Porcine skeletal muscle is an important source of protein, and a model for studying human muscle-related diseases [1
]. The development and regeneration of skeletal muscle mainly depends on the proliferation and differentiation of porcine skeletal muscle satellite cells [3
]. As an adult muscle stem cell, porcine satellite cells (PSCs) usually exist in a resting state between the sarcolemma and the basal lamina of the muscle tissue, and play an important role in maintaining its normal physiological function [4
]. The proliferation and differentiation of PSCs are regulated by a variety of factors, including histone modification, transcription factors, and long non-coding RNAs (lncRNAs) [6
]. How histone modifications control and affect the proliferation and differentiation of PSCs remains largely unknown.
Previous studies have shown that the differentiation ability of satellite cells was distinct in different species. The differentiation efficiency of PSCs was less than 65%, and the process was regulated by various myogenic factors such as MYOD1, MYF5, MYOG, and MYF6 [7
]. Second, epigenetic modifications like histone modifications are closely related to the transcriptional regulation of myogenic genes [3
]. In mouse C2C12 cells, H3K27me3 histone modification regulates myogenic differentiation of satellite cells by regulating the expression of myogenic transcription factors, such as Myog, while H3K4me3 at the transcription initiation sites of Myf5
would respectively promote the proliferation and differentiation in C2C12 cells [12
]. In addition, chromatin conformation rearrangements are also required for differentiation and muscle fiber formation. The polycomb protein Ezh2 mediates H3K27me3 to maintain the inhibitory chromatin conformation of myogenic genes, and to inhibit differentiation [15
]. The methyltransferase UTX
knockout mice showed that abnormal H3K27me3 modification leads to disordered muscle fiber regeneration [16
]. In addition, decrease in H4K20me2 marks in Suv-20h1
gene knockout mice resulted in genome-wide H3K27me3 depletion, thereby activating the resting satellite cells to commit myogenic differentiation [3
]. However, whether the differentiation of PSCs is associated with histone modifications, especially the H3K27me3, is not well established.
In this study, PSCs were isolated from newborn piglets and cultured. Transcription and histone modifications were identified by genome-wide profiling of transcriptome and chromatin states, respectively. Our results identified 917 differentially expressed genes (DEGs), which were closely related to H3K4me3, H3K27ac, and H3K27me3. Furthermore, the genome-wide histone modification level significantly decreased during differentiation. Further, H3K27me3 was reduced by 50%, which largely led to the upregulation of 139 myogenic DEGs during PSC differentiation. The results of this study will provide clues for further studies on the relationship between myogenic DEGs and histone modifications, especially H3K27me3 underlying differentiation mechanism of PSCs, to further supplement the explanation for chromatin state and dynamics during skeletal muscle myogenesis in pigs.
2. Materials and Methods
2.1. Isolation, Culture, and Differentiation of Porcine Satellite Cells (PSCs)
Satellite cells were primarily isolated from hind leg muscles of one-week-old Yorkshire male piglets. Piglets were slaughtered according to a standard procedure approved by guidelines from the Regulation of the Standing Committee of Hubei People’s Congress (Hubei Province, China, HZAUSW-2017-008). Skeletal muscles were minced into pieces and digested in 0.2% type I collagenase (Sigma, USA, V900891) in a shaking water bath at 37 °C for 2 h. The supernatant cell suspension was washed with DMEM high glucose medium (Life, USA, 10569) supplemented with 1% antibiotic-antimycotic (Life, 15240) and 50 μg/mL gentamycin (Life, 15750), and sequentially passed through 100 μm, 70 μm, and 40 μm filters (BD, USA, 352360, 352350, 352340) to remove tissue debris. The cells were resuspended in RPMI-1640 medium (Life, A10491) supplemented with 20% FBS (Gibco, USA, 10099-141), 1% non-essential amino acids (Gibco, 11140-050), 0.5% chicken embryo extract (GEMINI, USA, 100-163P), 1% GlutaMax (Gibco, 35050), 1% Antibiotic-Antimycotic, 50 μg/mL Gentamycin, and 2.5 ng/mL bFGF (Life, 13256). The mixed cells were cultured in uncoated plates for 2 h to remove fibroblasts using differential adhesion property. The purified satellite cells were transferred into the Matrigel (BD, 356234) coated plates for proliferation cultures. At ~60% confluence, the proliferation medium was replaced by the DMEM high glucose medium (Life, 10569) supplemented with 5% horse serum (HyClone, USA, SH30074.02), 1% antibiotic-antimycotic, and 50 μg/mL gentamycin to induce PSCs from proliferation to differentiation state. Three time points of differentiation, 1-day (D1), 2-day (D2), and 4-day (D4) were chosen to investigate the differentiation efficiency on three replicates from 3 independent piglets.
