Full text

Turn on search term navigation
Introduction

Skeletal muscle atrophy is referred to as the loss of skeletal muscle mass.1,2 Its major characteristics include the reduction of the myofibre cross-sectional area (CSA) and protein content of skeletal muscle and loss of muscle strength.1,2 Skeletal muscle atrophy is a condition associated with chronic disease conditions, including neuromuscular diseases, cancer and cachexia.1,2 Skeletal muscle atrophy is associated with increased morbidity and mortality, increased duration of hospitalization and a poor quality of life.3 Peripheral nerve injuries (PNIs) are a major cause of skeletal muscle atrophy and usually associated with stab wounds, fall, fracture and obstetric trauma.4 The treatment options for skeletal muscle atrophy are limited. Anabolic-androgenic steroids (AAS), growth hormone (GH) and insulin-like growth factor 1 (IGF-1) were shown to increase muscle mass; however, they were associated with various adverse effects and not approved by FDA.5–8

The triggers and the molecular pathogenic mechanisms underlying skeletal muscle atrophy are not well understood. Two muscle-specific E3 ubiquitin ligases, MAFbx (muscle atrophy F-box) and MuRF1, were found to be up-regulated with atrophy resulting from cachexia and disuse.9 Transcription factor NF-κB was found to be a key intracellular signalling molecule in muscle atrophy resulting from denervation10 and cachexia.11 NF-κB plays an important role in skeletal muscle atrophy under tumour growth, denervation and unloading.10 NF-κB activation can induce MuRF1 expression, which causes the breakdown of MyHC and other components of the thick filament of the sarcomere during atrophy.12,13 Some studies suggested that the pro-inflammatory cytokine TWEAK is an inducer of skeletal muscle atrophy.14 TWEAK causes degradation of thick filament protein MyHC in C2C12 differentiated myotubes.12,14 TWEAK is a type II transmembrane glycoprotein with 249 amino acid residues.15 The only known receptor of TWEAK is Fn14.16 TWEAK can activate the ubiquitin-proteasome system and autophagy, whereas TWEAK-induced skeletal muscle atrophy was reduced by specific inhibitors of these systems.12 Moreover, the TWEAK/Fn14 signalling was shown to activate MAPK, JNK, AP-1(transcription factor activator protein-1) and NF-κB signalling in various cell types, including skeletal muscle.17 The role of autophagy in skeletal muscle atrophy is controversial. TWEAK was shown to increase autophagy, and autophagy inhibitor 3-MA (3-methyladenine) increased the diameter of TWEAK-treated myotubes.12 Skeletal muscle-specific Atg7 knockout (KO) mice with inhibition of autophagy showed severe skeletal muscle atrophy.18 More studies are needed to characterize the key mechanisms for the pathogenesis of skeletal muscle atrophy.

Our laboratory cloned the AGGF1 gene while studying Klippel–Trenaunay syndrome (KTS) characterized by capillary malformations, venous malformations or varicose veins, lymphatic malformations and affected tissue or limb hypertrophy, and showed that it encoded an angiogenic factor with a G-patch domain and a forkhead-associated domain (FHA) 1.19 We further showed that genetic variants of AGGF1 increased susceptibility to KTS.19 Aggf1 is essential for embryogenesis.20 Aggf1+/− mice showed defective vascular development and angiogenesis in embryos and adult organs.20 Our studies with zebrafish further showed that aggf1 is involved in blood vessel development and specification of veins by activating AKT (protein kinase B) signalling.21 AGGF1 can treat coronary artery disease and myocardial infarction by activating autophagy via JNK signalling.22 Through characterization of AGGF1, we discovered a non-canonical endoplasmic reticulum (ER) stress pathway and attenuation of cardiac hypertrophy and heart failure by AGGF1.23

We reported that AGGF1 binds to TWEAK.19 As TWEAK plays an important role in skeletal muscle atrophy, we investigated the role of AGGF1 in skeletal muscle atrophy in this study. We demonstrated that the expression level of AGGF1 was altered in patients and mouse models with skeletal muscle atrophy and Aggf1+/− mice showed aggravated skeletal muscle atrophy. We then identified the underlying molecular mechanism and tested whether AGGF1 protein treatment attenuated skeletal muscle atrophy in mouse models and isolated leg muscle from patients with lumbar disc herniation.

Methods Animal models

Animal numbers were determined using the PASS software. For denervation, the sciatic nerve of 8-week-old mice was exposed and cut off, and the wound was sutured. Cancer cachexia was established by Lewis lung carcinoma cell (LLC) implantation (1 × 106 cells/mouse) in 10-week-old mice. AGGF1 (0.25 mg/kg body weight in 100 μL) or elution buffer (EB, 100 μL) was injected to mice intramuscularly 12 h after denervation or intraperitoneally 7 days after LLC implantation and then once per day until 24 h before experiments.

