Genome-Wide Analysis of Circular RNA-Mediated ceRNA Regulation in Porcine Skeletal Muscle Development

doi:10.21203/rs.3.rs-2095653/v1

Download PDF

Research Article

Genome-Wide Analysis of Circular RNA-Mediated ceRNA Regulation in Porcine Skeletal Muscle Development

https://doi.org/10.21203/rs.3.rs-2095653/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 12 Apr, 2023

Read the published version in BMC Genomics →

You are reading this latest preprint version

Background: As a diverse and abundant class of endogenous RNAs, circular RNAs (circRNAs) participate in various biological processes including cell proliferation and apoptosis. Nevertheless, few researchers have investigated the role of circRNAs in muscle development in cultivated pigs.

Results: In this study, we used RNA-seq to construct circRNA expression profiles in skeletal muscle of Jinfen White pigs at the age of 1, 90, and 180 days. Among the 16,990 identified circRNAs, 584 were differentially expressed, with 255, 477, and 63 DE circRNAs in the 90 d vs. 1 d, 180 d vs. 1 d, and 180 d vs. 90 d groups, respectively. Moreover, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses of DE circRNA host genes revealed them to be mainly involved in skeletal muscle fiber-related processes (e.g., muscle contraction, muscle organ development, and muscle system processes) and skeletal muscle fiber-related signaling pathways (e.g., AMPK and cAMP pathways). We also constructed circRNA–miRNA–mRNA co-expression network to screen out circRNAs many involved in the regulation of porcine skeletal muscle growth and development through the competitive endogenous RNA (ceRNA) mechanism. In this network, we predicted circ_0018595 may as a potential sponge of miR-1343 to regulate PGM1 expression, in turn promoting the proliferation of pig skeletal muscle satellite cells. The structure and expression of circ_0018595 were confirmed using convergent and divergent primer amplification, RNase R digestion, and qRT-PCR.

Conclusions: This study has identified 584 candidate circRNAs, especially circ_0018595, which may be involved in the growth and development of porcine skeletal muscle, and will therefore serve as a valuable resource for further in-depth study of circRNA regulatory mechanisms in skeletal muscle development.

pig

skeletal muscle

RNA-seq

circRNA

ceRNA

Skeletal muscle is the main tissue of the animal body, typically accounting for 30–50% of weight, and it has important functions in exercise, heat production, maintenance of body shape, storage of protein, and protection of organs[1, 2]. The development of mammalian skeletal muscle comprises two main stages: an increase in the number of muscle fibers before birth, and an increase in their volume after birth. Muscle fibers are derived from mesenchymal stem cells, which proliferate into myoblasts, differentiate and fuse into myotubes, fuse into myofibers, and finally form mature muscle tissue[3]. The quantity, diameter, and composition of muscle fibers are closely related to muscle yield and meat quality. Therefore, the growth and development of skeletal muscle directly affect meat production and growth rate in pigs[4, 5]. Regulation of the activity of myogenic stem cells involves many transcription factors, such as myoblast determination protein 1 (MyoD1), myogenin (MyoG), myogenic factor 5 (Myf5), and myogenic factor 6 (Myf6), as well as other myogenesis-related regulatory transcription factors, including the paired box 3 (PAX3), paired box 7 (PAX7), and myocyte enhancer factor 2 (MEF2) families[6–9]. Among these, MyoD1 and Myf5 are involved in the first stage of skeletal muscle development, which promoting the proliferation and differentiation of myogenic progenitor cells into myoblasts, while MyoG determines the generation of myotubes, and Myf6 is involved in cell differentiation and cell fate[10]. Myf5 and MyoD1 are myogenic determinants, as has been demonstrated using double- or triple-knockout (KO) mice[11, 12]. In contrast, MyoG is a differentiation factor, so MyoG-knockout mice lack differentiated muscles and die at birth[13]. Furthermore, since expression of MyoG is regulated by MRF4 and MyoD, these other factors are, in turn, involved in myogenic differentiation[14].

In addition to the above roles of transcription factors, a growing number of studies have established that non-coding RNAs (ncRNAs) are also involved in the regulation of myogenesis. These include microRNAs (miRNAs), long non-coding RNAs (lncRNAs), and circular RNAs (circRNAs), whose regulatory roles in the growth of skeletal muscle are gradually being revealed[15–17]. With the development of high-throughput sequencing technology, circRNA has received increasing attention in recent years. As the most recently emerging class of ncRNA, circRNA is essentially a closed RNA transcript that is produced by reverse splicing of precursor mRNA (pre-mRNA). The 3′ and 5′ ends are covalently linked, and they differ from their linear counterparts in the absence of a 5′ cap and polyadenylated [poly(A)] tail[18]. Moreover, circRNAs have been identified in most species, and are generally expressed at low levels, but partly more abundantly than their host genes (also referred to as source genes)[19–21]. Compared with lncRNAs, circRNAs display highly conserved expression patterns and transcript sequences among species, and they are widely expressed in various tissues and organs[22]. Furthermore, due to their nonlinear structure, circRNAs are more stable than linear RNA species[23, 24].

Much current research on circRNAs focuses on aspects of cancer, including cell proliferation, apoptosis, invasion, and chemoresistance[25]. For example, circANAPC7 acts through the CREB/miR-373/PHLPP2 axis, resulting in dephosphorylation of AKT and downregulation of Cyclin D1 and TGF-β, thereby inhibiting tumor growth and muscle atrophy in pancreatic cancer[26]. The splicing factor-derived circular RNA, circCAMSAP1, accelerates nasopharyngeal carcinoma tumorigenesis via a SERPINH1/c-Myc positive feedback loop[27]. Conversely, METTL14-mediated m^6A modification of circORC5 suppresses gastric cancer progression by regulating a miR-30c-2-3p/AKT1S1 axis[28].

