Skeletal muscle plays important roles in activities of daily life and health. Abnormal regulation of skeletal muscle‐specific genes leads to various muscle diseases, such as sarcopenia, myosarcoma, and muscle metabolic disorder.1,2 Skeletal muscle is the main agricultural animal protein source consumed by humans, and the growth and development of skeletal muscle directly impact the quantity and quality of meat.3 Skeletal myogenesis is a process in which mesenchymal stem cells differentiate into myoblasts, which then proliferate, exit the cell cycle, and fuse together to form mature myofibres, which determine the protein characteristics of muscle tissue.4 When skeletal muscles are damaged postnatally, the muscle repair and regeneration processes parallel developmental myogenesis, with satellite cells located in niches on the myofibre surface being responsible for the generation of myoblasts.5 The determination and terminal differentiation of muscle cells are primarily controlled by four myogenic regulatory factors (MRFs): myogenic factor 5, muscle‐specific regulatory factor 4, myoblast determination factor (MyoD), and myogenin (MyoG).6 MRFs activate numerous downstream genes to initiate muscle cell differentiation by cooperating with other transcription factors, such as myocyte enhancer factor 2 (MEF2), and in concert with epigenetic regulatory mechanisms.7,8 In‐depth exploration of MRF‐mediated regulatory networks could contribute greatly to identifying molecular targets for treating muscle diseases and improving meat production.
In recent years, a large number of non‐coding RNAs (ncRNAs) have been discovered and shown to be involved in the regulation of biological processes in almost all species. Among the ncRNAs, long non‐coding RNAs (lncRNAs) are a class of RNA transcripts greater than 200 nt in length with little or no protein‐coding capacity.9,10 Substantial evidence has shown that lncRNAs play critical regulatory roles in diverse biological processes and diseases, such as X‐chromosome inactivation,11 genome imprinting,12 stem cell pluripotency,13 organ development,14 immune responses,15 drug resistance,16 and tumorigenesis.17 These lncRNAs function through multiple regulatory mechanisms, including chromatin modification, transcription activation, microRNA (miRNA) sponging, and mRNA splicing and translation.18 In terms of myogenesis, tens of thousands of lncRNAs have been discovered in skeletal muscles using high‐throughput technologies and bioinformatics analyses, but only a few of them have been identified as functional regulators of skeletal muscle development and regeneration; these include H19,19 SYISL,20 linc‐RAM,21 Linc‐MD1,22 lnc‐mg,23 m1/2sbsRNAs,24 Dum,25 LncMyoD,26 MAR1,27 and LncIRS1.28 Therefore, the identification of novel functional lncRNAs involved in myogenesis, and elucidation of their mechanisms of action, requires more work.
In this study, we identified the muscle‐enriched and MyoD‐activated lncRNA AK003290 in mice and its homologous lncRNA AK394747 in pigs and MT510647 in humans. Functional analyses revealed that AK003290, AK394747, and MT510647 significantly promote myogenic differentiation and muscle growth; therefore, we named them lncMGPF (lncRNA muscle growth promoting factor). lncMGPF acts as a molecular sponge of miR‐135a‐5p, thus promoting MEF2C gene expression, and also regulates human antigen R (HuR; also known as ELAV‐like RNA binding protein 1)‐mediated mRNA stabilization of MyoD and MyoG genes. Our observations show that lncMGPF is a conserved positive regulator of skeletal muscle growth in mice, pigs, and humans.
All procedures involving animals were performed in accordance with the guidelines of good laboratory practice, and animals were supplied with nutritious food and sufficient water. Animal feeding and tests were conducted based on the National Research Council Guide for the Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee at Huazhong Agricultural University. Piglets were slaughtered according to a standard procedure based on guidelines in the Regulation of the Standing Committee of Hubei People's Congress (Hubei Province, China, HZAUSW‐2017‐008). All mice were obtained from the experimental animal centre of Huazhong Agricultural University. All pigs were obtained from the experimental pig farm of Huazhong Agricultural University.
Mouse primary myogenic progenitor cells were isolated from 5‐week‐old C57BL mice; pig primary myogenic progenitor cells were isolated from 1‐day‐old large white male piglets. Primary myogenic progenitor cells were isolated and cultured as described previously.27,29 Briefly, muscles were minced and digested in 2 mg/mL type I collagenase (C0130; Sigma‐Aldrich, St. Louis, MO, USA). Digestion was stopped with RPMI 1640 medium containing 20% foetal bovine serum (FBS). Cells were cultured in growth medium [RPMI 1640 supplemented with 20% FBS, 4 ng/mL basic fibroblast growth factor, 1% chicken embryo extract, and 1% penicillin–streptomycin] on collagen‐coated cell culture plates at 37°C and 5% CO2. C2C12 myoblasts, pig kidney (PK) cells, 293T cells, and HeLa cells were obtained from the Chinese Academy of Sciences Cell Bank and grown in incubators at 37°C and 5% CO2, and proliferating cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% FBS (Gibco, Grand Island, NY, USA). Human skeletal muscle myoblasts (Cat. #3501) were obtained from ScienCell Research Laboratories and grown in incubators at 37°C and 5% CO2, and proliferating cells were cultured in DMEM supplemented with 5% FBS (Gibco, Grand Island, NY, USA). For myogenic differentiation, cells were cultured in DMEM containing 2% horse serum (Gibco). All cells were grown to 80–90% confluence before differentiation was induced.
Single myofibres were isolated from extensor digitorum longus (EDL) muscles of mice, as described previously.30 Briefly, EDL muscles were separated and digested with 2 mg/mL collagenase I (C0130) in DMEM for 1 h at 37°C. Digestion was stopped by carefully transferring EDL muscles to a pre‐warmed Petri dish (10 cm) with 6 mL of DMEM supplemented with 10% horse serum; the muscles were then lightly flushed with a P200 micropipette to release single myofibres. The released single myofibres were transferred and cultured in a Matrigel‐coated Petri dish (10 cm) in DMEM supplemented with 20% FBS, 4 ng/mL basic fibroblast growth factor, and 1% penicillin–streptomycin at 37°C and 5% CO2.
Total RNA was isolated using TRIzol reagent (Invitrogen, USA) according to the manufacturer's instructions. The Mouse 4 × 44 K Gene Expression Microarray containing 34,456 probes was used to identify differentially expressed genes in C2C12 myoblasts that had differentiated for 2 days after knockdown of lncMGPF as described below. Six microarrays were used in the experiment, with three biological replicates of each treatment. Microarray hybridization was conducted according to the Expression Analysis Technical Manual (KangChen Bio‐technology, China). Microarray data were analysed according to a previously reported method.20 Differentially expressed mRNAs with statistically significant changes were identified through fold‐change filtering (≥1.5), unpaired t‐tests (P < 0.05), and multiple hypothesis testing (false discovery rate < 0.05). Pathway analyses were based on Kyoto Encyclopedia of Genes and Genomes, and P values were used to determine the significance of the pathways (threshold P < 0.05).
Total RNA was reverse‐transcribed using RevertAid Reverse Transcriptase (Thermo Scientific, USA). All miRNAs were reverse‐transcribed using specific loop primer. Quantitative real‐time PCR (qRT‐PCR) analyses was performed using the Applied Biosystems StepOnePlus real‐time PCR system. Relative RNA expression was calculated using the Ct (2–ΔΔCt) method.31 All primers used in qRT‐PCR are presented in Supporting Information, Table S1.