Cell culture and transfection

Mouse myoblasts (C2C12) cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS) and transfected with siRNA using the Lipofectamine™ RNAiMAX (Invitrogen) or plasmids (HA-CMV-AGGF1, pFLAG-CMV-TNFRSF12 or pEGFP-N1-TNFRSF12a) using FuGene HD (Promega, Madison, WI, USA). LLC cell culture and TNF-α or IFN-γ treatments were performed as described previously.24

Western blotting, co-immunoprecipitation (Co-IP) and real-time RT-PCR analysis

Western blot analysis was carried out using antibodies in Table S1 as previously described.25 Mouse muscle or C2C12 cells were lysed with lysis buffer (50 mM Tris/HCl, pH 7.5, 150 mM NaCl, 2 mM EDTA, 1% Nonidet P-40 and a proteinase inhibitor cocktail) and centrifuged. The supernatant was used for western blot analysis. Co-IP analysis and real-time RT-PCR analysis were performed as described.26

Histological and immunostaining analyses

Muscle was isolated, fixed with 4% paraformaldehyde, embedded in paraffin and sectioned (0.5 μm). Muscle fibrosis was determined by Sirius red staining (collagens) as described.26 Immunostaining analysis was carried out as described.22 Macrophage infiltration was determined by immunostaining with an antibody against F4/80.20 Myogenesis was determined by immunostaining for Ki67 as described.27 The images were captured by a BX53 Olympus light microscope (Olympus, USA) and analysed using ImageJ software.

The muscle CSA was characterized by immunostaining for dystrophin, and H&E staining was as described28 and analysed using ImageJ software.

Ethical statement

Gastrocnemius muscles from the back part of the leg were isolated from patients with severe lumbar disc herniation. The patients in this study had stage II–IV lumbar disc herniation lasting for 5–10 years, and various degrees of leg muscle atrophy and claudication, and needed surgeries. The muscle granules were isolated and cultured in Ham's F-10 medium with 20% FBS. The human subject studies were approved by the Ethics Committees on Human Subject Research at Huazhong University of Science and Technology and the Institutional Ethics Committee of Renmin Hospital of Wuhan University. This study conforms to the guidelines set forth by the Declaration of Helsinki, and study participants provided written informed consent. Animal care and experimental procedures were approved by the Ethics Committee on Animal Research of Huazhong University of Science and Technology.

Statistical analysis

The data were presented as mean ± standard deviation (SD) and analysed using a two-tailed paired or unpaired Student's t-test for comparison of two groups. One-way analysis of variance (ANOVA) was used for data with normal distribution, and the Kruskal–Wallis test was used for data with non-normal distribution for comparison of more than two groups. A P value of <0.05 was considered to be statistically significant.

Results Expression remodelling of AGGF1 in atrophy muscle

Lumbar disc herniation often leads to sciatic nerve injury and unilateral leg skeletal muscle atrophy. The myofibre cross-sectional area (CSA) was significantly decreased in impaired leg muscle sections from patients compared with the unimpaired side (Figure 1A). Western blot and immunostaining analyses showed that the level of AGGF1 was significantly increased in patients with lumbar disc herniation (Figure 1B,C). Fn14, but not TWEAK, was dramatically up-regulated in atrophy muscle from patients (Figure 1D). Similar results were obtained in mouse atrophy muscle (Figure S1). AGGF1 protein treatment did not have apparent effect on the levels of Fn14 and TWEAK in mouse atrophy muscle (Figure S1F).

Figure 1. AGGF1 protein treatment reverses down-regulation of myoproteins in impaired leg muscles from patients with lumbar disc herniation. (A) Representative H&E staining images for leg muscle fibres from patients (n = 10/group, impaired group: impaired leg muscle from patients; control group: leg muscle from the normal side from patients). Scale bar = 100 μm. (B) Western blot analysis for AGGF1 in leg muscles from patients and quantification of Western blot images on the right (n = 10/group, C, control group; I, impaired group). (C) Representative immunostaining images for AGGF1 in leg muscles of patients and quantification of AGGF1-positive signals in the sections on the right (n = 10/group). (D) Western blot analysis for Fn14 and TWEAK in leg muscles from patients and quantification of Western blot images on the right (n = 4/group). (E) Western blot analysis of MyHC and α-actin, MuRF1 and MAFbx in isolated and cultured leg muscle myotubes from patients following AGGF1-treatment and quantification of western blot images on the right (n = 4/group). Data are shown as mean ± SD. Statistical significance was evaluated using a Student's test (B–D) and a one-way ANOVA test (E). *P [less than] 0.05, **P [less than] 0.01, n.s., not significant.
AGGF1 protein therapy reverses the loss of myoproteins in leg muscle from patients with lumbar disc herniation

We isolated and cultured the impaired and control unimpaired leg muscle from patients and treated the samples with the AGGF1 protein or EB (elution buffer) for 3 or 10 days. H&E staining showed that the muscle fibres were still alive in samples treated with AGGF1 for 3 days, and AGGF1 treatment significantly improved muscle fibres by increasing the CSA (Figure S2A). Immunostaining for Ki67 did not detect Ki67-positive signals in muscle myotubes treated with either AGGF1 or EB, suggesting that AGGF1 did not lead to myogenesis in atrophy muscle (Figure S2A). The levels of MuRF1 and MAFbx were dramatically up-regulated in impaired sides compared with normal counterparts (Figure 1E). AGGF1 protein treatment significantly reduced the level of MuRF1, but not MAFbx, in impaired muscle (Figure 1E). The levels of myoproteins MyHC and α-actin were significantly decreased in the impaired side compared with unimpaired side (Figure 1E). Interestingly, treatment with the AGGF1 protein partially reversed the down-regulation of skeletal muscle proteins (Figure 1E). Similar results were obtained in cultured muscle samples treated with AGGF1 for 10 days (Figures S2B and S3A). H&E staining showed that muscle fibres cultured for 10 days were severely damaged, and AGGF1 treatment dramatically improved muscle fibres by increasing the CSA (Figure S2B). These data show that AGGF1 protein therapy can attenuate the loss of myoproteins in patients with skeletal muscle atrophy.