2.2. Immunofluorescence Assay of PSCs
Cells were seeded in Matrigel coated 6-well plates overnight to approximately 60% confluence, and washed with cold PBS (HyClone, SH30256.01) twice. The cells were then fixed in 4% paraformaldehyde for 15 min, and incubated in 0.25% Triton X-100 (Amresco, USA, 0694) for 15 min at room temperature. The primary monoclonal antibodies, PAX7 (ABclonal, China, A7335), MYOD1 (ABclonal, A0671), DES (Abcam, UK, ab8976), MYOSIN (Abcam, ab15), and MYOG (Abcam, ab1835) were hatched with the prepared cells at 4 °C overnight. The secondary antibodies, anti-mouse IgG alexa fluor 555 (CST, USA, 4409s) and anti-rabbit IgG alexa fluor 488 (CST, 4412s), were incubated with the cells for 1 h in dark room. Cell nuclei were stained with Hoechst33342 (Sigma, B2261). Images were captured using the OLYMPUS IX73 TH4-200 system (OLYMPUS, Japan).
2.3. Validation of Differentially Expressed Genes (DEGs) by Quantitative PCR (qPCR)
Total RNA was extracted using RNeasy Mini kit (QIAGEN, Germany, 74104) and DNase I (QIAGEN, 79254). Reverse transcription was performed using PrimeScript RT reagent kit with gDNA Eraser (Takara, Japan, RR047A). Random primers and oligo-dT primer were added to initiate cDNA first strand synthesis. The quantitative PCR reactions were carried out using the SYBR Green master mix (TOYOBO, Japan, QPK-201) on a LightCycler 480 II (Roche, Switzerland) system. Data were calculated by the Delta-Delta-Ct algorism, using GAPDH as internal calibration on 3 replicate samples. All primer sequences are listed in the supplementary data (Supplementary Materials Table S2
2.4. Chromatin Immunoprecipitation (ChIP) and ChIP-Seq Library Preparation
Cells (107) were cross-linked with 1% formaldehyde (EMD Millipore, USA, 344198) at room temperature for 10 min. The chromatin was obtained by cell lysis (50 mM HEPES, pH 7.5; 150 mM NaCl; 1mM EDTA; 1% Triton X-100; 0.1% sodium deoxycholate; 0.1% SDS; and proteinase inhibitors) and was followed by sonication (Sonics VCX130, USA) on ice for 6 min. ChIP-grade antibodies (10 μg), anti-H3K44me3 (Abcam, ab8580), anti-H3K27ac (Abcam, ab4729), anti-H3K27me3 (Millipore, 07-499), and anti-RNA polymerase II (Abcam, ab817), were incubated with 100 μL Protein G DynaBeads (Life, 10009D) at 4 °C for 2 h. The chromatin and the antibody coated beads were co-immunoprecipitated on a rotator at 4 °C overnight. The ChIP DNA was de-crosslinked by protein K (Ambion, AM2546) digestion at 55 °C overnight. DNA (1 μg) was used for the preparation of the Chromatin Immunoprecipitation Sequencing (ChIP-Seq) library by NEBNext UltraⅡ DNA Library Prep Kit for Illumina (NEB, USA, E7645 and E7335), according to the manufacturer’s instructions. Each ChIP-Seq library was high-throughput sequenced, ~50M paired-end reads (2 × 150 bp) on Illumina HiSeq3000 platform.
2.5. Library Construction and Sequencing
For each sample, total RNA was extracted and sent to Shanghai Biotechnology Corporation (Shanghai, China) for library construction. The RNA Sequencing (RNA-Seq) library of each sample was prepared by the [email protected]
II DNA Library Prep Kit (Illumina). And the library was sequenced paired-end reads (2 × 150 bp) on Illumina HiSeq3000 platform.
2.6. Data Preprocessing and Alignment
The sequencing raw data was first filtered to remove low-quality reads by fastQC, and adapters were removed using Cutadapt. The clean reads were then mapped to the pig genome (Ensembl Sscrofa 11.1) by Bowtie2 with default parameters [17
]. The unique mapping reads were further processed to remove duplications based on genomic coordinates. The reproducibility of de-duplication reads was analyzed by calculating the coefficients of two samples. Briefly, the genome is first divided into a number of 10 kb bins, then the reads of two samples (de-duplication unique mapping bam file) on each bin are taken as the coordinates of the x-axis and y-axis, respectively. Draw the reads on all bins in the form of a scatter plot and calculate the Pearson correlation coefficient. The closer reads number in a bin are to the samples, the more accumulation points are on the diagonal, the greater the correlation coefficient, indicating that the better repeatability of two samples [18
2.7. RNA-Seq Data Analysis
HTSeq software was used to count the reads for RNA-Seq data, after pre-processing. Further data mining was subsequently conducted including: DEGs analysis by DESeq2 package with default parameters, gene ontology (GO) analysis by DAVID and clusterProfiler package [19
], further analysis was performed using GOplot with the results from DAVID [20
], which are all available R packages [21
2.8. ChIP-Seq Data Analysis
The ChIP-Seq peak calling was performed by MACS2 with a 3 kb extension of promoter regions. The parameters of regular mode were used for peak calling of H3K4me3 H3K27ac and RNA polymerase Ⅱ (RNAPII), which were ‘macs2 callpeak -t ChIP.bam -c Control.bam -f BAMPE -g 2.7e+9 -n name -B -q 0.05′. While broad mode was used as ‘macs2 callpeak -t ChIP.bam -c Control.bam -f BAMPE -g 2.7e+9 -n name --broad --broad-cutoff 0.05′ for peak calling of H3K27me3 [22
]. The modifications of three histone marks (H3K4me3, H3K27ac, and H3K27me3) and RNAPII were normalized with input files as control using bamCompare of deepTools based on log2 fold change of the reads per kilobase million (RPKM) [23
]. Principal component analysis (PCA) was used to compare the correlation of ChIP-Seq data with multiBigwigSummary and plotPCA of deepTools with normalized files using spearman model with default parameters, and the enrichment profiles were performed using computeMatrix, plotProfile and plotHeatmap of deepTools. The annotation of the modification regions of three histone modifications and RNAPII were illustrated by ChIPseeker ± 3 kb of the transcription start site (TSS) region [24
2.9. Statistical Analysis
All results were presented as mean ± standard error of mean (SEM). Unpaired Student’s t-test was used to show the statistical significance. p <0.05 indicated the significant difference.
