Adipose-derived exosomes ameliorate skeletal muscle atrophy via miR-146a-5p/IGF-1R signaling

Abstract

The study of muscle disorders has gained popularity, with a particular emphasis on the relationship between adipose tissue and skeletal muscle. In our investigation, we discovered that the deletion of miR-146a-5p specifically in adipose tissue (aKO) led to a notable rise in mice’s mass and adiposity. In contrast, it led to a decline in lean mass, ability to exercise, diameter of muscle fibers, and the levels of genes associated with differentiation. The co-culture experiment showed that the transfection of miR-146a-5p mimics to 3T3-L1 significantly suppressive cell growth and promotes myotube differentiation in C2C12 cells. Exosomes from white adipose tissue (WAT) of aKO mice (aKO-WAT-Exos) significantly promoted muscle atrophy and inhibited differentiation of C2C12 cells but were reversed by co-incubation with miR-146a-5p-mimics. The miR-146a-5p can specifically target IGF-1R to improve skeletal muscle wasting. In this process, the PI3K/AKT/mTOR pathway is activated or the FoxO3 pathway is inhibited to enhance the synthesis of skeletal muscle proteins. Significantly, miR-146a-5p serves a crucial function as a microRNA in the communication of the fat-muscle connection. It can be transported through the pathway of exosomes derived from adipose tissue, ultimately ameliorating skeletal muscle atrophy and modulating body mass index (BMI).

Graphical abstract

Introduction

The communication between myoblast and adipocytes may have a significant impact on muscle development, the synthesis and degradation of proteins [1]. Additionally, co-culture models using both myocytes and adipocytes have demonstrated crosstalk between the two cell types [2, 3]. In various models of skeletal muscle atrophy, such as Fbx32 (also referred to as Atrogin1/Mafbx), an increase in expression has been noted, indicating the loss of muscle protein [4]. Fbx32 transcription is regulated by FoxO transcription factors, which also contribute to muscle atrophy [5]. Muscle atrophy development is associated with the suppression of the IGF1-PI3K-AKT signaling pathway [6, 7]. Muscle atrophy results in a loss of strength and mass, as well as several signaling pathways that influence the synthesis and degradation of protein [8,9,10,11]. The mechanisms behind muscle atrophy are not fully understood, and there are currently no clinical therapies available [12].

Exosomes facilitate the transmission of a diverse array of molecular signals by exchanging lipids, proteins, and nucleic acids [13,14,15]. MicroRNAs (miRNAs) are abundant in exosomes derived from cells, which can act locally or distally [16,17,18]. MiRNAs derived from adipocytes are a novel class of adipokines that function as metabolic regulators in distant tissues, providing novel pathways for cell-to-cell communication [19,20,21,22]. Moreover, adipokines are becoming increasingly important in explaining related muscle disorders [23,24,25,26].

Previous studies have suggested that exosomal miR-146a-5p may have a role in the communication between adipose tissue and skeletal muscle [27]. However, the detailed crosstalk mechanism remains unclear as to whether fat-specific deletion of miR-146a-5p affects muscle development. The study utilized the Cre/loxP method was employed to create mice with adipose-specific knockout of miR-146a-5p (referred to as aKO). Various techniques including transgenic animal models, transwell assay, in vitro exosome incubation test, in vivo internalized transport test of the exosome, and muscle atrophy models test were utilized to examine the function of exosomal miR-146a-5p as an adipokine released from adipose tissue, which plays an essential role in modulating skeletal muscle development through muscle-fat signaling.

Methods

Animals

Every single animal was kept in separate ventilated cages (IVC) within a specific pathogen-free (SPF) facility for animals. Indoor temperatures were controlled at 24 °C and light cycles were 12 h long. Throughout the entire duration of the experiment, the subjects were provided with a conventional diet from the Guangdong Province Animal Experiment Center located in Guangzhou, China. The miR-146a-5p flox/flox (miR-146a-5pflox^+/+, Flox) and adipose-Cre mice were generated using the Cre-loxP technique (Cyagen, Suzhou, China). The skeletal muscle development was examined using miR-146a-5p knockouts (aKO) specific to adipose tissue. Initially, mice with control (miR-146a-5p flox^+/+, Cre^−/−) were crossed with adipose-Cre mice (miR-146a-5p flox^−/−, Cre^+/−) to generate F1 mice (miR-146a-5p flox^+/−, Cre^+/−). Subsequently, the F1 mice were crossbred with flox+/+ mice to produce the aKO mice (miR-146a-5p flox^+/+, Cre^+/− mice), as well as the Flox mice (flox^+/+, control mice). Unless otherwise stated, all experiments used male 8-12-week-old mice. The guidelines for the care of laboratory animals issued by the Ministry of Science and Technology of the People’s Republic of China were adhered to at South China Agricultural University, and all animals and procedures were approved by the university’s Animal Subjects Committee.

Analysing whole body composition using NMR

Following the guidelines provided by the manufacturer, the mice underwent analysis to measure body composition, including lean mass and fat mass. Utilizing a quantitative magnetic resonance method (MiniQMR23-060 H-I, Niumag Corporation, Shanghai, China) facilitated the completion of this task.

Exercise endurance, weight lifting, and grip strength

Mice were trained on mouse treadmills (FT-200, Taimen Co., Chengdu, China) for two days at a speed of 10 m per minute (10 m/min) for ten minutes each before treadmill endurance tests. The treadmill was set at an incline of 10° and the speed was gradually raised (After 10 min of 10 m/min, follow up with 5 min of 14 m/min, then finish with 18 m/min). Upon reaching exhaustion (immobility for more than 30 s), tests were terminated.

In short, every mouse was grasped by its mid-tail and gradually lowered to seize the initial weight. Once the mice had firmly gripped the gauze, they were once again lifted until their weight reached the surface of the bench. To deem the trial successful, the mouse must maintain its grip on the weight for 3 s. Alternatively, a 10-second break would be given to the mouse before conducting another attempt. Before moving on to the next weight load, a maximum of five successful attempts were conducted [28]. The grip strength of mice was measured using digital grip strength meters (YLS-13 A, China) in compliance with a recognized protocol. After a 10-minute acclimatization period, the mice underwent a test to measure their grip strength. A metal bar was allowed to be grasped by the mice and then pulled with force. When the force was released, the peak tension was recorded. A 30-second break was taken between each test, and the mice were tested five times. A blind experiment was performed by an investigator who was unaware of the allocation to the groups.

ELISA

ELISA kits were used to measure the levels of IL-6 and TNF-α in serum, following the instructions provided by Nanjing Jiancheng Bioengineering Institute in Nanjing, China.

Measurement of oxygen consumption

Sable Systems International’s promotion metabolism measurement system from the USA was utilized to determine respiratory exchange ratio (RER) and oxygen consumption (VO₂) in Flox and aKO mice.

IPITT and IPGTT

After fasting for 6 h, mice underwent insulin tolerance tests (IPITT) and were then injected with 0.5 U/kg of insulin (Sigma). The Accu-Check Active glucometer from Roche was used to measure the glucose levels of 10 µL of blood at 0, 30, 60, 90, and 120 min. Before the IPGTT, mice were subjected to a 12-hour fasting period and then given an injection of 1 g/kg of glucose (Sigma). Glucose concentrations were assessed at 0, 30, 60, 90, and 120 min by analyzing a 10 µL blood sample collected from the tip of the tail, following the method described earlier.

Muscle injury model

To cause muscle injury, cardiotoxins (CTX, ZX0005, Shanghai Boyao Biological Technology Co., Ltd) are injected intramuscularly into the tibialis anterior (TA) muscle following the previously mentioned method. PBS was used to dissolve CTX to a concentration of 10 µM. In the case of TA muscle, three injections were administered to each mouse, with each injection site receiving 30 µL. The skin was pierced with a 28 gauge needle and inserted at a 90-degree angle. The TA muscle received two injections at its extremities, while a third injection was administered in its center. During the surgery, mice were anesthetized using a small animal anesthesia machine (TAIJI-IE, Rayward Life Technologies Inc, Shenzhen, China), utilizing 3–4% isoflurane to induce anesthesia and 1-1.5% isoflurane to maintain anesthesia.

C2C12 myoblasts were cultured in 12-well plates (Corning) using DMEM high glucose medium (Gibco, Gaithersburg, MD), supplemented with 10% FBS (HyClone, Logan, UT) and 1% P/S (Gibco). Upon achieving confluency, the culture medium was substituted with DMEM containing 2% horse serum (HS) (Gibco) for 6 days to stimulate differentiation. The medium was refreshed every 24 h. Afterward, they were denied FBS, in DMEM with high glucose (Gibco, Gaithersburg, MD) for 6 h. Subsequently, they were exposed to 0.08 µM CTX for 24 h.