Aggf1 haploinsufficiency (Aggf1+/−) exacerbates skeletal muscle atrophy by inducing inflammation and fibrosis and inhibiting autophagy in skeletal muscle 7 days after denervation

We investigated the role of Aggf1 in skeletal muscle atrophy using Aggf1+/− mice as homozygous Aggf1−/− mice are embryonically lethal.20 There was no significant difference in the body weight and lean mass between Aggf1+/− mice and WT littermates (Figure S4). We created a mouse model for skeletal muscle atrophy by denervation in Aggf1+/− and wild-type (WT) littermates. Seven days after denervation, the weight of hindlimb muscle was reduced compared with the sham group and was further reduced in Aggf1+/− mice as compared with WT mice (Figure 2A,B). Immunostaining for dystrophin showed that Aggf1+/− mice had significantly reduced myofibre CSA compared to WT littermates (Figures 2C and S5A). H&E staining showed the same results (Figures S6A and S7A). Western blot analysis showed that the levels of skeletal muscle proteins MyHC and α-actin were significantly reduced after denervation and further reduced in Aggf1+/− mice compared with WT mice (Figures 2D and S5B). Immunostaining with anti-F4/80 (a marker for macrophages) showed markedly elevated F4/80 signals in Aggf1+/− mice compared with WT mice in both gastrocnemius and soleus muscles, but not in the sham group (Figures 2C and S5A). Sirius red staining demonstrated that Aggf1 haploinsufficiency significantly aggravated collagen accumulation (indicator of skeletal muscle fibrosis, a major pathologic change during skeletal muscle atrophy) in gastrocnemius and soleus muscle sections after denervation (Figures 2C and S5A). Consistent with the immunostaining data (Figures 2C and S5A), western blot analysis showed that denervation significantly increased the levels of collagen I and collagen III (markers for fibrosis), and the effect was further aggravated in Aggf1+/− mice compared with WT mice (Figures S6B and S7B). The data suggest that Aggf1 haploinsufficiency aggravates inflammation and fibrosis in skeletal muscle after denervation.

Figure 2. Aggf1 haploinsufficiency (heterozygous Aggf1+/−) exacerbates skeletal muscle atrophy 7 days after denervation in male mice. (A) and soleus muscle (B) (n = 6/group; scale bar = 5 mm). (C) Immunostaining for dystrophin (left, scale bar = 20 μm) and F4/80 (middle, scale bar = 50 μm) and Sirius red staining (right, scale bar = 50 μm) of gastrocnemius muscle sections. The images were quantified and shown at the bottom (n = 6/group). (D) Western blot analysis for MyHC, α-actin, p-p65, MuRF1 and MAFbx in gastrocnemius muscle of WT and Aggf1+/− mice and quantification of western blot images (n = 6/group). Data are shown as mean ± SD. Statistical significance was evaluated using a one-way ANOVA test (A–D). *P [less than] 0.05, **P [less than] 0.01, ***P [less than] 0.001, n.s., not significant.

NF-κB is the key transcriptional factor regulating skeletal muscle atrophy.13 Western blot analysis showed that the level of phosphorylated-p65 (p-p65; a subunit of transcription factor NF-κB) was significantly increased in gastrocnemius and soleus muscles after denervation and further increased in Aggf1+/− mice compared with WT mice (Figures 2D and S5B), suggesting that Aggf1 haploinsufficiency aggravates the activation of NF-κB-mediated inflammation in muscle after denervation. MuRF1 and MAFbx were up-regulated in denervated muscle (Figures 2D and S5B). The levels of MuRF1, but not MAFbx, were further enhanced in Aggf1+/− mice compared with WT mice (Figures 2D and S5B). Western blot analysis showed that the levels of cleaved caspase-3 and cleaved PARP1 (indicators of apoptosis) were significantly increased in muscle after denervation and further increased in Aggf1+/− mice compared with WT mice (Figures S6B and S7B).

As AGGF1 induced autophagy by activating JNK,22 we examined the effects of Aggf1 haploinsufficiency on autophagy.29 Western blot analysis showed that levels of LC-3B II and phosphorylated JNK were significantly induced in gastrocnemius and soleus muscles from WT mice 7 days after denervation; however, the effect was lost in Aggf1+/− mice (Figures S6B and S7B).

All above results from male mice were also obtained from female Aggf1+/− mice (Figures S8–S11). Altogether, our data suggest that Aggf1 haploinsufficiency exacerbates skeletal muscle atrophy after denervation by inducing phosphorylated p65, inflammation and fibrosis, inhibiting autophagy via reducing JNK phosphorylation, increasing MuRF1 and reducing MyHC and α-actin.

