Pharmacological inhibition of TLR4 ameliorates muscle and liver ceramide content after disuse in previously physically active mice

Abstract

Toll-like receptor 4 (TLR4) is a proposed mediator of ceramide accumulation, muscle atrophy, and insulin resistance in skeletal muscle. It is currently unknown whether pharmacological inhibition of TLR4, using the TLR4-specific inhibitor TAK-242 during muscle disuse, is able to prevent changes in intracellular ceramide species and consequently preserve muscle size and insulin sensitivity in physically active mice. To address this question, we subjected running wheel-conditioned C57BL/6 male mice (13 wk old; ∼10/group) to 7 days of hindlimb suspension (HS), 7 days of continued wheel running (WR), or daily injections of TAK-242 during HS (HS + TAK242) for 7 days. We measured hindlimb muscle morphology, intramuscular and liver ceramide content, HOMA-IR, mRNA proxies of ceramide turnover and lipid trafficking, and muscle fatty acid and glycerolipid content. As a result, soleus and liver ceramide abundance was greater (P < 0.05) in HS vs. WR but was reduced with TLR4 inhibition (HS + TAK-242 vs. HS). Muscle mass declined (P < 0.01) with HS (vs. WR), but TLR4 inhibition did not prevent this loss (soleus: P = 0.08; HS vs. HS + TAK-242). HOMA-IR was impaired (P < 0.01) in HS versus WR mice, but only fasting blood glucose was reduced with TLR4 inhibition (HS + TAK-242 vs HS, P < 0.05). Robust decreases in muscle Spt2 and Cd36 mRNA and muscle lipidomic trafficking may partially explain reductions in ceramides with TLR4 inhibition. In conclusion, pharmacological TLR4 inhibition in wheel-conditioned mice prevented ceramide accumulation during the early phase of hindlimb suspension (7 days) but had little effect on muscle size and insulin sensitivity.

INTRODUCTION

Brief periods of physical inactivity (e.g., 5–14 days) are often accompanied by rapid muscle loss and metabolic dysfunction (2, 14, 22, 27, 28, 36). Recurring inactivity periods may result in the development or further complications of physical inactivity-related diseases (e.g., CVD, T2DM) and contribute to early mortality (7). A furthered understanding of the mechanisms responsible for mediating disuse-related muscle loss and metabolic dysfunction is critical for developing preventative treatments that may confer protection from disuse events.

The mechanisms responsible for inactivity-related muscle atrophy and metabolic dysfunction remain incompletely understood. However, the accumulation of bioactive sphingolipids (i.e., ceramides) driven by muscle-specific inflammation during disuse, has emerged as one possibility (19, 21). An overabundance of ceramide within muscle is hypothesized to disrupt metabolic function by interfering with insulin signaling, thereby contributing to insulin resistance and impairing muscle growth (for review see Ref. 9). Using muscle disuse models of short-term bed rest in humans and hindlimb suspension in mice (14, 22), muscle loss and metabolic dysfunction are often accompanied by ceramide accumulation (21, 22), possibly by activation of the pro-inflammatory pathway Toll-like receptor 4 (TLR4). Holland et al. (19) previously demonstrated a link between TLR4 signaling and ceramide synthesis as a result of high-fat feeding in mice, yet this connection is less understood in the context of muscle disuse. We recently have shown modest muscle and metabolic protection after 2 wk hindlimb suspension in mice absent of TLR4 (21). However, use of a pharmacological approach to block TLR4 signaling during the early phase of muscle loss (7 vs. 14 days of hindlimb suspension) (5) has not been investigated, nor has evaluation of the possible multiorgan (e.g., liver) effects of the inhibitor on ceramides during disuse. Taken together, it stands to reason that pharmacological delivery of a TLR4 inhibitor during a period of muscle disuse may be an effective strategy to mitigate muscle loss and metabolic dysfunction, particularly by targeting the accumulation of muscle and liver ceramides.

Regular physical activity is regarded as an important criterion to lessen future disease and disability (7). As such, several laboratories, including our own, have conducted muscle disuse studies (i.e., hindlimb suspension) in sedentary rodents and made comparisons with sedentary control animals. However, it has become increasingly apparent that this methodology is likely problematic, as mice in regular housing environments are robustly inactive and likely already under sedentary stress, thereby artificially deviating from the disuse stimulus that occurs in human inactivity (21, 22). To circumvent this limitation, we implemented a research design comparing physically active mice (i.e., regular wheel running activity) to wheel run-conditioned mice subjected to a sudden cessation of muscle disuse (i.e., hindlimb suspension). This allowed us to observe under two different “control” conditions, allowing us to better interpret the implication of TLR4 inhibition on disuse and improve the translational ability of research to a clinical setting.

Therefore, the purpose of this study was to evaluate the efficacy of pharmacological inhibition of TLR4 signaling during hindlimb suspension in previously wheel run-conditioned mice in preventing muscle and liver ceramide accumulation, muscle loss, and metabolic dysfunction. We hypothesized that hindlimb suspension following wheel run conditioning will result in muscle and liver ceramide accumulation, muscle loss, and metabolic dysfunction and that these negative consequences will be mitigated by TLR4 inhibition.

METHODS

Animals.

