CircPLXNA2 Affects the Proliferation and Apoptosis of Myoblast through circPLXNA2/gga-miR-12207-5P/MDM4 Axis

CircRNAs are newly identified special endogenous RNA molecules that covalently close a loop by back-splicing with pre-mRNA. In the cytoplasm, circRNAs would act as molecular sponges to bind with specific miRNA to promote the expression of target genes. However, knowledge of circRNA functional alternation in skeletal myogenesis is still in its infancy. In this study, we identified a circRNA–miRNA–mRNA interaction network in which the axis may be implicated in the progression of chicken primary myoblasts’ (CPMs) myogenesis by multi-omics (i.e., circRNA-seq and ribo-seq). In total, 314 circRNA–miRNA–mRNA regulatory axes containing 66 circRNAs, 70 miRNAs, and 24 mRNAs that may be relevant to myogenesis were collected. With these, the circPLXNA2-gga-miR-12207-5P-MDM4 axis aroused our research interest. The circPLXNA2 is highly differentially expressed during differentiation versus proliferation. It was demonstrated that circPLXNA2 inhibited the process of apoptosis while at the same time stimulating cell proliferation. Furthermore, we demonstrated that circPLXNA2 could inhibit the repression of gga-miR-12207-5p to MDM4 by directing binding to gga-miR-12207-5p, thereby restoring MDM4 expression. In conclusion, circPLXNA2 could function as a competing endogenous RNA (ceRNA) to recover the function of MDM4 by directing binding to gga-miR-12207-5p, thereby regulating the myogenesis.


Introduction
Chicken is an economically significant domestic animal that plays an indispensable role in human meat consumption as the most consumed food source. China is the largest country for commercial broil farming in the world, with an annual broil slaughter number of 11.83 billion, and the second ranking-country for chicken production, having produced 1878.2 tons in 2021 (http://www.stats.gov.cn/) (accessed on 1 March 2023). According to the USDA, world broiler production will be 10,108.6 million tons in 2022, with a growth rate of 0.57%. The top four ranking countries and regions' broiler production, i.e., the US, China, Brazil, and the EU are 21.5 million tons, 14.3 million tons, 14.25 million tons, and 10.92 million tons, respectively (https://apps.fas.usda.gov/psdonline/app/index.html# /app/downloads, accessed on 1 March 2023) Due to its economic value, the muscle growth and development of chicken are of great importance and have become the industry's primary focus. As an essential component of the body, skeletal muscle plays a crucial role in maintaining movement, regulating metabolism, etc. Importantly, the growth and

The circRNA-miRNA-mRNA Axes Network Is Constructed by Multi-Omics
We used circRNA-seq and ribo-seq on chicken primary myoblasts (CPMs) to identify candidate circRNA involved in muscle growth regulation by examining differentially expressed circRNA and regulatory genes. First, circRNA-seq identified 14,241 circRNAs distributed among different chromosomes, with a centralized length span of 0-1000 bp, in which 360 were significantly differentially expressed (Figures 1A-C and S1-S3). After filtration, 200 circRNAs were identified as up-regulating, and 160 circRNAs were identified as down-regulating in differentiation medium (DM) relative to the growth medium (GM) of CPMs ( Figure 1B,C, Table S1). We mapped differentially expressed circRNAs to their parental genes and performed GO and KEGG enrichment analysis. GO enrichment analysis suggested a large variety of functional gene enrichment in the process of cellular development and differentiation, and KEGG biological pathway identification indicated that the genes were involved in certain signal pathways, such as MAPK, Wnt, and au-tophagy ( Figure 1D,E). In addition, a total of 3342 differentially expressed genes (DEGs) were identified by ribo-seq profile, in which 2185 DEGs were up-regulated and 1157 DEGs were down-regulated (Figure 2A,B, Figure S4). GO enrichment analysis enriched the most genes into biological processes, such as biological regulation, signaling, and cell proliferation. KEGG signaling pathway analysis indicated that most signaling pathways associated with muscle development were regulated, such as the p53, MAPK, mTOR, and JAK-STAT pathways ( Figure 2C,D). dium (GM) of CPMs ( Figure 1B,C, Table S1). We mapped differentially expressed circ NAs to their parental genes and performed GO and KEGG enrichment analysis. GO e richment analysis suggested a large variety of functional gene enrichment in the proc of cellular development and differentiation, and KEGG biological pathway identificati indicated that the genes were involved in certain signal pathways, such as MAPK, W and autophagy ( Figure 1D,E). In addition, a total of 3342 differentially expressed gen (DEGs) were identified by ribo-seq profile, in which 2185 DEGs were up-regulated a 1157 DEGs were down-regulated (Figure 2A,B, Figure S4). GO enrichment analysis e riched the most genes into biological processes, such as biological regulation, signalin and cell proliferation. KEGG signaling pathway analysis indicated that most signali pathways associated with muscle development were regulated, such as the p53, MAP mTOR, and JAK-STAT pathways ( Figure 2C,D). Combined with the newly identified circRNA and mRNA with differentiated abu dance, we characterized a circRNA-miRNA-mRNA interaction network that may be volved in regulating the myogenesis process. To be specific, 314 circRNA-miRNA-mRN regulatory networks with 66 circRNAs, 70 miRNAs, and 24 mRNAs which may be relat to myogenesis were acquired as candidate regulation network for further study (Figu 2E, Table S2). Combined with the newly identified circRNA and mRNA with differentiated abundance, we characterized a circRNA-miRNA-mRNA interaction network that may be involved in regulating the myogenesis process. To be specific, 314 circRNA-miRNA-mRNA regulatory networks with 66 circRNAs, 70 miRNAs, and 24 mRNAs which may be related to myogenesis were acquired as candidate regulation network for further study ( Figure 2E, Table S2).