Studies on circRNAs in animals are less common, but have mainly looked at regulatory roles of circRNAs in different species, particularly with respect to myoblast proliferation and differentiation. During the development of chicken muscle, for example, circRBFOX2 can sponge and negatively regulate miR-206, thereby increasing CCND2 expression and promoting myoblast proliferation[29]. Similarly, circFAM188B encodes a novel protein named circFAM188B-103aa that promotes the proliferation of chicken skeletal muscle satellite cells but suppresses differentiation[30]. In addition, circFGFR2—produced from exons 3–6 of FGFR2 acts as a sponge for miR-133a-5p and miR-29b-1-5p[31], both of which are important regulators of myogenesis, and can inhibit proliferation and promote differentiation of myoblasts by targeting AKT[32, 33]. During bovine muscle development, circFUT10, generated from FUT10, is expressed at higher levels in embryonic skeletal muscle than adult skeletal muscle. Mechanistically, circFUT10 sponges miR-133a, leading to inhibition of cell proliferation and induction of apoptosis in myoblasts; but it also finally enhances myoblast differentiation[34]. At present, research on circRNA regulation of porcine skeletal muscle growth mainly relies on in-depth mining of sequencing data, with the relevant mechanisms typically being explored in C2C12 cells (a mouse myoblast cell line). This approach was used to identify the role of circFGFR2 in inhibiting myoblast proliferation, and promoting differentiation and skeletal muscle regeneration in pigs[35].

Jinfen White pigs are cultivated from four parental breeds, including Mashen, Taihu, Danish Landrace, and Large White pigs, which have the advantages of good meat quality, strong resistance to stress, and rapid growth. However, in Jinfen White pigs, the characteristic and mechanism of regulation by circRNAs at different growth stages are still unclear. The existing research on circRNAs in muscle development is far from sufficient, and their regulatory mechanisms have not yet been fully clarified. Thus, further exploration is needed[36–38].

In this study, we performed RNA-seq on the longissimus dorsi muscle of Jinfen White pigs at three different developmental stages (1, 90, and 180 d; N = 3). As a result, we found that most circRNAs were specifically expressed at different stages. Furthermore, we constructed a competitive endogenous RNA (ceRNA) interaction network, which suggested that these circRNAs, especially circ_0018595, may act as key regulators of muscle development in Jinfen White pigs. To summarize, this study provides a comprehensive temporal atlas of circRNA expression in Jinfen White pig skeletal muscle that should benefit further studies on the molecular mechanisms of pig skeletal muscle development at a non-coding RNA level.

Overview of Sequencing Data

To understand the expression characteristics of circRNA in porcine muscle at different developmental stages, we performed ribosomal RNA-depleted RNA-seq on muscle sampled at three time points. Firstly, we constructed nine non-coding RNA libraries named JFW_1d_1, JFW_1d_2, JFW_1d_3, JFW_90d_1, JFW_90d_2, JFW_90d_3, JFW_180d_1, JFW_180d_2, and JFW_180d_3. Secondly, we purified and sequenced the RNA using the Illumina paired-end RNA-seq method. For these libraries, 908,657,602 raw reads were obtained from the platform, and after filtering, a final dataset of 902,652,752 clean reads was obtained (Table 1). The JFW_180d_3 group contained the greatest number of reads (106,567,824), and the JFW_1d_1 group had the least (91,942,766). The Q30 quality threshold (Phred quality score >30) was met for 95.2% of the data. Taken together, these results confirmed that the sample and sequencing data generated in this study were of good quality and were reliable for subsequent bioinformatic analysis.

Table 1. Overview of RNA-seq results for each sample.

Sample name	Raw reads	Clean reads	Raw_bases (G)	Clean_bases (G)	Error rate (%)	Q20 (%)	Q30 (%)
JFW_1d_1	92,655,706	91,942,766	13.9	13.79	0.02	98.62	95.72
JFW_1d_2	107,050,232	106,444,968	16.06	15.97	0.02	98.64	95.78
JFW_1d_3	104,392,040	103,814,406	15.66	15.57	0.02	98.58	95.45
JFW_90d_1	96,202,328	95,550,798	14.43	14.33	0.03	98.08	94.2
JFW_90d_2	106,311,518	105,563,842	15.95	15.83	0.02	98.56	95.75
JFW_90d_3	95,622,436	94,984,938	14.34	14.25	0.02	98.67	95.89
JFW_180d_1	94,853,354	94,375,978	14.23	14.16	0.03	97.81	93.78
JFW_180d_2	104,189,042	103,407,232	15.63	15.51	0.02	98.59	95.54
JFW_180d_3	107,380,946	106,567,824	16.11	15.99	0.02	98.40	95.04
Total	908,657,602	902,652,752	15.15	15.04	0.02	98.44	95.24

Characterization of circRNAs in Porcine Skeletal Muscle

After mapping the clean data to the pig genome (Sscrofa11.1) using the find_circ and CIRI software, we identified a total of 16,990 highly credible circRNAs from the nine libraries (summarized in more detail in Supplementary Table S1). In addition, we mapped the chromosomal distribution of these circRNAs, finding them to be transcribed from all chromosomes but with an uneven distribution (Figure 1A). In agreement with other earlier work[39], we observed that the highest number of circRNAs originated from chromosomes 1, 6, and 13 in this study (Figure 1A). Furthermore, since circRNAs are primarily derived from annotated exons[40], we also found that >90% identified circRNAs in this study contained exon sequences (Figure 1B). The length distribution of the predicted circRNAs showed that most of them were less than 1,000 nt in length (Figure 1C), which was also consistent with previous reports[41].

Differential Expression of circRNAs

To better understand the differences in circRNA expression patterns in pig muscle at 1, 90, and 180 days, we used DESeq software to identify DE circRNAs between sample pairs (i.e., JFW_90d vs. JFW_1d, JFW_180d vs. JFW_1d, and JFW_180d vs. JFW_90d). We also performed hierarchical clustering analysis of the DE circRNAs. This revealed distinct differential expression patterns for the identified circRNAs between the three time points in muscle, with JFW_1d showing significant differences compared to the other two groups (Figure 2A). In total, 584 DE circRNAs were detected using the above pairwise comparisons. The number of differentially expressed genes was 477, 63, and 255 for the JFW_180d vs. JFW_1d, JFW_180d vs. JFW_90d, and JFW_90d vs. JFW_1d comparisons, respectively (Figure 2B). We further constructed a Venn diagram of the identified DE circRNAs in the three pairwise comparisons to find the shared circRNAs (Figure 2C). It revealed that only two circRNAs (novel_circ_0009188 and novel_circ_0018581) were consistently differentially expressed in all comparisons (Figure 2C). One of these, novel_circ_0009188, is generated by circularization of two exons of tubulin tyrosine ligase-like 7 (TTLL7) and is 311 nt in length. The other, novel_circ_0018581, arises through circularization of two exons of glutamate decarboxylase-like protein 1 (GADL1), and has a length of 608 nt. The expression levels of these two circRNAs increased significantly with age in Jinfen White pigs, across the three growth stages.