Male, seven-week old, C57BL/6J wild-type mice (Jackson Laboratories) were housed in a conventional animal house, maintained on a 12:12-h light-dark cycle and temperature-controlled environment (22–23°C). Before any experimentation, mice were acclimatized to the animal facility for one week. All experimental procedures were conducted in accordance with the guidelines set by The University of Utah IACUC (protocol no. 16–11003).

Wheel running protocol.

After being acclimatized to an animal facility, mice were randomly divided into two groups (Fig. 1). One group of mice were singly housed and provided access to a running wheel for 5 wk (n = 30), whereas a group of sedentary mice (Sed) were positioned on the same cage rack, with no access to a running wheel for the duration of the study (n = 10). Running wheel design was as follows: standard mouse cages (31 cm length × 19 cm width × 15 cm height) were modified to allow a metal running wheel (11.5 cm in diameter). Wheel counts were determined by the movement of a magnet adhered to the outside of the running wheel, which was registered by a Reed switch LCD counter (Starr Life Sciences, Oakmont, PA). Wheel counts were transmitted to a wired counting device positioned outside the animal cage and were recorded regularly (1- to 3-day intervals). Running wheels were inspected regularly to ensure that the wheels were spinning smoothly and not clogged with cage litter.

Fig. 1.

Experimental design. See figure for abbreviation definitions.

Hindlimb suspension protocol and serum and tissue collection.

After 5 wk of voluntary wheel running, animals with access to running wheels were allocated into three separate groups: wheel run (WR; n = 10), 7 days of hindlimb suspension (HS; n = 10), or 7 days of HS combined with sterile intravenous (iv) injections [100-µl dose, extraorbital (39)] of TAK-242 (4 mg/kg, cat. no. 614316; Millipore Sigma; inhibitor specific to TLR4 signaling (24)] (HS + TAK-242, n = 10). This dose is effective to inhibit TLR4 signaling in mice (15, 16, 38). Additional cohorts of mice were used for a follow-up gene expression and lipidomic analysis in soleus [WR (n = 5), HS (n = 7), and HS + TAK-242 (n = 8)]. During the hindlimb suspension period, WR and HS animals were given daily intravenous injections of sterile 1× phosphate-buffered saline (PBS) as a control. All animals had ad libitum access to food (standard chow) and water throughout the course of the study. HS and HS + TAK-242 animals underwent hindlimb suspension (2 animals/cage), using a modified unloading method based on our previously published studies (21, 22). Briefly, animals were anesthetized and fitted with a piece of elastic tape. The tape was then glued to their tail, and this tape was then affixed to the ceiling of the cage at a height that suspended the animals’ hindlimbs from the cage bottom (30° between cage bottom and animal’s body). HS and HS + TAK242 animals had access to a 360° perimeter within the cage and were able to reach food and water. All mice were monitored at least once daily for behavior and for food and water intake. Body weights were monitored daily to ensure that mice were not experiencing excessive weight loss due to malnutrition or dehydration, and there was no evident weight loss in the HS and HS + TAK-242 animals.

On day 6 of HS, all animals were fasted for 5 h for a 2-h glucose tolerance test [GTT; intraperitoneal (ip) injection of glucose (1 mg/g body wt)]. During this period, WR animals had their wheels locked (zip-tied) to prevent transient physical activity effects. Blood glucose levels (Bayer Contour) were measured via tail vein at time intervals of 0, 15, 30, 60, and 120 min. On day 7 of HS, animals were fasted for 5 h, and again WR animals had their wheels locked. Thirty minutes before tissue harvest, animals were administered an intraperitoneal injection of insulin (0.75 U/kg body wt; Novo Nordisk, Novolin) in order for tissue to be collected under insulin-stimulated conditions. Liver-specific data was collected from a separate subset of mice that were not insulin stimulated before dissection (WR n = 7, HS n = 5, HS + TAK-242 n = 8). Before tissue harvest, animals were euthanized by anesthetization via isoflurane followed by cervical dislocation. Afterward, the hindlimb muscles [soleus and gastrocnemius (gastroc)] and livers were rapidly dissected, weighed, frozen in liquid nitrogen, and stored in −80°C for subsequent analysis. One gastroc from each animal was mounted on a foil-covered cork and frozen in liquid nitrogen-cooled isopentane for histological analysis. Whole blood was collected via heart puncture, and 100–500 μL of blood was centrifuged at 10,000 rpm to collect serum.

Insulin and triglycerides.

Serum insulin was measured (Insulin ELISA Kit; Crystal Chem, Chicago, IL) in animals that did not receive an insulin injection before tissue harvest. Serum levels of triglycerides were measured using Infinity Triglycerides Liquid Stable Reagent from ThermoScientific (cat. no. TR22421) according to the manufacturer specifications. Briefly, plasma samples were collected then stored at −80°C. Two microliters of serum were quantified per mouse. Samples were incubated for 5 min at 37°C in the reagent and then analyzed at 500 and 660 nM wavelength. Sample triglyceride concentrations were calculated from a standard curve.

Physical function testing.