We performed 5′ and 3′ rapid amplification of complementary DNA ends (RACE) using the Takara SMARTer RACE complementary DNA amplification Kit (Clontech, USA) according to the manufacturer's instructions. The sequences of gene‐specific primers used for lncMGPF RACE were 5′‐CCTGGGCTTCATTTACTCCTGAAGACAGTC‐3′ (5′ RACE) and 5′‐GCAGCCATCTTGGTGGAGAAGTTCCTGGC‐3′ (3′ RACE). The sequences of gene‐specific primers used for AK394747 RACE were 5′‐CCTGCCTCAAACCCCTCCTTTTACTCATCA‐3′ (5′ RACE) and 5′‐ATTGCCAGCATTTGTAAGTGATTGTGCAAT‐3′ (3′ RACE). For the hlncMGPF RACE assay, homologous sequences in the human genome were retrieved from NCBI database using Basic Local Alignment Search Tool, and confirmed by RT‐PCR using forward (5′‐ACAGTTTACAAGGTGTAGGT‐3′) and reverse (5′‐ TGAAAATGTGATTCTCAGAG‐3′) primers and sequencing. From the sequencing results, 5′ and 3′ RACE primers were designed. The sequences of gene‐specific primers were 5′‐GGGGTGCAAACCTACACCTTGTAAACTG‐3′ (5′ RACE) and 5′‐GCAAAAAGTCCTTTCCAGGGCTCCC‐3′ (3′ RACE).
Northern blot was performed according to a published method.32 Total RNA from differentiated C2C12 myoblasts was extracted using a total RNA kit (Omega bio‐tek, Norcross, GA, USA). Then, transferred the RNA to a positively charged nylon membrane and UV cross‐linked, the p32‐labelled probe was hybridized to the membrane at 60°C overnight. The lncMGPF RNA probe sequence was as follows: 5′‐GCAGCCATCTTGGTGGAGAAGTTCCTGGC‐3′.
Nuclear and cytoplasmic RNA fractionation was performed according to a previously published method,21 and distribution quantification was analysed via qRT‐PCR. Nuclear and cytoplasmic protein fractionation was performed using the Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime Biotechnology, China) according to the manufacturer's instructions, and the distribution was quantified using Western blotting.
RNA fluorescence in situ hybridization (FISH) was performed using the lncRNA FISH Kit (Guangzhou RiboBio, China) according to the manufacturer's instructions. Briefly, cells were fixed with 4% formaldehyde for 10 min at room temperature. After washing, cells were permeabilized with 0.5% Triton X‐100 for 30 min at 37°C. Then, cells were incubated with RNA probes in hybridization buffer overnight at 37°C. The RNA probes were directly conjugated with a fluorophore. Then, the cells were washed three times using saline sodium citrate buffer, stained with 4,′6‐diamidino‐2‐phenylindole (DAPI) for 10 min at room temperature, and examined using a fluorescence microscope.
lncMGPF promoter regions were cloned into a PGL3‐basic luciferase reporter vector (Addgene, USA), and the constructed luciferase reporter vectors were transfected into C2C12 myoblasts in a 24‐well plate. The 3′ untranslated region (UTR) of MEF2C, sequences containing the miR‐135a‐5p binding site of lncMGPF or AK394747, and the sequences of lncMGPF or AK394747 with mutated miR‐135a‐5p binding sites, were cloned into the pmirGLO Dual‐Luciferase miRNA Target Expression Vector (Promega, USA) to construct the corresponding reporter vectors. Those reporter vectors were transfected into C2C12 myoblasts, HeLa cells, PK cells, and myogenic progenitor cells using Lipofectamine 2000. The luciferase assay was performed using the Dual‐Luciferase Reporter Assay System (Promega, USA) and the enzymatic activity of luciferase measured using a PerkinElmer 2030 Multilabel Reader (PerkinElmer, USA).
Small interfering RNAs (siRNAs) targeting lncMGPF, AK394747, HuR, and MyoD, as well as the inhibitor and mimics of miR‐135a‐5p, were synthesized by Genepharma (Genepharma, China); all siRNA sequences are presented in Supporting Information, Table S2. For cell transfection, we transfected C2C12 myoblasts, PK cells, HeLa cells, and mouse and pig myogenic progenitor cells with approximately 160 μM siRNA oligonucleotides using 9 μL Lipofectamine 2000 (Invitrogen, USA) in each well of a 6‐well plate.
To construct the lncMGPF, MyoD, MEF2C, HuR, and AK394747 overexpression plasmids, the full‐length or coding sequence of lncMGPF, MyoD, and MEF2C were cloned into the pcDNA3.1 plasmid (Addgene, USA). For in vitro translation assays, the coding sequence of HuR and predicted open reading frame sequence were cloned into the pCMV‐C‐enhanced green fluorescent protein (EGFP) vector to produce fusion proteins with EGFP (Beyotime Biotechnology, China). For in vitro transcription assays, full‐length and truncated fragments of lncMGPF and AK394747 were cloned into the pGEM‐3Z vector (Promega, USA). Full‐length sequences of lncMGPF, MyoD, MEF2C, HuR, and AK394747 were amplified using specific F/R primers (Supporting Information, Table S3); the primers used to amplify the truncated fragments of lncMGPF and AK394747 are listed in Supporting Information, Table S5. For cell transfection, we transfected C2C12 myoblasts, HeLa cells, PK cells, and mouse myogenic progenitor cells with approximately 4 μg plasmid using 9 μL Lipofectamine 2000 (Invitrogen, USA) in each well of a 6‐well plate.
To construct lentivirus‐mediated overexpression vectors for lncMGPF, AK394747, and hlncMGPF, sequences of lncMGPF, AK394747 and hlncMGPF were separately subcloned into the lentivirus vector PCDH‐CMV‐copGFP (Addgene, USA). To construct the lentivirus‐mediated interference vector for AK394747, siRNA targeting AK394747 was inserted into the SuperSilencing short hairpin RNA expression vector pLKO.1‐TRC (Addgene, USA). We packaged the lentivirus in 293T cells using three vectors: 10.7 μg pLKO.1‐TRC or PCDH‐CMV‐copGFP, 8.0 μg psPAX2 (Addgene, USA), and 5.3 μg PDM2.G (Addgene, USA). For lentivirus infection of cells, 10 μL virus and 1 μg polybrene were added to each millilitre of medium and then replaced with fresh medium after 24 h. For mouse muscle infection, we injected 50 μL lncMGPF, AK394747, or hlncMGPF overexpression lentivirus vector and empty lentivirus vector into the Gas muscles of the left and right legs, respectively, of five 1‐month‐old lncMGPF knockout (KO) mice or wild‐type (WT) mice every 7 days. For pig muscle infection, we injected 1 mL AK394747 overexpression lentivirus vector and empty lentivirus vector into the biceps femoris muscles of the left and right legs, respectively, of three 1‐month‐old pigs every 7 days. Gas and biceps femoris muscles were sampled after 4 weeks of weekly injections. The lentivirus concentration used in all the assays was above 1 × 108 transducing units per millilitre. The multiplicity of infection was 10, 15, and 25 plaque forming unit per cell for the infections of C2C12 myoblasts, pig myogenic progenitor cells, and human skeletal muscle myoblasts, respectively. All primers used for plasmid construction and lentivirus production are presented in Table S3. The hlncMGPF sequence was synthesized by Beijing TSINGKE Biotechnology Corporation and constructed into the lentivirus vector PCDH‐CMV‐copGFP (Addgene, USA).
Chromatin immunoprecipitation (ChIP) was performed using a ChIP Kit (Millipore, USA) according to the manufacturer's protocol. We performed each ChIP assay using 1 μg antibodies. The antibodies included MyoD (sc‐760; Santa Cruz Biotechnology, USA). Immunoglobulin G was used as the negative control. We conducted qRT‐PCR using the retrieved DNA and primers to determine the enrichment levels of MyoD at certain DNA loci. The primers used for ChIP‐qRT‐PCR are presented in Table S1.
Cell proliferation was determined through 5‐ethynyl‐2′‐deoxyuridine (EdU) staining, real‐time cell proliferation monitoring, and Cell Counting Kit‐8 (CCK‐8) assays. EdU staining was conducted using the BeyoClick™ EdU Cell Proliferation Kit (Beyotime Biotechnology, China) according to the manufacturer's instructions. The cells cultured in DMEM supplemented with 50 μM EdU and 20% FBS for 2 h, at 37°C and 5% CO2. Real‐time cell proliferation monitoring was performed using the xCELLigence real‐time cell analyses (RTCA) system as described previously.33 CCK‐8 assay was conducted using the CCK‐8 (BOSTER, China) according to the manufacturer's instructions.