Immunofluorescence and HE staining

Muscles sample were freshly embedded in Tissue-Tek OCT (Fisher Scientific/Thermo Scientific) and dipped into isopentane cooled by liquid nitrogen. The muscle specimens were cut into cryosections that were 10 μm in thickness, ensuring that the entire embedding and slicing procedure was conducted at a temperature of -25℃. Immunofluorescent staining sections were first fixed in 4% paraformaldehyde (PFA) for 10 min. Sections were rinsed three times with PBS, each rinse lasting 5 min. The blocking buffer (10 mL blocking buffer was prepared with PBS containing 0.5 mL goat serum, 0.2 g BSA, 0.2 mL 10% Triton X-100 and 0.01 g sodium azide) was supplemented with the primary antibodies and left to incubate overnight at a temperature of 4℃. Mouse monoclonal anti-MyHC (1:2000, MAB4470, R&D Systems), mouse monoclonal anti-MyoD antibody (1:100, sc-377460, Santa Cruz), and rabbit anti-laminin (1:1000, PA1-16730, Thermo Fisher) were the primary antibodies employed. The sections were rinsed thrice with PBS, with each rinse lasting for a duration of 5 min. Next, the sections were treated with goat anti-rabbit FITC antibody (bs-0295G- FITC, Bioss) and goat Anti-Mouse IgM/Alexa Fluor 555 antibody (bs-0368G-AF555, Bioss) at a concentration of 1:2000 for 1 h. Sudan black 0.05% was used to suppress the background for 5 min. A Nikon Eclipse Ti inverted microscope situated in Tokyo, Japan was utilized to capture fluorescent images.

C2C12 cells underwent treatment with 4% PFA at ambient temperature for a duration of 20 min, followed by permeabilization utilizing 0.4% Triton X-100 in PBS for a period of 20 min. Following a 1-hour blocking step using goat serum containing 5%, the cells were then incubated overnight at 4 °C with anti-MyHC (MAB4470, R&D System), and monoclonal anti-MyoD antibody (sc-377460, Santa Cruz). Afterward, the goat anti-rabbit FITC antibody (bs-0295G- FITC, Bioss) and goat Anti-Mouse IgM/Alexa Fluor 555 antibody (bs-0368G-AF555, Bioss) were administered and left to incubate at room temperature, shielded from light, for a duration of 1 h. Additionally, the nucleus was stained with DAPI (1:2,000, #MF234-01, Mei5 Biotechnology). To capture images, a fluorescent microscope from Nikon, Tokyo, Japan (Eclipse Ti) was employed, and the diameter of myotubes was assessed using Image J.

To conduct HE staining, BAT, iWAT, eWAT, and liver were first treated with a 10% formalin solution for fixation and subsequently enclosed in paraffin. Subsequently, the samples from BAT, iWAT, eWAT, and liver were divided into sections and subjected to staining with hematoxylin-eosin (HE).

Exosomes isolation, identification, and treatment

To isolate the exosomes, the white adipose tissue (WAT) from 3-month-old mice was rinsed with PBS three times and then sliced into tiny fragments measuring less than 3 mm in diameter. These dishes (100 mm) were filled with 20 mL of DMEM (Gibco) supplemented with 1% Penicillin-Streptomycin (PS) (Gibco) and 10% fetal bovine serum (FBS) that had undergone a 24-hour depletion in a cell incubator maintained at a temperature of 37℃. After collecting the supernatants from the cell culture, they were subjected to centrifugation at a speed of 300×g for a duration of 10 min to eliminate suspended cells. Afterward, the mixture was subjected to centrifugation at a speed of 2,000×g the force of gravity for a duration of 10 min to remove nonviable cells, and finally at 10,000×g for 30 min to get rid of cellular debris. After utilizing a protein concentrator with a 100-kDa cutoff (Macrosep), the exosomes underwent filtration using a 0.22 μm PVDF filter (Millipore). Afterward, the screened exosomes were moved to a 38.5 mL ultracentrifugal tube (Beckman Coulter) and subjected to ultracentrifugation at a speed of 120,000×g for 90 min. After enriching the exosomes, they were rinsed with PBS, resuspended, and then preserved in 100–150 µL PBS at -80℃. The exosomes derived from iWAT/eWAT of Flox and aKO mice were designated as Flox-iWAT-Exos/Flox-eWAT-Exos (also known as Flox-WAT-Exos/Flox-Exos) and aKO-iWAT-Exos/aKO-eWAT-Exos (referred to as aKO-WAT-Exos/aKO-Exos) respectively.

To identify the isolated exosomes, the presence of exosome marker proteins Alix, TSG101, CD63, and CD9, along with the endoplasmic reticulum membrane protein Calnexin, was detected using Western blotting. The concentration and size distribution of exosomes derived from white adipose tissue (WAT-Exos) were measured using the Nanosight instrument. To examine the shape of WAT-Exos, imaging using transmission electron microscopy (TEM) was conducted.

To administer exosomes to mice, recipient mice were injected with WAT-Exos through TA injection. This process involved three injections, with each injection containing 100 µg and administered every 7 days. To treat the cells, WAT-Exos were added to the culture medium at a concentration of 10 µg/mL of exosomes per 1.2 × 10⁵ recipient cells. To monitor the motion of exosomes, exosomes were labeled with a PKH67 fluorescent cell linker kit (Sigma-Aldrich) following the manufacturer’s instructions.

Cell culture

C2C12 myoblasts were cultured in 12-well plates (Corning) using DMEM high glucose medium (Gibco, Gaithersburg, MD), supplemented with 10% FBS (HyClone, Logan, UT) and 1% P/S (Gibco). Upon achieving confluency, the culture medium was substituted with DMEM containing 2% horse serum (HS) (Gibco) for 6 days to stimulate differentiation. The medium was refreshed every 24 h.

C2C12 cells were seeded at a density of 1.0 × 10⁵ cells per well in 12-well plates and then treated with miR-146a-5p mimics (40 nM), miR-146a-5p inhibitor (80 nM), si-IGF-1R (50 nM), or exosomes (10 µg/mL) for introduction purposes. Transfection was initiated once the cell confluency reached 60 to 70%. The siRNAs (obtained from GenePharma and Tsingke Biological Technology in China) and lipofectamine 2000 (supplied by Thermo Fisher) were employed following the instructions provided by the manufacturer.

3T3-L1 cells were cultured using DMEM containing 10% FBS and 1% P/S as a supplement. To promote differentiation, cells that had reached confluence in 12-well plates were cultured in a differentiation medium for 2 days. Differentiation began on Day 0, marking the initial day. The medium utilized was DMEM supplemented with 10% FBS, 0.5 mM IBMX, 1 µM Dex, and 10 µg/mL of insulin. From the second day onwards, the cells were given a medium containing 10% fetal bovine serum along with 10 µg/mL of insulin. The medium was changed every 24 h until the point where fully formed lipid droplets became detectable.

The co-culture test was conducted using 12 well Transwell chambers (BIOFIL, TCS016012). C2C12 cells (20,000 cells per well) were added to the upper layer of the cell chamber, whereas the lower layer consisted of 3T3-L1 cells (800,000 cells per well). Transfection was performed when the 3T3L-1 cell confluency reached 60 to 70%, using either a 40 nM concentration of miR-146a-5p mimic or an 80 nM concentration of miR-146a-5p inhibitor. The temperature of all these cells was kept at 37℃ along with 5% CO₂.

Cell counting kit-8 assay (CCK-8)

The evaluation of cell proliferation was conducted by employing a CCK-8 kit (Beyotime, Haimen, China) following the manufacturer’s instructions. C2C12 cells were evenly allocated into 96-well plates, with a density of 1.0 × 10⁴ cells per well and 6 replicates for each group. We monitored cell proliferation 24 h after transfection 10 µL of CCK-8 solution was added to each well and left to incubate for 1 h. Using a BioTek microplate reader, the absorbance at 450 nm was determined, with the reference being the empty wells.

EdU incorporation assay

The proliferation of cells was detected using BeyoClick™ EdU Cell Proliferation Kit with Alexa Fluor 488 (Beyotime, C0071) according to the manufacturer’s instructions. C2C12 cells were seeded in 96-well plates at a concentration of 1.0 × 10⁴ cells per well and exposed to 10 µM EdU for 2 h in a 37℃ CO₂ incubator. Afterward, the cells were fixed with 4% paraformaldehyde for 20 min at room temperature. Then permeated with 0.3% Triton X-100 for an additional 15 min. The cells were then incubated with the Click Reaction Mixture for 30 min at room temperature in a dark place and incubated with Hoechst 33,342 for 10 min. Image J software was used to perform the cell counting for EdU-positive cells.

Flow cytometry

The cell cycle was examined using flow cytometry.C2C12 cells were seeded into a 6-well culture plate (Corning) at a density of 3 × 10⁵ cells per well. For 24 h, the cells were transfected with either a mimic or an inhibitor of miR-146a-5p. After that, the cells were treated with 70% alcohol, followed by washing with a solution containing 1% BSA. Afterward, the cells were kept in the dark at a temperature of 4℃ and incubated overnight with a staining mixture containing propidium iodide (PI). In the end, flow cytometry (BD FACScan; BD Biosciences, Franklin Lakes, NJ, USA) was utilized to examine the fluorescence emitted by individual nuclei, which amounted to around 10,000 occurrences.

Dual-luciferase experiments

To summarize, we introduced HEK293T cells into 96-well plates (Corning) at a concentration of 2.5 × 10⁴ cells per well and permitted them to achieve a growth of 60-70% confluence. miRNAs were co-transfected with the dual-luciferase reporter plasmid in HEK293T cells. In every well, there were 3 pmol of miR-146a-5p negative control/mimic and 100 ng of the dual-luciferase gene reporter vector with either wild-type, mutation, or deletion. Cells were collected and luciferase activity was measured after 24 h of transfection using the Dual-GLO luciferase reporter assay system (Promega) following the manufacturer’s instructions.