Physical function testing was performed 1 h before tissue collection. Grip strength was quantified by measuring maximal fore/hindlimb tension (kg force/body wt) and hang latency duration (seconds) (13a, 23). Animals performed three repetitions of each of the three performance tests, with each test separated by 10 min of rest. Fore/hindlimb peak strength was measured by a force transducer (Grip Strength Meter, no. 160163; Columbus Instruments, Columbus, OH), whereby an investigator allowed the animals to reflexively grasp the force transducer, applied even tension to the animals’ tail parallel to the transducer, and recorded the maximal force produced until the animals’ grip released. Hindlimb strength was measured similarly, with the exception that the animals were provided a 1-mm hex key to grasp with their forelimbs while their hindlimbs gripped the force transducer. Hanging grip strength measurements were acquired by placing the animal onto a piece of sheet metal comprised of perforations separated by 1.5-mm of sheet metal. Once the animal was positioned on the metal, the apparatus was turned over (∼220°/s) and positioned 21 cm above a padded container, at which hanging duration time began and then concluded at the moment the mouse fell into the padded container.

Body composition.

Tissue composition was assayed with a Bruker (Rheinstetten, Germany) Minispec MQ20 NMR analyzer using the manufacturer’s guidelines.

Liquid chromatography-mass spectrometric sphingolipid quantification.

Lipid extracts were separated on a Waters Acquity UPLC CSH C18 1.7 µm 2.1 × 100 mm column maintained at 65°C and connected to an Agilent HiP 1290 Sampler, Agilent 1290 Infinity pump, equipped with an Agilent 1290 Flex Cube and Agilent 6545 Accurate Mass Q-TOF dual AJS ESI mass spectrometer. For positive mode, the source gas temperature was set to 250°C, with a gas flow of 12 L/min, nebulizer pressure of 35 psig, sheath gas temperature of 325°C, and sheath gas flow of 11 L/min. VCap voltage was set at 3,500 V, nozzle voltage was 250 V, fragmentor was 100 V, skimmer was set at 65 V, and Octopole RF peak was set at 750 V. For negative mode, the source settings were the, same except that the VCap was 3,000 V and nebulizer pressure was 30 psig. Samples were analyzed in a randomized order in both positive and negative ionization modes in separate experiments acquired with the scan range m/z 100–1,700. Mobile phase A consisted of ACN:H₂O (60:40 vol/vol) in 10 mM ammonium formate and 0.1% formic acid, and mobile phase B consisted of IPA:ACN:H₂O (90:9:1 vol/vol) in 10 mM ammonium formate and 0.1% formic acid. The chromatography gradient for both positive and negative modes started at 15% mobile phase B and then increased to 30% mobile phase B over 2.4 min; it then increaseed to 48% mobile phase B from 2.4 to 3.0 min and then increased to 82% mobile phase B from 3 to 13.2 min; then it increased to 99% mobile phase B from 13.2 to 13.8 min, where it was held until 15.4 min and then returned to the initial conditioned and equilibrated for 4 min. Flow was 0.5 mL/min throughout, and injection volume was 1 µL for positive and 5 µL for negative mode. Tandem mass spectrometry was conducted using the same liquid chromatography (LC) gradient at collision energy of 25 V. Lipid analyses were performed whereby quality control (QC) samples (n = 8) and blanks (n = 4) were injected throughout the sample queue and ensured the reliability of acquired lipidomics data. Results from liquid chromatography-mass spectrometry (LC-MS) experiments were collected using a Agilent Mass Hunter (MH) Workstation and analyzed using the software packages MH Qual, MH Quant, and Lipid Annotator (Agilent Technologies, Inc.) to prepare the data set. The data table exported from MHQuant was evaluated using Excel, where initial lipid targets were parsed based on the following criteria. Only lipids with relative standard deviations (RSD) of <30% in QC samples and were used for data analysis. Additionally, only lipids with background area under the curve (AUC) counts in process blanks that are <30% of QC were used for data analysis. The parsed excel data was normalized to sum. Lipids were quantitated based on peak area ratios relative to internal standards (C17 ceramide d18:1/17:0; Avanti Polar Lipids, Alabaster, AL; SPLASH LIPIDOMIX no. 330707; Avanti Polar Lipids, Alabaster, AL; and n-Heneicosanoic acid no. R-420210; Supleco, Bellefonte, PA).

Immunoblotting.

The relative abundance of proteins associated with insulin and TLR4 signaling and mitochondrial content was determined via immunoblotting, as we have done previously (21, 22). Briefly, soleus or ∼30 mg of red/oxidative gastroc and liver tissues was homogenized 1:10 (tissue weight/volume) using a glass tube and mechanically driven pestle grinder in an ice-cold buffer containing 50 mM Tris (pH 7.5), 250 mM mannitol, 40 mM NaF, 5 mM pyrophosphate, 1 mM EDTA, 1 mM EGTA, and 1% Triton X-100 with a protease inhibitor cocktail. Homogenates were centrifuged at 10,000 rpm for 10 min at 4°C. After centrifugation, the supernatant was collected and the protein concentration determined using the Bradford technique. Proteins from the supernatant fraction were separated via polyacrylamide gel electrophoresis, transferred onto a polyvinylidene difluoride membrane (PVDF), and incubated with primary and secondary antibodies directed against the proteins of interest. Membranes were exposed on a ChemiDoc XRS (Bio-Rad) and quantified with Image Laboratory software (Bio-Rad). Protein lysates were Western blotted for phospho-Akt (Akt Ser⁴⁷³, no. 9271; Cell Signaling Technology, Danvers, MA), total Akt (no. 9272; Cell Signaling Technology), phospho-Akt substrate 160 (AS160; Ser⁵⁸⁸, no. 8730; Cell Signaling Technology), total nuclear factor of kappa light polypeptide gene enhancer in B cells inhibitor-α (IΚΒα, no. 9242; Cell Signaling Technology), total oxidative phosphorylation protein complexes (OXPHOS, no. ab110413; Abcam, Cambridge, MA), and total glyceraldehyde 3-phosphate dehydrogenase (GAPDH, no. 2118, Cell Signaling Technology). Secondary antibodies were purchased from Cell Signaling Technology. Phospho-Akt was normalized to total Akt, AS160 Ser⁵⁸⁸ and IΚΒα were normalized to GAPDH, and OXPHOS protein was normalized to an unchanged band of protein from a Ponceau stain. An internal control was also included on each gel to normalize protein levels, to prevent gel-to-gel variability. We were not able to detect phosphorylated AS160 in the liver due to methodological issues with the AS160 Ser⁵⁸⁸ antibody.

