Conservative analysis of Synaptopodin-2 intron sense-overlapping lncRNA reveals its novel function in promoting muscle atrophy.

BACKGROUND: Dissection of the regulatory pathways that control skeletal muscle development and atrophy is important for the treatment of muscle wasting. Long noncoding RNA (lncRNA) play important roles in various stages of muscle development. We previously reported that Synaptopodin-2 (SYNPO2) intron sense-overlapping lncRNA (SYISL) regulates myogenesis through an interaction with enhancer of zeste homologue 2 (EZH2). However, it remains unclear whether SYISL homologues exist in humans and pigs, and whether the functions and mechanisms of these homologues are conserved among species.
METHODS: Bioinformatics, cell fractionation, and quantitative real-time polymerase chain reaction (qRT-PCR) analyses were used for the identification and molecular characterization of SYISL homologues in humans and pigs. Effects on myogenesis and muscle atrophy were determined via loss-of-function or gain-of-function experiments using C2C12 myoblasts, myogenic progenitor cells, dexamethasone (DEX), and aging-induced muscle atrophy models. RNA pulldown, RNA immunoprecipitation, dual luciferase reporting, and co-transfection experiments were used to explore the mechanisms of SYISL interactions with proteins and miRNAs.
RESULTS: We identified SYISL homologues in humans (designated hSYISL) and pigs (designated pSYISL). Functional experiments demonstrated that hSYISL and pSYISL regulate myogenesis through interactions with EZH2. Interestingly, we showed that SYISL functions to regulate muscle atrophy and sarcopenia through comparative analysis. SYISL is significantly up-regulated after muscle atrophy (P < 0.01); it significantly promotes muscle atrophy in DEX-induced muscle atrophy models (P < 0.01). SYISL knockdown or knockout alleviates muscle atrophy and sarcopenia in DEX-induced and aged mice. The tibialis anterior (TA) muscle weight of 3-month-old wild-type (WT) mice decreased by 33.24% after DEX treatment (P < 0.001), while the muscle weight loss of 3-month-old SYISL knockout mice was only 18.20% after DEX treatment (P < 0.001). SYISL knockout in 18-month-old WT mice significantly increased the weights of quadriceps (Qu), gastrocnemius (Gas), and TA muscles by 10.45% (P < 0.05), 13.95% (P < 0.01), and 24.82% (P < 0.05), respectively. Mechanistically, SYISL increases the expression levels of the muscle atrophy genes forkhead box protein O3a (FoxO3a), muscle ring finger 1 (MuRF1), and muscle atrophy-related F-box (Atrogin-1) via sponging of miR-23a-3p/miR-103-3p/miR-205-5p and thus promotes muscle atrophy. Additionally, we verified that human SYISL overexpression in muscles of 18-month-old WT mice significantly decreased the weights of Gas, Qu, and TA muscles by 7.76% (P < 0.01), 12.26% (P < 0.05), and 13.44% (P < 0.01), respectively, and accelerates muscle atrophy through conserved mechanisms.
CONCLUSIONS: Our results identify SYISL as a conserved lncRNA that modulates myogenesis in mice, pigs, and humans. We also demonstrated its previously unknown ability to promote muscle atrophy.

Introduction

Normal muscle development is controlled by multiple strict regulatory networks and disordered regulation of gene expression leads to various muscle diseases. , Muscle atrophy or sarcopenia occurs in specific muscles under conditions of inactivity, denervation, aging, and diseases, , and it is characterized by progressive and extensive loss of skeletal muscle mass and strength, as well as a decrease in protein synthesis. Muscle atrophy leads to protein degradation through activation of two proteolytic systems: the ubiquitin‐proteasome system (UPS) and the autophagy‐lysosome system. , Most muscle proteins are degraded by ubiquitin ligases in the UPS system, such as MuRF1 and Atrogin‐1, which are the core components of the UPS ; these proteins are excellent markers of muscle atrophy. , Autophagy is another proteolysis system with a crucial role in the turnover of specific cell organelles (e.g. mitochondria) and abnormal proteins. , , Activation of the UPS and autophagy during muscle atrophy is regulated by multiple genes and signalling pathways, including the insulin‐like growth factor‐1 (IGF‐1)‐phosphatidylinositol3‐kinase (PI3K)‐protein kinase B (PKB; also called Akt)‐forkhead box protein O (FoxO), nuclear factor κB (NF‐κB) signalling pathways, and so on. , Recent studies have revealed that FoxOs promote the expression of atrophy‐related and autophagy‐related genes, which then activate the UPS and autophagy, leading to muscle atrophy. , Most eukaryotic genome contents comprise noncoding RNA; only a few regions represent protein‐coding genes. Long noncoding RNAs (LncRNAs) have emerged as crucial regulators of numerous biological processes. , Although lncRNAs have low conservation among species, particularly in terms of nucleotide sequences, many lncRNAs (e.g. Neat1, Linc‐YY1, H19, and lncMGPF ) exhibit cross‐species conservation of genomic positions, functions, and regulatory mechanisms. We previously showed that Synaptopodin‐2 (SYNPO2) intron sense‐overlapping lncRNA (SYISL) promotes myoblast proliferation but inhibits myogenic differentiation. SYISL knockout (KO) in mice results in the increased muscle fibre density and muscle mass. Mechanistically, SYISL recruits polycomb repressive complex 2 (PRC2) to the promoters of the cell‐cycle inhibitor gene p21 and muscle‐specific genes such as myogenin (MyoG), leading to H3K27 trimethylation and epigenetic silencing of target genes. Bioinformatics analysis revealed that AK021986 (renamed hSYISL) and AK238214 (renamed pSYISL) are transcribed from the intron of the SYNPO2 gene in humans and pigs, respectively. However, it is not clear whether the homologous transcripts of SYISL are functionally and mechanistically conserved in humans and pigs. In this study, we investigated the roles of human and pig SYISL homologue transcripts in muscle development and atrophy. Our results revealed that the functions of SYISL in myogenesis regulation are conserved among mice, pigs, and humans. Interestingly, comparative analysis of SYISL among mice, pigs, and humans revealed a novel function in promoting muscle atrophy. SYISL increases the expression levels of muscle atrophy genes (FoxO3a, MuRF1, and Atrogin‐1) via sponging of miR‐23a‐3p/miR‐103‐3p/miR‐205‐5p, activates the UPS and autophagy‐lysosome system, and thus aggravates muscle atrophy. Our findings highlight the conservation of SYISL among different species and its potential as a therapeutic target for sarcopenia.