Validation of DE circRNAs at Different Developmental Stages of Muscle

To validate the reliability of temporal circRNA expression profiles derived from RNA-seq data, divergent primers were designed for four randomly selected DE circRNAs, and the expression levels of that were quantified by qRT-PCR (Figure 3A). The results were consistent with the expression patterns of these circRNAs obtained through RNA-seq data. In addition, the expected size PCR products were obtained by amplification using cDNA as the PCR template. The post-splice sites were verified by Sanger sequencing (Figure 3B). These results suggested that the identified circRNAs in this study are credible.

Functional Enrichment Analysis of DE circRNA Host Genes

Previous studies have shown that circRNAs regulate host gene transcription by competing with linear pre-mRNA splicing[42,43]. To explore the potential functions of the identified DE circRNAs in the development of skeletal muscle, we performed GO and KEGG pathway enrichment analyses of their host genes. Notably, we found that these host genes mainly have known roles in muscle biology, including muscle contraction, muscle organ development, muscle system processes, and muscle structure development (Figure 4A). The host genes CHD2, SMAD3, BMPR1A, HOMER1, DMD, and MYL2 are involved in muscle organ development; HOMER1, DMD, and ATP11A are involved in myotube differentiation; and NFATC1, BMPR1A, and PDCD4 are involved in the negative regulation of muscle cell differentiation.The KEGG pathway enrichment analysis showed that the DE circRNA host genes were enriched in 20 pathways, including skeletal muscle fiber-related signaling pathways, and the cAMP and AMPK signaling pathways et al (Figure 4B). PDE4B, NFATC1, PLCE1, ATP2B4, PPP1R12A, and CAMK2B were enriched in the cAMP signaling pathway, and EEF2K, FBP2, PPP2CB, PFKFB1, and PRKAA2 were enriched in AMPK signal pathway. The circRNAs generated from these host genes might act as promoters or repressors during the development of porcine skeletal muscle, thereby providing candidate genes for future exploration of the regulatory functions of these genes and the circRNAs derived from them.

Construction of a potential circRNA-miRNA-mRNA regulatory network

CircRNAs often act as miRNA molecular sponges to indirectly regulate the expression of miRNA target genes, thereby meaning that they participate in the related biological processes[44]. Therefore, we integrated miRNA and mRNA library data (The sequencing results obtained at the same time, the accession number is consistent with the circRNA) with the results from the present study to constructed a ceRNA network and identified key circRNAs associated with pork quality and regulation of cell proliferation and differentiation. This ceRNA network contained 229 circRNAs, 55 miRNAs and 143 mRNAs (Supplementary Table S2). Because the ceRNA network are too large, we selected the top 20 DE circRNAs with the most obvious up- or down-regulation profiles to present two ceRNA network maps (Figure 5). In ceRNA network, novel_circ_0018595 (hereafter, circ_0018595) is predicted to function as a sponge for ssc-miR-1343, which binds to the mRNAs transcribed from the PGM1, TNNT3, and IFNAR2 genes (Figure 5A).

To explore the functional significance of mRNAs in the circRNA–miRNA–mRNA regulatory network, we performed functional enrichment analysis on 143 targeted mRNAs. As a result the GO enrichment analysis showed that these target mRNAs were significantly enriched in pyruvate metabolism, hexose metabolism, striated muscle contraction, glycolysis process, muscle system process, and muscle contraction. More specifically, PGAM2, ALDOA, ACTN3, PKM, and GPI are involved in the process of glycolysis, and ACTN3, PI16, WDR1 are involved in striated muscle development. It is worth noting that ACTN3 is also involved in the positive regulation of fast-twitch skeletal muscle fiber contraction, skeletal muscle fiber development, skeletal muscle tissue growth, and other important processes (Supplementary Figure S1A). Furthermore, KEGG enrichment analysis showed that the same target mRNAs were mainly enriched in carbon metabolism, glycolysis/gluconeogenesis, and regulation of the actin cytoskeleton. Specifically, PGAM2, ALDOA, GPI, and EFCAB7 are involved in glycolysis and gluconeogenesis, HOMER3 and PDPK1 are involved in the FOXO signaling pathway, and PDPK1, ITGA5, and ITGB4 are associated with the PI3K-AKT signaling pathway (Supplementary Figure S1B). These result suggested that these circRNAs in the ceRNA network many play important role in the muscle development.

Experimental Validation of circ_0018595

From the above network, we selected circ_0018595 for follow-up study, speculating that it might be involved in regulating growth and development of porcine skeletal muscle via the predicted circ_0018595/miR-1343/PGM1 axis (Figure 5A). Circ_0018595is an exonic circRNA generated from exons 3 and 4 of STT3B. It was amplified from longissimus dorsi cDNA using specific divergent primers, and the product was subjected to Sanger sequencing to validate the ligation site (Figure 6A). We also processed the total RNA extract with RNase R and performed qRT-PCR, revealing that circ_0018595 was more resistant to RNase R than STT3B mRNA and 18S rRNA as expected (Figure 6B). The circ_0018595 was expressed in various tissues, especially in lung, liver, skeletal muscle, and kidney (Figure 6C). We also examined the expression of circ_0018595 in muscle of Jinfen White pigs at 1, 90, and 180 days, and found it was significantly associated with developmental stage, with strong upregulation between 1 day and 90 days (P <0.05), and maintained at 180 days (Figure 6D). Furthermore, the results of a nucleocytoplasmic separation assay showed that circ_0018595 is primarily localized in the cytoplasm, which was consistent with its proposed role as a sponge for miRNAs (Figure 6E). Finally, the RNAhybrid website found that circ_0018595 and miR-1343 had potential binding sits, which was also found between miR-1343 and PGM1 mRNA transcript (Figure 6F). The function and mechanism experiment of circ_0018595in the skeletal muscle development will be carried out in the future work.