Gene expression.

Expression of genes associated with sphingolipid synthesis and metabolism was measured to aid in the interpretation of LC-MS results. RNA was isolated from soleus, red/oxidative gastroc, and liver tissues, as we have reported previously (22, 25). Briefly, total RNA was DNase treated, and cDNA was synthesized using commercially available kits (DNA-free; Ambion, Austin, TX, and iScript; Bio-Rad, Hercules, CA). Real-time quantitiative PCR was carried out with a CFX Connect real-time PCR cycler (Bio-Rad) combined with SYBR Green fluorescence. Cycle threshold values were normalized to the geometric mean of GAPDH and 36B4, and then fold change values (vs. Sed gene expression levels) were calculated using the $2^{{}^{–\Delta \Delta C_T}}$ method. Primers were obtained from Bio-Rad for desaturase-1 (Degs1; qMmuCID0020427), serine palmitoyltransferase-2 (Spt2; qMmuCID0017980), ceramide synthase-2 (Cers2; qMmuCED0044992), sphingomyelinase-1 (Smpd1; qMmuCED0048806) and -2 (Smpd2; qMmuCED0004019), acid ceramidase-1 (Asah1; qMmuCID0009385), sphingosine kinase-2 (Sphk2; qMmuCED0039969), sphingosine-1-phosphate phosphatase 1 (Sgpp1; qMmuCID0008550), cluster of differentiation 36 (Cd36; qMmuCID0014852), carnitine palmitoyltransferase-1 (Cpt1; qMmuCID0021095), and Toll-like receptor 4 (Tlr4; qMmuCID0023548).

Immunohistochemistry.

To assess myofiber-type specific cross-sectional area (CSA) experimental differences, gastroc samples frozen in liquid nitrogen-cooled isopentane were cut into 8-µm sections using a cryostat (Microtome Plus), mounted on slides at −25°C, left to air-dry overnight, and stored at −20°C. Myofiber-type immunofluorescent staining was performed using traditional techniques, resulting in staining for fiber borders myosin heavy chain I (MyHC I) and IIa and (MyHC IIa). Fibers that stained for neither for MyHC I or IIa were considered MyHC IIb/x fibers. Briefly, samples mounted on slides were rehydrated in PBS and blocked in a mouse-on-mouse blocking reagent (cat no. MKB-2231; Vector) for 1 h at room temperature. After a PBS wash, slides were incubated for 90 min at room temperature in the appropriate primary antibodies diluted in 2.5% normal horse serum (NHS). Primary antibodies included anti-laminin (1:200, L9393; Sigma), anti-MyHC I (1:100, BA.D5 concentrate; Developmental Studies Hybridoma Bank), and and anti-MyHC IIa, (1:100, SC.71 concentrate; Developmental Studies Hybridoma Bank). Slides were then washed in PBS and incubated with the appropriate secondary antibodies and diluted in PBS for 1 h at room temperature. Secondary antibodies included AF555 for laminin (1:250, cat. no. A21429; Invitrogen), AF647 for MyHC I (1:250, cat. no. A21242; Invitrogen), and AF488 for MyHC IIa (1:500, cat. no. A21121; Invitrogen). Slides were then washed in PBS, then they underwent a postfixation in methanol for 5 min, and then they were washed once more in PBS and finally mounted with fluorescent mounting media (cat. no. H-1000; Vector).

After staining was completed, slides were imaged using a fully automated wide-field light microscope (Nikon; Nikon, Tokyo, Japan) with the ×10 objective lens, and images were captured using a high-sensitivity Andor Clara CCD. Captured images were analyzed for fiber type-specific CSA using semiautomatic muscle analysis, using segmentation of histology, a MATLAB application (SMASH) (32). The integration of the automated microscope stage and SMASH analysis software permitted the imaging and analysis of the entire stained cross-section. As such, an average of 2,407 ± 876 total, 47 ± 43 MyHC I, 463 ± 350 MyHC IIa, and 1,967 ± 824 MyHC IIb/x fibers were analyzed per cross-section.

Statistical analyses.

Data are reported as means ± SE. To evaluate the effectiveness of the wheel-running model, soleus and gastroc masses from the Sed, WR, and HS groups were assessed for equal variances and normality, compared using a one-way analysis of variance (ANOVA) and followed by a Tukey post hoc test as appropriate to test for group effects. For our primary outcomes, unpaired Student’s t tests were used to compare WR versus HS and then HS versus HS + TAK242. For all statistical comparisons, significance was set at the level of P < 0.05 and trends as 0.05< P < 0.08. Nonparametric Spearman correlations were conducted to identify possible associations between ceramide abundance and other dependent measures (i.e., muscle size and plasma triglycerides), with statistical significance set at P < 0.05. All analyses and figures were conducted with Graph Pad Prism 7 (La Jolla, CA).