Materials and methods

An extended materials and methods section can be found online in the Supporting Information.

Animals

All C57BL/6J wild‐type (WT) mice and SYISL KO mice used for the study were from the Experimental Animal Center of Huazhong Agricultural University. SYISL KO mice were originally generated by our lab. All pigs were from the experimental pig farm of Huazhong Agricultural University.

Measurement of phenotypes

The whole leg, gastrocnemius (Gas), tibialis anterior (TA), and quadriceps (Qu) muscles of KO and WT mice were collected and weighed. Data were normalized to the body weight (mg/g). The strength test was performed using a grip strength metre (BIO‐GS3; Bioseb, France).

Dual‐luciferase reporter assay

The sequences of 3′untranslated regions (3′UTRs) of MuRF1, Atrogin‐1, and FoxO3a containing miR‐23a‐3p, miR‐103‐3p, and miR‐205‐5p binding sites, and SYISL, hSYISL, and pSYISL with unmutated or mutated binding sites were cloned into the pmirGLO Dual‐Luciferase miRNA Target Expression Vectors (E1330; Promega, USA), respectively. Those reporter vectors were transfected or co‐transfected into C2C12 or HeLa cells using Lipofectamine 2000 and then tested according to a previously published method.

Muscle atrophy models

For dexamethasone (DEX)‐induced muscle atrophy model, 3‐month‐old WT and SYISL KO mice were intraperitoneally injected with DEX (HY‐14648; 25 μg/g weight, MedChemExpress, China) once a day for 10 days. For the starvation‐induced muscular atrophy model, 3‐month‐old WT mice were fasted for 5 days but given free water. For the sarcopenia model, WT mice with the same birth date and conditions grew up to 18 months old.

Cell immunofluorescence staining

Cell immunofluorescence staining was performed according to the previous literature. Antibodies included a primary antibody MyHC (sc‐376157; 1:200; Santa Cruz Biotechnology, USA) and a secondary antibody (A0521; goat anti‐mouse CY3; Beyotime Biotechnology, China). The images were visualized with a fluorescence microscope (IX51‐A21PH; Olympus, Japan).

Statistical analysis

Statistical analyses among different groups were performed using unpaired or paired Student's t‐test. For all analyses, * P < 0.05 and ** P < 0.01 were considered to be statistically significant. All data were presented as mean ± standard deviation (SD).

Results

SYNPO2 intron sense‐overlapping lncRNA exhibits conserved genomic position and expression pattern among mice, pigs, and humans

Mouse SYISL is transcribed from the fourth intron of the SYNPO2 gene and negatively regulates muscle growth. To explore whether SYISL homologues occur in pigs and humans, we searched for lncRNAs transcribed from the intron of SYNPO2 in human and pig genomes using the University of California Santa Cruz database (UCSC database). We identified transcript AK238214 (i.e. pSYISL) in pigs; it is transcribed from the first intron of the pig SYNPO2 gene. We also identified transcript AK021986 (i.e. hSYISL) in humans; it is transcribed from the fourth intron of the human SYNPO2 gene (Figure S1A). Bioinformatics analysis showed that they have low coding potential (Figure S1B), short conserved sequences with SYISL (Figure S1C), and conserved loop structures that may bind to proteins (Figure S1D). In our previous study, we showed that SYISL could interact with enhancer of zeste homologue 2 (EZH2) to form a larger structure. Therefore, we predicted the capacities of pSYISL and hSYISL to bind EZH2 by using the machine learning classifier RNA‐protein interaction prediction (RPISeq). Notably, EZH2 had high affinity for SYISL, pSYISL, and hSYISL (Figure S1E). Moreover, RegRNA2 prediction of miRNA binding sites showed that mouse, pig, and human SYISL contain 49, 108, and 68 potential miRNA‐binding sites, respectively, among which binding sites for five miRNAs (miR‐103‐3p, miR‐23a‐3p, miR‐24a‐3p, miR‐27a‐3p, and miR‐205‐5p) were common (Figure 1A). Binding interactions with EZH2 and SYISL‐specific miRNAs were confirmed via subsequent analysis.

Figure 1

SYNPO2 intron sense‐overlapping lncRNA exhibits conserved genomic position, subcellular distribution, and expression pattern among mice, pigs, and humans. (A) Venn diagram shows potential miRNAs that can binding to SYISL, hSYISL, and pSYISL. The overlapping results show that miR‐103‐3p, miR‐23a‐3p, miR‐24a‐3p, miR‐27a‐3p, and miR‐205‐5p are the predicted common miRNAs. (B,C) qRT‐PCR results showing the distribution of pSYISL and hSYISL in the cytoplasm and nucleus of proliferating myoblasts (D0), myotubes differentiated for 6 days (D6), and D6 atrophy myotubes after dexamethasone (DEX) treatment (D6 + DEX). (D) qRT‐PCR results show that pSYISL is highly expressed in muscle tissues including the longissimus dorsi, leg muscle, and heart. (E) qRT‐PCR results show that the expression levels of hSYISL and pSYISL gradually increase during myoblast differentiation in human skeletal muscle myoblasts and pig primary myoblasts, respectively. (F) qRT‐PCR results show that pSYISL expression level gradually increases during postnatal longissimus dorsi muscle development. The relative RNA levels are normalized to GAPDH. The data are presented as mean ± SD.