AGGF1 protein therapy attenuates skeletal muscle atrophy induced by denervation after 7 days

Because Aggf1 haploinsufficiency aggravates skeletal muscle atrophy, we hypothesized that injection of AGGF1 protein can be used as a therapy. Interestingly, injection of AGGF1 protein or EB into gastrocnemius muscle once a day after denervation dramatically increased the mass of muscle in mice with skeletal muscle atrophy and restored the mass to 60–65% of the sham WT mice (AGGF1-denervated/WT-innervated) (Figure 3A,B). Immunostaining for dystrophin and H&E staining showed that AGGF1 protein treatment attenuated the reduction of myofibre CSA and restored CSA to 70% of sham WT mice (Figures 3C, S12A, S13A and S14A). Immunostaining analysis for F4/80 showed that AGGF1 protein therapy significantly reduced macrophage infiltration (inflammation) in skeletal muscle atrophy mice (Figures 3C and S12A). Sirius red staining demonstrated that AGGF1 protein significantly reduced collagen accumulation in muscle induced by denervation (Figures 3C and S12A).

Figure 3. AGGF1 protein therapy attenuates skeletal muscle atrophy in a 7-day denervation male mouse model. (A,B) Representative photography and mass of gastrocnemius and soleus muscle (n = 6/group; scale bar = 5 mm). (C) Immunostaining for dystrophin (top, scale bar = 20 μm) and F4/80 (middle, scale bar = 50 μm),and Sirius red staining (bottom, scale bar = 50 μm) of gastrocnemius muscle sections. The images were quantified and shown on the right (n = 5/group). (D) Western blot analysis for MyHC, α-actin, p-p65, MuRF1 and MAFbx in gastrocnemius muscle and quantification for western blot images (n = 5/group). Data are shown as mean ± SD. Statistical significance was evaluated using a one-way ANOVA test (A–D). *P [less than] 0.05, **P [less than] 0.01, ***P [less than] 0.001. n.s., not significant.

Western blot analysis showed that AGGF1 was up-regulated in gastrocnemius and soleus muscles after denervation and further enhanced by AGGF1 protein therapy (Figures S13B and S14B). Denervation significantly reduced the level of MyHC and α-actin; however, the effect was reversed by AGGF1 (Figures 3D and S12B). Similar analysis showed that AGGF1 protein treatment reversed the increased levels of phosphorylated-p65, cleaved Caspase-3 and cleaved PARP1 in denervated muscles (Figures 3D, S12B, S13B and S14B). For autophagy, AGGF1 significantly increased the levels of autophagy-related protein LC-3B and p-JNK in muscles compared with EB group after denervation (Figures S13B and S14B). Denervation significantly increased the level of MuRF1. However, the effect was reversed by AGGF1 (Figures 3D and S12B). Although denervation also significantly increased the level of MAFbx, AGGF1 protein did not have any significant effect on this (Figures 3D and S12B).

All results from male mice were confirmed in female mice (Figures S15–S18). These data suggest that AGGF1 protein therapy attenuates skeletal muscle atrophy in a denervation mouse model by inhibiting inflammation, apoptosis and fibrosis.

Aggf1 haploinsufficiency induces inflammation and fibrosis and inhibits autophagy in skeletal muscle 2 days after denervation

Because the phenotype of skeletal muscle atrophy was too overt 7 days after denervation, we analysed the effects of Aggf1 haploinsufficiency 2 days after denervation. There were no significant changes in the morphology and mass of mouse hindlimb gastrocnemius and soleus muscles in Aggf1+/− mice (Figures 4A,B). Immunostaining analysis for dystrophin and H&E staining showed no significant changes in the CSA of myofibres in gastrocnemius and soleus muscles from Aggf1+/− mice (Figures 4C, S19A, S20A and S21A). Western blot analysis also showed no significant changes of MyHC and α-actin in gastrocnemius and soleus muscles of Aggf1+/− mice or WT littermates 2 days after denervation (Figures 4D and S19B).

Figure 4. Aggf1 haploinsufficiency exacerbates macrophage infiltration and fibrosis in skeletal muscle 2 days after denervation in male mice. (A,B) Representative photography and mass of gastrocnemius muscle (A) and soleus muscle (B) (n = 6/group; scale bar = 5 mm). (C) Immunostaining for dystrophin (left, scale bar = 20 μm) and F4/80 (middle, scale bar = 50 μm) and Sirius red staining (right, scale bar = 50 μm) of gastrocnemius muscle sections. The images were quantified and shown at the bottom (n = 6/group). (D) Western blot analysis for MyHC, α-actin, p-p65, MuRF1 and MAFbx in gastrocnemius muscle of WT and Aggf1+/− mice and quantification of western blot images (n = 6/group). Data are shown as mean ± SD. Statistical significance was evaluated using a one-way ANOVA test (A–D). *P [less than] 0.05, **P [less than] 0.01, ***P [less than] 0.001, n.s., not significant.

