test. All data is shown as mean ± SD unless otherwise noted. Significance was set a priori at p < 0.05 all exact statistical outcomes are presented in Supplementary Table 1.

For metabolomics analyses, metabolite measurements with greater than 20% missing values among the sample cohort were excluded from the analysis. Missing values were imputed using a K-Nearest Neighbors (k = 10) approach using the impute R package⁵⁴. Metabolomic concentrations were Log transformed to improve the normality assumption for statistical modeling. Statistical differences between injury-movement systems were analyzed using a generalized linear model approach controlling for batch-wise variation. Model effect estimates and statistical p-values were estimated for each metabolite independently. A Benjamini-Hochberg p-value adjustment was applied to control for multiple comparisons⁵⁵.