Next, we examined the subcellular distribution and expression profiles of hSYISL and pSYISL. Cell fractionation assays demonstrated that pSYISL is mainly localized to the nuclei, while hSYISL is localized to both nuclei and cytoplasm in proliferating myoblasts (D0), myotubes differentiated for 6 days (D6), and D6 atrophy myotubes after DEX treatment (D6 + DEX) (Figure 1B,C); the subcellular localizations of hSYISL and pSYISL were confirmed by RNA FISH (Figure S1F,G). Our previous lncRNA database results showed that hSYISL is highly expressed in human muscles. Further analysis of the enrichment pathways of genes co‐expressed with hSYISL in the AnnoLnc2 database indicated that hSYISL may be involved in pathways related to muscle development and muscle contraction (Figure S1H). Similar to mouse and human SYISL, pSYISL was highly expressed in skeletal muscles (Figure 1D), and the expression levels of hSYISL and pSYISL increased with the differentiation of human skeletal muscle myoblasts and pig primary myoblasts, respectively (Figure 1E). Moreover, the expression of pSYISL increased during skeletal muscle growth after birth (Figure 1F; Figure S1I). Prediction of the binding sites of transcription factors showed that the pSYISL promoter region contains multiple potential binding sites of transcription factors such as MyoD, MyoG, Myf5, MEF2C, and FoxO3, which are involved in embryonic and postnatal muscle development, as well as muscle atrophy (Figure S1J). Therefore, we speculated that the increased expression levels of pSYISL after birth and during myogenic differentiation may be induced by these transcription factors. These hypotheses need further validation in future. In summary, we concluded that SYISL exhibits conservation of its genomic position and expression pattern among mice, pigs, and humans.

Regulation of myogenesis by SYNPO2 intron sense‐overlapping lncRNA is conserved in pigs and humans

To explore whether the SYISL myogenesis function is conserved in pigs and humans, we used loss‐of‐function and gain‐of‐function experiments to study the functions of hSYISL and pSYISL. Compared with the control, overexpression of hSYISL and pSYISL promoted the proliferation capacities of human skeletal muscle myoblasts (Figure 2A; Figure S2A,C), and pig primary myoblasts (Figure 2B; Figure S2B,D), respectively, while hSYISL and pSYISL knockdown inhibited myoblast proliferation (Figure S2E–I). Furthermore, we investigated the influences of hSYISL and pSYISL on myoblast differentiation in human skeletal muscle myoblasts and pig primary myoblasts, respectively. The results showed that overexpression of hSYISL and pSYISL inhibited myoblast differentiation (Figure 2C,D; Figure S2J–M), while knockdown of hSYISL and pSYISL promoted myoblast differentiation (Figure S2N–Q). These results were consistent with our previous findings in mice. To investigate whether pSYISL and hSYISL exhibit cross‐species functional conservation, we expressed pSYISL and hSYISL in C2C12 cells. The results showed that overexpression of hSYISL and pSYISL significantly promoted cell proliferation (Figure 2E; Figure S3A–C) and inhibited C2C12 cell differentiation (Figure 2F; Figure S3D–F), which was consistent with the effects of SYISL overexpression. We further examined the expression of SYNPO2 and no significant change was found after overexpression of hSYISL and pSYISL (Figure S3G–J). Taken together, these results demonstrated that SYISL functions to regulate myogenesis in mice, pigs, and humans.

Figure 2

Regulation of myogenesis by SYNPO2 intron sense‐overlapping lncRNA is conserved in pigs and humans. (A,B) Representative photographs of EdU staining in human skeletal muscle myoblasts and pig primary myoblasts and quantification show that overexpression of hSYISL in human skeletal muscle myoblasts (A) and pSYISL in pig primary myoblasts (B) can significantly increase myoblast proliferation. Nuclei are stained with DAPI, scale bars, 100 μm. (C,D) Representative photographs of MyHC immunofluorescence staining and quantification show that hSYISL and pSYISL overexpression inhibits myoblasts differentiation in human skeletal muscle myoblasts (C) and pig primary myoblasts (D), respectively. Positively stained cells are quantified; nuclei are stained with DAPI. Scale bars, 100 μm. (E) Representative photographs of EdU staining and quantification show that hSYISL, SYISL, and pSYISL overexpression in C2C12 cells significantly increases EdU incorporation and cell proliferation. Nuclei are stained with DAPI. Scale bars, 100 μm. (F) Representative photographs of immunofluorescence staining for MyHC and quantification show that overexpression of hSYISL, SYISL, and pSYISL in C2C12 cells significantly inhibits myoblast differentiation. Nuclei are stained with DAPI. Scale bars, 100 μm. The data of three independent experiments are presented as mean ± SD; *P < 0.05, **P < 0.01. N.S. indicates statistical non‐significance.

Next, we further verified the conservative functional of hSYISL and pSYISL on muscle growth and development in 1‐month‐old WT mice. Compared with control group, overexpression of hSYISL and pSYISL in mice muscles significantly reduced the muscle sizes and weights (Figure 3A,B; Figure S4A,B). EGFP immunofluorescence staining in TA muscles showed that the infection efficiency of lentivirus was 100% in both control and overexpression groups, excluding the effect of transfection efficiency on muscle growth (Figure S4C,D). Moreover, haematoxylin and eosin (H&E) and Lamin immunofluorescence staining showed that hSYISL or pSYISL overexpression significantly reduced the cross‐sectional areas of the Qu, Gas, and TA muscles (Figure 3C,D; Figure S4E,F). Quantification of the number of centrally localized nuclei after infection showed that there were no centrally localized nuclei after infection of lentivirus, indicating the reduction in muscle size is due to the reduced muscle growth, not to the myopathy. MyoD and MyoG immunofluorescence staining in muscles showed that overexpression of hSYISL significantly decreased the number of MyoG positive cells but did not significantly affect the number of MyoD positive cells in the interstitial cell matrix (Figure S4G,H); these results were also confirmed by Western blotting (Figure S4I,J). Then, we further verified the functional conservation of pSYISL during pig muscle development, the inhibiting effects of pSYISL on pig muscle growth, and myogenic differentiation were consistent with the results in mice (Figure 3E,F; Figure S4K,L). These results indicated that hSYISL and pSYISL inhibit muscle growth and development in vivo.