Co-immunoprecipitation (Co-IP)

The instructions display the precise procedures. The antibody targeting IGF-1R was affixed to Pierce™ Protein A/G Magnetic Beads (88803, Thermo Scientific) and subsequently added to the lysed samples. Following the spin column’s centrifugation of the antibody resin, the cell lysates were subsequently added. After being eluted with Elution Buffer, the immunoprecipitation product underwent analysis through western blotting.

Quantitative real-time PCR

Total RNA was extracted using TRIzol (Thermo Fisher) according to the manufacturer’s instructions. Afterward, the complete RNA sample was subjected to DNase I digestion (EZB, Shanghai, China). Following the manufacturer’s instructions, EZB 4×EZscript Reverse Transcription Mix II (EZB, Shanghai, China) was used to convert 1–2 µg of RNA into complementary deoxyribonucleic acid (cDNA). The Bio-Rad C1000 Touch, known as the QuantStudio Real-Time PCR System, was employed for performing quantitative real-time PCR (qPCR) using 2×RealStar Fast SYBR qPCR Mix (GenStar, Cat No.A301) following the instructions provided by the manufacturer. For mRNA, Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal reference, while U6 RNA served as the internal reference for miRNA.

Western blot

Functional gene expression levels were assessed using Western blotting. The protein samples were prepared by utilizing the Radioimmunoprecipitation assay (RIPA) lysis solution, and the protein concentration was assessed using Thermo Fisher’s Rapid Gold BCA Protein Assay Kit. The Western blot was performed by loading 15 µg of lysate onto SDS-PAGE gels and then transferring them onto PVDF membranes (Millipore). The primary antibodies utilized in the experiment included rabbit anti-Alix (D262028, Sangon Biotech), rabbit anti-TSG101 (381538, ZEN BIO), rabbit anti-CD63 (D160973, Sangon Biotech), rabbit anti-Calnexin (D262986, Sangon Biotech), rabbit anti-Wnt 10b (bs-3662R, Bioss), rabbit anti-CD9 (AP68-965, Abcepta), mouse anti-C-myc (D199941, Sangon Biotech), rabbit anti-Cyclin A2 (GTX103042, Gene Tex), rabbit anti-Cyclin D1 (2978 S, CST), rabbit anti-Cyclin E1 (340298, ZEN BIO), mouse anti-PCNA (200947, ZEN BIO), rabbit anti-MyoD (252249, ZEN BIO), mouse anti-MyHC (MAB4470, R & D Systems), rabbit anti-Fbx32 (A3193, Abclonal), rabbit anti-MyoG (382257, ZEN BIO), rabbit anti-MuRF (A3101, Abclonal), rabbit anti-phospho-PI3K (4228, CST, P-PI3K), rabbit anti-phospho-AKT (AP0140, Abclonal, P-AKT), rabbit anti-phospho-mTOR (2972, CST, P-mTOR), rabbit anti-phospho-S6 (AP0537, Abclonal, P-S6), rabbit anti-phospho-FoxO3 (9466, CST, P-FoxO3), mouse anti-PI3K (MAB2686, R&D Systems), rabbit anti-AKT (9272 S, CST), rabbit anti-mTOR (5536, CST), rabbit anti-S6 (2217, CST), rabbit anti-FoxO3 (2497, CST), and rabbit anti-Tubulin (AP0064, Bioworld). Afterward, the membranes were incubated with the secondary antibodies (Bioworld) at room temperature for 1 h. The grayscale was scanned using a gel imaging system and then normalized to Tubulin expression using Image J software for calculation.

Statistical analysis

A minimum of three biological replicates was included in every experiment. SPSS 25 and GraphPad Prism 9.0 were employed to examine and visualize one-way ANOVA and independent sample t-test. The findings were displayed as mean ± SEM (standard error of the mean). The importance of the distinction was assessed using a confidence level of *p < 0.05 or **p < 0.01.

Results

Effects of the deletion of miR-146a-5p specifically in adipose tissue on composition, exercise capacity, metabolism, and glucose homeostasis in mice

Our lab previously conducted an investigation that revealed a significant difference in the expression of miR-146a-5p between exosomes derived from skeletal muscle and those derived from adipose tissue. Moreover, the lack of miR-146a-5p was discovered to promote adipogenesis [27]. However, the effect of reducing miR-146a-5p levels in adipose tissue on skeletal muscle development is still unclear. As a result, we utilized the CRISPR/Cas9 system (Fig. 1a, b) to generate a mouse model in which miR-146a-5p was specifically deleted in adipose tissue (aKO). In aKO mice, the expression of miR-146a-5p in brown adipose tissue (BAT), inguinal white adipose tissue (iWAT), and epididymal white adipose tissue (eWAT) exhibited significant reduction when compared to the control group (Flox) (Fig. 1c). Furthermore, the level of miR-146a-5p in GAS, SOL, TA, and EDL was notably decreased (Figure S1a). In this study, it was observed that aKO mice exhibited higher body weight (Fig. 1d), fat mass (Fig. 1e), fat accumulation (Fig. 1f, g), adipocyte size (Fig. 1h), and adipose tissue weight (BAT, iWAT, eWAT) (Fig. 1i) in comparison to Flox mice. Additionally, aKO mice showed mild fatty liver disease (Figure S1b) but no significant variation in the cumulative feed intake (Figure S1c). Significantly, we observed that aKO also efficiently decreased lean mass (Fig. 1e), the weight of muscle tissue (GAS, SOL, TA, EDL) (Fig. 1i), exercise capabilities such as running duration or distance (Fig. 1j, Figure S1d), weightlifting (Fig. 1k), and the grip of muscles (Fig. 1l). During our investigation, we noticed a notable reduction in the size of the muscle fibers in GAS, SOL, TA, and EDL (as shown in Fig. 1m-n, S1e-j) of aKO. Furthermore, the expression of MyoD and MyoG, which are genes associated with muscle differentiation, along with the levels of MyHC protein, were observed to have decreased. Conversely, there was an increase in the mRNA and protein quantities of the muscle-specific F-box protein 32 (Fbx32, alternatively referred to as Mafbx/Atrogin1) and muscle RING finger (MuRF, also known as Trim63), as well as the proliferation genes Cyclin A2, Cyclin D1, Cyclin E1, PCNA (Fig. 1o-q, S1k-s). Furthermore, the levels of IL-6 and TNF- α were also increased (Fig. 1r-s). Furthermore, aKO mice exhibited elevated oxygen consumption (Fig. 1t-u), reduced respiratory exchange ratio (RER) (Fig. 1v-w), and notable enhancement in insulin and glucose tolerance levels (Fig. 1x-y). To summarize, these results indicate that miR-146a-5p originating from adipose tissue may have a potential role in governing muscle growth, preserving glucose equilibrium, enhancing insulin responsiveness, promoting oxidative metabolism, and regulating inflammation by facilitating communication between adipose tissue and muscle.

Fig. 1

Effects of adipose-specific miR-146a-5p knockout on endurance exercise capacity, metabolism, and glucose homeostasis in mice. (a) The schematic diagram for the development strategy of miR-146a-5p-knockout mice. (b) aKO mouse WAT tissue gDNA PCR result with a sequence of only 200 bp. (c) The expression of miR-146a-5p gene in BAT, iWAT, and eWAT of Flox and aKO mice (n = 6). (d) Body weight (n = 8). (e) Body composition (n = 8). (f) Representative images of body imaging. (g) Representative images of mice. (h) Representative H&E staining of iWAT, eWAT, and BAT from mice (scale bar = 50 μm). (i) Tissue weight in BAT, iWAT, eWAT, GAS, SOL, TA, and EDL of mice (n = 7). (j) Running distance at low speed (n = 6). (k) Score of weight lifting (n = 7). (l) Muscle grip strength (n = 7). (m) Representative cross sections TA fiber immunofluorescent MyHC staining (scale bar = 100 μm). (n) Frequency histogram of fiber cross-sectional area (n = 6). (o) RT-qPCR analysis for Cyclin A2, Cyclin D1, Cyclin E1, PCNA, MyoD, MyoG, Fbx32, and MuFR in the TA muscles of Flox and aKO mice (n = 6). (p, q) The protein levels and statistical analyses of Cyclin A2, Cyclin D1, Cyclin E1, PCNA, MyHC, MyoD, MyoG, Fbx32, and MuFR measured by Western blot in the TA muscles of Flox and aKO mice (n = 3). (r, s) ELISA analysis for IL-6 and TNF-α in Flox and aKO mice (n = 6). (t, u) The O₂ consumption (VO₂) (n = 6). (v, w) RER (n = 6). (x, y) IPITT and IPGTT blood glucose changes in Flox and aKO mice (n = 8). Values are presented as means ± SEM, *P <0.05, and **P <0.01, according to the non-paired Student’s t-test or one-way ANOVA between individual groups

The transfer of miR-146a-5p between 3T3-L1 cells hinders cell growth and encourages cell maturation in C2C12 cells