CircPLXNA2 Is a Novel-Identified circRNA Regulated by PLXNA2
The circPLXNA2-gga-miR-12207-5P-MDM4 axis was the one that piqued our interest for research purposes among these interaction axes. The circPLXNA2, novel_circ_010465, was termed on the basis of its transcriptional precursor (i.e., pre-PLXNA2-mRNA). Sanger sequencing was initially used to validate the circPLXNA2 back-splicing junction ( Figure 3A,D). We then validated the circular structure of circPLXNA2. We designed specific divergent and convergent primers and used genomic DNA (gDNA) and cDNA as templates, respectively, for a PCR reaction to verify the genomic structure. The expected double bands were inclusively shown in the cDNA group visualized by agarose gel electrophoresis, while when we selected gDNA as an amplifying template, the expected band was merely shown in the convergent primers' lane with weak detection ( Figure 3B). In addition, we employed RNase R treatment to detect the resistance of circPLXNA2, which showed successful amplification of circular transcripts by divergent primers using the RNase R-treated cDNA as a template, whereas it failed to be amplified by using convergent primers. In contrast, targeted bands were successfully amplified by both primers using the cDNA template without RNase R digestion ( Figure 3C). We also quantified the RNA expression of the circPLXNA2 after RNase R treatment, which showed no significant effect on the expression level of circPLXNA2 after digestion, whereas the expression of β-actin was significantly decreased after digestion ( Figure 3E). Together, both findings supported circPLXNA2 s actual existence as a circular structure. Nuclear and cytoplasmic localization experiments of circPLXNA2 were performed to further identify its distribution, and the results showed that circPLXNA2 was present in both the nucleus and the cytoplasm, with its primary localization being in the latter ( Figure 3F). The circRNA-miRNA-mRNA regulatory axes network associated with muscle development, with blue circles representing circRNAs, red triangles representing miRNAs, and green rectangles representing mRNAs.

CircPLXNA2 Is a Novel-Identified circRNA Regulated by PLXNA2
The circPLXNA2-gga-miR-12207-5P-MDM4 axis was the one that piqued our interest for research purposes among these interaction axes. The circPLXNA2, novel_circ_010465, was termed on the basis of its transcriptional precursor (i.e., pre-PLXNA2-mRNA). Sanger sequencing was initially used to validate the circPLXNA2 back-splicing junction ( Figure  3A,D). We then validated the circular structure of circPLXNA2. We designed specific divergent and convergent primers and used genomic DNA (gDNA) and cDNA as templates, respectively, for a PCR reaction to verify the genomic structure. The expected double bands were inclusively shown in the cDNA group visualized by agarose gel electrophoresis, while when we selected gDNA as an amplifying template, the expected band was merely shown in the convergent primers' lane with weak detection ( Figure 3B). In addition, we employed RNase R treatment to detect the resistance of circPLXNA2, which showed successful amplification of circular transcripts by divergent primers using the RNase R-treated cDNA as a template, whereas it failed to be amplified by using conver-