Figure 3

Human and pig SYNPO2 intron sense‐overlapping lncRNA inhibit the muscle growth and development in vivo. (A,B) Representative photographs of the muscle mass for quadriceps (Qu), gastrocnemius (Gas), and tibialis anterior (TA) of 1‐month‐old mice injected with lentivirus‐mediated hSYISL overexpression (LV‐hSYISL) vector (A) or pSYISL overexpression (LV‐pSYISL) vector (B). Quantification of six independent experiments shows that lentivirus‐mediated hSYISL (A) or pSYISL (B) overexpression significantly decreases the weights of Gas, TA, and Qu muscles. P values are determined by paired t‐test. Data are normalized to the body weight (BW) (mg/g). (C,D) Representative images of Lamin immunofluorescence staining for Qu, Gas, and TA muscles of mice. Quantification of three independent experiments indicates that lentivirus‐mediated hSYISL (C) or pSYISL (D) overexpression in muscles of mice significantly decreases the average cross‐sectional areas of individual myofibre. At least 150 myofibres are analysed in an independent experiment. Scale bar, 50 μm. (E) Western blotting results of pig biceps femoris muscles show that overexpression of pSYISL significantly increases the expression of Ki67 and CDK6 and decreases the protein expression levels of MyoG, MyHC, and p21 but does not affect MyoD expression. (F) Representative images of Lamin immunofluorescence staining for the left and right biceps femoris muscles of 1‐month‐old piglets injected with LV‐pSYISL vector or LV‐control vector. Quantification of four independent experiments shows that pSYISL overexpression decreases average cross‐sectional areas of individual myofibre. At least 150 myofibres are analysed in an independent experiment. Scale bar, 50 μm. The relative protein levels are normalized to GAPDH. The data of three independent experiments are presented as mean ± SD; *P < 0.05, **P < 0.01. N.S. indicates statistical non‐significance.

We previously verified that SYISL regulates myogenesis and inhibits the expression of target genes such as p21, MyoG, and MyHC by recruiting EZH2/H3K27me3 to its target gene promoters; it has been unclear whether this mechanism is conserved in other species. Therefore, we investigated the conservation of this mechanism in pig primary myoblasts. RNA pulldown and immunoprecipitation results confirmed the interaction between EZH2 and pSYISL (Figure S5A,B), and pSYISL overexpression significantly increased the enrichment of H3K27me3 at the proliferation and differentiation stages (Figure S5C). Chromatin immunoprecipitation (ChIP)‐quantitative real‐time polymerase chain reaction (qRT‐PCR) results showed that pSYISL overexpression could significantly increase the binding of EZH2 and H3K27me3 to the p21, MyoG, and myosin heavy chain 4 (MyH4) gene promoters (Figure S5D,E); knockdown of pSYISL could significantly reduce these binding capacities (Figure S5F,G). These results suggested that the molecular mechanism whereby SYISL regulates myogenesis via EZH2/H3K27me3 is conserved.

Comparative analysis of SYNPO2 intron sense‐overlapping lncRNA among mice, pigs, and humans reveals its novel function in promoting muscle atrophy

The bioinformatics analysis described above showed that the overlapping miRNAs among mice, pigs, and humans were miR‐103‐3p, miR‐23a‐3p, miR‐24a‐3p, miR‐27a‐3p, and miR‐205‐5p. Dual luciferase reporter experiments and qRT‐PCR results showed that miR‐103‐3p, miR‐23a‐3p, and miR‐205‐5p could simultaneously bind to SYISL, pSYISL, and hSYISL and significantly decreased the expression level of SYISL (Figure 4A,B; Figure S6A,B). We verified the binding of SYISL to Ago2 via RNA immunoprecipitation (Figure S6C), suggesting that SYISL functions via miRNA targeting. To confirm that these three miRNAs could directly bind to SYISL, pSYISL, and hSYISL, we performed dual luciferase reporter analysis of SYISL sequences containing the miR‐23a‐3p, miR‐103‐3p, and miR‐205‐5p binding sites, as well as their mutated binding sites. The results showed that overexpression of miR‐23a‐3p, miR‐103‐3p, and miR‐205‐5p significantly reduced the relative luciferase activities of SYISL, pSYISL, and hSYISL, respectively; no significant effects were observed upon mutation of these binding sites (Figure S6D–L). Among these three miRNAs, miR‐23a‐3p is known to regulate muscle atrophy. Additionally, we performed pathway analysis of SYISL target genes using our reported gene chip results (GSE102087). The differentially expressed genes mainly participated in signalling pathways related to muscle atrophy, such as PI3K‐Akt, FoxO, transforming growth factor‐β (TGF‐β), and ubiquitin‐mediated proteolysis (Figure S6M,N). Therefore, we hypothesized that SYISL may be involved in muscle atrophy. To investigate this hypothesis, we examined the expression patterns of SYISL and related miRNAs in muscle atrophy models induced by aging, DEX treatment, and fasting. Among these models, the expression levels of MuRF1, SYISL, miR‐205‐5p, and other muscle atrophy‐associated E3 ubiquitin ligases significantly increased, while the expression levels of miR‐23a‐3p and miR‐103‐3p significantly decreased; no significant change for the expression of SYNPO2 was observed in different muscle atrophy models (Figure 4C–E). Next, we examined the expression changes of these three miRNAs and SYISL at various stages of muscle growth. The expression levels of SYISL increased with age (Figure 4F), consistent with the pattern of MuRF1 expression (Figure S6O). In addition, the expression level of miR‐23a‐3p gradually decreased with age, and the expression of miR‐103‐3p remained stable during postnatal muscle development, while the expression level of miR‐205‐5p peaked at the age of 12 months (Figure S6P–R). These results suggest that SYISL is involved in the process of muscle atrophy.