Immunostaining of gastrocnemius and soleus muscles showed markedly elevated F4/80 signals 2 days after denervation compared with the sham group, which was further increased in Aggf1+/− mice compared with WT mice (Figures 4C and S19A). The data suggest that Aggf1+/− haploinsufficiency aggravates macrophage infiltration and inflammation after denervation. Sirius red staining demonstrated that Aggf1 haploinsufficiency further enhanced collagen accumulation in gastrocnemius and soleus muscle after denervation (Figures 4C and S19A). Interestingly, the collagen I level, but not collagen III, was elevated 2 days after denervation and further increased in Aggf1+/− mice (Figures S20B and S21B). MuRF1 was elevated only in Aggf1+/− mice after denervation. However, MAFbx was not affected by denervation or changed in Aggf1+/− mice (Figures 4D and S19B). Cleaved caspase-3 and cleaved PARP1 were too low to be detected. Phosphorylated p65 was elevated in mice 2 days after denervation and was further elevated in Aggf1+/− mice (Figures 4D and S19B). In addition, denervation increased LC-3B II and p-JNK 2 days after denervation, but the effect was lost in Aggf1+/− mice (Figures S20B and S21B).

All above results from male mice were also obtained from female Aggf1+/− mice except that the MuRF1 level was not elevated (Figures S22–S25). Altogether, our data suggest that Aggf1 haploinsufficiency increases p-p65, inflammation and fibrosis and inhibits autophagy via reducing JNK phosphorylation 2 days after denervation.

AGGF1 protein therapy attenuates skeletal muscle inflammation and fibrosis 2 days after denervation

We also examined the therapeutic effects of AGGF1 in mice 2 days after denervation. Two-day denervation did not affect gastrocnemius and soleus muscle mass or CSA of myofibres, and no effect was observed for AGGF1 protein treatment (Figures 5A–C, S26A, S27A and S28A). Immunostaining for F4/80 and Sirius red staining showed that AGGF1 inhibited macrophage infiltration and fibrosis in mice 2 days after denervation (Figures 5C and S26A). Western blot analysis showed that 2-day denervation did not affect the level of MyHC, α-actin and MAFbx in gastrocnemius and soleus muscles, and no effect was observed for AGGF1, either (Figures 5D and S26b). However, p-p65 was induced by denervation but inhibited by AGGF1 (Figures 5D and S26B). Two-day denervation did not affect the level of MuRF1 in gastrocnemius and soleus muscles; however, AGGF1 significantly reduced the MuRF1 level (Figures 5D and S26B). For autophagy, western blotting showed that the intramuscular injection of AGGF1 protein increased the levels of LC-3B and phosphorylated JNK in the gastrocnemius and soleus muscles 2 days after denervation (Figures S27B and S28B). Similar results were obtained in female mice (Figures S29–S32). These data suggest that AGGF1 protein therapy attenuates inflammation and fibrosis in mice 2 days after denervation.

Figure 5. AGGF1 protein treatment inhibits inflammation and fibrosis in skeletal muscle 2 days after denervation in male mice. (A,B) Representative photography and mass of gastrocnemius and soleus muscle (n = 6/group). Scale bar = 5 mm. (C) Immunostaining for dystrophin (top, scale bar = 20 μm) and F4/80 (middle, scale bar = 50 μm) and Sirius red staining (bottom, scale bar = 50 μm) of gastrocnemius muscle sections. The images were quantified and shown on the right (n = 5/group). (D) Western blot analysis for MyHC, α-actin, p-p65, MuRF1 and MAFbx in gastrocnemius muscle and quantification for western blot images (n = 5/group). Data are shown as mean ± SD. Statistical significance was evaluated using a one-way ANOVA test (A–D). *P [less than] 0.05, **P [less than] 0.01, ***P [less than] 0.001. N.s., not significant.
Aggf1 haploinsufficiency (Aggf1+/−) exacerbates skeletal muscle atrophy induced by cancer cachexia

Cancer cachexia can also cause severe muscle atrophy30,31; thus, we established a cachexia mouse model by implantation of LLC to validate the effects of Aggf1 haploinsufficiency on muscle atrophy. At 21 days after injection of LLC, the mouse body weight was significantly reduced, and the reduction was aggravated in Aggf1+/− mice compared with WT littermates (Figure 6A). Similar results were obtained for the weight of gastrocnemius and soleus muscles (Figure 6B). Immunostaining for dystrophin and F4/80, H&E staining, Sirius red staining and western blot analysis for MyHC, α-actin, cleaved caspase-3, cleaved PARP1, phosphorylated p65, collagen I, collagen III, MuRF1, MAFbx, LC-3B II and p-JNK confirmed and validated the results obtained from mice 7 days after denervation (Figures 6C,D and S33–S35). Similar results were obtained in female Aggf1+/− mice (Figures S36–S39). Together, these results suggest that Aggf1 haploinsufficiency also aggravates the development of muscle atrophy in a mouse model of cancer cachexia by inducing p-p65, inflammation and fibrosis, inhibiting autophagy via reducing p-JNK, increasing MuRF1 and reducing MyHC and α-actin.