To validate the relationship between muscle and fat, an experiment was carried out where C2C12 cells and 3T3-L1 cells were co-cultured (Fig. 2a). In an in vitro environment, a mimic or inhibitor of miR-146a-5p was added to 3T3-L1 cells. To achieve over-expression or down-regulation of miR-146a-5p in 3T3-L1 cells (Fig. 2b), the transfection of miRNA mimics or inhibitors was employed. The same trend also appeared in co-cultured C2C12 cells, indicating an intercellular transfer of miR-146-5p (Fig. 2c). In the proliferative stage of C2C12 cells, it was noted that the application of miR-146a-5p mimics to 3T3-L1 cells decreased the survival rate of C2C12 cells (Fig. 2d), the proportion of EdU-positive cells (Fig. 2e-f), as well as the mRNA and protein quantities of cell cycle stimulators Cyclin A2, Cyclin D1, Cyclin E1, and PCNA proliferation genes (Fig. 2g-i). In contrast, the cellular characteristics and gene expressions mentioned above were reversed by administering an inhibitor of miR-146a-5p. Afterward, after the initiation of C2C12 cells to undergo differentiation into myotubes, the co-cultivation with 3T3-L1 cells transfected with miR-146a-5p mimics significantly improved the mRNA and protein expressions of MyoD and MyoG, raised the protein quantities of MyHC, enlarged the size of myotubes, and notably decreased the mRNA and protein quantities of Fbx32 and MuRF in C2C12 cells. In contrast, the outcomes of inhibitor incubation were completely different (Fig. 2j-n). The aforementioned findings indicated that the communication of C2C12 and 3T3-L1 cells had the potential to hinder the growth and enhance the maturation of C2C12 cells utilizing transmitting miR-146a-5p between cells.

Fig. 2

3T3-L1 cell-derived miR-146a-5p participates in C2C12 cell proliferation and differentiation. (a) 3T3-L1 cells were co-cultured with C2C12 cells, and the cells were grown in a transwell. (b) The expression of miR-146a-5p gene in 3T3-L1 cells following transfection with mimics and inhibitors (n = 6). (c) The expression of miR-146a-5p gene in co-cultured C2C12 cells following transfecting 3T3-L1 cells with mimics and inhibitors (n = 6). (d) CCK-8 result of co-cultured C2C12 cells (n = 9). (e, f) EdU image and statistical analyses of C2C12 cells (scale bar = 50 μm) (n = 7). (g) RT-qPCR analysis for Cyclin A2, Cyclin D1, Cyclin E1, and PCNA of C2C12 cells (n = 6). (h, i) The protein levels of Cyclin A2, Cyclin D1, Cyclin E1, and PCNA by Western blot and the statistical analyses results of C2C12 cells (n = 3). (j) RT-qPCR analysis fo MyoD, MyoG, Fbx32 and MuFR of C2C12 cells (n = 6). (k, l) The protein levels of MyHC, MyoD, MyoG, Fbx32, and MuFR by Western blot and the statistical analyses results of C2C12 cells (n = 3). (m, n) Representative muscle fiber immunofluorescent MyHC staining of C2C12 cells (scale bar = 50 μm) (n = 4). Values are presented as means ± SEM, *P <0.05, and **P <0.01, according to the non-paired Student’s t-test or one-way ANOVA between individual groups

Transient transfection of miR-146a-5p suppresses the growth of C2C12 cells and enhances their differentiation

During the expansion of C2C12 cells, there was a progressive increase in the expression of miR-146a-5p (Fig. 3a). Introducing miR-146a-5p mimics to C2C12 cells resulted in a decrease in the number of EdU-positive cells (Fig. 3c-d), cell viability (Fig. 3e), cells in S and G2 phases (Fig. 3f-g), the mRNA levels of Cyclin A2, Cyclin B1, Cyclin D1, and PCNA, and the protein levels of Cyclin A2, Cyclin D1, Cyclin E1, and PCNA. Additionally, it led to an increase in the mRNA level of P21, an inhibitor of cell cycle-dependent kinase (Fig. 3h-j). In contrast, the results of the miR-146a-5p inhibitor treatment varied. Furthermore, the treatment group with miR-146a-5p mimics showed a significant decrease in the mRNA levels of β-catenin, Axin, TCF4, LEF1, Snail, and C-myc when compared to the negative control (NC) in C2C12 cells, as evidenced by the RT-qPCR findings. Furthermore, the miR-146a-5p inhibitor effectively restored the expression levels of these genes (Figure S2a). As expected, the use of miR-146a-5p imitations led to a significant decrease in the levels of Wnt 10b and β-catenin proteins in C2C12 cells. In contrast, the use of a miR-146a-5p inhibitor showed a different effect (Figure S2b-c).

Fig. 3

miR-146a-5p inhibits proliferation and promotes differentiation in C2C12 cells. (a) RT-qPCR to detect the expression of miR-146a-5p in proliferating C2C12 cells (n = 12). (b) The expression of miR-146a-5p gene in proliferating C2C12 cells after transfection with mimics and inhibitors (n = 12). (c, d) EdU image and statistical analyses of C2C12 (scale bar = 50 μm) (n = 12). (e) CCK-8 result of C2C12 (n = 8). (f, g) Cell cycle analysis of C2C12 by flow cytometry and statistical results (n = 3). (h) RT-qPCR analysis for Cyclin A2, Cyclin B1, Cyclin D1, PCNA and P21 in C2C12 (n = 6). (i, j) The protein levels of Cyclin A2, Cyclin D1, Cyclin E1, and PCNA by Western blot and the statistical analyses results in C2C12 (n = 3). (k) The expression of miR-146a-5p gene in differentiating C2C12 (n = 12). (l-n) Representative muscle fiber immunofluorescent MyHC staining in C2C12 (scale bar = 100 μm) (n = 4). (o) RT-qPCR analysis for MyoD, MyoG, Pax7, Fbx32, and MuFR in C2C12 (n = 9). (p-q) The protein levels of MyHC, MyoD, MyoG, Fbx32, and MuFR by Western blot and the statistical analyses results in C2C12 (n = 3). (r) RT-qPCR analysis for MyHC I, MyHC IIa, MyHC IIb and MyHC IIx in C2C12 (n = 9). (s) RT-qPCR analysis for IL-1β, IL-6 and TNF-α in C2C12 (n = 9). Values are presented as means ± SEM, *P <0.05, and **P <0.01, according to the non-paired Student’s t-test or one-way ANOVA between individual groups

Since myogenic differentiation plays a vital role in the development of skeletal muscle, our study focused on exploring the involvement of miR-146a-5p in myoblast differentiation. After undergoing the process of differentiation, the immunofluorescence results of MyHC indicated that C2C12 cells, which were transfected with miR-146a-5p mimics, displayed an increased diameter of myotubes and rate of nuclear fusion (Fig. 3l-n). Furthermore, there was an elevation in the mRNA expression of MyoD, MyoG, and paired box7 (Pax7), along with a rise in the protein abundance of MyHC, MyoD, and MyoG. In contrast, there was a decline in the levels of Fbx32 and MuRF mRNA and protein (Fig. 3o-q). In C2C12 cells transfected with a miR-146a-5p inhibitor, the reverse was noted. Significantly, the overexpression of miR-146a-5p simultaneously increased the mRNA quantities of MyHC I, MyHC IIa, MyHC IIb, and MyHC IIx, without inducing any alteration in muscle fiber (Fig. 3r). Furthermore, the miR-146a-5p mimics also notably decreased the expression of pro-inflammatory cytokines IL-1β, IL-6, and TNF-α (Fig. 3s). In summary, the previously mentioned data indicated that the application of miR-146a-5p overexpression inhibited cell proliferation and improved the progression of myotube differentiation in C2C12 cells. The WNT/β-catenin pathway also showed a potential regulatory influence, in line with controlling the intercellular transfer of miR-146a-5p derived from 3T3-L1 cells.

Exosomal transfer and replenishment of miR-146a-5p attenuates myotube atrophy in vitro by miR-146a-5p-null adipose exosomes

Exosome (Exos) secreted from cells is a major mechanism for miRNA transport. Thus, we postulated that the previously mentioned in vivo transmission of miR-146a-5p among cells is linked to the transport mechanism of exosomes. Consequently, we extracted exosomes from white adipose tissue (WAT) of aKO and Flox mice. Subsequently, through the utilization of electron microscopy and NanoSight analysis, it was discovered that the diameter of WAT-Exos varied between 20 and 180 nm (Fig. 4a). Afterward, we verified the WAT-Exos using Western blot to detect the exosome-specific proteins Alix, TSG101, CD63, and CD9. Successful isolation of exosomes was confirmed by the absence of calnexin protein, in the exosome fraction (Fig. 4b). The research found that the amounts of miR-146a-5p were significantly decreased in aKO-iWAT-Exos and aKO-eWAT-Exos (exosomes derived from aKO mice) when compared to Flox-iWAT-Exos and Flox-eWAT-Exos (exosomes derived from Flox mice). This finding aligned with the outcomes observed at the cellular level (Fig. 4c, S3c, e). After analyzing the CCK8 findings, it was concluded that a concentration of 10 µg/mL provided the best results for treating C2C12 with exosomes (Figure S3a-b). In C2C12 cells, the viability of cells was enhanced by aKO-iWAT-Exos and aKO-eWAT-Exos, while it was inhibited by Flox-iWAT-Exos and Flox-eWAT-Exos (Fig. 4d). The EdU incorporation assay provided additional confirmation of this phenotype (Fig. 4e-f). After 48 h of treatment, the levels of mRNA and protein for Cyclin A2, Cyclin D1, Cyclin E1, and PCNA were significantly higher in the aKO-iWAT-Exos and aKO-eWAT-Exos group compared to the Flox-iWAT-Exos and Flox-eWAT-Exos group, which further supported the evidence (Fig. 4g-h, S3d). Furthermore, by conducting immunofluorescence analysis on MyHC, it was observed that the myotube diameters and nuclear fusion ratios were reduced in the myoblasts of the aKO-Exos group, in comparison to the Flox-Exos group (Fig. 4i-k). In addition, MyHC, MyoD, and MyoG were dramatically lower than Flox-iWAT-Exos and Flox-eWAT-Exos groups. On the other hand, the expression levels of Fbx32 and MuRF mRNA were significantly increased in the aKO-Exos groups (Fig. 4l). Additionally, the protein levels demonstrated a similar trend (Fig. 4m, S3f). A lipophilic fluorescent dye called PKH67 (Sigma-Aldrich, St.Louis, MO) was utilized for labeling aKO-WAT-Exos and Flox-WAT-Exos. Fluorescence microscopy showed that PKH67-labelled exosomes were effectively absorbed into the C2C12 cells after incubation for 12 h and 24 h (Figure S3i-k).