Figure 4

Comparative analysis of SYNPO2 intron sense‐overlapping lncRNA among mice, pigs, and humans reveals its potential function in promoting muscle atrophy. (A) Dual luciferase reporter assays show that overexpression of miR‐23a‐3p, miR‐103‐3p, and miR‐205‐5p can respectively reduce the dual luciferase activities of SYISL. (B) qRT‐PCR results show that overexpression of miR‐23a‐3p, miR‐103‐3p, and miR‐205‐5p can reduce the expression level of SYISL, respectively. (C) qRT‐PCR results show that DEX‐induced myotube atrophy can increase the expression levels of MuRF1, Atrogin‐1, FoxO3a, SYISL, and miR‐205‐5p, decrease the expression levels of miR‐23a‐3p and miR‐103‐3p, but does not affect the expression level of SYNPO2. (D,E) qRT‐PCR results show that the expression levels of MuRF1, Atrogin‐1, FoxO3a, SYISL, and miR‐205‐5p are up‐regulated while miR‐23a‐3p and miR‐103‐3p are down‐regulated in DEX‐induced muscle atrophy (D) and starvation‐induced muscle atrophy models (E), while there is no significant change for the expression level of SYNPO2. Muscle atrophy and starvation‐induced muscle atrophy are induced in 3‐month‐old wild‐type mice by intraperitoneal injection of DEX for 10 days and fasted for 5 days, respectively. (F) qRT‐PCR results show that the expression level of SYISL is up‐regulated with aging. The relative miRNA levels are normalized to U6. The relative lncRNA and mRNA levels are normalized to GAPDH. The data of three independent experiments are presented as mean ± SD; *P < 0.05, **P < 0.01. N.S. indicates statistical non‐significance.

To explore the roles of SYISL in muscle atrophy in vitro, we overexpressed and knocked down SYISL in C2C12 myotubes and myotubes after DEX treatment. The results showed that SYISL overexpression significantly increased the expression levels of FoxO3a, Atrogin‐1, and MuRF1 (Figure S7A,B); SYISL knockdown inhibited the expression of these genes (Figure S7C,D). Meanwhile, SYISL overexpression further promoted the expression of DEX‐induced muscle atrophy genes (Figure S7E) and reduced the diameter of myotubes (Figure 5A). In contrast, SYISL knockdown alleviated DEX‐induced myotube atrophy (Figure S7F,G). To determine the roles of SYISL in muscle atrophy in vivo, we performed SYISL KO, lentivirus‐mediated overexpression, or knockdown experiments in DEX‐induced muscle atrophy animal models. We found no significant differences in the expression levels of FoxO3a, Atrogin‐1, and MuRF1 between 3‐month‐old WT and SYISL KO mice (Figure S7H), mainly because 3‐month‐old mice do not exhibit an atrophy phenotype. Therefore, we treated 3‐month‐old mice with DEX to induce muscle atrophy. SYISL overexpression aggravated the DEX‐induced loss of TA weight (Figure 5B). H&E, Lamin immunofluorescence, and Sirius red staining showed SYISL overexpression further reduced the TA muscle cross‐sectional area and aggravated DEX‐induced muscle fibrosis (Figure 5C). Moreover, SYISL overexpression significantly increased the expression of ubiquitinated proteins (Figure 5D), as well as the expression levels of MuRF1, Atrogin‐1, and FoxO3a (Figure S7I). In contrast, SYISL knockdown alleviated DEX‐induced TA muscle weights loss and atrophy phenotype and reduced the expression levels of DEX‐induced muscle atrophy genes (Figure S7J–L). Quantification of the number of centrally localized nuclei after infection showed that there was no significant difference between control and SYISL overexpression (Figure S7M), or knockdown groups (Figure S7N), indicating the muscle atrophy by SYISL is not caused by the myopathy. To further characterize the contribution of SYISL to the regulation of DEX‐induced muscle atrophy, 3‐month‐old WT and SYISL KO mice were treated with DEX to induce muscle atrophy. The results showed that TA muscle weight of 3‐month‐old WT mice decreased by 33.24% after DEX treatment (P < 0.001), while the muscle weight loss of 3‐month‐old SYISL KO mice was only 18.20% after DEX treatment (P < 0.001) (Figure 5E). SYISL KO significantly alleviated the DEX‐induced reduction in muscle fibre cross‐sectional area (Figure 5F) and inhibited the expression of muscle atrophy genes (Figure S7O). Taken together, these results indicated that SYISL aggravates DEX‐induced muscle atrophy in vivo and in vitro.

Figure 5

SYNPO2 intron sense‐overlapping lncRNA promotes myotube atrophy in vitro and aggravates DEX‐induced muscle atrophy in mice. (A) Representative photographs of MyHC immunofluorescence for overexpression and knockdown SYISL in DEX‐treated myotubes. Quantification of three independent experiments indicates that SYISL overexpression can further reduce the diameter of myotubes. Scale bar, 100 μm. (B) Quantification of six independent experiments shows that lentivirus‐mediated SYISL overexpression can aggravate muscle weight loss caused by DEX‐induced muscle atrophy. (C) Representative photographs of H&E, Lamin immunofluorescence, and Sirius red staining for TA muscles. Quantification of the three independent experiments shows that overexpression of SYISL further reduces the cross‐sectional areas of TA muscles in DEX‐induced muscle atrophy and increases the muscle fibrosis. At least 150 myofibres are analysed in an independent experiment. Scale bar, 50 μm (Lamin), 100 μm (H&E). (D) Western blotting results show that SYISL overexpression can further promote protein ubiquitination levels in DEX‐induced muscular atrophy. Muscle atrophy is induced by intraperitoneal injection of DEX for 10 days in 3‐month‐old wild‐type and KO mice, respectively. Intraperitoneal injection of PBS is used as control. (E) Quantification of five independent experiments shows that SYISL knockout counteracts the muscle weight loss caused by DEX‐induced muscle atrophy. (F) Representative photographs of H&E and Lamin immunofluorescence staining for WT and KO TA muscles. Quantification of five independent experiments shows that knockout of SYISL alleviates muscle weight loss caused by DEX‐induced muscle atrophy. At least 150 myofibres are analysed in an independent experiment. Scale bar, 50 μm (Lamin), 100 μm (H&E). The relative protein levels are normalized to GAPDH. The data of at least three independent experiments are presented as means ± SD. *P < 0.05, **P < 0.01. N.S. indicates statistical non‐significance.