2.10. Data Availability
RNA-Seq and ChIP-Seq data were submitted to NCBI Sequence Read Archive database (SRP180031 and SRP180432).
In this study, satellite cells of porcine skeletal muscle were isolated, primarily cultured, and differentiated into myotubes in vitro, presenting an advantage for research on myogenesis in pigs. Three histone modification marks, H3K4me3, H3K27ac, and H3K27me3, together with RNAPII, were interrogated in myoblasts and myotubes to illustrate the relationship between transcriptomic profiles and histone modifications during differentiation. This study provides evidences on H3K27me3 modification depletion in PSCs, which promote myogenic differentiation. The genome-wide integrated analysis that combined epigenomics and transcriptomics to illustrate the relationship between H3K27me3 and myogenesis will be helpful for better understanding myogenesis in pigs.
Primary PSCs were purified by pre-plating and fused into the myotubes within 48h of culture in differentiation medium, less than the 72 h previously reported [9
]. Moreover, Transcriptomic analysis revealed that during cell differentiation a large number of muscle development-related genes were upregulated, in addition to metabolism-related genes including SHISA2
]. Cell cycle-related genes were significantly downregulated promoting PSCs differentiation, which were consistent with previous studies [27
]. Our results show that these important genes were differentially expressed during PSC differentiation, which were also confirmed by qPCR.
Intriguingly, previous studies mentioned that epigenetics played important roles in shifting heterochromatin to open chromatin for myogenic differentiation [28
]. The dramatic depletion of H3K27me3 histone modification was observed during PSCs differentiation (~50%). Furthermore, 139 myogenic DEGs were upregulated through H3K27me3 depletion, which enriched not only the terms necessary for myotube formation such as muscle tissue development, but negative biological processes of cell proliferation and neural development pathways as well. The rapid reduction in H3K27me3 mark was also found during specific stages of embryogenesis and stem cell differentiation [13
], as demonstrated by Dilworth using UTX
conditional knockout mice, suggesting the necessity of H3K27me3 depletion [16
]. In addition, it was believed that the dynamic balance between methyltransferase EZH2,
regulated the H3K27me3 depletion during cell differentiation [33
]. Consistently, the expression of EZH2
was shown to be significantly downregulated—barely detected in myotubes of PSCs. It was also found that the genes related to cell fate development, as well as maintenance and differentiation, were marked with downregulated H3K27me3 histone modifications, such as the gene regions of HOXD
cluster regulating the biomorphic development.
Furthermore, molecular mechanisms have increasingly uncovered that histone modification promote myogenesis. Our results emphasized the dramatic decrease in H3K27me3 modifications during cell differentiation, and genome-wide mapping of three epigenetic marks together with RNAPII demonstrated that the myogenic transcription factors were repressed with H3K27me3 modifications during proliferation phase, such as MYOG, which would drive myogenesis through H3K27me3 depletion. While the reduction of H3K27me3 modification was also reported at the TSS region of Myog
in C2C12, that lied at the key position in myogenic differentiation [13
]. Consistent with previous results, the terminal differentiation genes, which were induced by MYOD1
such as TNNC1
, were modified with positive marks (H3K4me3, H3K27ac, and RNAPII) instead of H3K27me3 for promoting myotube formation [13
]. In addition, significant myogenic characteristics were demonstrated in PSCs, which confirmed that myogenic transcription factors escaped from H3K27me3 histone modifications during cell differentiation, such as SIX4 and MYOD1 [35
]. The critical transcription factors of adipogenesis, osteogenesis, and muscle inhibitory genes such as PPAR-γ and PRDM16, were labeled with Polycomb repression at the TSS region to show tissue specificity, which would convert cell identity [13
]. Our result suggested the spatiotemporal specificity of H3K27me3 modification in releasing myogenic factors, which would activate downstream genes to regulate cell differentiation and myotube formation.
This study revealed that the significant depletion of H3K27me3 modification promotes PSCs differentiation. Our results extend the genome-wide understanding of the roles of H3K27me3 marks in regulating myogenic differentiation in pigs. However, further studies are required to fully dissect the epigenetic remodeling and epigenetic mechanism during PSCs differentiation.