Figure 6. Aggf1 haploinsufficiency exacerbates skeletal muscle atrophy in male mice injected with LLC cells. (A,B) Body weight and mass of skeletal muscles (n = 6/group). (c) Immunostaining for dystrophin (left, scale bar = 20 μm) and F4/80 (middle, scale bar = 50 μm) and Sirius red staining (right, scale bar = 50 μm) of gastrocnemius muscle sections. The images were quantified and shown at the bottom (n = 6/group). (D) Western blot analysis for MyHC, α-actin, p-p65, MuRF1 and MAFbx in gastrocnemius muscle of WT and Aggf1+/− mice and quantification of western blot images (n = 6/group). Data are shown as mean ± SD. Statistical significance was evaluated using a one-way ANOVA test (A–D). *P [less than] 0.05, **P [less than] 0.01, ***P [less than] 0.001, n.s., not significant.
AGGF1 protein therapy attenuates skeletal muscle atrophy induced by cancer cachexia

AGGF1 protein attenuated atrophy induced by cachexia in male mice by partially reversing the reduced body weight, muscle weight (gastrocnemius and soleus) and CSA and restored these parameters to 80–95% of the sham WT mice (Figures 7A–C, S40A, S41A and S42A). AGGF1 reduced macrophage infiltration, fibrosis, p-p65, MuRF1, cleaved caspase-3, cleaved PARP1 and increased MyHC and α-actin (Figures 7C,D, S40B, S41B and S42B). For autophagy, AGGF1 increased the levels of LC-3B and p-JNK in muscles (Figures S41B and S42B). Similar results were obtained in female mice (Figures S43–S46). These data suggest that AGGF1 protein therapy attenuates skeletal muscle atrophy in a mouse model of cancer cachexia.

Figure 7. AGGF1 protein therapy attenuates skeletal muscle atrophy in a 21-day cachexia male mouse model. (A,B) Mass of gastrocnemius and soleus muscle (n = 6/group). (C) Immunostaining for dystrophin (top, scale bar = 20 μm) and F4/80 (middle, scale bar = 50 μm) and Sirius red staining (bottom, scale bar = 50 μm) of gastrocnemius muscle sections. The images were quantified and shown on the right (n = 5/group). (D) Western blot analysis for MyHC, α-actin, p-p65, MuRF1 and MAFbx in gastrocnemius muscle and quantification for western blot images (n = 5/group). Data are shown as mean ± SD. Statistical significance was evaluated using a one-way ANOVA test (A–D). *P [less than] 0.05, **P [less than] 0.01, ***P [less than] 0.001. n.s., not significant.
AGGF1 interacts with TWEAK and inhibits TWEAK–Fn14 signalling

To identify the molecular mechanism by which AGGF1 attenuates skeletal muscle atrophy, we examined the interaction of AGGF1 with TWEAK as TWEAK was shown to be associated with skeletal muscle atrophy.13 We hypothesized that AGGF1 inhibited the interaction of TWEAK with its ligand Fn14 through competitive binding with TWEAK. First, Co-IP analysis showed that AGGF1 interacted with TWEAK in C2C12 cells (Figure 8A). Similarly, in vivo in atrophy muscles, endogenous AGGF1 interacted with endogenous TWEAK (Figure 8C). Second, Co-IP analysis showed that knockdown of Aggf1 expression enhanced the interaction between TWEAK and Fn14 in C2C12 cells (Figure 8B). On the other hand, AGGF1 protein treatment in C2C12 cells significantly reduced the interaction between TWEAK and Fn14 in a dose dependent manner (Figure 8B). Similarly, in vivo in atrophy muscles, the interaction between TWEAK and Fn14 was enhanced in Aggf1+/− mice, and the effect was reversed by AGGF1 protein therapy (Figure 8D).

Figure 8. AGGF1 inhibits the TWEAK-Fn14 signalling pathway by interacting with TWEAK in vitro and in vivo. (A) Co-IP analysis for AGGF1 and TWEAK in C2C12 cells. (B) Co-IP analysis with C2C12 cells transfected with siRNA for Aggf1 (siAggf1) or scramble siRNA (siNC) or treated with different concentrations of AGGF1 or EB (n = 3/group). (C) Co-IP analysis for AGGF1 and TWEAK in denervated muscles. (D) Co-IP analysis with atrophy muscles isolated from Aggf1+/− versus Aggf1+/+ (WT) mice or mice treated with AGGF1 or EB (n = 6/group). (E,F) Western blot analysis for p-p65 in C2C12 cells and quantification (n = 3/group). Data are shown as mean ± SD. Statistical significance was evaluated using a one-way ANOVA test (B,D,E,F). *P [less than] 0.05, **P [less than] 0.01, n.s., not significant.

Phosphorylated-p65 is a major downstream signalling molecule in TWEAK-mediated loss of skeletal muscle mass.13 Thus, we also determined the role of AGGF1 in NF-κB activation in C2C12 cells. C2C12 cells were treated with the AGGF1 or EB under a normal condition or under starvation (serum-free) for 24 h. Similar to AGGF1 therapy in atrophy mice (Figures 3D, 5D, 7D, S12B, S26B and S40B), AGGF1 significantly reduced the level p-p65 in C2C12 cells under starvation, but not under a normal condition (Figure 8E). Interestingly, overexpression of TWEAK in C2C12 cells abrogated the inhibitory effect of AGGF1 on p-p65 under starvation (Figure 8E). Similar results were obtained in C2C12 cells treated with TNF-α or IFN-γ (Figure 8F). Altogether, these data support that AGGF1 inhibits TWEAK-Fn14 signalling, NF-κB (p-p65) activation and MuRF1 expression by competitive binding with TWEAK.