Based on the results presented above, we investigated the roles of SYISL in muscle atrophy of aged mice. Compared with WT mice, 18‐month‐old SYISL KO mice had significantly greater grasping force (Figure S8A), and SYISL KO significantly increased the weights of Qu, Gas, and TA muscles by 10.45% (P < 0.05), 13.95% (P < 0.01), and 24.82% (P < 0.05), respectively (Figure 6A). Compared with 18‐month‐old WT mice, the expression levels of MuRF1, FoxO3a, and Atrogin‐1 in 18‐month‐old SYISL KO mice were significantly decreased (Figure S8B,C), while no significant difference for the expression of SYNPO2 was found (Figure S8D,E). Lamin immunofluorescence and Sirius red staining results showed that the degree of muscle atrophy in SYISL KO mice was significantly reduced, compared with WT mice (Figure 6B; Figure S8F). The results of Lamin immunofluorescence staining showed that the cross‐sectional areas of muscle fibres were significantly smaller in SYISL KO mice than in WT mice at 3 and 8 months; they were significantly larger at 18 months (Figure 6C). Notably, no significant differences in the expression of MuRF1, FoxO3a, or Atrogin‐1 were found between 3‐month‐old SYISL KO and WT mice; however, their expression levels were significantly lower in 18‐month‐old SYISL KO mice than WT mice (Figure S8G). Next, we performed lentivirus‐mediated SYISL interference experiments to determine whether SYISL knockdown in 18‐month‐old WT mice had a similar effect. The results showed SYISL knockdown significantly increased muscle sizes (Figure S8H) and weights (Figure 6D) and alleviated the muscle atrophy phenotype (Figure 6E; Figure S8I); it significantly decreased the expression levels of MuRF1, FoxO3a, and Atrogin‐1 (Figure 6F; Figure S8J). In conclusion, SYISL promotes muscle atrophy in aged mice; knockdown or KO of SYISL alleviates sarcopenia.

Figure 6

Knockdown or knockout of SYNPO2 intron sense‐overlapping lncRNA alleviates muscle atrophy in aged mice. (A) Quantification of five independent experiments for Qu, Gas, and TA muscles in 18‐month‐old KO and WT mice shows that compared with WT, knockout of SYISL can maintain muscle mass in aged mice. (B) Representative photographs of Sirius red staining for Gas and TA muscles show that SYISL knockout can significantly reduce muscle fibrosis in aged mice. (C) Representative photographs of Lamin immunofluorescence for TA muscles of 1‐, 3‐, 8‐, 12‐, and 18‐month‐old mice. Quantification of three independent experiments shows that SYISL knockout can maintain muscle fibre size during muscle atrophy. At least 150 myofibres are analysed in an independent experiment. Scale bars, 50 μm. (D) Representative photographs of Qu, Gas, and TA muscles of left and right legs of 18‐month‐old mice. Quantification of eight independent experiments shows that knockdown of SYISL increases muscle size and maintains muscle weight in aged mice. (E) Representative photographs of Sirius red staining show that SYISL knockdown reduces muscle fibrosis in aged mice. Scale bars, 200 μm. (F) Western blotting results show that knockdown of SYISL significantly decreases the expression levels of atrophy genes FoxO3a, MuRF1, and Atrogin‐1 in 18‐month‐old mice. The relative protein levels are normalized to GAPDH. The data of at least three independent experiments are presented as means ± SD. *P < 0.05, **P < 0.01. N.S. indicates statistical non‐significance.

miR‐23a‐3p/miR‐205‐5p/miR‐103‐3p are required for SYNPO2 intron sense‐overlapping lncRNA to regulate muscle atrophy and sarcopenia

miR‐23a‐3p inhibits muscle atrophy by targeting MuRF1 and Atrogin‐1. Because the binding of SYISL to miR‐23a‐3p was verified by the results described above, we tested whether SYISL promotes muscle atrophy via miR‐23a‐3p. We introduced the 3′UTRs of MuRF1 and Atrogin‐1 into dual luciferase reporter vectors and then co‐transfected them with SYISL in C2C12 and HeLa cells. The results showed that miR‐23a‐3p significantly reduced the luciferase activities of Luc‐MuRF1‐3′UTR and Luc‐Atrogin‐1‐3′UTR, while SYISL competed with MuRF1‐3′UTR and Atrogin‐1‐3′UTR to combine with miR‐23a‐3p. Mutation of the binding sites between SYISL (SYISL‐Mut) and miR‐23a‐3p eliminated this competitive activity (Figure S9A–D). Next, we injected a lentivirus‐mediated SYISL overexpression vector with mutated binding sites of SYISL and miR‐23a‐3p into the leg muscles of 3‐month‐old mice, which were then treated with DEX to induce muscle atrophy. qRT‐PCR and Western blotting results showed that overexpression of SYISL‐Mut had no effect on the expression of Atrogin‐1, although it promoted the expression of FoxO3a and MuRF1 (Figure S9E,F). These results indicated that, in addition to the miR‐23a‐3p‐mediated Atrogin‐1 and MuRF1 pathway, SYISL promotes muscle atrophy through other pathways. As shown by the above experiments, miR‐103‐3p and miR‐205‐5p could directly bind to SYISL; the expression levels of miR‐103‐3p and miR‐205‐5p significantly varied after muscle atrophy. Moreover, RNAhybrid 2.12 prediction showed that miR‐103‐3p could bind to MuRF1, while miR‐205‐5p could bind to MuRF1 and FoxO3a (Figure S9G). These results suggested that miR‐103‐3p and miR‐205‐5p may participate in SYISL‐mediated regulation of muscle atrophy. Therefore, we selected miR‐103‐3p and miR‐205‐5p for further investigation. Firstly, we overexpressed these two miRNAs in C2C12 cells; the results showed that miR‐103‐3p significantly decreased the expression level of MuRF1 (Figure 7A; Figure S9H), and miR‐205‐5p significantly reduced the expression levels of MuRF1 and FoxO3a (Figure 7B; Figure S9I). Next, we introduced the 3′UTRs of MuRF1 and FoxO3a into the dual luciferase reporter vectors and then co‐transfected them with SYISL or SYISL‐Mut into C2C12 and HeLa cells. The results showed that miR‐103‐3p and miR‐205‐5p significantly reduced the luciferase activity of Luc‐MuRF1‐3′UTR; SYISL competed with MuRF1‐3′UTR to combine with miR‐103‐3p and miR‐205‐5p (Figure 7C,D; Figure S9J,K,M,N,P,Q). In addition, miR‐205‐5p significantly reduced the luciferase activity of Luc‐FoxO3a‐3′UTR (Figure S9L,O); SYISL competed with FoxO3a‐3′UTR to combine with miR‐205‐5p (Figure 7E; Figure S9R). Mutation of the binding sites between SYISL and miR‐103‐3p or miR‐205‐5p eliminated their competitive activities (Figure 7C–E; Figure S9J–R). These results suggested that SYISL functions as a sponge of miR‐103‐3p and miR‐205‐5p to promote the expression of MuRF1 and FoxO3a, as well as muscle atrophy.