RESULTS

Mouse muscle characteristics.

To demonstrate the effectiveness of our physical activity preconditioning model (i.e., 5 wk of WR), we included a group of sedentary (Sed) cage control animals not exposed to wheel running. As expected, 5 wk of WR activity (∼10 km/day) increased soleus muscle size (+19%, P < 0.05) compared with Sed (Table 1), and these differences occurred irrespective of normalization to body weight (BW) or lean tissue (LT). Animals HS for 7 days did not exhibit significant alterations in body weight or food consumption. However, HS animals did have lower LT and marginally greater adipose tissue (AT) compared with WR animals (−9.9% LT, P < 0.05; +36% AT, P = 0.05). TLR4 inhibition during HS had no effect on either body composition or body weight. HS had lower soleus (−34%, P < 0.01) and gastroc masses (−18%, P < 0.01) compared with WR. HS animals (vs. WR) had smaller MyHC IIa gastroc fiber sizes (−19%, P = 0.03; Supplemental Table S1; data used to inform the interpretation of the present article may be accessed upon reasonable request from the authors; Supplemental Material is publicly available at https://figshare.com/articles/Raw_data/6025748/2). Finally, there were no differences in grip strength or hanging grip strength duration across any of the groups.

Table 1.

Morphological and physiological changes in body composition, muscle size and function, and insulin sensitivity in response to WR, HS, and TLR4 inhibition

	Sed	WR	HS	HS + TAK-242
LT, g	21.05 ± 0.39	20.23 ± 0.44	18.21 ± 0.17*	18.36 ± 0.27
AT, g	3.33 ± 0.50	1.43 ± 0.25	1.94 ± 0.12‡	1.93 ± 0.19
BW, g	28.40 ± 0.67	25.83 ± 0.60	25.11 ± 0.45	24.86 ± 0.62
Week 6 food intake, g	24.5 ± 2.4	29.9 ± 3.2	29.5 ± 3.7	25.8 ± 1.2
Soleus mass, mg	8.68 ± 0.24	10.72 ± 0.41†	7.07 ± 0.36*	7.84 ± 0.29
Soleus mass mg/g body wt	0.32 ± 0.01	0.42 ± 0.01†	0.30 ± 0.02*	0.33 ± 0.01§
Soleus mass, mg/g lean tissue	0.41 ± 0.01	0.53 ± 0.02†	0.39 ± 0.02*	0.43 ± 0.02
Gastroc mass, mg	0.41 ± 0.01	142.79 ± 2.33	116.87 ± 3.10*	124.95 ± 5.85
Gastroc mass, mg/g body wt	5.18 ± 0.15	5.66 ± 0.17	4.94 ± 0.08*	5.35 ± 0.26
Gastroc mass, mg/g lean tissue	6.69 ± 0.08	7.07 ± 0.12	6.42 ± 0.18*	6.83 ± 0.36
Forelimb grip strength, gForce/mg lean tissue	6.2 ± 0.4	7.3 ± 1.4	5.7 ± 0.2	5.4 ± 0.3
Hindlimb grip strength, gForce/mg lean tissue	4.5 ± 0.4	4.6 ± 0.3	4.7 ± 0.2	4.4 ± 0.3
Hanging grip strength duration, s/g body wt	2.0 ± 1.0	1.1 ± 0.3	1.1 ± 0.5	1.5 ± 0.5
Average week 6 daily running distance, km		10.5 ± 1.1
Fasting blood glucose, mg/dL	141.8 ± 8.4	135.5 ± 8.9	151.6 ± 14.2	109.5 ± 6.4#
GTT AUC	20,372 ± 1,245	22,326 ± 1,215	21,182 ± 1,755	25,288 ± 1,161
Fasting serum insulin, ng/mL	0.50 ± 0.09	0.29 ± 0.03	0.57 ± 0.05*	0.56 ± 0.06
HOMA-IR	0.18 ± 0.03	0.09 ± 0.01	0.20 ± 0.02*	0.15 ± 0.02§
Insulin-stimulated serum TG, mg/dL		20.8 ± 1.9	39.7 ± 3.2*	35.1 ± 2.1

Data are reported as means ± SE; n = 10 mice. AT, adipose tissue; AUC, area under the curve; GTT, glucose tolerance test; HOMA-IR, homeostatic model of insulin resistance; HS, hindlimb suspension; Sed, sedentary; TG, triglycerides; TLR4, Toll-like receptor 4; WR, wheel running.

^*Different from WR (P < 0.05);

^#different from HS,

^†different from Sed;

^‡P = 0.05–0.10 vs. WR;

^§P = 0.05–0.10.

Mouse metabolic characteristics.

HS and WR had similar fasting blood glucose levels (Table 1 and Supplemental Fig. S1); however, HS fasting insulin levels (+97%, P < 0.05) and HOMA-IR values (+122%, P < 0.05) were nearly twice the level of WR. HS did not alter the blood glucose excursion during the GTT compared with WR, whereas HS + TAK242 had slightly elevated glucose AUC compared with HS (+20%, P = 0.06). HS + TAK242 had fasting insulin levels similar to HS; however, HS + TAK242 resulted in lower fasting blood glucose levels compared with HS (−27%, P < 0.05). HS animals were found to have elevated serum triglyceride levels compared with WR (+91%, P < 0.05), with no difference between HS- and HS + TAK242-treated mice.