CircRNAs are abundant in transcriptomes and participate in various biological activities by regulating transcription, splicing, mRNA stability, and translation. At present, many studies have investigated circRNAs as potential therapeutic targets for various diseases[45–47]. With the ongoing development of related research strategies and methods, many biological functions of circRNAs in animals have also been emerging, and circRNAs have been found to regulate the proliferation and differentiation of animal skeletal muscle cells by acting as ceRNAs and encoding small peptides[48]. For example, Shen et al[49], found that circTMTC1 can bind miR-128-3p and regulate expression of myostatin (MSTN), thereby inhibiting the proliferation and differentiation of chicken skeletal muscle satellite cells. Pigs are important farm animals that provide meat for humans and are often used for disease models in medical research. In 2017, Liang and coworkers constructed the first publicly available Sus scrofa circRNA database through analysis of nine organs of Guizhou pigs at three developmental stages[39]. Current research on circRNAs in porcine skeletal muscle mainly relies on the mining and discussion of sequencing data, and the specific mechanisms of circRNA involvement in skeletal muscle growth and fat deposition are rarely studied[50–52]. Here, we used RNA-seq to construct circRNA expression profiles for Jinfen White pig skeletal muscle at three developmental stages (1, 90, and 180 d). Compared with previous studies, our sequencing data are robust, with > 97% of bases meeting the Q20 threshold.

We identified a total of 16,990 expressed circRNAs, which is much higher than the number reported in one recent study [50]. We also analyzed the chromosomal distribution, composition, and length of these circRNAs. We found that most are generated by exon circularization and are less than 1,000 nt in length, with comparatively greater numbers originating from chromosomes 1, 6, and 13. Temporal changes in the relative abundance of circRNAs were also analyzed across the three different age groups. To a large extent, qRT-PCR confirmed the expression patterns derived from RNA-seq data for four examples, further establishing the accuracy of our sequencing results, along with the developmental stage specificity of circRNA expression profiles. These findings are also consistent with previous studies[50–52]. Our results therefore enrich the known porcine circRNA library and confirm that circRNAs are consistent across pig species. In addition, we identified a total of 584 circRNAs that were differentially expressed between the three developmental stages. This may be helpful in identifying circRNAs that regulate skeletal muscle growth and development, and opening up new avenues of future exploration to elucidate the specific mechanisms involved.

Usually, circRNAs are generated by back-splicing of pre-mRNAs. They have notable structural stability and spatiotemporal specificity, and can positively or negatively regulate transcription of their parent genes[53]. We performed enrichment analysis on the host genes of all DE circRNAs identified in this study, and found that these host genes were mainly involved in the AMPK and cAMP signaling pathways. The AMPK signaling pathway is known as a regulator of oxidative metabolism in skeletal muscle, and it promotes the conversion of glycolytic muscle fibers into oxidative fibers[54, 55]. For example, in C2C12 myotubes, AMPK signaling drives expression of skeletal slow-twitch muscle fiber genes[56]. On the other hand, the cAMP signaling pathway affects protein synthesis rate, inhibits autolysosomal proteolysis, promotes skeletal muscle hypertrophy, and is expected to become a new target for the treatment of muscle mass loss during atrophic diseases[57–59].Interestingly, more than one circRNA can arise from a single gene due to alternative circularization and splicing mechanisms. In this study, we found that lim domain only protein 7 (Lmo7) produced 15 circRNAs, but among these, only novel_circ_0016243 and novel_circ_0016253 had high expression levels, suggesting that these two novel circRNAs may play important roles in cell differentiation in Jinfen White pigs. LMO7 is essential for skeletal muscle development and encodes a transcription factor that binds to the PAX3, MyoD, and Myf5 promoters to activate expression of key myogenic differentiation genes in C2C12 cells[60]. Thus, downregulation of LMO7 was shown to inhibit myogenesis in chicken myoblasts[37]. In addition, circLMO7 (formed by circularization of exons 3–5 of bovine LMO7), can target miR-378a-3p to inhibit the differentiation and apoptosis of myoblasts, and the sequence of bovine circLMO7 is completely different from that of porcine novel_circ_0016243 and novel_circ_0016253 in our study[61].

As mentioned above, circRNAs can act as miRNA sponges to regulate the expression of target genes, through the ceRNA mechanism. We constructed a circRNA-miRNA-mRNA co-expression network using high-throughput sequencing of mRNA and miRNA expression network data. For some of the miRNAs in this ceRNA network, a role in regulating proliferation and differentiation of skeletal muscle satellite cells has already been reported in other animals. These include miR-370[62], miR-214[63], miR-133[64], and miR-885[65]. We predicted that circ_0000576 could act as a sponge for ssc-miR-15a, ssc-miR-98, and ssc-miR-885-3p, modulating the target genes of these miRNAs to regulate muscle growth and development. In addition, circ_0004213 is an exon-derived circRNA from the polypyrimidine tract binding protein 1 (PTBP1) gene, which itself encodes a splicing factor involved in regulatory processes related to cell differentiation, cell cycle and apoptosis, cell motility, cell metabolism, and the immune response[66]. We also predicted that circ_0004213 could target miR-370, which in turn targets α-actinin-3 (ACTN3). Involvement of miR-370 has been reported in many cellular functions, including cell proliferation, migration, and differentiation[62], while ACTN3 encodes the sarcomeric α-actinin-3 protein that is a component of the Z line in mammalian skeletal muscle fibers[67].We also identified circ_0018595, which is generated by cyclization of exons 3–4 of STT3B and is predicted to bind miR-1343. This miRNA is involved in cell proliferation and apoptotic processes, as well as affecting the pluripotency of stem cells. In the ovaries of highly fertile sows, miR-1343 suppresses expression of transforming growth factor-β receptor type 1 (TGFBR1) to inhibit pig granule cell proliferation and promote apoptosis, thus affecting sow fecundity[68]. Furthermore, miR-1343 specifically binds to the 3´UTR of Orthodenticle homeobox 2 (OTX2) and inhibits endogenous OTX2 expression in pig pluripotent stem cells (piPSCs), enhancing the expression of pluripotency genes and thus maintaining piPSCs pluripotency in pigs[69].Finally, we chose to focus specifically on PGM1, which has the highest expression level in porcine longissimus dorsi and biceps femoris. It is involved in the glycolytic pathway, which may be related to pork quality and productivity[70, 71], since reducing the proportion of glycolytic fibers in porcine longissimus dorsi can slow down the rate and extent of muscle pH decline after slaughter[72]. Thus, PGM1 may be involved in regulating skeletal muscle growth and lipid deposition. Given these considerations, the circ_0018595/miR-1343/PGM1 axis may play a key role in the growth and development of skeletal muscle in Jinfen White pigs.