Figure 7

miR‐23a‐3p/miR‐205‐5p/miR‐103‐3p are required for SYNPO2 intron sense‐overlapping lncRNA to regulate muscle atrophy and sarcopenia. (A) Western blotting results show that overexpression of miR‐103‐3p inhibits the expression of MuRF1 at protein levels in C2C12 myoblasts differentiated for 5 days. (B) Western blotting results show that in C2C12 myoblasts differentiated for 5 days, overexpression of miR‐205‐5p inhibits the expression of MuRF1 and FoxO3a at protein levels. (C,D) Dual luciferase reporter assays of Luc‐MuRF1‐3′UTR in C2C12 myoblasts show that SYISL competes with MuRF1‐3′UTR to combine with miR‐103‐3p (C) and miR‐205‐5p (D). (E) Dual luciferase reporter assays of Luc‐FoxO3a‐3′UTR in C2C12 myoblasts show that SYISL competes with FoxO3a‐3′UTR to combine with miR‐205‐5p. (F) Representative photographs of H&E staining and Lamin immunofluorescence staining for TA muscles. Quantification of three independent experiments shows that overexpression of SYISL‐WT can further reduce muscle fibre cross‐sectional areas and aggravate muscle fibrosis in DEX‐induced muscle atrophy mice, while overexpression of SYISL‐Mut‐All has no significant effect on muscle fibre cross‐sectional areas and muscle fibrosis. At least 150 myofibres are analysed in an independent experiment. Scale bar, 100 μm (H&E), 50 μm (Lamin). The relative protein levels are normalized to GAPDH. The data of at least three independent experiments are presented as means ± SD. *P < 0.05, **P < 0.01. N.S. indicates statistical non‐significance.

To determine whether miR‐23a‐3p, miR‐103‐3p, and miR‐205‐5p function cooperatively in SYISL‐mediated regulation of muscle atrophy, we mutated all binding sites of SYISL with miR‐23a‐3p, miR‐103‐3p, and miR‐205‐5p (SYISL‐Mut‐All) and then conducted lentivirus injection into the leg muscles of 3‐month‐old WT mice. The mice were intraperitoneally injected with DEX to induce muscle atrophy. The results showed that overexpression of SYISL significantly promoted the muscle atrophy, while overexpression of SYISL‐Mut‐All had no effect (Figure 7F; Figure S9S–X). Overall, these data indicated that miR‐23a‐3p/miR‐205‐5p/miR‐103‐3p are required for SYISL‐mediated regulation of muscle atrophy.

Human SYNPO2 intron sense‐overlapping lncRNA accelerates muscle atrophy via conserved mechanisms

To confirm that hSYISL has a conserved ability to promote muscle atrophy, we performed lentivirus‐mediated hSYISL overexpression experiments in the muscles of 3‐month‐old DEX‐induced mice and 18‐month‐old aged mice, respectively. There was no significant difference for the lentivirus infection efficiency between control and hSYISL overexpression groups (Figure S10A), which excluded the effect of transfection efficiency on muscle atrophy. As expected, overexpression of hSYISL aggravated the DEX‐induced loss of TA weight (Figure S10B), further reduced the cross‐sectional area of the TA muscles, and increased muscle fibrosis (Figure S10C). The expression levels of MuRF1, Atrogin‐1, and FoxO3a were significantly increased after overexpression of hSYISL (Figure S10D). Quantification of the number of centrally localized nuclei after infection showed that there was no significant difference between control and hSYISL overexpression groups (Figure S10E), indicating the muscle atrophy by hSYISL overexpression is not caused by the myopathy. Similarly, hSYISL overexpression significantly decreased the weights of Gas, Qu, and TA muscles of 18‐month‐old WT mice by 7.76% (P < 0.01), 12.26% (P < 0.01), and 13.44% (P < 0.01), respectively (Figure S10F). hSYISL overexpression also significantly decreased muscle fibre cross‐sectional area and increased muscle fibrosis and expression levels of MuRF1, Atrogin‐1, FoxO3a; no significant change for the expression level of SYNPO2 was observed (Figure 8A; Figure S10G–I). To verify that hSYISL promotes muscle atrophy via sponging of miR‐23a‐3p/miR‐205‐5p/miR‐103‐3p, we performed dual luciferase reporter experiments and the results were consistent with the findings described above (Figure 8B–F; Figure S10J–L). Taken together, these results indicated that human SYISL has a conserved ability to promote muscle atrophy through interactions with miR‐23a‐3p/miR‐205‐5p/miR‐103‐3p.

Figure 8

Human SYNPO2 intron sense‐overlapping lncRNA accelerates muscle atrophy via conserved mechanisms. (A) Representative photographs of H&E staining, Lamin immunofluorescence staining, and Sirius red staining for Gas muscles. Quantification of three independent experiments shows that overexpression of hSYISL reduces the muscle fibre cross‐sectional area and increases the muscle fibrosis in 18‐month‐old WT mice. At least 150 myofibres are analysed in an independent experiment. Scale bar, 200 μm (H&E), 50 μm (Lamin), 200 μm (Sirius red). (B,C) Dual luciferase reporter assays results show that hSYISL can competitively bind to miR‐23a‐3p with MuRF1 (B) and Atrogin‐1 (C) in HeLa cells. (D,E) Dual luciferase reporter assays of Luc‐MuRF1‐3′UTR in HeLa cells show that hSYISL competes with MuRF1 to combine with miR‐103‐3p (D) and miR‐205‐5p (E). (F) Dual luciferase reporter assays of Luc‐FoxO3a‐3′UTR in HeLa cells show that hSYISL competes with FoxO3a to combine with miR‐205‐5p. The data of at least three independent experiments are presented as means ± SD. *P < 0.05, **P < 0.01. N.S. indicates statistical non‐significance.