Fig. 4

miR-146a-5p reversed the myotube atrophy of C2C12 cells induced by aKO-WAT-Exos. (a) Electron microscopy results and nanoparticle tracking analysis was used to determine the size distribution of adipose-derived exosomes (scale bar = 200 nm). (b) Calnexin in adipose cells and Alix, TSG101, CD9, and CD63 in adipose-derived exosomes of aKO and Flox mice were detected by Western Blot. (c) The expression of miR-146a-5p gene in Flox-iWAT-Exos, aKO-iWAT-Exos, Flox-eWAT-Exos, aKO-eWAT-Exos (n = 6). (d) CCK-8 result of C2C12 cells treated with Control (PBS), Flox-iWAT-Exos, aKO-iWAT-Exos, Flox-eWAT-Exos, aKO-eWAT-Exos (n = 8). (e, f) EdU and statistical analyses image of C2C12 (scale bar = 50 μm) (n = 6). (g) RT-qPCR analysis for Cyclin A2, Cyclin D1, Cyclin E1, PCNA in C2C12 cells (n = 6). (h) The protein levels of Cyclin A2, Cyclin D1, Cyclin E1, and PCNA by Western blot in C2C12 cells (n = 3). (i-k) Representative muscle fiber immunofluorescent MyHC staining, myotube diameter, and myotube fusion rate in C2C12 cells (scale bar = 100 μm) (n = 6). (l) RT-qPCR analysis for MyoD, MyoG, Fbx32, and MuFR in differentiated C2C12 (n = 6). (m) The protein levels of MyHC, MyoD, MyoG, Fbx32, and MuFR by Western blot in differentiated C2C12 (n = 3). (n) The expression of miR-146a-5p gene in proliferating C2C12 treated with aKO-WAT-Exos + NC and aKO-WAT-Exos + Mimics (n = 6). (o) CCK-8 result of C2C12 (n = 8). (p, q) EdU image and statistical results of C2C12 (scale bar = 50 μm) (n = 6). (r) RT-qPCR analysis for Cyclin A2, Cyclin D1, Cyclin E1, PCNA in C2C12 cells (n = 6). (s) The protein levels of Cyclin A2, Cyclin D1, Cyclin E1, PCNA in C2C12 (n = 3). (t, u) Representative muscle fiber immunofluorescent MyHC staining in C2C12 (n = 4) (scale bar = 50 μm). (v) RT-qPCR analysis for MyoD, MyoG, Fbx32 and MuFR in C2C12 (n = 6). (w) The protein levels of MyHC, MyoD, MyoG, Pax7, Fbx32, and MuFR by Western blot in C2C12 (n = 3). Values are presented as means ± SEM, *P <0.05, and **P <0.01, according to the non-paired Student’s t-test or one-way ANOVA between individual groups

The above-mentioned results suggested that these characteristics might be associated with the lack of exosomal miR-146a-5p derived from white adipose tissue (WAT). Following that, we experimented to enhance the function, in which miR-146a-5p mimics were introduced into C2C12 cells co-incubated with aKO-WAT-Exos, known as aKO-WAT-Exos + mimics.The results in C2C12 cells showed a significant rise in the concentrations of miR-146a-5p (Fig. 4n). These cells demonstrated decreased cell viability and a smaller percentage of EdU-positive cells when compared to the control group that was co-treated with NC and aKO-WAT-Exos (aKO-WAT-Exos + NC) (Fig. 4o-q). In addition, the aKO-WAT-Exos + mimics group significantly inhibited the mRNA and protein expression of Cyclin A2, Cyclin D1, Cyclin E1, and PCNA to a higher degree in comparison to the control group (Fig. 4r-s, S3g). The findings from the MyHC immunofluorescence analysis demonstrated that the miR-146a-5p complement effectively reversed the muscle atrophy induced by aKO-WAT-Exos on C2C12 cells (Fig. 4t-u). Furthermore, there was a notable rise in the mRNA and protein expression of MyoD and MyoG, along with the protein levels of MyHC. In contrast, there was a notable reduction in the expression of Fbx32 and MuRF at both mRNA and protein levels throughout the differentiation of C2C12 cells (Fig. 4v-w, S3h). Therefore, the above-mentioned results indicated that the delivery and recovery of miR-146a-5p via exosomes could relieve the muscle cell atrophy induced in vitro by adipose exosomes lacking miR-146a-5p.

Intramuscular injection of the exosomal miR-146a-5p into aKO mice ameliorates skeletal muscle atrophy

Afterward, the aforementioned two WAT-Exos were administered via intramuscular injection into the TA muscle of aKO mice. Following a 24-hour injection, the distribution of PKH67-labeled exosomes was primarily observed in the GAS, SOL, TA, EDL, iWAT, eWAT, and VAT (visceral adipose tissue, intestinal fat) according to live imaging results (Fig. 5a). Internalizing Flox-WAT-Exos led to an increase in the expression of miR-146a-5p in the TA muscle (Fig. 5b, e, o). Furthermore, there was a notable rise in the mRNA and protein quantities of MyoD and MyoG, along with the protein level of MyHC, the area of muscle fibers (Fig. 5r-s, S4b-c, i-j), and the intensity of fluorescence in MyOD (Fig. 5t-u, S4d-e, k-l). Moreover, there was a decline in the mRNA and protein concentrations of Cyclin A2, Cyclin D1, Cyclin E1, PCNA, Fbx32 (a gene associated with muscle wasting), and MuRF (Fig. 5c-d, f-g, p-q, S4a, h, p). The concentrations of IL-6 and TNF-α were additionally decreased (Figure S4f-g, l-m, q-r). After 24 days of injection, Flox-Exos significantly reduced the body weight gain (Fig. 5h), fat mass (Fig. 5i), fat enrichment (Fig. 5j), oxygen consumption (Figure S5a), and significantly increased the muscle mass (Fig. 5i), muscle content (Fig. 5k), running time or distance (Fig. 5l, S4o), weight lifting (Fig. 5m), muscle grip (Fig. 5n), Pax7 fluorescence intensity of mice (Fig. 5v-w), RER (Figure S5b), insulin and glucose tolerance levels (Figure S5c-d). Nonetheless, there were no noticeable variances in the consumption of food throughout the complete duration of the experiment (Figure S4s). The aforementioned findings suggest that the intracellular transportation of miR-146a-5p into the aKO mice can enhance muscle wasting and substantial reduction in adipose tissue and body weight while promoting muscle growth in vivo.