Soleus and liver ceramides.

Soleus from HS (vs. WR; Fig. 2 and Supplemental Fig. S3) had higher total ceramide levels (+78%, P < 0.05) and several individual ceramide species (Cer12:0: +118%, P < 0.01; Cer14:0: +290%, P < 0.01; Cer16:0: +81%, P < 0.01; Cer18:0: +172%, P < 0.01; and Cer20:0: +122%, P = 0.01). In the soleus of HS + TAK-242 (vs. HS) mice, total ceramides (−33%, P < 0.05), as well as multiple individual ceramide species (Cer18:0: −32%, P < 0.05; Cer22:0 −54%, P = 0.02; Cer24:0: −78%, P < 0.02; and Cer24:1: −45%, P = 0.01), were reduced. Additionally, soleus ceramide abundance was found to be inversely related to soleus mass (r = −0.57, P < 0.01), whereas total soleus ceramide, as well as several other species, was found to be positively correlated with plasma triglyceride levels (total ceramide, r = 0.42, P = 0.03; Cer12:0, r = 0.49, P < 0.01; Cer14:0, r = 0.55, P < 0.01; Cer16:0, r = 0.52, P < 0.01; Cer18:0, r = 0.60, P < 0.01; Cer20:0, r = 0.43, P = 0.02). In contrast, HS or TAK-242 interventions did not alter ceramides in gastroc.

Fig. 2.

Soleus (A) and liver (B) ceramide (Cer) abundance. Data are reported as means ± SE; n, number of mice used from each group for the analyses reported in each figure. Soleus: sedentary (Sed) n = 8 mice, wheel running (WR) n = 7 mice, hindlimb suspension (HS) n = 11 mice, HS + TAK-242 n = 10 mice; liver (noninsulin stimulated): Sed n = 3 mice, WR n = 7 mice, HS n = 5 mice, HS + TAK242 n = 8 mice. *Different from WR (P < 0.05); #different from HS (P < 0.05); ^P = 0.05–0.10 vs. WR; $P = 0.05–0.10 vs. HS. Open bars, Sed; black bars, WR; dark gray bars, HS; light gray bars, HS + TAK-242.

The lowered fasting blood glucose for HS + TAK-242 mice directed us to investigate ceramide abundance in the liver under fasting conditions (not insulin stimulated). Liver Cer18:0, Cer20:0, and Cer22:1 were elevated in HS (vs. WR) (+35%, P = 0.05; +85%, P = 0.01; and +77%, P < 0.01, respectively). HS + TAK-242 (vs. HS) had lower liver ceramides, particularly the longer-chain species Cer20:0, Cer22:0, Cer22:1, and Cer24:1 (−32%, P < 0.05; −36%, P = 0.05; −36%, P < 0.03; and −28%, P < 0.04, respectively). This may explain the effect of TAK-242 to lower fasting blood glucose.

Ceramide biosynthesis and lipid trafficking mRNA in soleus.

In an effort to explore the possible mechanisms for altered ceramide synthesis/metabolic pathways by HS and TLR4 inhibition, we measured gene expression of several genes involved in various sphingolipid pathways in the soleus of a new cohort of mice containing our primary groups of interest (WR, HS, and HS + TAK-242; Fig. 3). Soleus genes associated with de novo ceramide synthesis, Spt2 and Degs1, were not altered in response to HS (vs. WR). However, TLR4 inhibition during HS significantly lowered soleus Spt2 and Cd36 mRNA expression (P = 0.02 and P < 0.05, respectively) compared with HS. HS reduced Cpt1 expression compared with WR (P < 0.01), although TAK-242 had no effect. As expected, TAK-242 treatment resulted in a significant decrease in Tlr4 expression in HS + TAK-242 mice (∼87% decrease, P = 0.04; n = 8) compared with mice treated with saline (WR and HS; n = 12; data not shown).

Fig. 3.

Soleus transcripts for lipid trafficking and ceramide biosynthesis. Data are means ± SE; n, number of mice used from each group for the analyses reported in figure. Wheel running (WR) n = 5 mice, hindlimb suspension (HS) n = 7 mice, HS + TAK-242 n = 8 mice. *Different from WR (P < 0.05); #different from HS (P < 0.05); $P = 0.05–0.10 vs. HS. Black bars, WR; dark gray bars, HS; light gray bars, HS + TAK-242. Cd36, cluster of differentiation 36; Cpt1, carnitine palmitoyltransferase-1; Degs1, desaturase-1; Spt2, serine palmitoyltransferase-2.

Fatty acid composition and glycerolipid content in soleus.

To follow up on the robust decrease in soleus Cd36 gene expression with TAK-242, we conducted an untargeted lipidomic analysis on soleus to determine whether the effect of TAK-242 to lower ceramide was similarly observed in intracellular fatty acids and in other complex lipids (Fig. 4). HS resulted in an increase in most of the intracellular fatty acid content (vs. WR) (total fatty acid content: +88.9%, P < 0.05). HS also elevated (vs. WR) intramuscular triglyceride (TG) content (+114.3%, P < 0.05). TLR4 inhibition (vs. HS) significantly reduced both PA and DG content without changes in TG content. Together, these findings suggest that TAK-242 was able to prevent the HS-induced increase in skeletal muscle ceramides and glycerolipids that may be explained by its effect on Cd36 expression.