Proteins were extracted from muscle tissue and cells using radioimmunoprecipitation assay (RIPA) buffer with 1% (v/v) phenylmethylsulfonyl fluoride (Beyotime Biotechnology, China). Western blotting was performed according to a previously reported method.34 The antibodies used included MyoG (sc‐12732; 1:200; Santa Cruz Biotechnology, USA), MyoD (sc‐760; 1:200; Santa Cruz Biotechnology, USA), MyHC (sc‐376157; 1:1000; Santa Cruz Biotechnology, USA), β‐actin (sc‐4777; 1:1000; Santa Cruz Biotechnology, USA), MEF2C (sc‐365862; 1:200; Santa Cruz Biotechnology, USA), N‐Ras (sc‐31; 1:200; Santa Cruz Biotechnology, USA), Ki67 (ab16667; 1:1000; Abcam, UK), HuR (sc‐5261; 1:200; Santa Cruz Biotechnology, USA), Histone H3 (17168‐1‐AP; 1:1000; Proteintech, USA), α‐tubulin (1E4C11; 1:1000; Proteintech, USA), GFP (50430‐2‐AP; 1:1000; Proteintech, USA), and GAPDH (sc‐47724; 1:1000; Santa Cruz Biotechnology, USA). All protein levels were normalized to that of the housekeeping protein β‐actin, and densitometric quantification of the western blotting bands was performed using ImageJ software.
Cell immunofluorescence staining was performed according to a previously published method.19 Immunofluorescence staining antibodies included MyHC (sc‐376157; 1:200; Santa Cruz Biotechnology, USA), MyoG (sc‐12732; 1:200; Santa Cruz Biotechnology, USA), HuR (sc‐5261; 1:200; Santa Cruz Biotechnology, USA), myosin (M4276; 1:1000; Sigma, USA), and a secondary antibody (anti‐mouse CY3; Beyotime Biotechnology, USA). DAPI was used to visualize cell nuclei with a fluorescence microscope (DP80; Olympus, Japan). The differentiation index was calculated as the percentage of nuclei in MyHC positive cells, and the fusion index was calculated as the percentage of nuclei in fused myotubes which have two or more nuclei out of the total nuclei.35
We generated lncMGPF KO mice using the CRISPR genome‐editing system with the C57BL/6 background according to a previous report.36 Two single‐guide RNAs (sgRNAs) (sgRNA1: 5′‐CCTGTTGGATAATGGGGTCATGT‐3′; sgRNA2: 5′‐CTTAGCTTCTAGTTCGAATTCGG‐3′) were designed using an online CRISPR design tool (
Muscle injury was performed as reported previously.38 Briefly, we injected 50 μL, 10 mM cardiotoxin (CTX) (Sigma, USA) in phosphate‐buffered saline into the TA muscles of 8‐week‐old male mice and injected 100 μg EdU (Thermo Fisher Scientific, USA) intraperitoneally 4 h before harvesting muscles.
Haematoxylin and eosin staining of muscle sections was performed according to a previously reported method23 and visualized using an optical microscope (BX53; Olympus, Japan). The cross‐sectional areas of individual myofibres were quantified using ImageJ software. Immunohistochemical staining was performed as described previously39 and visualized using a confocal laser scanning microscope (LSM800; Zeiss, Germany). For immunofluorescence staining, samples were repaired in 0.01 M sodium citrate solution (pH 6.0) for 30 min at 70°C, and incubated in blocking buffer (5% goat serum, 2% bovine serum albumin, 0.1% Triton X‐100 and 0.1% sodium azide in phosphate‐buffered saline) for 2 h. Samples were then incubated with primary antibodies diluted in blocking buffer overnight at 4°C. After washing with phosphate‐buffered saline, the samples were incubated with secondary antibodies for 1 h at 37°C. And lastly, the samples were incubated with DAPI for 10 min at room temperature. The antibodies used included dystrophin (ab‐15277; 1:100; Abcam, UK), embryonic MyHC (eMyHC) (BF‐G6; 1:100; Developmental Studies Hybridoma Bank, USA), MyoG (sc‐12732; 1:200; Santa Cruz Biotechnology, USA), Pax7 (1:100; Developmental Studies Hybridoma Bank, USA), COX4 (1:1000, ABclonal, A6564) and a secondary antibody (anti‐mouse CY3 and anti‐rabbit FITC; Beyotime Biotechnology, USA). To detect EdU incorporation, we incubated the slides using the Life Technologies Click‐iT Kit (Invitrogen, USA) according to the manufacturer's instructions. Myosin ATPase staining of muscle sections was performed according to a previously reported method.39 Briefly, a frozen muscle section was incubated for 2 min in pre‐hybridization buffer containing 50 mM potassium acetate and 70 mM CaCl2 (pH 10.4). Then, the frozen muscle section was washed three times in 100 mM Tris/HCl buffer containing 20 mM CaCl2 (pH 7.2). Finally, the frozen section was incubated for 25 min in hybridization buffer containing 2.5 mM ATP disodium salt, 40 mM glycine, and 18 mM CaCl2 (pH 9.4).
We predicted protein‐coding capability using coding potential calculator (CPC) (
RNA immunoprecipitation (RIP) was performed using the Magna RIP RNA‐Binding Protein Immunoprecipitation Kit (Millipore, USA) according to the manufacturer's instructions. The Ago2 antibody (ab3748; Abcam, UK) and HuR (sc‐5261; 1:200; Santa Cruz Biotechnology, USA) were used for RIP. Co‐precipitated RNA was detected using reverse transcription PCR (RT‐PCR) or qRT‐PCR. The qRT‐PCR data are presented as a percentage of the input, as described previously.40
Biotin‐labelled RNA pulldown was performed according to a previously published method.20 For mass spectrometry, the proteins pulled down with lncMGPF were separated using a 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS‐PAGE) gel and then subjected them to the silver staining. The differentially expressed bands were cut out and analysed using mass spectrometry (Novogene, China). The procedures for the lncMGPF and miR‐135a‐5p binding assay were the same as those for RNA pulldown, except that the captured RNA was purified and analysed via qRT‐PCR.
RNA stability analyses were performed as described previously.32 Briefly, C2C12 cells, mouse primary myoblasts, and pig primary myoblasts were induced to differentiate for 3 days. Then, the differentiated cells were treated with actinomycin D (ActD) (5 μg/mL), and samples were collected at 0, 1, 2, 3, and 4 h after ActD treatment. Total RNAs were extracted using Trizol (Invitrogen, USA), and reverse‐transcribed using RevertAid Reverse Transcriptase (Thermo Scientific, USA). The relative RNA expression at different time points was determined using qRT‐PCR.
We calculated correlations between the expression of various lncRNAs and the MyoD gene using the Pearson correlation method based on previously reported microarray data (accession number: GSE102098). All differences between groups were analysed using unpaired or paired Student's t‐test. P < 0.05 was considered statistically significant; significance is denoted as *P < 0.05 and **P < 0.01. All data are presented as mean ± standard deviation.
The microarray data generated in this study have been deposited in the Gene Expression Omnibus database (accession number: GSE138022).