Discussion

In this study, we demonstrate that AGGF1 increased the muscle mass and myofibre CSA, partially reversed the characteristic loss of myoproteins such as MyHC and α-actin in denervated muscles and attenuated inflammation, apoptosis and fibrosis in atrophy mice (Figures 3, 5 and 7). Muscle myotubes isolated and cultured from impaired leg muscles from patients with lumbar disc herniation showed severely reduced CSA, MyHC and α-actin and increased MuRF1, and these effects were attenuated by AGGF1 treatment (Figures 1E, S2 and S3). These mouse and human data show that AGGF1 may be developed into drugs to treat patients with skeletal muscle atrophy in the future. AAS, GH and IGF-1 were shown to increase muscle mass in patients with muscle atrophy,6 although these treatment options were not officially approved by FDA. These agents may cause adverse effects. For example, AAS is a carcinogen that may cause cancer, stroke and hormonal disorders,6 whereas AGGF1 is a tumour-suppressor gene.32 GH treatment may cause soft tissue oedema, arthralgias, carpal tunnel syndrome and gynecomastia.7 Moreover, agents that increase the level of GH or IGF-1 may cause diabetes or insulin resistance.6 MK0773 is a small chemical agent that increases IGF-1 levels, but its clinical trial was terminated because it increased cardiovascular risk in female patients with sarcopenia.33 Therefore, AGGF1 therapy offers a new potential for treatment of muscle atrophy.

Our study identifies a novel molecular mechanism for the pathogenesis of skeletal muscle atrophy. We show that Aggf1 plays an important role in the progression of atrophy. Aggf1+/− mice showed aggravated skeletal muscle atrophy after denervation and implantation of LLC. Mechanistically, we found that AGGF1 interacted with TWEAK and disrupted the interaction between TWEAK and Fn14. This led to the inhibition of NF-κB activation and a reduced level of MuRF1, which inhibits degradation of MyHC and α-actin and attenuates skeletal muscle atrophy (Figures 3, 5, 7 and 8). These findings identify a molecular signalling pathway of AGGF1-TWEAK/Fn14-NF-κB-MuRF1-MyHC/α-actin for skeletal muscle atrophy (Figure S47).

Accumulating evidence suggests that TWEAK-Fn14 signalling affects multiple steps not only in the process of myogenesis but also in the degeneration of muscle.34 Low levels of exogenous TWEAK may transiently enhance myoblast fusion through predominately activating a non-canonical arm of NF-κB signalling.35 A small amount of TWEAK reduced the survival of differentiated myotubes and caused atrophy.14 The activation of NF-κB inhibited skeletal muscle regeneration in response to injury.10 Activation of NF-κB in skeletal muscle may up-regulate MuRF1 in response to a variety of catabolic stimuli, including denervation and tumour growth.13 However, it is unknown what regulates the TWEAK-Fn14 signalling pathway. Interestingly, by disrupting the interaction between TWEAK and Fn14, AGGF1 inhibited NF-κB activation (p65 phosphorylation) and thus reduced the level of MuRF1, but did not have any significant effect on MAFbx (Figures 3, 5 and 7). Cai et al. showed that activation of NF-κB through muscle-specific transgenic level of activated IκB kinase β increased the level of MuRF1, but did not have any effect on MAFbx.36 As AGGF1 inhibited NF-κB activation, it reduced the level of MuRF1, but did not have any significant effect on MAFbx.

In muscle atrophy, two muscle-specific E3 ubiquitin ligases, MAFbx and MuRF1, that label the target proteins for degradation by 26S proteasome, are highly up-regulated.9,37 Muscle atrophy is inhibited in mice null for MuRF1 or MAFbx, suggesting that these ubiquitin ligases are the master drivers of muscle proteolysis.38 Consistent with these findings, our study suggests that one of the mechanisms by which AGGF1 inhibits degradation of muscle proteins is through the reduced level of MuRF1 in denervated muscle (Figures 3, 5 and 7). Our data are in agreement with a recent report that MuRF1 targets thick filament proteins MyHC in skeletal muscle.39

AGGF1 promotes angiogenesis,22 and recent studies reported that angiogenesis could promote muscle regeneration.40 Therefore, it is possible that reduced angiogenesis in Aggf1+/− mice may contribute to exacerbation of skeletal muscle atrophy in these mice. Meanwhile, increased angiogenesis by AGGF1 may attenuate muscle atrophy.

AGGF1 was significantly increased in skeletal muscles from denervated and cachexia mice models, and patients with lumbar disc herniation. This may reflect an adaptive expression remodelling effect by AGGF1. AGGF1 may be induced in atrophy muscles to prevent a more severe muscle atrophy. One limitation of the present study should be noted. Although we showed that injection of AGGF1 attenuated muscle atrophy, our mouse models could not allow us to test whether attenuated muscle atrophy by AGGF1 resulted in increased muscle strength. This question may be addressed using different models in the future.