Fig. 4.

Fatty acid composition and glycerolipid content in soleus. Data are means ± SE; n, number of mice used from each group for the analyses reported in figure. Wheel running (WR) n = 5 mice, hindlimb suspension (HS) n = 7 mice, HS + TAK242 n = 8 mice. *Different from WR (P < 0.05); #different from HS (P < 0.05); ^P = 0.05–0.10 vs. WR. Black bars, WR; dark gray bars, HS; light gray bars, HS + TAK242. A: soleus fatty acid (FA) content. B: soleus phosphatidic acid (PA), diacylglycerol (DG), and triglyceride (TG) content.

Skeletal muscle and liver insulin signaling and OXPHOS.

Akt phosphorylation (p-Akt Ser⁴⁷³) was not different between groups in soleus, gastric, or liver. However, AS160 phosphorylation (pAS160 Ser⁵⁸⁸) was reduced following HS in the soleus (−50%, P = 0.05) and gastroc (−37%, P = 0.01). AS160 phosphorylation in soleus or gastroc from HS + TAK-242 mice was not different compared with HS. OXPHOS and IΚΒα were not different between groups in gastroc or liver (Supplemental Fig. S4).

DISCUSSION

The purpose of this study was to determine whether pharmacological inhibition of TLR4 in preconditioned wheel run mice would prevent ceramide accumulation, muscle loss, and metabolic dysfunction during the early phase of muscle disuse (7 days of hindlimb suspension; HS). The major finding was that increased soleus and liver ceramides caused by HS was blocked with TLR4 inhibition. This finding was supported by lower soleus Spt2 and Cd36 gene expression and diminished soleus glycerolipid content in mice treated with a TLR4 inhibitor during HS. However, TLR4 inhibition did not offset muscle atrophy or insulin resistance during the early period of muscle disuse.

The novel finding of our study was that TLR4 inhibition using the well-known TLR4 antagonist TAK-242 reduced soleus and liver ceramide abundance during the early period of muscle disuse (7 days of HS). The effects of TLR4 inhibition are in line with our previous work that demonstrated the regulatory role of TLR4 signaling contributing to muscle ceramide accumulation following 2 wk of HS in TLR4-deficient mice (21) as well as a recent paper showing that rats treated with an inhibitor of ceramide de novo synthesis (myriocin) during HS were protected from inactivity-induced soleus ceramide abundance (30). Our results extend these findings by demonstrating the efficacy of an FDA-approved TLR4 inhibitor that has been tested in septic patients (29) but has yet to be used in the context of short-term muscle disuse, a milder inflammatory insult.

Ceramide accumulation and insulin sensitivity are typically viewed as being inversely related (10, 26, 35). Indeed, 7 days of HS in the present study resulted in reduced insulin sensitivity (via HOMA-IR) and increased tissue ceramides. In contrast to our hypothesis, TLR4 inhibition did not improve insulin sensitivity despite diminished ceramide abundance inferring a disconnect between ceramides and insulin sensitivity with disuse. On the other hand, TLR4 inhibition did reduce fasting blood glucose compared with HS mice, suggesting that TLR4 signaling had minor effects on glucose metabolism using this short-term model of muscle disuse. Moreover, TLR4 inhibition did not prevent hindlimb suspension-induced muscle or fiber loss, although there was a trend to have a positive effect (P = 0.08). Although TLR4 inhibition did not protect muscle size, it is important to note that intramuscular ceramides were inversely related to soleus size. Thus, TLR4 inhibition during HS may have subtle positive effects with regard to muscle size preservation, and we were possibly underpowered to observe them and/or other mechanisms concurrently driving muscle atrophy (e.g., ubiquitin proteasome). Nonetheless, the muscle size and metabolic discrepancies between the present data and our previous work (using a whole body TLR4-knockout mouse) (21) may be explained by the fact that a genetic deletion of TLR4 at birth constitutively prevents TLR4 downstream signaling, whereas daily injections, resulting in fluctuating levels of circulating TAK-242 (20), may result in a partial resurgence of TLR4 signaling several hours following the daily injection, contributing to the milder metabolic and muscle size protection, or lack thereof, seen here. However, this is unlikely given that TLR4 mRNA was suppressed in mice treated with TAK-242. Alternately, it is possible that TLR4 inhibition for 7 days (vs. 14 days) of HS was not an adequate duration of time to observe more robust protective responses.