To screen for potential functional lncRNAs related to MRF‐mediated regulatory network for myogenesis, we analysed the correlations between the expression of lncRNAs and the myogenic regulatory gene MyoD during C2C12 myoblast differentiation based on our previously reported microarray data (accession number: GSE102098). We found that the expression of 19 uncharacterized lncRNAs was highly correlated with that of the MyoD gene (|r| ≥ 0.9, P < 0.05; Supporting Information, Table S4). From among these lncRNAs, we selected AK003290 (lncMGPF), which had the highest expression during cell differentiation, for further investigation (Figure 1A). RACE and northern blot assays confirmed that lncMGPF is a 1795 nt polyadenylated lncRNA (Supporting Information, Figure S1A and S1B). Protein‐coding potential prediction using CPC and an in vitro translation experiment of the predicted open reading frame indicated that lncMGPF is a non‐coding RNA similar to the known lncRNAs HOTAIR and NEAT1 (Figure S1C and S1D). Results of qRT‐PCR showed that lncMGPF was highly expressed in muscle tissues such as the longissimus dorsi, leg muscle, tongue, and heart when β‐actin and GAPDH were used as reference genes, respectively (Figures 1B and S1E). Furthermore, lncMGPF expression increased during C2C12 myoblast differentiation as well as during embryonic and postnatal muscle development (Figures 1C and 1D and S1F and S1G). Cell fractionation assays and RNA FISH demonstrated that lncMGPF is mainly localized to the nucleus of proliferating myoblasts, whereas it accumulates in the cytoplasm during myogenic differentiation (Figure 1E and 1F).
To explore the mechanisms through which lncMGPF is regulated at the transcriptional level, we conducted luciferase assays with four reporter constructs containing different fragments of the lncMGPF promoter (located between −2000 and +200 bp). The region between −345 and +200 bp had the highest transcriptional activity, which suggests that this region is the core promoter (Figure S1H). Two potential MyoD binding sites (E‐box) in the core promoter of lncMGPF were predicted through bioinformatics analyses (Figure 1H). Overexpression of MyoD increased the luciferase activity of reporter constructs containing predicted MyoD binding sites (−345 ~ +200 bp) and full‐length reporter constructs (−2000 ~ +200 bp) (Figure S1I). MyoD knockdown and overexpression decreased significantly and increased the expression of endogenous lncMGPF, respectively (Figures 1G and S1J). Moreover, results of a ChIP assay confirmed that MyoD could physically bind to the E‐box between +65 and +69 bp downstream the transcription start site of lncMGPF (Figure 1H). Collectively, these data reveal that lncMGPF is positively regulated by the MyoD gene.
To explore the possible role of lncMGPF in myogenesis, we used microarrays to analyse changes in gene expression after lncMGPF knockdown in C2C12 myoblasts differentiated for 2 days. We designed two siRNAs to knock down lncMGPF and found that cotransfection of siRNA‐1 and siRNA‐2 was more efficient than transfection of siRNA‐1 or siRNA‐2 alone (Supporting Information, Figure S2A). lncMGPF was successfully knocked down when β‐actin, GAPDH, and 18S rRNA were used as reference genes (Figure S2B), as shown by the results of qRT‐PCR. Compared with the controls, 407 down‐regulated genes (including MYL2, MYH7, MYH8, and MYO7B) and 136 up‐regulated genes (including MYF6, MYL1, and WNT7B) were identified in lncMGPF knockdown cells (fold change ≥ 1.5, P < 0.05) (Figure S2C and Supporting Information, Data Set 1). To validate the results of the microarray, we used qRT‐PCR to examine changes in the expression of some differentially expressed genes after lncMGPF overexpression or knockdown. The results from these two techniques were generally consistent (Figure S2D–S2G). Pathway analyses indicated that the differentially expressed genes were enriched in some signalling pathways related to muscle disease and metabolism, such as cardiac muscle contraction, and hypertrophic cardiomyopathy, which suggests that lncMGPF might be involved in regulating muscle development (Figure S2H and S2I).
To confirm the role of lncMGPF in myogenesis, we used functional gain and loss assays to investigate the effects of lncMGPF on myoblast proliferation, differentiation, and fusion in C2C12 myoblasts. Results of RTCA using the xCELLigence assay and EdU staining for proliferating myoblast cells showed that overexpression or knockdown of lncMGPF had no significant effect on cell proliferation (Supporting Information, Figure S3A, S3B, S3D, and S3E). mRNA expression of the proliferation marker genes Ki67 and N‐Ras was likewise unchanged after overexpression and knockdown of lncMGPF, as demonstrated through qRT‐PCR (Figure S3C and S3F). Results of qRT‐PCR, western blotting, and immunofluorescence staining for differentiated C2C12 myoblasts indicated that lncMGPF knockdown significantly decreased mRNA and protein expression of the myogenic marker genes MyoD, MyoG, and MyHC (Figures 2A–2C and S4A–S4D), whereas lncMGPF overexpression significantly increased that of the MyoD, MyoG, and MyHC genes (Figures 2D–2F and S4E–S4H). Results of qRT‐PCR and MyHC immunofluorescence staining for differentiated myoblasts showed that lncMGPF knockdown significantly decreased mRNA expression of β‐1 integrin, a myoblast fusion marker gene (Figure S3I), and myoblast fusion (Figure 2C), whereas lncMGPF overexpression significantly increased that of the β‐1 integrin (Figure S4J) and myoblast fusion (Figure 2F). Taken together, these data indicated that lncMGPF promotes C2C12 myoblast differentiation and fusion but has no significant effects on C2C12 myoblast proliferation.
To determine the function of lncMGPF in muscle development in vivo, we used a CRISPR/Cas9‐mediated genome editing approach to inactivate lncMGPF in C57BL/6 mice by deleting a 1896 bp genomic region containing the complete lncMGPF transcript. Then we validated the genotypes using genomic DNA PCR (Supporting Information, Figure S5A). lncMGPF KO mice were healthy but experienced significant decreases in weight and growth rate compared with WT mice (Figures 3A, 3B, and S5B). Likewise, the weights of the whole leg, gastrocnemius (Gas), tibialis anterior (TA), and quadriceps (Qu) muscles of 2‐month‐old KO mice were significantly lower than those of 2‐month‐old WT mice in both male and female mice (Figures 3C and S5C–S5E). Haematoxylin and eosin staining and dystrophin immunofluorescence staining of muscles from 2‐month‐old mice showed that KO mice had significantly smaller mean cross‐sectional areas of individual myofibres and higher proportions of smaller myofibres compared with WT mice (Figures 3D and S5F–S5I). DAPI staining results indicated that WT mice had ~20% more myonuclei per myofibre than KO mice (Figure 3E). In addition, there was no significant difference in the total number of satellite cells between WT and KO mice, as demonstrated by the results of Pax7 (a specific marker of muscle satellite cells) and DAPI staining of single myofibres, as well as by Pax7 and dystrophin staining of normal muscles (Figures 3F and S5J). To confirm the effects of lncMGPF depletion on the expression of related genes, we examined changes in the expression of marker genes for proliferation and myogenic differentiation in the skeletal muscles of 2‐month‐old WT and KO mice. Compared with WT mice, KO mice showed less mRNA and protein expression of MyoD, MyoG, MyHC, and β‐1 integrin, but no significant differences were found for the proliferation marker genes Ki67 and N‐Ras (Figure S5K and S5L). Although lncMGPF is also expressed in the heart, no significant changes in the size or weight of the heart were observed between WT and KO mice (Figure S5M). Moreover, muscle grip and forced swimming tests were used to determine muscle contractility and endurance, respectively. The results showed that both the muscle grip and endurance of KO mice were significantly lower than those of WT mice (Figure 3G and 3H). We detected cytochrome c oxidase subunit 4 (COX4; the terminal enzyme of the mitochondrial oxidative respiratory chain) in protein by immunohistochemistry and analysed the fibre type composition by myosin ATPase staining. KO mice had lower COX4 protein levels, and lower percentages of type I and oxidative fibres (Figure S5N–S5P), compared with WT mice, indicating that the reduced endurance was due to a decrease in oxidative myofibres and COX4 enzyme. On the basis of these results, we concluded that lncMGPF promotes growth rate, muscle mass, muscle contractility, and endurance.
To further confirm the roles of lncMGPF in cell proliferation and differentiation in myogenic progenitor cells, we isolated primary myogenic progenitor cells from the leg muscles of WT and KO mice. Consistent with the results for C2C12 myoblasts, lncMGPF KO in myogenic progenitor cells had no effect on cell proliferation (Supporting Information, Figure S6A–S6D), as determined using the CCK‐8, xCELLigence RTCA, EdU staining, and qRT‐PCR, whereas the deletion of lncMGPF in myogenic progenitor cells significantly decreased myogenic progenitor cell differentiation and fusion according to results from qRT‐PCR, western blotting, and immunofluorescence staining (Figures 3H and 3I and S6E and S6F). This indicates that lncMGPF promotes myogenic progenitor cell differentiation and fusion.