Fig. 5

WAT-derived exosomes miR-146a-5p may be involved in muscle atrophy. (a) The fluorescence signal distribution of PKH67-labeled adipose-derived exosomes in aKO organs for 24 h after TA injection. The isolated organs from left to right are as follows: heart, liver, spleen, lung, kidney, BAT, iWAT, eWAT, GAS, SOL, TA, EDL, and intestinal fat. (b) The expression of miR-146a-5p gene in aKO TA muscles 12 h after injected with aKO-Exos and Flox-Exos (n = 6). (c) RT-qPCR analysis for Cyclin A2, Cyclin D1, Cyclin E1, PCNA, MyoD, MyoG, Fbx32, and MuFR in aKO TA muscles (12 h) (n = 6). (d) The protein levels of Cyclin A2, Cyclin D1, Cyclin E1, PCNA, MyHC, MyoD, MyoG, Fbx32, and MuFR were measured by Western blot in aKO TA muscles (12 h) (n = 3). (e) The expression of miR-146a-5p gene in aKO TA muscles 24 h after injected with aKO-Exos and Flox-Exos (n = 6). (f) RT-qPCR analysis for Cyclin A2, Cyclin D1, Cyclin E1, PCNA, MyoD, MyoG, Fbx32, and MuFR in aKO mice TA muscles (24 h) (n = 6). (g) The protein levels of Cyclin A2, Cyclin D1, Cyclin E1, PCNA, MyHC, MyoD, MyoG, Fbx32, and MuFR were measured by Western blot in aKO TA muscles (24 h) (n = 3). (h) Body weight gain (n = 4). (i) Body composition (n = 4). (j) Representative images of body imaging. (k) Tissue weight in GAS, SOL, TA, and EDL of mice (n = 4). (l) Running distance at low speed (n = 3). (m) Score of weight lifting (n = 4). (n) Muscle grip strength (n = 4). (o) The expression of miR-146a-5p gene in the aKO TA muscles injected with aKO-Exos and Flox-Exos for 24 d (n = 6). (p) RT-qPCR analysis for Cyclin A2, Cyclin D1, Cyclin E1, PCNA, MyoD, MyoG, Fbx32, and MuFR in the aKO TA muscles injected with aKO-Exos and Flox-Exos for 24 d (n = 3). (q) The protein levels of Cyclin A2, Cyclin D1, Cyclin E1, PCNA, MyHC, MyoD, MyoG, Fbx32, and MuFR measured by Western blot in the aKO TA muscles injected with aKO-Exos and Flox-Exos for 24 d (n = 3). (r, s) Representative cross sections TA fiber immunofluorescent MyHC staining and statistical results in the aKO TA muscles injected with aKO-Exos and Flox-Exos for 24 d (scale bar = 100 μm) (n = 4). (t, u) Representative cross sections TA fiber immunofluorescent MyoD staining and statistical results in the aKO TA muscles injected with aKO-Exos and Flox-Exos for 24 d (scale bar = 100 μm) (n = 4). (v, w) Representative cross sections TA fiber immunofluorescent Pax7 staining and statistical results in the aKO TA muscles injected with aKO-Exos and Flox-Exos for 24 d (scale bar = 100 μm) (n = 4). Values are presented as means ± SEM, *P <0.05, and **P <0.01, according to the non-paired Student’s t-test or one-way ANOVA between individual groups

Adipose-derived exosomal miR-146a-5p attenuates cardiotoxins-induced myocyte atrophy in vitro

To enhance understanding of the role of miR-146a-5p in muscle atrophy, a model of muscle atrophy was created by utilizing cardiotoxins (CTX) to investigate the regenerative function of miR-146a-5p in muscle atrophy. Differentiating cells treated with CTX at high concentrations (0.5 µM and 1 µM) experienced severe abscission and death (Figure S6a). Furthermore, the levels of miR-146a-5p noticeably decreased in different CTX concentrations in comparison to the control group (Figure S6b). In addition, the CTX-induced group exhibited a notable decrease in the levels of MyoD and MyoG mRNA compared to the control group. Conversely, the mRNA levels of Fbx32 and MuRF, which are genes associated with muscle atrophy, were significantly elevated after CTX treatments of 0.1 µM and 0.08 µM (Figure S6c). Therefore, the dose of 0.08 µM was used in the following experiments. Following CTX-induced muscle atrophy, we observed a significant increase in the diameter of myotubes transfected with miR-146a-5p mimics compared to NC. In addition, the size of myotubes incubated with aKO-Exos (aKO-WAT-Exos) was smaller than that of Flox-Exos (Flox-WAT-Exos). Nevertheless, the introduction of a miR-146a-5p suppressor (Flox-Exos + I) resulted in the acceleration of myotubular atrophy. In contrast, the co-transfection of miR-146a-5p mimics (aKO-Exos + M) with aKO-Exos resulted in the alleviation of muscle atrophy, as opposed to the treatment of aKO-Exos alone (Fig. 6a-b) The above-mentioned results suggested that miR-146a-5p possessed the capacity to mitigate or reverse muscle wasting. Furthermore, immunofluorescence staining confirmed that cells transfected with miR-146a-5p mimics exhibited elevated levels of fluorescence intensity for MyoD compared to the negative control (NC). Additionally, treatment with Flox-Exos also resulted in higher fluorescence intensity levels of MyoD compared to aKO-Exos. However, the fluorescence intensity of MyoD was notably reduced when Flox-Exos was co-transfected with a miR-146a-5p inhibitor (Flox-Exos + I). Conversely, the fluorescence intensity of MyoD and Pax7 was significantly increased when aKO-Exos and miR-146a-5p mimics were co-transfected (Fig. 6c-f). The results indicated that miR-146a-5p may have the ability to relieve muscle atrophy.

Fig. 6

miR-146a-5p alleviates muscle atrophy in vitro. (a, b) Representative muscle fiber immunofluorescent MyHC staining of CTX-induced C2C12 cells after transfection with miR-146a-5p mimics/inhibitors or co-treatment with Flox-Exos, aKO-Exos (scale bar = 100 μm) (n = 4). (c, d) Representative muscle fiber immunofluorescent MyoD staining of CTX-induced C2C12 cells after transfection with miR-146a-5p mimics/inhibitors or co-treatment with Flox-Exos, aKO-Exos (scale bar = 100 μm) (n = 4). (e, f) Representative muscle fiber immunofluorescent Pax7 staining of CTX-induced C2C12 cells after transfection with miR-146a-5p mimics/inhibitors or co-treatment with Flox-Exos, aKO-Exos (scale bar = 100 μm) (n = 4). Values are presented as means ± SEM, *P <0.05, and **P <0.01, according to the non-paired Student’s t-test or one-way ANOVA between individual groups

Adipose-derived exosomal miR-146a-5p is indispensable for muscle atrophy and repair in vivo

Furthermore, the present research examined the controlling impact of miR-146a-5p on muscle atrophy in live animals by utilizing the adipose-specific miR-146a-5p deficiency animal model. To begin with, the TA muscle of Flox and aKO mice was injected with CTX to cause muscle injury. TA muscles were harvested on the 5th, 7th, and 15th days to study muscle regeneration and repair after induced damage. During the muscle injury, the expression level of miR-146a-5p consistently stayed diminished in aKO mice in comparison to Flox mice (Fig. 7a). On the 5th day following injury, aKO mice had a more chaotic TA muscle structure than Flox mice. MyHC fiber size decreased significantly in regenerated TA muscles of aKO mice at 7 d and 15 d after induced injury (Fig. 7b-c), indicating the differentiation potential of TA muscle in aKO mice is lower. Immunofluorescence staining of MyoD and Pax7 showed significantly reduced fluorescence intensity in aKO mice on the 5th, 7th, and 15th days (Fig. 7d-g). Moreover, there was a notable rise in the expression of Cyclin A2, Cyclin D1, Cyclin E1, PCNA, Fbx32, and MuRF genes in aKO mice. Conversely, the expression of MyoD, MyoG, and Pax7 genes, along with the MyHC protein level, experienced a marked decline on the 5th, 7th, and 15th days after the injury occurred (Fig. 7f-i, S7a-b). The results indicate that the lack of adipose-miR-146a-5p impedes the skeletal muscle regeneration process.

Fig. 7

Adipose-derived miR-146a-5p is indispensable in muscle atrophy and repair. (a) RT-qPCR of miR-146a-5p in TA muscle cross-sections from injured (5, 7, and 15 days after CTX injection) Flox and aKO mice (n = 4). (b, c) Representative of MyHC immunofluorescent staining of TA muscle cross-sections from uninjured (Day 0) and injured (5, 7, and 15 days after CTX injection) Flox and aKO mice (scale bar = 100 μm) (n = 4). (d, e) Representative of MyoD immunofluorescent staining of TA muscle cross-sections from uninjured and injured Flox and aKO mice (scale bar = 100 μm) (n = 4). (f, g) Representative of Pax7 immunofluorescent staining of TA muscle cross-sections from both uninjured and injured Flox and aKO mice (scale bar = 100 μm) (n = 4). (h) RT-qPCR analysis for Cyclin A2, Cyclin D1, Cyclin E1, and PCNA in TA muscle cross-sections from injured Flox and aKO mice (n = 4). (i) RT-qPCR analysis for MyoD, MyoG, Pax7, Fbx32, and MuFR in TA muscle cross-sections from injured (5, 7, and 15 days after CTX, aKO-Exos and Flox-Exos injection) aKO mice (n = 4). (j) The protein levels of MyHC, MyoD, MyoG, Pax7, Fbx32, and MuFR measured by Western blot in TA muscle cross-sections from injured aKO mice (n = 3). (k) RT-qPCR of miR-146a-5p in TA muscle cross-sections from injured aKO mice treated with Exos (aKO-Exos and Flox-Exos injection) (n = 4). (l, m) Representative of MyHC immunofluorescent staining of TA muscle cross-sections from both uninjured and injured aKO mice treated with Exos (scale bar = 100 μm) (n = 4). (n, o) Representative of MyoD immunofluorescent staining of TA muscle cross-sections (scale bar = 100 μm) (n = 4). (p, q) Representative of Pax7 immunofluorescent staining of TA muscle cross-sections from both uninjured (Day 0) and injured (5, 7, and 15 days after CTX, aKO-Exos, and Flox-Exos injection) aKO mice (scale bar = 100 μm) (n = 4). (r) RT-qPCR analysis for Cyclin A2, Cyclin D1, Cyclin E1, and PCNA in TA muscle cross-sections from injured (5, 7, and 15 days after CTX, aKO-Exos, and Flox-Exos injection) aKO mice (n = 4). (s) RT-qPCR analysis for MyoD, MyoG, Pax7, Fbx32, and MuFR in TA muscle cross-sections from injured (5, 7, and 15 days after CTX, aKO-Exos and Flox-Exos injection) aKO mice (n = 4). (t) The protein levels of MyHC, MyoD, MyoG, Pax7, Fbx32, and MuFR measured by Western blot in TA muscle cross-sections from injured (5, 7, and 15 days after CTX, aKO-Exos, and Flox-Exos injection) aKO mice (n = 3). Values are presented as means ± SEM, *P <0.05, and **P <0.01, according to the non-paired Student’s t-test or one-way ANOVA between individual groups