Discussion

There is increasing evidence that lncRNAs are functionally conserved among species. , , The functional conservation of lncRNAs among species can be explained as follows: lncRNAs have conserved binding domains and secondary or higher structures. For example, the 2D structure of the A region of Xist can recruit components of PRC2 that are needed for X‐chromosome inactivation. The 3D structure formed by the 98–153 nt sequence of the lncRNA EDAL is necessary for binding to EZH2 protein and subsequent EDAL function. The binding sites of miRNAs require only 6–10 nt of base pairing, so most lncRNAs are highly conserved among species as molecular sponges of miRNAs. Our previous study showed that the binding sites of miR‐135‐5p on lncMGPF are highly conserved among humans, mice, and pigs. Here, we identified conserved lncRNA transcripts of SYISL in humans and pigs. Sequence alignment revealed that their nucleotide sequences are dissimilar. However, bioinformatics analysis showed that SYISL can bind to EZH2 and has five common binding sites in mice, pigs, and humans. Further experiments verified that the regulation of myogenesis via SYISL–EZH2 interaction was conserved between pigs and humans. In addition, miR‐23a‐3p/miR‐205‐5p/miR‐103‐3p has vital roles in SYISL‐mediated promotion of muscle atrophy among mice, humans, and pigs. These results indicate that SYISL is a conserved regulator of myogenesis and muscle atrophy.

Increasing numbers of lncRNAs have been identified as important regulators of muscle development, muscle regeneration, and muscle wasting. , , In this study, we revealed the novel function of SYISL in muscle atrophy and found that SYISL promotes muscle atrophy through FoxO signalling pathways and ubiquitin‐mediated proteolysis. SYISL KO or knockdown could alleviate sarcopenia in aged mice. Sarcopenia is a serious muscle wasting disease; it also incurs a substantial economic burden related to medical costs. , , Currently, targeting of muscular dystrophy‐related pathways (e.g. myostatin/TGF‐β/activin scavengers or bone morphogenetic protein agonists, IGF‐1 signalling, β2‐adrenoreceptor agonists, and inhibitors of the UPS ) has become an important therapeutic strategy for sarcopenia. To explore the possibility of SYISL as a therapeutic target for sarcopenia, we investigated whether hSYISL could regulate muscle atrophy. The results suggest that targeting of SYISL is a promising novel therapeutic strategy for muscle atrophy and sarcopenia.

LncRNAs can function in various manners, such as binding to chromosomal modification complexes and transcription factors, and serve as molecular sponges for miRNAs. Our study demonstrated that sponging of miR‐23a‐3p, miR‐103‐3p, and miR‐205‐5p by SYISL attenuates the inhibitory effects of these miRNAs on target gene expression and muscle atrophy. miR‐23a‐3p targets MuRF1 and Atrogin‐1 and inhibits muscle atrophy. Moreover, multiple reports have described sponging of miR‐103‐3p and miR‐205‐5p, mainly in the context of cancer. , However, the effects of miR‐103‐3p and miR‐205‐5p on muscle atrophy have been unclear. Here, we found that miR‐103‐3p and miR‐205‐5p inhibit muscle atrophy by targeting MuRF1 and FoxO3a; sponging of miRNAs by SYISL weakens the inhibitory effects of miR‐103‐3p and miR‐205‐5p on target gene expression. Thus, miRNA sponging is the main mechanism by which SYISL regulates muscle atrophy; miR‐23a‐3p, miR‐103‐3p, and miR‐205‐5p function cooperatively in this process.

In summary, we identified SYISL homologues in humans and pigs; we found that the SYISL‐mediated regulation of myogenesis is conserved among mice, pigs, and humans. Comparative analysis of SYISL in mice, pigs, and humans revealed its novel function in muscle atrophy and sarcopenia. Mechanistic analysis showed that SYISL simultaneously binds to miR‐23a‐3p, miR‐103‐3p, and miR‐205‐5p; it promotes muscle atrophy by attenuating their inhibitory effects on the expression levels of the muscle atrophy‐related genes MuRF1, Atrogin‐1, and FoxO3a. In addition, we verified that human SYISL accelerates muscle atrophy via conserved mechanisms. Our study provides novel perspectives regarding potential therapeutic targets for muscle atrophy and sarcopenia.

Conflict of interest

The authors declare no conflict of interest.

Figure S1. Conservative analysis of SYISL among humans, pigs and mice.

Figure S2. Regulation of myogenesis by SYISL is conserved in pigs and humans.

Figure S3. Regulation of myogenesis by SYISL in mice, pigs, and humans is not caused by SYNPO2.

Figure S4. hSYISL and pSYISL inhibit skeletal muscle growth and development in vivo.

Figure S5. The molecular mechanism of SYISL regulating myogenesis through EZH2/H3K27me3 is conserved in pigs.

Figure S6. Comparative analysis of SYISL among mice, pigs, and humans reveals its potential function in promoting muscle atrophy.

Figure S7. SYISL promotes myotube atrophy in vitro and SYISL knockdown alleviates muscle atrophy induced by DEX.

Figure S8. Knockdown or knockout of SYISL alleviates mucle atrophy in aged mice.

Figure S9. miR‐23a‐3p/miR‐205‐5p/miR‐103‐3p are required for SYISL to regulate muscle atrophy and sarcopenia.

Figure S10. Human SYISL accelerates muscle atrophy via conserved mechanisms.

Table S1. Sequences of primers for qRT‐PCR, RIP and ChIP qRT‐PCR

Table S2. Sequences of miRNA oligonucleotides

Table S3. Sequences of primers used for plasmid construction

Click here for additional data file.