We also have some limitations of this study. All circRNAs and their target miRNAs or mRNAs were predicted computationally, and the specific mechanisms of these circRNAs in regulating skeletal muscle still require experimental verification. A series of complex potential mechanisms, such as circRNAs interacting with proteins, regulating parental gene transcription, and participating in protein coding, which remain to be explored.

We have used RNA-seq to construct circRNA expression profiles of skeletal muscle in Jinfen White pigs at 1 d, 90 d, and 180 d of age. Of the 16,990 circRNA candidates identified, 584 were characterized as DE circRNAs, which may be particularly beneficial to future research on circRNA regulatory mechanisms that control skeletal muscle growth and development. In addition, GO and KEGG pathway enrichment analyses revealed that the host genes of these DE circRNAs are mainly involved in skeletal muscle fiber-related signaling pathways (such as the AMPK and cAMP pathways). We also constructed a circRNA-miRNA-mRNA co-expression network, providing possible mechanistic insights into regulation of skeletal muscle fiber development. A novel axis within this network, circ_0018595/miR-1343/PGM1, may particularly be involved in the regulation of skeletal muscle growth. In conclusion, we have identified a series of novel candidate circRNAs with potential roles in regulation of growth and development of porcine skeletal muscle. This study provides an important reference point for future investigations to improve our understanding of circRNA regulation in porcine skeletal muscle growth and development.

Experimental Animals and Samples

All experimental animals and procedures in this experiment were approved by the Institutional Animal Care and Use Committee of Shanxi Agricultural University (Shanxi, China). The 9 Jinfen White pigs used in the experiment came from Datong pig farm (Shanxi, China) and were raised under standard conditions without restriction on feeding and drinking. Three heads per group (denoted as samples JFW_1d_1, JFW_1d_2, JFW_1d_3; JFW_90d_1, JFW_90d_2, JFW_90d_3; JFW_180d_1, JFW_180d_2, JFW_180d_3) were sacrificed by corona and neck exsanguination on days 1, 90, and 180, respectively. Tissues such as heart, liver, spleen, lung, kidney, pancreas, longissimus dorsi and subcutaneous fat were quickly collected and immediately snap-frozen in liquid nitrogen, and stored at -80 °C for later use.

Library Preparation and RNA Sequencing

Total RNA was isolated and purified using the Trizol reagent (Takara, Japan), following the manufacturer’s instructions. The concentration and purity of isolated RNA samples were tested by electrophoresis and further examined with a nucleic acid analysis system (Experion System, Bio-Rad, USA). The A₂₆₀/A₂₈₀ ratio was between 1.9 and 2.1, indicating that the extracted RNA samples were of good quality. Sequencing was done by Beijing Novogene, using the Illumina HiSeq 2500 sequencing platform for two-end (pair-end), high-throughput sequencing (2 × 150 bp), and the amount of sequencing data (clean reads) per library was about 10 Gb.

Identification of differentially expressed circRNAs

To obtain high quality clean reads for subsequent bioinformatic analysis, raw data were filtered to remove low quality reads, adapter contamination, high unknown base content, and all rRNA-mapping reads. The Q20 and Q30 scores of the clean reads were calculated simultaneously, and circRNAs were then identified by aligning clean reads to the pig reference genome (Sscrofa 11.1) using find_circ and CIRI. The true expression levels of junction reads and circRNAs extracted from the sequencing data were affected by other factors, such as sequencing depth. Therefore, in order to accurately estimate true circRNA expression levels, the readcount data for predicted circRNAs were corrected using the TPM method. The filtered sequences were compared with the pig reference genome, the genomic features of each identified circRNA were analyzed, then the wider genomic and sequence characteristics of circRNAs in Jinfen White pigs were analyzed; mainly including the distribution among chromosomes, genomic location, and exon number of circRNAs.

Gene Ontology (GO) and Pathway Enrichment Analysis (KEGG)

Using GO analysis (http://www.geneontology.org/) and KEGG pathway enrichment analysis (https://www.kegg.jp/), the host genes of DE circRNAs at the three growth and developmental stages were categorized in order to understand their biological functions. In the GO analysis, using the pig reference genome as background, the host genes were mapped to different functional entries. Thereafter, the enrichment significance of each entry was expressed as the corrected q-value, where entries with P <0.05 were considered to be significantly enriched. In the KEGG pathway enrichment analysis, the hypergeometric test was applied—again using the pig reference genome as the analysis background—to screen the significantly enriched pathways (P <0.05) for the host genes, with the results plotted as a bubble chart.

Target miRNA and Gene Prediction, and Network Analysis

To reveal functional interactions between ncRNAs and mRNAs, we used circRNA–miRNA interaction network analysis to construct a competing endogenous RNA network. Here, the public project TargetScan 7.0 (http://www.microrna.org/microrna/home.do) and miRanda software were used to predict potential regulatory relationships between circRNA, mRNA, and miRNA. Cytoscape (http://cytoscape.org/) was used for network visualization.

Quantitative Reverse Transcription Polymerase Chain Reaction (qRT-PCR)

Total RNA was extracted from the nine muscle samples used in the RNA-seq experiments, then reverse transcribed using PrimeScript ® RT Master Mix (Takara, Japan) with divergent primers designed using Premier 3.0 (including qRT-PCR primers and PCR primer-specific alignment by NCBI Primer-BLAST). Details of primers used in this study are listed in Supplementary Table S3. The expression of each circRNA, as well as other parameters, was detected using a CFX96 quantitative PCR instrument (Bio-Rad). The relative expression of target circRNAs was calculated using the 2^−ΔΔCt method.

Circularization Site Verification of circRNA

Four selected circRNAs were amplified by PCR using the designed qRT-PCR primers and the longissimus dorsi cDNA as template. Then, 2% agarose gel electrophoresis and Sanger sequencing of the PCR products were used to confirm reverse splicing of each circRNA by identifying the junction sequence.