Is it possible for the muscle damage caused by intramuscular injection of CTX to be repaired by the introduction of exosomes derived from white adipose tissue (WAT), with or without miR-146a-5p (Flox-Exos and aKO-Exos), as previously stated? Consequently, the TA muscle of aKO mice received injections of Flox-Exos and aKO-Exos and was subsequently collected on the 5th, 7th, and 15th days post-injection. Following the internalization of Flox-Exos, a significant increase in the miR-146a-5p expression was observed in the TA muscle (Fig. 7j). On Day 5 and Day 7, compared with Flox-Exos injection, more myofibres with centrally located nuclei appeared after aKO-Exos injection, and on Day 15 after injury, the size of regenerated muscle fibers was significantly smaller (Fig. 7k-l). At the same time, Pax7 and MyoD immunofluorescence labeling in TA muscle also demonstrated higher fluorescence intensity after Flox-Exos injection (Fig. 7n-q). Moreover, the injection of Flox-Exos led to an enhancement in mRNA and protein quantities for MyoD, MyoG, and Pax7, along with a rise in MyHC protein manifestation, which was validated through RT-qPCR and Western blot examination. On the other hand, the levels of Cyclin A2, Cyclin D1, Cyclin E1, PCNA, Fbx32 atrophy gene, and MuRF mRNA and protein levels were markedly down-regulated (Fig. 7o-r, S7c-d). The data indicated that the administration of Flox-Exos through intramuscular injection successfully reversed muscle atrophy resulting from both external CTX induction and internal deletion of adipose miR-146a-5p. This finding further suggests that this particular miRNA might play a crucial role in the in vivo muscle regeneration and repair process.

The PI3K/AKT/mTOR signaling pathway is activated by miR-146a-5p, which improves muscle atrophy by targeting the IGF-1R

The regulation of muscle protein synthesis is controlled by the signaling pathway PI3K/AKT/mTOR, which includes the involvement of the receptor Insulin-like growth factor 1 (IGF-1R). Bioinformatics software (Fig. 8a) indicated that miR-146a-5p could potentially target IGF-1R. To further evaluate the targeting process of miR-146a-5p in muscle atrophy, a luciferase reporter construct was generated containing the 3′-UTR region of IGF-1R. Luciferase expression was effectively suppressed by miR-146a-5p in the reporter assay. Nevertheless, once a mutation transpired in the 3′-UTR of IGF-1R, the hindrance was entirely annulled (Fig. 8b). Therefore, miR-146a-5p was initially identified as a target of IGF-1R. The construction of si-IGF-1R was achieved through siRNA interference and its transfection into C2C12 cells was successful (Figure S8a-b). The si-IGF-1R group experienced a significant increase in the amount of miR-146a-5p, which was later decreased when co-transfected with the miR-146a-5p inhibitor (si-IGF-1R + miR-146a-5p I) (Fig. 8c). Moreover, it was noted that the application of si-IGF-1R led to a significant reduction in cell viability (Fig. 8d), the percentage of EdU-labeled cells (Fig. 8e-f), and a decrease in the expression of genes linked to cell growth (Fig. 8g-i). In contrast, the simultaneous introduction of si-IGF-1 and inhibitor of miR-146a-5p promoted cell growth. Concurrently, following the induction of differentiation in C2C12 cells, treatment with si-IGF-1R enhanced the size of myotubes, the fusion ratio of their nuclei (Fig. 8j-l), the intensity of fluorescence for MyoD and Pax7 (Fig. 8m-p), the levels of mRNA and protein for MyoD, MyoG, and Pax7, as well as the protein level of MyHC. However, it significantly suppressed the levels of mRNA and protein in IGF-1R, as well as the atrophy genes Fbx32 and MuRF (Fig. 8o-p, S8c). However, the concurrent transfection of si-IGF-1 and miR-146a-5p inhibitor (si-IGF-1R + miR-146a-5p I) impeded the muscle differentiation process and promoted the contraction of myotubes. According to the Western blot analysis, si-IGF-1R treatment led to increased levels of P-mTOR, P-PI3K, P-AKT, and P-S6, while suppressing P-FoxO3 (Fig. 8q, S8d). To further verify whether IGF-1R directly regulates PI3K/AKT/mTOR signaling, an immunoprecipitation (Co-IP) assay was carried out. According to the results, the introduction of miR-146a-5p (mimics) resulted in an elevation in the levels of P-mTOR, P-PI3K, P-AKT, and P-S6, whereas P-FoxO3 demonstrated a decline (Fig. 8r).

Fig. 8

miR-146a-5p prevents muscle atrophy by targeting IGF-1R. (a) miR-146a-5p has a target interaction with the 3’UTR of IGF-1R. (b) Relative luciferase activity was calculated by firefly luminescence/renilla luminescence (WT: pmiGLO- IGF-1R -WT, mut site 1 + 2: pmirGLO-IGF-1R -Mut1 + pmirGLO-IGF-1R-Mut2, mut site 1: pmirGLO- IGF-1R -Mut1, mut site 2: pmirGLO-IGF-1R -Mut2) (n = 10). (c) The expression of miR-146a-5p gene in C2C12 cells following transfection with siRNA-IGF-1R and co-treatment with siRNA-IGF-1R and miR-146a-5p inhibitor (n = 4). (d) CCK-8 result of C2C12 cells (n = 8). (e, f) EdU image and statistical analyses of C2C12 cells (scale bar = 100 μm) (n = 6). (g) RT-qPCR analysis for Cyclin A2, Cyclin D1, Cyclin E1, and PCNA in C2C12 cells (n = 4). (h, i) The protein levels of Cyclin A2, Cyclin D1, Cyclin E1, and PCNA by Western blot and the statistical analyses results in C2C12 cells (n = 3). (j-l) Representative muscle fiber immunofluorescent MyHC staining, myotube diameter, and myotube fusion rate of C2C12 cells (scale bar = 100 μm) (n = 4). (m, n) Representative muscle fiber immunofluorescent MyoD staining and statistical results of C2C12 cells (scale bar = 100 μm) (n = 4). (o, p) Representative muscle fiber immunofluorescent Pax7 staining and statistical results of C2C12 cells (scale bar = 100 μm) (n = 4). (q) RT-qPCR analysis for IGF-1R, MyoD, MyoG, Pax7, Fbx32 and MuFR in C2C12 cells (n = 4). (r) The protein levels of MyHC, MyoD, MyoG, Pax7, Fbx32, and MuFR by Western blot in C2C12 cells. (s) The protein levels of IGF-1R, P-PI3K, PI3K, P-AKT, AKT, P-mTOR, mTOR, P-S6, S6, P-FoxO3 and FoxO3 measured by Western blot in C2C12 cells. (n = 3). (t) Immunoprecipitation assay revealed an enrichment of P-PI3K, PI3K, P-AKT, AKT, P-mTOR, mTOR, P-S6, S6, P-FoxO3 and FoxO3 when introduced with IGF-1R. (u) The protein levels of IGF-1R, P-PI3K, PI3K, P-AKT, AKT, P-mTOR, mTOR, P-S6, S6, P-FoxO3, and FoxO3 measured by Western blot in the TA muscles of Flox and aKO mice (n = 3). (v) RT-qPCR analysis for IGF-1R in C2C12 transfected with miR-146a-5p mimics/inhibitor (n = 6). (w) The protein levels of IGF-1R, P-PI3K, PI3K, P-AKT, AKT, P-mTOR, mTOR, P-S6, S6, P-FoxO3 and FoxO3 by Western blot in C2C12 transfected with miR-146a-5p mimics/inhibitor (n = 3). (x) RT-qPCR analysis for IGF-1R in aKO TA muscles injected aKO-Exos and Flox-Exos for 24 d (n = 3). (y) The protein levels of IGF-1R, P-PI3K, PI3K, P-AKT, AKT, P-mTOR, mTOR, P-S6, S6, P-FoxO3, and FoxO3 measured by Western blot in aKO TA muscles injected aKO-Exos and Flox-Exos for 24 d (n = 3). (z) The protein levels of IGF-1R, P-PI3K, PI3K, P-AKT, AKT, P-mTOR, mTOR, P-S6, S6, P-FoxO3, and FoxO3 measured by Western blot in TA muscle cross-sections from injured (5, 7 and 15 days after CTX injection) Flox and aKO mice (n = 3). Values are presented as means ± SEM, *P <0.05, and **P <0.01, according to the non-paired Student’s t-test or one-way ANOVA between individual groups