To determine whether overexpression of lncMGPF could rescue the muscle phenotype of KO mice, we injected the lentivirus‐mediated overexpression of lncMGPF (LV‐lncMGPF) vector and empty control (LV‐control) vector intramuscularly into the left and right legs of 1‐month‐old KO mice, respectively (Supporting Information, Figure S7A). The EGFP immunofluorescence staining results showed no significant difference in infection efficiency between the LV‐control and LV‐lncMGPF groups (Figure S7B). The qRT‐PCR results showed that lncMGPF RNA expression was significantly increased after lentivirus infection (Figure S7C). Overexpression of lncMGPF significantly increased the weights of the whole leg, TA, Qu, and Gas muscles, as well as the mean cross‐sectional areas of individual myofibres (Figures 3J and S7D–S7G). In addition, overexpression of lncMGPF in KO muscles significantly increased mRNA and protein expression of MEF2C, MyoD, MyoG, MyHC, and resulted in an increased number of myonuclei per fibre (Figure S7H and S7I). These results confirmed that the decrease in muscle mass in KO mice is due to deletion of the lncMGPF transcript and overexpression of lncMGPF can rescue the muscle phenotype of KO mice. Finally, to determine the effect of lncMGPF overexpression in WT mice, we injected the LV‐lncMGPF and LV‐control vectors intramuscularly into the left and right legs of 1‐month‐old and 3‐month‐old growing WT mice. Consistent with the results for KO mice, there were significant increases in muscle mass, mean cross‐sectional area, number of myonuclei per fibre, and myogenic gene expression after lncMGPF overexpression in WT muscles (Supporting Information, Figures S8 and S9). Therefore, increasing lncMGPF via lentivirus causes young growing muscle to grow more, but the effect of lncMGPF on mature muscle fibres is not known and needs to be further investigated.
To evaluate the role of lncMGPF in muscle regeneration, we performed a CTX‐induced TA muscles injury experiment with WT and KO mice. At Day 21, muscle regeneration and repair were almost complete in WT mice, but new formation of myofibres was ongoing in KO mice (Figure 3K). Immunofluorescence results for MyoG and eMyHC, a marker of muscle regeneration, showed that WT mice had increased numbers of MyoG positive (MyoG+) cells and eMyHC at 3 days post‐injection with CTX compared with KO mice (Supporting Information, Figure S10A and S10B), which indicates that lncMGPF KO decreases myogenic differentiation and weakens muscle regeneration. To further test whether lncMGPF affects proliferation of satellite cells, we used Pax7 and EdU to compare the proliferation capacities of satellite cells in regenerating myofibres between WT and KO mice 3 days after CTX injury. As expected, no significant changes were found between WT and KO mice in the percentage of proliferating muscle satellite cells (Pax7+/EdU+) (Figure S10C). In summary, these results suggest that lncMGPF promotes muscle regeneration by enhancing satellite cell differentiation.
Because localization of lncMGPF shifts to the cytoplasm during the course of myoblast differentiation, we speculated that lncMGPF might function as a molecular sponge of miRNA in myogenic differentiation. To confirm the association between lncMGPF and the miRNA‐protein complex, we performed an RIP assay for Ago2, which is the core component of the RNA‐induced silencing complex. The results showed that lncMGPF was present in Ago2‐enriched RNA (Figure 4A), which indicates that lncMGPF can interact with the Ago2‐based miRNA‐induced repression complex. Next, we identified binding sites for seven myogenesis‐related miRNAs in lncMGPF sequences using RNAhybrid 2.12 (
To further confirm that lncMGPF is a target gene of miR‐135a‐5p, we established a WT luciferase construct of lncMGPF (Luc‐lncMGPF) as well as a form in which the putative miR‐135a‐5p binding site was mutated (Luc‐lncMGPF‐Mut). Transfection of miR‐135a‐5p suppressed the luciferase activity of Luc‐lncMGPF but had no significant effects on the luciferase activity of Luc‐lncMGPF‐Mut in C2C12 myoblasts or HeLa cells (Figure 4E). Then, we investigated whether lncMGPF can bind to miR‐135a‐5p competitively with MEF2C. Transfection of miR‐135a‐5p mimics in C2C12 myoblasts and HeLa cells led to decreased Luc‐lncMGPF luciferase activity, whereas luciferase activity was substantially elevated after cotransfection of MEF2C‐3′ UTR (Figure 4F). Similarly, transfection of lncMGPF attenuated the inhibitory effects of miR‐135a‐5p on Luc‐MEF2C‐3′ UTR luciferase activity (Figure 4G). These data demonstrate that lncMGPF competes with MEF2C‐3′ UTR to bind miR‐135a‐5p. To further confirm that lncMGPF interacts directly with miR‐135a‐5p, we incubated different amounts of biotin‐labelled lncMGPF (1, 3, and 5 μg) with cytoplasmic lysates from C2C12 myoblasts transfected with miR‐135a‐5p, and we purified and analysed the RNA captured using qRT‐PCR. Compared with lncMGPF‐Mut, in which the putative miR‐135a‐5p binding site was mutated, biotin‐labelled lncMGPF specifically pulled down miR‐135a‐5p in a dose‐dependent manner (Figure 4H). Taken together, these data indicate that lncMGPF is a target gene of miR‐135a‐5p.
In addition, we performed cotransfection experiments using lncMGPF and miR‐135a‐5p overexpression vectors in C2C12 myoblasts and WT myogenic progenitor cells. Overexpression of miR‐135a‐5p decreased mRNA and protein expression of MEF2C, MyoD, MyoG, and MyHC, whereas cotransfection of lncMGPF eliminated miR‐135a‐5p activity. By contrast, cotransfection of lncMGPF‐Mut did not alter the inhibitory effects of miR‐135a‐5p on MEF2C gene expression but still counterbalanced miR‐135a‐5p‐induced inhibition of MyoD, MyoG, and MyHC expression (Figure 4I–L), which suggests the involvement of mechanisms other than sponging of miR‐135a‐5p. To further verify this possibility, we transfected the empty vector and the lncMGPF and lncMGPF‐Mut vectors separately into C2C12 myoblasts. As predicted, overexpression of lncMGPF‐Mut significantly increased expression of the MyHC gene but had no effect on the expression of MEF2C, which confirms the existence of other regulatory mechanisms (Figure 4M and 4N). Taken together, these data demonstrate that lncMGPF regulates myogenesis through sponging of miR‐135a‐5p and other unknown mechanisms.
As lncRNAs can regulate gene expression by interacting with multiple proteins,18,42 we performed a biotin‐labelled RNA pulldown assay along with SDS‐PAGE staining to identify lncMGPF‐specific binding proteins. Three specific protein bands were identified on the silver‐stained SDS‐PAGE gel (Supplementary Figure S12A). These specific protein bands were cut out and subjected to mass spectrometry (MS). A total of 225 binding proteins verified to contain Ago2 were identified through MS (Supporting Information, Data Set 2). From among these proteins, we selected HuR for further investigation as previous studies have shown that HuR promotes myogenic differentiation by increasing the mRNA stability of MyoD and MyoG in the cytoplasm.32,43 To validate the interaction between lncMGPF and HuR, we performed RIP experiments in C2C12 myoblasts using the HuR antibody. The results showed that HuR interacted with lncMGPF in addition to known interacting RNAs for MyoD, MyoG, and Linc‐MD1 (Figure 5A). This interaction was validated through bio‐labelled RNA pulldown (Figure 5B). Although HuR interacted with lncMGPF, no mutual regulation of lncMGPF and HuR expression was observed (Figure S12B–S12E).