RNase R Enzyme Digestion Test

To test the circular nature of selected circRNAs, total RNA from porcine longissimus dorsi tissue was treated with RNase R enzyme (Epicentre) followed by qRT-PCR analysis.

Statistical Analysis

Data processing and production of graphics were carried out using SPSS and GraphPad Prism 6 software. To test for statistically significant differences between two averages, data were analyzed by Student’s t test, and P <0.05 was considered to indicate statistical significance.

CircRNAs: circular RNAs; DE: differentially expressed; GO: Gene Ontology; KEGG: Kyoto Encyclopedia of Genes and Genomes; ceRNA: competitive endogenous RNA; ncRNAs: non-coding RNAs; miRNAs：microRNAs; lncRNAs: long non-coding RNAs; pre-mRNA: precursor mRNA; piPSCs: pig pluripotent stem cells; MyoD1: myoblast determination protein 1; Myf5: myogenic factor 5; MyoG: myogenin; Myf6: myogenic factor 6; PAX3: paired box 3; PAX7: paired box 7; MEF2: myocyte enhancer factor 2; TTLL7: tubulin tyrosine ligase like 7; GADL1: glutamate decarboxylase-like protein 1; MSTN: myostatin; Lmo7: LIM domain only protein 7; PTBP1: polypyrimidine tract binding protein 1; ACTN3: α-actinin-3; TGFBR1: transforming growth factor-β receptor type 1; OTX2: Orthodenticle homeobox 2; PGM1: phosphoglucomutase 1 .

Acknowledgements

We thank Novogene for providing helpful technical support in bioinformatics analysis. We also thank all the staff of the Datong Pig Farm.

Authors’ contributions

JLY did the draft writing. MYS, WXL, XYH, CL, JN conducted experiments. CBC, YY contributed data analysis. BGL, XHG, PFG supervise and coordinate experiments. CL and GQC designed the study and revised the manuscript. All authors read and approved the final manuscript.

Funding

This research was jointly supported by the National Natural Science Foundation of China (31872336), Special Funds for Scholars Support Program of Shanxi Province (2016; 2017), and the Basic Research Project of Shanxi Province (20210302124639).

Availability of data and materials

The circRNA, miRNA and mRNA raw sequence data reported in this paper have been uploaded to NCBI and are publicly accessible at PRJNA867525.

Ethics approval and consent to participate

The nine pigs used in this study were all from the Datong Pig Breeding Farm (Shanxi, China). We obtained the written consent of the director of Datong pig farm and provided a proof. All methods are reported in accordance with ARRIVE guidelines (https://arriveguidelines.org) for the reporting of animal experiments, and all experimental procedures were carried out in accordance with the guidelines of the Regional Animal Ethics Committee and approved by the Animal Ethics Committee of Shanxi Agricultural University (Approval No. SXAU-EAW-2021Sus.AB.0926001).