The results from the present study reveal greater ceramide content after 7 days of HS, specific to several ceramide species lengths, in both the soleus and liver but not in the gastroc. The elevated soleus Cer18:0 following HS is particularly noteworthy given that C18:0 is the most abundant species in skeletal muscle, and this species has gained recent attention for its association with impaired insulin sensitivity in humans (4) and rodents (40). Elevated Cer18:0 is in line with our previous work following 14 days of HS (21) and with 7 days of HS (30) but not following cessation of WR (wheel lock) (1). The lack of ceramide alterations in gastroc in the present study may be partially explained by its difference in fiber-type composition (oxidative vs. glycolytic) (12). Prior studies reporting liver ceramide content following disuse are not known. However, adult female rats recovering from a spinal cord injury (SCI), a condition partly akin to disuse and known to contribute to metabolic complications (13), were shown to have greater liver ceramide abundance (and species; i.e., Cer18:0, 20:0, and 22:1) 1–3 days following SCI (31). Moreover, circulating low-density lipoproteins (LDL) containing ceramide have been shown to impair skeletal muscle metabolic function (6). This information might suggest that ceramides originally synthesized in liver and delivered to skeletal muscle could be an additional source of deleterious skeletal muscle ceramide during periods of inactivity. HS also resulted in increased serum triglyceride levels compared with WR, which may infer that triglycerides synthesized in the liver may be transported to muscle and contribute to de novo ceramide synthesis. This is supported by the positive correlations between soleus ceramide abundance and serum triglyceride levels reported in the present study and others (11). Taken together, ceramide accumulation as a result of short-term muscle disuse is likely driven by multiple organ systems and may be muscle specific (12).

Changes in muscle ceramides was not explained by alterations in transcripts associated with de novo sphingolipid biosynthesis (Spt2 and Degs1). However, we observed that TLR4 inhibition resulted in a robust decrease in Spt2 and Cd36 gene expression in soleus. SPT2 is essential for ceramide synthesis (9), and therefore, lowered soleus ceramide abundance may be partially explained by diminished de novo ceramide synthesis. However, TLR4 inhibition during HS did not prevent the increased circulating triglycerides, and therefore, this led us to investigate the soleus gene expression of lipid transport proteins in an attempt to explore an alternate route for ceramide accumulation during short-term hindlimb suspension. Given that CD36 is a transporter responsible for long-chain fatty acid trafficking into muscle (18) and a membrane protein shown to participate in the initiation of TLR4-mediated inflammatory pathways (34), we speculate herein that TAK-242 may reduce fatty acid influx, and thus ceramide biosynthesis, by possibly lowering CD36. To further explore this possibility, we performed a nontargeted lipidomic analysis to quantify fatty acids and glycerolipids in muscle. Consistent with previous literature implicating lipotoxicity as an important mediator of ceramide-induced metabolic dysfunction (9, 12), HS resulted in elevated intramuscular fatty acid and TG content. The diminished soleus Cpt1 expression in HS mice also suggest an impairment in lipid oxidative phosphorylation (33), which may further exacerbate the deleterious lipid content observed herein. The reduction in PA and DG content following TLR4 inhibition during HS lends support to our hypothesis that TLR4 inhibition during HS may alter intramuscular fatty acid flux in general, including toward ceramide biosynthesis and glycerolipid biosynthesis (9, 17).

Although not a primary objective of this study, we were surprised that 5 wk of exercise preconditioning in the form of WR was not capable of preventing ceramide accumulation, muscle loss, or metabolic disruption in the form of impaired Akt signaling and elevated HOMA-IR values. Two recent publications have demonstrated the efficacy of physical activity around the time of hindlimb suspension (7–14 days) in young mice, whereby 14 days of voluntary wheel running immediately following hindlimb suspension improved indices of muscle regeneration (8), and 14 days of high-intensity treadmill training prevented muscle atrophy before 7 days of hindlimb suspension (37). Therefore, it is possible that the preconditioning conducted in the current study may not have been long or intense enough to protect from the potent negative consequences of hindlimb suspension.

Perspectives and Significance

Blockade of TLR4 using TAK-242 during HS reduced muscle and liver ceramide accumulation, but this was not accompanied by a rescue of muscle size and metabolic function during the early phase of muscle disuse (7 days). Follow-up analyses determined that TLR4 inhibition resulted in lower muscle glycerolipids as well as reduced muscle gene expression of Spt2 and Cd36, suggesting a role for TLR4 in lipid trafficking and ceramide synthesis. Future research investigating strategies to prevent muscle loss and metabolic dysfunction should consider the potential involvement of the TLR4/CD36 signaling axis and the contribution of multiple organ systems (muscle, liver, adipose, etc.). Together, the results from the present study provide new information on the role of TLR4 during short-term disuse and inform future clinical trials aimed at muscle and metabolic health promotion.

GRANTS

Funding was provided by the National Institute on Aging (R01-AG-050781 and F31 AG059438), and metabolomics analysis was supported by 1-S10-OD016232-01, 1-S10-OD021505-01, and 1 U54 DK110858-01 awarded to the University of Utah’s Metabolomics Core Facility.

DISCLOSURES

SAS is a cofounder, consultant, and shareholder for Centaurus Therapeutics.

AUTHOR CONTRIBUTIONS

A.I.M. and M.J.D. conceived and designed research; A.I.M., P.T.R., D.S.N., J.L.M., N.M.Y., J.J.P., and T.S.T. performed experiments; A.I.M., D.S.N., and M.J.D. analyzed data; A.I.M., P.T.R., Z.S.M., S.A.S., K.F., and M.J.D. interpreted results of experiments; A.I.M. and M.J.D. prepared figures; A.I.M., P.T.R., and M.J.D. drafted manuscript; A.I.M., P.T.R., D.S.N., J.L.M., N.M.Y., J.J.P., Z.S.M., T.S.T., K.F., and M.J.D. edited and revised manuscript; A.I.M., P.T.R., D.S.N., J.L.M., N.M.Y., J.J.P., Z.S.M., T.S.T., S.A.S., K.F., and M.J.D. approved final version of manuscript.