As HuR enhances the mRNA stability of MyoD and MyoG by binding to AU‐rich elements (AREs) in the mRNA 3′ UTR,32,43 we speculated that lncMGPF may promote myogenic differentiation by enhancing HuR‐mediated mRNA stabilization of MyoD and MyoG. To validate this assumption, we transfected the overexpression vector and siRNA of lncMGPF into C2C12 myoblasts and then treated cells with ActD, which can inhibit the synthesis of RNA and interfere with the transcription process, to detect mRNA stability. Results of qRT‐PCR showed that overexpression of lncMGPF increased the mRNA stability of MyoD and MyoG (Figure 5C and 5D). By contrast, knockdown of lncMGPF reduced the mRNA stability of MyoD and MyoG (Figure 5E and 5F). Moreover, lncMGPF KO in myogenic progenitor cells significantly reduced the mRNA stability of MyoD and MyoG (Supporting Information, Figure S13A and S13B). To further investigate whether lncMGPF enhances mRNA stability through increasing the capacity of HuR to bind to the 3′ UTR of MyoD and MyoG mRNA, we performed HuR RIP assays in C2C12 myoblasts after overexpression or knockdown of lncMGPF. Overexpression and knockdown of lncMGPF significantly increased and decreased the capacity of HuR protein to bind to MyoD and MyoG mRNAs (Figure 5G and 5H), respectively. Consistent with these results, KO of lncMGPF in myogenic progenitor cells significantly decreased the ability of HuR to bind to MyoD and MyoG mRNAs (Figure S13C). Finally, we performed cotransfection experiments with the lncMGPF overexpression vector and HuR siRNAs in C2C12 myoblasts to explore the effects of HuR on lncMGPF‐mediated mRNA stabilization. The results indicated that HuR knockdown eliminated the effects of lncMGPF promotion of the mRNA stability of the MyoD and MyoG genes (Figure 5I and 5J). However, the mRNA stability of the MyoD and MyoG genes increased significantly with HuR overexpression in both WT and lncMGPF KO myogenic progenitor cells (Figure 5K and 5L). These results suggest that lncMGPF regulation of target mRNA stability depends on HuR protein, but the role of HuR is not entirely dependent on lncMGPF.
To identify the HuR binding region of lncMGPF, we constructed three truncated fragments of lncMGPF and used them in RNA pulldown experiments. The results showed that fragment F3 (1176–1795 bp) efficiently pulled down HuR, whereas fragments F1 (1–555 bp) and F2 (556–1175 bp) rarely pulled down HuR (Figure 6A). Next, we constructed six truncated fragments of lncMGPF and found that fragment F7 (1630–1795 bp) was the core region involved in the interaction between lncMGPF and HuR (Figure 6B). Bioinformatics analyses showed that the sequence of fragment F7 contains predicted classic HuR binding motifs (GUUG, GUUUUUC) and multiple A‐enriched or U‐enriched sequences, which are not classic binding motifs of HuR (Figure S13D). To investigate whether the classic binding motif is necessary for lncMGPF to interact with HuR, we constructed the fragment lacking the classic binding motifs (F7Δ1764–1779) and performed RNA pulldown. Unexpectedly, deletion of the classic binding motifs did not affect the interaction between lncMGPF and HuR (Figure 6C). Next, we examined the effects of the various truncated fragments of lncMGPF on the mRNA stability of target genes in differentiated C2C12 myoblasts. The results showed that overexpression of full‐length lncMGPF (FL) and truncated fragments (F7 and F7Δ1764–1779) significantly increased the stability of MyoD and MyoG mRNAs, whereas overexpression of the truncated fragment lacking the HuR binding region (F8; 1–1629 bp) did not increase the mRNA stability of the MyoD and MyoG genes (Figure 6D and 6E). To further confirm the core motif of lncMGPF binding to HuR, we deleted three potential binding motifs, including one classic HuR binding motif (F7Δ1764–1779) and two A/U‐enriched sequences (F7Δ1730–1743; F7Δ1747–1763). The deletion of any single binding motif did not affect the binding of lncMGPF and HuR, while no binding of lncMGPF and HuR was observed after deletion of all three binding motifs (Figure 6F). These results suggested that the three motifs are independent of each other and that their binding functions are complementary. On the basis of these results, we conclude that lncMGPF enhances the stability of MyoD and MyoG mRNA via its HuR binding region.
Previous studies have reported that HuR protein is transferred from the nucleus to the cytoplasm during cell differentiation, and accumulation of HuR protein in the cytoplasm is required for target mRNA stabilization and pro‐myogenic processes.44–46 Therefore, we performed nuclear and cytoplasmic fractionation experiments to examine the effects of overexpression and knockdown of lncMGPF on HuR protein localization in differentiated C2C12 myoblasts. We found that cytoplasmic HuR increased significantly with lncMGPF overexpression (Figure 6G and 6H), whereas knockdown of lncMGPF reduced HuR accumulation in the cytoplasm (Figure 6I and 6J). These observations were validated in lncMGPF KO and WT myogenic progenitor cells (Figure S13E and S13F). Previous studies have indicated that HuR is cleaved into two fragments, known as HuR‐cleavage products 1 and 2 (HuR‐CP1/2), where HuR‐CP1 increases the accumulation of HuR in the cytoplasm.44,46 To verify whether lncMGPF affects the cleavage of HuR, we determined the influence of lncMGPF overexpression or knockdown on protein levels of HuR‐CP1 by western blotting and found that overexpression of lncMGPF significantly increased HuR‐CP1 levels (Figure 6K), while knockdown of lncMGPF significantly reduced HuR‐CP1 levels (Figure 6L). Taken together, these results demonstrated that lncMGPF enhances the stability of MyoD and MyoG mRNAs by increasing HuR cleavage and accumulation in the cytoplasm.
Recent studies have indicated that lncRNA exhibits cross‐species conservation in terms of genomic position and RNA sequence.47,48 For example, linc‐YY1 is transcribed from upstream the YY1 promoter in the human and mouse genomes and is functionally conserved in human and mouse myogenesis.49 In the UCSC Genome Browser database, we found that pig lncRNA AK394747 has a conserved genomic position with mouse lncMGPF (Supporting Information, Figure S14). Both AK394747 and lncMGPF are transcribed from the genomic region between the H2AFZ and MTTP genes (Figure S14A). Bioinformatics analyses showed that pig AK394747 and mouse lncMGPF each have one miR‐135a‐5p binding site and one conserved sequence containing the HuR binding region (Figure S14B and S14C). 5′ and 3′ RACE assays showed that pig AK394747 is a transcript 1586 bp in length (Supporting Information, Figure S15A). CPC analyses indicated that AK394747 is a non‐coding RNA transcript similar to the described lncRNA HOTAIR, NEAT1, and MEG3 (Figure S15B). In addition, AK394747 is highly expressed in skeletal muscles and is gradually up‐regulated during porcine myogenic progenitor cells differentiation and postnatal muscle development (Figures 7A and 7B and S15C–S15F). Cell fractionation assays demonstrated that AK394747 occurs mainly in the nuclei of proliferating pig myogenic progenitor cells and in the cytoplasm of differentiated cells (Figure 7C).