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Horak M, Novak J, Bienertova-Vasku J. Muscle-specific microRNAs in skeletal muscle development [J]. Developmental Biology, 2016, 410(1): 1-13.
Yin H, Price F, Rudnicki M. Satellite cells and the muscle stem cell niche [J]. Physiological Reviews, 2013, 93(1): 23-67.
Kuang S, Kuroda K, Le Grand F, et al. Asymmetric self-renewal and commitment of satellite stem cells in muscle [J]. Cell, 2007, 129(5): 999-1010.
Verma N, Rettenmeier A, Schmitz-Spanke S. Recent advances in the use of Sus scrofa (pig) as a model system for proteomic studies [J]. Proteomics, 2011, 11(4): 776-793.
Sanger H, Klotz G, Riesner D, et al. Viroids are single-stranded covalently closed circular RNA molecules existing as highly base-paired rod-like structures [J]. Proceedings of the National Academy of Sciences of the United States of America, 1976, 73(11): 3852-3856.
Brand-Saberi B. Genetic and epigenetic control of skeletal muscle development [J]. Annals of Anatomy, 2005, 187(3): 199-207.
Buckingham M, Relaix F. The role of Pax genes in the development of tissues and organs: Pax3 and Pax7 regulate muscle progenitor cell functions [J]. Annual Review of Cell and Developmental Biology, 2007, 23: 645-673.
Sabourin L, Rudnicki M. The molecular regulation of myogenesis [J]. Clinical Genetics, 2000, 57(1): 16-25.
Bi P, Ramirez-Martinez A, Li H, et al. Control of muscle formation by the fusogenic micropeptide myomixer [J]. Science, 2017, 356(6335): 323-327.
Ito Y, Kayama T, Asahara H. A systems approach and skeletal myogenesis [J]. Comparative and Functional Genomics, 2012, 2012: 759407.
Rudnicki M, Schnegelsberg P, Stead R, et al. MyoD or Myf-5 is required for the formation of skeletal muscle [J]. Cell, 1993, 75(7): 1351-1359.
Kassar-Duchossoy L, Gayraud-Morel B, Gomès D, et al. Mrf4 determines skeletal muscle identity in Myf5:Myod double-mutant mice [J]. Nature, 2004, 431(7007): 466-471.
Hasty P, Bradley A, Morris J, et al. Muscle deficiency and neonatal death in mice with a targeted mutation in the myogenin gene [J]. Nature, 1993, 364(6437): 501-506.
Myer A, Olson E, Klein W. MyoD cannot compensate for the absence of myogenin during skeletal muscle differentiation in murine embryonic stem cells [J]. Developmental Biology, 2001, 229(2): 340-350.
Li J, Zhao W, Li Q, et al. Long Non-Coding RNA H19 promotes porcine satellite cell differentiation by interacting with TDP43 [J]. Genes, 2020, 11(3): 259-271.
Geng T, Liu Y, Xu Y, et al. H19 lncRNA promotes skeletal muscle insulin sensitivity in part by targeting AMPK [J]. Diabetes, 2018, 67(11): 2183-2198.
Raza S, Kaster N, Khan R, et al. The role of micrornas in muscle tissue development in beef cattle [J]. Genes, 2020, 11(3): 295-308.
Vicens Q, Westhof E. Biogenesis of circular RNAs [J]. Cell, 2014, 159(1): 13-14.
Yuan X, Diao J, Du A, et al. Circular RNA expression profiles and features in NAFLD mice: a study using RNA-seq data [J]. Journal of Translational Medicine, 2020, 18(1): 476-488.
Hao Z, Zhou H, Hickford J, et al. Identification and characterization of circular RNA in lactating mammary glands from two breeds of sheep with different milk production profiles using RNA-Seq [J]. Genomics, 2020, 112(3): 2186-2193.
Li M, Xie X, Zhou J, et al. Quantifying circular RNA expression from RNA-seq data using model-based framework [J]. Bioinformatics, 2017, 33(14): 2131-2139.
Rybak-Wolf A, Stottmeister C, Glažar P, et al. Circular RNAs in the mammalian brain are highly abundant, conserved, and dynamically expressed [J]. Molecular Cell, 2015, 58(5): 870-885.
Li X, Yang L, Chen L. The Biogenesis, functions, and challenges of circular rnas [J]. Molecular Cell, 2018, 71(3): 428-442.
Memczak S, Jens M, Elefsinioti A, et al. Circular RNAs are a large class of animal RNAs with regulatory potency [J]. Nature, 2013, 495(7441): 333-338.
Tao M, Zheng M, Xu Y, et al. CircRNAs and their regulatory roles in cancers [J]. Molecular Medicine, 2021, 27(1): 94-116.
Shi X, Yang J, Liu M, et al. Circular RNA ANAPC7 inhibits tumor growth and muscle wasting via PHLPP2-AKT-TGF-β signaling axis in pancreatic cancer [J]. Gastroenterology, 2022, 162: 2004-2017.
Wang Y, Yan Q, Mo Y, et al. Splicing factor derived circular RNA circCAMSAP1 accelerates nasopharyngeal carcinoma tumorigenesis via a SERPINH1/c-Myc positive feedback loop [J]. Molecular Cancer, 2022, 21(1): 62-82.
Fan H, Chen Z, Chen X, et al. METTL14-mediated mA modification of circORC5 suppresses gastric cancer progression by regulating miR-30c-2-3p/AKT1S1 axis [J]. Molecular Cancer, 2022, 21(1): 51-65.
Ouyang H, Chen X, Wang Z, et al. Circular RNAs are abundant and dynamically expressed during embryonic muscle development in chickens [J]. DNA Research, 2018, 25(1): 71-86.
Yin H, Shen X, Zhao J, et al. Circular RNA CircFAM188B encodes a protein that regulates proliferation and differentiation of chicken skeletal muscle satellite cells [J]. Frontiers in Cell and Developmental Biology, 2020, 8: 522588.
Chen X, Ouyang H, Wang Z, et al. A Novel Circular RNA generated by FGFR2 gene promotes myoblast proliferation and differentiation by sponging miR-133a-5p and miR-29b-1-5p [J]. Cells, 2018, 7(11): 7110199
Hitachi K, Tsuchida K. Role of microRNAs in skeletal muscle hypertrophy [J]. Frontiers in Physiology, 2013, 4: 408-415.
Wei W, He H, Zhang W, et al. MiR-29 targets Akt3 to reduce proliferation and facilitate differentiation of myoblasts in skeletal muscle development [J]. Cell Death & Disease, 2013, 4: 668-678.
Li H, Yang J, Wei X, et al. CircFUT10 reduces proliferation and facilitates differentiation of myoblasts by sponging miR-133a [J]. Journal of Cellular Physiology, 2018, 233(6): 4643-4651.
Yan J, Yang Y, Fan X, et al. circRNAome profiling reveals circFgfr2 regulates myogenesis and muscle regeneration via a feedback loop [J]. Journal of Cachexia, Sarcopenia and Muscle, 2022, 13(1): 696-712.
Cai B, Ma M, Zhou Z, et al. CircPTPN4 regulates myogenesis via the miR-499-3p/NAMPT axis [J]. Journal of Animal Science and Biotechnology, 2022, 13(1): 2-17.
Shen X, Wei Y, Liu W, et al. A novel circular RNA circITSN2 Targets the miR-218-5p/LMO7 Axis to Promote Chicken Embryonic Myoblast Proliferation and Differentiation [J]. Frontiers in Cell and Developmental Biology, 2021, 9: 748844.
Legnini I, Di Timoteo G, Rossi F, et al. Circ-ZNF609 is a circular rna that can be translated and functions in myogenesis [J]. Molecular Cell, 2017, 66(1): 22-37.
Liang G, Yang Y, Niu G, et al. Genome-wide profiling of Sus scrofa circular RNAs across nine organs and three developmental stages [J]. DNA Research, 2017, 24(5): 523-535.
Kristensen L, Andersen M, Stagsted L, et al. The biogenesis, biology and characterization of circular RNAs [J]. Nature Reviews Genetics, 2019, 20(11): 675-691.