In summary, our study identified an important role of AGGF1 in regulation of the pathogenesis of skeletal muscle atrophy. Moreover, our results uncovered a novel signalling pathway and a fundamental mechanism for the pathogenesis of skeletal muscle atrophy, which is consisted of AGGF1, TWEAK, NF-κB(p65), MuRF1, MyHC and α-actin. AGGF1 protein therapy may be developed into a novel treatment strategy for patients with skeletal muscle atrophy.

Acknowledgements

We appreciate Zhaoyi Gong for help with analysis of mouse lean mass data.

Conflict of interests

The authors declare no conflicts of interest with this article.

Funding

This study was supported by the National Natural Science Foundation of China (81630002, 32270662, 31430047 and 82000439) and the Fundamental Research Funds for the Central Universities HUST (2172020kfyXJJS116).

Word count: 5487

Show less

© 2023. This work is published under http://creativecommons.org/licenses/by-nc/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.

Abstract

Background

Skeletal muscle atrophy is a common condition without a pharmacologic therapy. AGGF1 encodes an angiogenic factor that regulates cell differentiation, proliferation, migration, apoptosis, autophagy and endoplasmic reticulum stress, promotes vasculogenesis and angiogenesis and successfully treats cardiovascular diseases. Here, we report the important role of AGGF1 in the pathogenesis of skeletal muscle atrophy and attenuation of muscle atrophy by AGGF1.

Methods

In vivo studies were carried out in impaired leg muscles from patients with lumbar disc herniation, two mouse models for skeletal muscle atrophy (denervation and cancer cachexia) and heterozygous Aggf1+/− mice. Mouse muscle atrophy phenotypes were characterized by body weight and myotube cross-sectional areas (CSA) using H&E staining and immunostaining for dystrophin. Molecular mechanistic studies include co-immunoprecipitation (Co-IP), western blotting, quantitative real-time PCR analysis and immunostaining analysis.

Results

Heterozygous Aggf1+/− mice showed exacerbated phenotypes of reduced muscle mass, myotube CSA, MyHC (myosin heavy chain) and α-actin, increased inflammation (macrophage infiltration), apoptosis and fibrosis after denervation and cachexia. Intramuscular and intraperitoneal injection of recombinant AGGF1 protein attenuates atrophy phenotypes in mice with denervation (gastrocnemius weight 81.3 ± 5.7 mg vs. 67.3 ± 5.1 mg for AGGF1 vs. buffer; P < 0.05) and cachexia (133.7 ± 4.7 vs. 124.3 ± 3.2; P < 0.05). AGGF1 expression undergoes remodelling and is up-regulated in gastrocnemius and soleus muscles from atrophy mice and impaired leg muscles from patients with lumbar disc herniation by 50–60% (P < 0.01). Mechanistically, AGGF1 interacts with TWEAK (tumour necrosis factor-like weak inducer of apoptosis), which reduces interaction between TWEAK and its receptor Fn14 (fibroblast growth factor-inducing protein 14). This leads to inhibition of Fn14-induced NF-kappa B (NF-κB) p65 phosphorylation, which reduces expression of muscle-specific E3 ubiquitin ligase MuRF1 (muscle RING finger 1), resulting in increased MyHC and α-actin and partial reversal of atrophy phenotypes. Autophagy is reduced in Aggf1+/− mice due to inhibition of JNK (c-Jun N-terminal kinase) activation in denervated and cachectic muscles, and AGGF1 treatment enhances autophagy in two atrophy models by activating JNK. In impaired leg muscles of patients with lumbar disc herniation, MuRF1 is up-regulated and MyHC and α-actin are down-regulated; these effects are reversed by AGGF1 by 50% (P < 0.01).

Conclusions

These results indicate that AGGF1 is a novel regulator for the pathogenesis of skeletal muscle atrophy and attenuates skeletal muscle atrophy by promoting autophagy and inhibiting MuRF1 expression through a molecular signalling pathway of AGGF1-TWEAK/Fn14-NF-κB. More importantly, the results indicate that AGGF1 protein therapy may be a novel approach to treat patients with skeletal muscle atrophy.

Details

Title
Protein therapy of skeletal muscle atrophy and mechanism by angiogenic factor AGGF1
Author
He, Zuhan 1 ; Song, Qixue 1 ; Yu, Yubing 1 ; Liu, Feng 2 ; Zhao, Jinyan 1 ; Waikeong Un 3 ; Da, Xingwen 1 ; Xu, Chengqi 1 ; Yao, Yufeng 1 ; Wang, Qing K 1   VIAFID ORCID Logo 

1  Center for Human Genome Research, Key Laboratory of Molecular Biophysics of the Ministry of Education, College of Life Science and Technology, Huazhong University of Science and Technology, Wuhan, China 
2  Department of Orthopedics, Renmin Hospital of Wuhan University, Wuhan, China 
3  Department of Urology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China 
Pages
978-991
Section
Original Articles
Publication year
2023
Publication date
Apr 2023
Publisher
John Wiley & Sons, Inc.
ISSN
21905991
e-ISSN
21906009
Source type
Scholarly Journal
Language of publication
English
DOI
https://doi.org/10.1002/jcsm.13179
ProQuest document ID
2793791020
Copyright
© 2023. This work is published under http://creativecommons.org/licenses/by-nc/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.