To determine whether AK394747 is involved in myogenesis, we used functional gain and loss experiments to study the effects of AK394747 on the proliferation and differentiation of pig myogenic progenitor cells. We designed three siRNAs to knock down AK394747 and found that siRNA‐2 has the highest interference efficiency among three designed siRNAs in PK cells (Supporting Information, Figure S16A). Then siRNA‐2 was subcloned into the lentivirus‐mediated shRNA vector to infect pig myogenic progenitor cells. Results of qRT‐PCR, western blotting, and immunofluorescence staining for differentiated myogenic progenitor cells showed that lentivirus‐mediated AK394747 knockdown (LV‐sh‐AK394747) inhibited myogenic progenitor cells differentiation and fusion (Figure S16B–ES16), whereas lentivirus‐mediated overexpression of AK394747 (LV‐AK394747) promoted myogenic progenitor cells differentiation and fusion (Figure 7D–7G). However, AK394747 did not significantly influence cell proliferation, as demonstrated by qRT‐PCR, western blotting, EdU staining, and CCK‐8 with proliferating myogenic progenitor cells (Supporting Information, Figure S17A–S17H). Moreover, we injected the LV‐AK394747 vector and empty control (LV‐control) vector into the left and right biceps femoris muscles of 1‐month‐old pigs, respectively. As expected, overexpression of AK394747 in pigs significantly increased expression of MEF2C, MyoD, MyoG, and MyHC and the cross‐sectional areas of individual myofibres (Figures 7H–7J and S17I). To further confirm that lncMGPF is functionally conserved between mice and pigs, we conducted the following experiments. First, we infected pig myogenic progenitor cells with LV‐lncMGPF vector and induced differentiation for 3 days. Results from qRT‐PCR, western blotting, and immunofluorescence staining showed that overexpression of mouse lncMGPF increased mRNA and protein expression of the myogenic marker genes MyoD, MyoG, and MyHC (Supporting Information, Figure S18A–S18D), which suggests that mouse lncMGPF promotes myogenic progenitor cells differentiation in pig. Next, we infected C2C12 myoblasts with LV‐AK394747 vector and induced differentiation. As expected, overexpression of pig AK394747 also significantly promoted C2C12 myoblast differentiation (Figure S18E–S18G). Finally, to further confirm whether pig AK394747 could facilitate satellite cell fusion and developmental growth, we injected LV‐AK394747 vector and LV‐control vector intramuscularly into the left and right leg muscles of 1‐month‐old and 3‐month‐old lncMGPF KO mice and WT mice, respectively. The results showed that overexpression of AK394747 significantly increased muscle weights, cross‐sectional areas of individual myofibres, and myonuclei number per fibre, as well as mRNA and protein expression of the MEF2C, MyoD, MyoG, and MyHC genes (Supporting Information, Figure S19–S21). All results indicated that pig AK394747 has functions consistent with those of mouse lncMGPF.
Next, we examined whether AK394747 regulates pig myogenesis via similar mechanisms to those of mouse lncMGPF. An RNA pulldown assay using biotinylated AK394747 showed that sense AK394747 could bind with HuR, whereas the antisense RNA strand pulled down no HuR protein (Figure 8A). To determine whether the conserved sequence (1209–1556 bp) in pig AK394747 is necessary for the interaction with HuR, we constructed four truncated AK394747 fragments and found that fragment F3 (1380–1556 bp) corresponding to HuR binding region of mouse lncMGPF is required in the interaction of AK394747 and HuR (Figure 8B). We confirmed that only overexpression of full‐length AK394747 and fragment F3 significantly increased mRNA stability of MyoD and MyoG genes in porcine myogenic progenitor cells (Figure 8C and 8D), which suggests that AK394747 regulates mRNA stability of these genes by interacting with HuR protein. To verify that AK394747 regulates myogenic differentiation via miRNA sponging, we established a WT luciferase construct of AK394747 (Luc‐AK394747) and a form in which the putative miR‐135a‐5p binding site was mutated (Luc‐AK394747‐Mut) and then investigated the effects of miR‐135a‐5p on luciferase activity in pig myogenic progenitor cells and PK cells. The miR‐135a‐5p mimics significantly reduced the luciferase activity of Luc‐AK394747 but had no significant effects on that of Luc‐AK394747‐Mut (Figure 8E), which indicates that AK394747 combined directly with miR‐135a‐5p as a miRNA decoy. Next, to investigate whether miR‐135a‐5p targets the MEF2C and AK394747 genes in pig myogenic progenitor cells, we transfected miR‐135a‐5p mimics and an inhibitor into myogenic progenitor cells and found that overexpression and knockdown of miR‐135a‐5p significantly reduced and promoted, respectively, the mRNA and protein expression of target genes, including MEF2C, MyoD, MyoG, MyHC, and AK394747 (Figure 8F and 8G and Supporting Information, Figure S22). To further demonstrate the relationship between AK394747 and miR‐135a‐5p in myogenic differentiation, we performed cotransfection experiments in pig myogenic progenitor cells. Overexpression of miR‐135a‐5p decreased the mRNA and protein expression of MEF2C, MyoD, MyoG, and MyHC genes, whereas cotransfection of AK394747 eliminated the effects of miR‐135a‐5p. When the miR‐135a‐5p binding site of AK394747 was mutated (AK394747‐Mut), cotransfection did not alter the inhibitory effects of miR‐135a‐5p on MEF2C gene expression but still counteracted miR‐135a‐5p inhibition of the expression of other genes (Figure 8H and 8I), consistent with the observations in mice. These results suggest that AK394747 promotes myogenesis by up‐regulating HuR‐mediated mRNA stability and by acting as a miR‐135a‐5p sponge to enhance expression of the MEF2C gene.
We identified a novel transcript in human skeletal muscle myoblasts from Basic Local Alignment Search Tool analyses of the NCBI database, RT‐PCR, and RACE assay verification. 5′ and 3′ RACE assays showed that the transcript is 1360 bp in length (Supporting Information, Figure S23A–S23C); it has been deposited in the GeneBank (accession number MT510647) and named human lncMGPF (hlncMGPF). hlncMGPF has a conserved genomic position with mouse and pig lncMGPF and contains a potential miR‐135a‐5p binding site and HuR protein binding motifs (Figure S23D–S23F). CPC2 analyses indicated that hlncMGPF is a non‐coding RNA transcript similar to lncRNA HOTAIR and Tsix (Figure S23G). The hlncMGPF expression gradually increases with human myoblast differentiation (Figure 9A). To determine whether hlncMGPF is involved in human myogenesis, we infected human skeletal muscle myoblasts with a lentivirus‐mediated hlncMGPF overexpression (LV‐hlncMGPF) vector and induced differentiation for 3 days. The overexpression of hlncMGPF significantly increased the expression of MyoD, MyoG, MyhC, MEF2C, and α‐actin (Figure 9B and 9C) and promoted human skeletal muscle myoblast differentiation and fusion (Figure 9D). Next, to elucidate whether hlncMGPF functions in skeletal muscle development, we injected the LV‐hlncMGPF vector and LV‐control vector into the left and right leg muscles of 1‐month‐old WT mice, respectively (Figure 9E). Consistent with the functions of lncMGPF in the mouse and pig, hlncMGPF overexpression significantly increased the muscle weights, cross‐sectional areas of individual myofibres, number of myonuclei per fibre, and expression of the MyoD, MyoG, MEF2C, and MyHC genes (Figures 9F–9L and S23H and S23I). The results indicated that lncMGPF is functionally conserved in humans.
Myogenesis is a complex process that is finely tuned by genetic and epigenetic regulatory networks. Over the past few years, researchers have identified a number of lncRNAs in skeletal muscles using high‐throughput technologies such as next‐generation sequencing and microarray, but few lncRNAs have been functionally annotated as key regulators of myogenesis and muscle growth. In this study, we functionally identified lncMGPF as a novel regulator of muscle differentiation. lncMGPF is highly expressed in muscles, and its expression increases gradually from embryonic to postnatal muscle development. KO of lncMGPF in mice substantially decreases their growth rate, reduces myofibre size and muscle mass, and impairs muscle regeneration. These data allow us to conclude that lncMGPF plays critical positive regulatory roles in muscle growth and regeneration. In addition, the function and mechanism of lncMGPF during myogenesis are conserved among mouse, pig, and human, which suggests that this lncRNA may have potential applications for treating muscle disease and improving the production of animal meat.