Das A, Shyamal S, Sinha T, et al. Identification of potential circRNA-microRNA-mRNA Regulatory Network in Skeletal Muscle [J]. Front Mol Biosci, 2021, 8: 762185.
Zhang Y, Zhang X, Chen T, et al. Circular intronic long noncoding RNAs [J]. Molecular Cell, 2013, 51(6): 792-806.
Li Z, Huang C, Bao C, et al. Exon-intron circular RNAs regulate transcription in the nucleus [J]. Nature Structural & Molecular Biology, 2015, 22(3): 256-264.
Chen B, Yu J, Guo L, et al. Circular RNA circHIPK3 promotes the proliferation and differentiation of chicken myoblast cells by sponging miR-30a-3p [J]. Cells, 2019, 8(2):177-186.
Cheng X, Ai K, Yi L, et al. The mmu_circRNA_37492/hsa_circ_0012138 function as potential ceRNA to attenuate obstructive renal fibrosis [J]. Cell Death & Disease, 2022, 13(3): 207-218.
Deng W, Zhou X, Zhu K, et al. Novel circular RNA circ_0086722 drives tumor progression by regulating the miR-339-5p/STAT5A axis in prostate cancer [J]. Cancer Letters, 2022, 533: 215606.
Wu X, Zhou J, Zhao L, et al. CircCYP24A1 hampered malignant phenotype of renal cancer carcinoma through modulating CMTM-4 expression via sponging miR-421 [J]. Cell Death & Disease, 2022, 13(2): 190-291.
Li H, Wei X, Yang J, et al. circFGFR4 promotes differentiation of myoblasts via binding miR-107 to relieve its inhibition of Wnt3a [J]. Molecular Therapy Nucleic Acids, 2018, 11: 272-283.
Shen X, Liu Z, Cao X, et al. Circular RNA profiling identified an abundant circular RNA circTMTC1 that inhibits chicken skeletal muscle satellite cell differentiation by sponging miR-128-3p [J]. International journal of biological sciences, 2019, 15(10): 2265-2281.
Li B, Yang J, He J, et al. Spatiotemporal regulation and functional analysis of circular RNAs in skeletal muscle and subcutaneous fat during pig growth [J]. Biology, 2021, 10(9): 841-862.
Li M, Zhang N, Zhang W, et al. Comprehensive analysis of differentially expressed circRNAs and ceRNA regulatory network in porcine skeletal muscle [J]. BMC Genomics, 2021, 22(1): 320-332.
Qi K, Liu Y, Li C, et al. Construction of circRNA-related ceRNA networks in longissimus dorsi muscle of Queshan Black and Large White pigs [J]. Molecular Genetics and Genomics, 2022, 297(1): 101-112.
Barrett S P, Salzman J. Circular RNAs: analysis, expression and potential functions [J]. Development, 2016, 143(11): 1838-1847.
Witczak C, Sharoff C, Goodyear L. AMP-activated protein kinase in skeletal muscle: from structure and localization to its role as a master regulator of cellular metabolism [J]. Cellular and Molecular Life Sciences, 2008, 65(23): 3737-3755.
Lee Y, Yun M, Kim H, et al. Exogenous administration of DLK1 ameliorates hepatic steatosis and regulates gluconeogenesis via activation of AMPK [J]. International Journal of Obesity, 2016, 40(2): 356-365.
Xu M, Chen X, Huang Z, et al. Procyanidin B2 promotes skeletal slow-twitch myofiber gene expression through the ampk signaling pathway in C2C12 myotubes [J]. Journal of Agricultural and Food Chemistry, 2020, 68(5): 1306-1314.
Lauterbach N, Gonçalves D, Silveira W, et al. Urocortin 2 promotes hypertrophy and enhances skeletal muscle function through cAMP and insulin/IGF-1 signaling pathways [J]. Molecular Metabolism, 2022, 60: 101492.
Machado J, Manfredi L, Silveira W, et al. Calcitonin gene-related peptide inhibits autophagic-lysosomal proteolysis through cAMP/PKA signaling in rat skeletal muscles [J]. The international journal of biochemistry & cell biology, 2016, 72: 40-50.
Tong T, Kim M, Park T. α-Ionone attenuates high-fat diet-induced skeletal muscle wasting in mice via activation of cAMP signaling [J]. Food & Function, 2019, 10(2): 1167-1178.
Dedeic Z, Cetera M, Cohen T, et al. Emerin inhibits Lmo7 binding to the Pax3 and MyoD promoters and expression of myoblast proliferation genes [J]. Journal of Cell Science, 2011, 124: 1691-1702.
Wei X, Li H, Yang J, et al. Circular RNA profiling reveals an abundant circLMO7 that regulates myoblasts differentiation and survival by sponging miR-378a-3p [J]. Cell Death & Disease, 2017, 8(10): e3153.
Zhang P, Du J, Guo X, et al. LncMyoD promotes skeletal myogenesis and regulates skeletal muscle fiber-type composition by sponging miR-370-3p [J]. Genes, 2021, 12(4): 589-595.
Duan Y, Li T, Zhang G, et al. Transcriptome sequencing to explore the effect of miR-214 on chicken primary myoblasts [J]. Animal Biotechnology, 2022, 1-10.
Feng Y, Niu L, Wei W, et al. A feedback circuit between miR-133 and the ERK1/2 pathway involving an exquisite mechanism for regulating myoblast proliferation and differentiation [J]. Cell Death & Disease, 2013, 4: e934.
Elsaeid Elnour I, Dong D, Wang X, et al. Bta-miR-885 promotes proliferation and inhibits differentiation of myoblasts by targeting MyoD1 [J]. Journal of Cellular Physiology, 2020, 235(10): 6625-6636.
Zhu W, Zhou B, Rong L, et al. Roles of PTBP1 in alternative splicing, glycolysis, and oncogenesis [J]. Journal of Zhejiang University Science B, 2020, 21(2): 122-136.
Harada N, Gotoda Y, Hatakeyama A, et al. Differential regulation of Actn2 and Actn3 expression during unfolded protein response in C2C12 myotubes [J]. Journal of Muscle Research and Cell Motility, 2020, 41:199-209.
Hu H, Fu Y, Zhou B, et al. [J]. Molecular and Cellular Endocrinology, 2021, 520: 111064.
Xie Y, Cao H, Zhang Z, et al. Molecular network of miR-1343 regulates the pluripotency of porcine pluripotent stem cells via repressing OTX2 expression [J]. RNA Biology, 2019, 16(1): 82-92.
Yu J, Shao S, Xiong Y. Molecular characterization, expression patterns, and promoter activity analysis of PGM1 in pigs [J]. Genetics and Molecular Research, 2015, 14(1): 2816-2824.
Xing K, Wang K, Ao H, et al. Comparative adipose transcriptome analysis digs out genes related to fat deposition in two pig breeds [J]. Scientific Reports, 2019, 9(1): 12925.
Kim G, Jeong J, Yang H, et al. Differential abundance of proteome associated with intramuscular variation of meat quality in porcine longissimus thoracis et lumborum muscle [J]. Meat Science, 2019, 149: 85-95.

No competing interests reported.

Download PDF

Journal Publication

published 12 Apr, 2023

Read the published version in BMC Genomics →

Editorial decision: Major revision
13 Feb, 2023
Reviews received at journal
08 Feb, 2023
Reviewers agreed at journal
08 Feb, 2023
Reviewers agreed at journal
08 Feb, 2023
Reviewers invited by journal
20 Oct, 2022
Editor assigned by journal
20 Oct, 2022
Editor invited by journal
10 Oct, 2022
Submission checks completed at journal
10 Oct, 2022
First submitted to journal
23 Sep, 2022

You are reading this latest preprint version

Genome-Wide Analysis of Circular RNA-Mediated ceRNA Regulation in Porcine Skeletal Muscle Development

Status:

Journal Publication

Version 1

Abstract

Figures

Background

Results

Discussion

Conclusion

Methods

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1