Additionally, we explored whether miR-146a-5p regulates muscle development at both living and cellular levels by analyzing its influence on the IGF-1R-PI3K/AKT/mTOR and FoxO3 signaling pathways. According to the results, the TA and GAS muscle of aKO mice showed an increase in the levels of IGF-1R and P-FoxO3 proteins, whereas a decrease was observed in the levels of P-PI3K, P-mTOR, P-AKT, and P-S6 proteins (Fig. 8s, S8e-g). During the differentiation of C2C12 cells, the presence of miR-146a-5p mimics impeded the mRNA and protein levels of IGF-1R and P-FoxO3, while boosting the phosphorylation protein levels of PI3K, mTOR, AKT, and S6. In contrast, inhibiting miR-146a-5p yielded the reverse outcome (Fig. 8t-u, S8h). In addition, the amounts of IGF-1R and P-FoxO3 proteins declined, while the levels of PI3K, mTOR, AKT, and S6 phosphorylation proteins rose after the injection of Flox-Exos during both brief (12 h/24 h) and extended (24 d) durations (Fig. 8w, S8i-k). In the experimental setup where CTX was injected into the TA muscle of Flox and aKO mice to induce injury, it was noticed that aKO reduced the levels of P-PI3K, P-mTOR, P-AKT, and P-S6 while increasing the level of IGF-1R and P-FoxO3 (Fig. 8x, S8l-m). The findings indicated that following the injection of Flox-Exos, the levels of P-PI3K, P-mTOR, P-AKT, and P-S6 increased, while IGF-1R and P-FoxO3 decreased (Figure S8n-p). Consequently, miR-146a-5p has the potential to relieve muscle atrophy through the targeting of IGF-1R and the activation of PI3K/AKT/mTOR and inhibition of FoxO3 pathways, ultimately enhancing the synthesis of muscle proteins and suppressing protein breakdown.

Discussion

Exosomes facilitate the transfer of miRNAs to enable communication among diverse cell types and microenvironments [29,30,31,32]. Recently, there has been an increased focus on adipose-derived exosomes as a medium for facilitating physiological communication between fat and skeletal muscle [33,34,35]. Although multiple studies have indicated a strong connection between muscle size or movement and the weight of fatty tissue [36,37,38,39,40], unfortunately, there is a lack of research investigating the influence of exosomes obtained from fatty tissue on the development of skeletal muscles [41, 42]. In this study, intramuscular injection of adipose-derived exosomes in mice, enabling them to target al.l skeletal muscles, would be more clinically relevant and it would be interesting to discover a protective effect against skeletal muscle atrophy after systemic intravenous injection.

Most studies have shown that downregulation of IGF-1R leads to IGF-1R signaling inactivation and muscle atrophy. It has been shown that atrophy-induced genes (e.g., Fbx32, MuRF) are inversely regulated by IGF-1R through the PI3K, AKT, and mTOR pathways, and that the PI3K/AKT/mTOR pathway represses the transcription factor FoxO3 in order to prevent the expression of skeletal muscle atrophy genes [43, 44]. However, the IGF-1R signaling pathway is a complex and tightly regulated network that is critical for cell proliferation, growth, and survival [45, 46], but the debate about the role of IGF signaling pathway and whether its inhibition is beneficial continues [47]. PI3K-Akt-mTORC1, as a downstream pathway of IGF-1R, is a highly dynamic network that is balanced and stabilized by a number of feedback inhibition loops, such as IGF1 signaling was desensitized by negative feedback loop through mTORC1 activation [48, 49]. Activation of mTORC1 triggers a number of feedback loops that converge on, and antagonize IGF-1 signaling [50]. It has been shown that a negative feedback inhibition loop is critical for IGF-induced signaling in cellular homeostasis [51]. Furthermore, the very limited efficacy of anti-IGF-1R drugs in clinical trials suggests that there is also a mechanism to antagonize IGF-1R inhibition in tumor cells [51,52,53,54]. Wang et al. found that treatment of colon tumor cells with IGF-1R inhibitors stimulated p70s6K1 activity to promote survival via MEK1/2, and inhibition of MEK1/2 enhanced Akt phosphorylation, and in turn, suppression of AKT activation results in stimulation of MEK1/2 activity [55]. Interestingly, P70s6K1, a serine/threonine kinase, is one of the key downstream targets of the AKT-mTOR pathway [56]. That is, activation of Pdcd4-inhibited p70s6k1 is essential for resistance of colon tumor cells to IGF-1R inhibitors and also provides a novel venue to overcome the resistance to the IGF-1R inhibitor in treating colorectal cancer [57]. Thus, evaluation of the regulatory mechanisms underlying the negative feedback of IGF-1R signaling is essential to fully understand the effects of IGF-1.

Generally, one miRNA is also part of signalling networks, can regulate multiple genes and serve as a mediator of crosstalk between signalling pathways, coordinating their activity [58]. Likewise, as a regulatory element, miRNA itself is coordinatively modulated by multifarious effectors when carrying out basic functions [59]. Many reports showed that miR-146-5p overexpression activated PI3K/AKT/mTOR signal pathways [60,61,62], with PI3K p110α, phosphorylated Akt (p-Akt T380, p-Akt S473) and phosphorylated mTOR (p-mTOR) were increased, which are consistent with our results. Meanwhile, the treatments of IGF-1R siRNA and miR-146a-5p mimics resulted in increased levels of P-mTOR, P-PI3K, P-AKT, and P-S6, while inhibiting P-FoxO3 in our study. Besides, IGF-1R signaling is aberrantly activated via disruption of the FOXO3a-miRNA negative feedback inhibition, such as some specific miRNAs (miR-128-3p, miR-30a-5p, and miR-193a-5p) that have been implicated in the regulation of negative feedback inhibition of IGF-1R signaling [51]. These data suggest that miRNAs including miR-146a-5p may be involved in the negative feedback regulation induced by IGF-1R inhibition.

Our study showed that overexpression of miR-146a-5p could ameliorate skeletal muscle atrophy. As a well-known immunomodulator, miR-146a-5p is involved in the regulation of inflammatory signaling [63], and inflammation is also a key factor causing skeletal muscle atrophy [64]. Numerous reports, including our previous study, showed that miR-146a-5p negatively regulates inflammatory response via downregulation of the IRAK1/TRAF6/NF-κB signaling pathway [65, 66]. And in turn, miR-146a-5p acts as an inhibitor targeting innate immune response signaling proteins, whose expression is regulated by NF-κB dependent induction [67], while NF-κB activation is dependent on IGF-1R signaling [68, 69]. IGF-1/AKT can inhibit muscle atrophy by inhibiting NF-κB pathways [70]. Thus, these data suggest that miR-146a-5p is involved in regulating the crosstalk between IGF-1R and NF-κB signaling and in the negative feedback loop of IGF-1R signaling, miR-146a-5p may prevent muscle atrophy by negatively modulating inflammatory processes.

Additionally, our previous research also found that miR-146a-5p participates in insulin signal regulation by targeting insulin receptors (IR) [71]. Insulin and IGF-1 receptors share many downstream signaling pathways, such as the PI3K/Akt/FoxO pathway [72]. Inhibition of IGF-1R results in a compensatory increase in IR signaling, but neutralizing IGF-1R did not decrease Akt phosphorylation, that is, IR maintains downstream signaling when IGF-1R is selectively inhibited [73]. The overlap of IR and IGF-1R signaling is critical to the regulation of muscle protein turnover, and this regulation depends on suppression of FoxO-regulated degradation [74, 75]. These data further demonstrate that miR-146a-5p is involved in the molecular crosstalk mechanism of two receptor signals.

Notably, the most suppressed gene by IGF-1R signaling was the ubiquitin ligase, Fbx32 (Atrogin-1, MAFbx) in skeletal muscle [76]. In atrophying muscles, Fbx32 is dramatically induced, and this response is necessary for rapid atrophy. In cultured myotubes undergoing atrophy, the activity of the PI3K/AKT pathway decreases, leading to activation of FoxO3 transcription factors and Fbx32 induction [77, 78]. Whereas FoxO3 transcription factor cause marked atrophy of adult skeletal muscle and induces the muscle-specific ubiquitin ligase Fbx32/Atrogin-1/MAFbx via the PI3K/AKT pathway [79, 80]. Interestingly, our colleagues found that miR-146a-5p was able to target FBX32 to inhibit muscle atrophy in post-receptor regulation of IGF-1R signaling (unpublished data).

Previous studies have shown that miR-146a-5p has the potential to be utilized as a therapeutic approach for reducing muscle fibrosis after an injury [81]. In recent studies, it has been found that miRNA-146a present in the bloodstream can serve as a diagnostic tool and potential biomarker to predict sarcopenia in older individuals [82]. In our recent investigation, we discovered that injecting exosomes derived from adipose tissue containing miR-146a-5p into TA muscle not only improved skeletal muscle atrophy but also reduced fat mass. Interestingly, when TA muscle was injected with PKH67 fluorescence-labeled adipose-derived exosomes, the fluorescence was mainly enriched in muscle tissue, and also in adipose tissue through the blood circulation. As per the reports, the overabundant manifestation of miR-146a-5p impeded the adipogenesis process [83, 84]. The aforementioned observation is consistent with the present finding, indicating that miR-146a-5p present in exosomes obtained from fat tissue can be absorbed by muscle cells and potentially transferred to adjacent fat tissue, possibly in a tissue-specific manner. The targeting approach of this miRNA not only offers a hopeful technique for treating skeletal muscle wasting but also has the potential to be utilized for other diseases associated with miRNA.