Recent evidence has revealed that lncRNAs may function at the post‐transcriptional level.18,50 In this work, we found that lncMGPF is transferred from the nucleus to the cytoplasm during cell differentiation and promotes myogenesis mainly through post‐transcriptional regulation. On the one hand, lncMGPF sponges miR‐135a‐5p and weakens miR‐135a‐5p‐mediated inhibition of MEF2C gene expression. miR‐135a‐5p plays critical roles in diverse biological processes and diseases, including osteoblast differentiation,51 myogenic differentiation,52 lung cancer,53 colorectal cancer,54 and prostate cancer.55 Previous studies have reported that Linc‐MD1 and MEG3 control muscle differentiation by functioning as sponges of miR‐135a‐5p to regulate the expression of MEF2C.22,56 Our results add the new evidence for the notion that the miR‐135a‐5p/MEF2C axis is an important pathway in the lncRNA‐mediated network of myogenesis regulation. On the other hand, lncMGPF modulates the mRNA stability of MyoD and MyoG genes by interacting with HuR protein. HuR is a member of the embryonic lethal abnormal vision—like family of RNA‐binding proteins and is widely expressed in a variety of tissues and cells.57,58 HuR increases the mRNA stability by specifically binding to AREs located in the 3′ UTRs of its target transcripts.59–61 LncRNAs also participate in stabilization of target mRNAs by interacting with HuR in tumorigenesis, such as B4GALT1‐AS1,62 MIR100HG,63 LINC00324,64 and LINC00707.65 In muscle cells, HuR regulates myogenic differentiation and muscle fibre formation by enhancing the stability of MyoD and MyoG mRNAs.32,43,45,46,66 Our data indicate that lncMGPF enhances the mRNA stability of myogenic genes by recruiting HuR to AREs of MyoD and MyoG genes. Bioinformatics analyses and RNA pulldown experiments identify three independent and complementary HuR binding motifs. Furthermore, we found that lncMGPF promotes the cleavage of HuR and increases HuR accumulation in the cytoplasm, which is necessary for HuR to function in myogenesis.44–46 Therefore, we conclude that lncMGPF affects the accumulation of HuR in the cytoplasm by regulating HuR cleavage.
Recent studies have revealed that lncRNAs play their roles by interacting with multiple proteins,42,67 so understanding how proteins bind to particular lncRNAs is crucial to elucidating their mechanisms of action. In this study, we used RNA pulldown and MS analyses to identify 225 proteins bound to lncMGPF. Aside from Ago2 and HuR, we observed proteins related to m6A methylation, including writer (Mettl3) and reader (IGF2BP2, HNRNPC, and HNRNPA2B1) proteins.68–70 Like mRNAs, m6A methylation of lncRNA is widespread and plays important regulatory roles in diverse biological processes; for example, the well‐known lncRNA Xist interacts with m6A modification proteins, including writer (Mettl14, BMP15/16, WAP) and reader (YTHDC1) proteins, leading to m6A methylation of Xist, which is necessary for Xist‐mediated transcriptional repression.71 In addition, HuR protein is involved in m6A methylation as a reader,72 and mettl3‐promoted MyoD mRNA maintenance is required for myogenic differentiation.73 Thus, it is interesting to further investigate whether the interactions of lncMGPF with Mettl3 and HuR are related to m6A methylation of lncMGPF or whether lncMGPF regulates MyoD mRNA stability through interaction and recruitment of Mettl3.
In conclusion, we have found that lncMGPF is a novel positive regulator of myogenesis in mice and pigs that regulates muscle differentiation at the post‐transcriptional level. The MyoD‐activated lncRNA lncMGPF is transferred from the nucleus to the cytoplasm during cell differentiation and promotes myogenesis mainly through post‐transcriptional regulation. lncMGPF functions as a miRNA sponge of miR‐135a‐5p, weakening the inhibitory effects of miR‐135a‐5p on MEF2C and thereby increasing expression of the MEF2C gene. Meanwhile, lncMGPF increases HuR accumulation in the cytoplasm and enhances stability of MyoD and MyoG mRNAs via recruitment of HuR to AREs in the 3′ UTRs of MyoD and MyoG mRNAs (Figure 9M). The lncMGPF/miR‐135a‐5p/MEF2C and lncMGPF/HuR/MyoD/MyoG pathways together constitute a novel MRF‐mediated regulatory network for myogenesis.
This work was financially supported by the National Key Project of Transgenic Research (Grant 2016ZX08006‐002), the National Natural Science Foundation of China (Grant 31900448), the Agricultural Innovation Fund of Hubei Province (2016‐620‐000‐001‐043), the Fundamental Research Funds for the Central Universities (Program 2662018PY045), and the China Postdoctoral Science Foundation (Program 590319103).
BZ, SZ, WL, and JJ conceived and designed the research; WL, JJ, ZX, HL, YG, XW, SW, JZ, HZ, WB, YP, and JT performed experiments; WL, JJ, ZX, and BZ analysed the data; WL, JJ, and BZ wrote the manuscript. All authors read and approved the final manuscript.
We thank Dr Wang FC for helping in the generation of gene knockout mice. We thank Cheng YJ and Qu WP for siRNAs screening. The authors certify that they comply with the ethical guidelines for authorship and publishing of the Journal of Cachexia, Sarcopenia and Muscle.74
The authors declare no conflict of interest.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
© 2020. 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
Long non‐coding RNAs (lncRNAs) play critical regulatory roles in diverse biological processes and diseases. While a large number of lncRNAs have been identified in skeletal muscles until now, their function and underlying mechanisms in skeletal myogenesis remain largely unclear.
Methods
We characterized a novel functional lncRNA designated lncMGPF (lncRNA muscle growth promoting factor) using RACE, Northern blot, fluorescence in situ hybridization and quantitative real‐time PCR. Its function was determined by gene overexpression, interference, and knockout experiments in C2C12 myoblasts, myogenic progenitor cells, and an animal model. The molecular mechanism by which lncMGPF regulates muscle differentiation was mainly examined by cotransfection experiments, luciferase reporter assay, RNA immunoprecipitation, RNA pull‐down, and RNA stability analyses.
Results
We report that lncMGPF, which is highly expressed in muscles and positively regulated by myoblast determination factor (MyoD), promotes myogenic differentiation of muscle cells in vivo and in vitro. lncMGPF knockout in mice substantially decreases growth rate, reduces muscle mass, and impairs muscle regeneration. Overexpression of lncMGPF in muscles can rescue the muscle phenotype of knockout mice and promote muscle growth of wild‐type mice. Mechanistically, lncMGPF promotes muscle differentiation by acting as a molecular sponge of miR‐135a‐5p and thus increasing the expression of myocyte enhancer factor 2C (MEF2C), as well as by enhancing human antigen R‐mediated messenger RNA stabilization of myogenic regulatory genes such as MyoD and myogenin (MyoG). We confirm that pig lncRNA AK394747 and human lncRNA MT510647 are homologous to mouse lncMGPF, with conserved function and mechanism during myogenesis.
Conclusions
Our data reveal that lncMGPF is a novel positive regulator of myogenic differentiation, muscle growth and regeneration in mice, pigs, and humans.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
Details
1 Key Laboratory of Swine Genetics and Breeding of Ministry of Agriculture and Rural Affairs & Key Laboratory of Agriculture Animal Genetics, Breeding and Reproduction of Ministry of Education, College of Animal Science and Technology, Huazhong Agricultural University, Wuhan, China
2 Key Laboratory of Swine Genetics and Breeding of Ministry of Agriculture and Rural Affairs & Key Laboratory of Agriculture Animal Genetics, Breeding and Reproduction of Ministry of Education, College of Animal Science and Technology, Huazhong Agricultural University, Wuhan, China; Department of Basic Veterinary Medicine, College of Veterinary Medicine, Huazhong Agricultural University, Wuhan, China
3 Hubei Key Laboratory of Embryonic Stem Cell Research, School of Basic Medicine Science, Hubei University of Medicine, Shiyan, China
4 Key Laboratory of Swine Genetics and Breeding of Ministry of Agriculture and Rural Affairs & Key Laboratory of Agriculture Animal Genetics, Breeding and Reproduction of Ministry of Education, College of Animal Science and Technology, Huazhong Agricultural University, Wuhan, China; The Cooperative Innovation Center for Sustainable Pig Production, Wuhan, China