CircPLXNA2 Promotes Cell Proliferation and Inhibits Cell Apoptosis
To identify the potential biological function of circPLXNA2 in myogenesis, we constructed the overexpression vector of circPLXNA2. As shown in Figure 4A, the relative mRNA expression level of circPLXNA2 was drastically increased after transfecting PCD25 circPLXNA2 ( Figure 4A). Overexpression of circPLXNA2 in chicken primary myoblasts (CPMs) increased the relative expression of proliferation biomarker genes (cyclin D1, cyclin D2, and cyclin B2), suggesting that it may contribute to proliferation ( Figure 4B). Furthermore, we detected the relative expression level of apoptosis-related genes, which revealed that overexpression of circPLXNA2 suppressed the relative expression of caspase 3, caspase 8, and caspase 9 at the mRNA level, indicating detectable apoptosis suppression by circPLXNA2 ( Figure 4C,F). In addition, more 5-ethynyl-2'-deoxyuridine (EdU)-stained cells were found in the circPLXNA2 overexpression group than in the control group (PDC25 negative Control, PDC25 NC) ( Figure 4D,E). These findings together suggested that circPLXNA2 could control the development of CPMs by regulating cell proliferation and apoptosis. As we predicted an interaction network involving miRNA (i.e., gga-miR-12207-5P), we hypothesized that circPLXNA2 could serve as an endogenous sponge to competitively interact with gga-miR-12207-5P, thereby altering the distribution of gga-miR-12207-5P against their target. The potential binding site between circPLXNA2 and gga-miR-12207-5P is predicted by the RNAhybrid online tool ( Figure 4G) (https://bibiserv.cebitec.uni-bielefeld.de/rnahybrid) (accessed on 1 November 2022). We used the dual-luciferase reporter assay and the pulldown assay to characterize their potential interaction. Compared with the gga-miR-12207-5P mimic and Pmir-GLO-circPLXNA2-MUT co-transfection group, gga-miR-12207-5P mimic co-transfected with the Pmir-GLO-circPLXNA2-WT vector significantly repressed the relative luciferase activity ( Figure 4H). Gel electrophoresis and biotin-labelled miRNA pulldown analysis showed that biotingga-miR-12207-5P mimic transfection significantly enriched circPLXNA2 mRNA levels in

CircPLXNA2 Promotes Cell Proliferation and Inhibits Cell Apoptosis
To identify the potential biological function of circPLXNA2 in myogenesis, we constructed the overexpression vector of circPLXNA2. As shown in Figure 4A, the relative mRNA expression level of circPLXNA2 was drastically increased after transfecting PCD25 circPLXNA2 ( Figure 4A). Overexpression of circPLXNA2 in chicken primary myoblasts (CPMs) increased the relative expression of proliferation biomarker genes (cyclin D1, cyclin D2, and cyclin B2), suggesting that it may contribute to proliferation ( Figure 4B). Furthermore, we detected the relative expression level of apoptosis-related genes, which revealed that overexpression of circPLXNA2 suppressed the relative expression of caspase 3, caspase 8, and caspase 9 at the mRNA level, indicating detectable apoptosis suppression by circ-PLXNA2 ( Figure 4C,F). In addition, more 5-ethynyl-2'-deoxyuridine (EdU)-stained cells were found in the circPLXNA2 overexpression group than in the control group (PDC25 negative Control, PDC25 NC) ( Figure 4D,E). These findings together suggested that circPLXNA2 could control the development of CPMs by regulating cell proliferation and apoptosis. As we predicted an interaction network involving miRNA (i.e., gga-miR-12207-5P), we hypothesized that circPLXNA2 could serve as an endogenous sponge to competitively interact with gga-miR-12207-5P, thereby altering the distribution of gga-miR-12207-5P against their target. The potential binding site between circPLXNA2 and gga-miR-12207-5P is predicted by the RNAhybrid online tool ( Figure 4G) (https://bibiserv.cebitec.uni-bielefeld.de/rnahybrid) (accessed on 1 November 2022). We used the dual-luciferase reporter assay and the pulldown assay to characterize their potential interaction. Compared with the gga-miR-12207-5P mimic and Pmir-GLO-circPLXNA2-MUT co-transfection group, gga-miR-12207-5P mimic co-transfected with the Pmir-GLO-circPLXNA2-WT vector significantly repressed the relative luciferase activity ( Figure 4H). Gel electrophoresis and biotin-labelled miRNA pulldown analysis showed that biotin-gga-miR-12207-5P mimic transfection significantly enriched circPLXNA2 mRNA levels in CPMs ( Figure 4I,J). Collectively, these findings suggested that circPLXNA2 and gga-miR-12207-5P have a targeted binding relationship. CPMs ( Figure 4I,J). Collectively, these findings suggested that circPLXNA2 and gga-miR-12207-5P have a targeted binding relationship.

gga-miR-12207-5P Inhibits Cell Proliferation and Promotes Apoptosis
To explore the biological function of gga-miR-12207-5P, we designed the overexpression vector of gga-miR-12207-5P and termed it the gga-miR-12207-5P mimic. After 48 h of transfection, the relative expression level of gga-miR-12207-5P was significantly higher than the control group, indicating reliable overexpression efficiency. ( Figure 5A). After transferring gga-miR-12207-5P mimic into CPMs 48 h, cyclin D1, cyclin D2, and cyclin B2 were all significantly downregulated compared to the mimic negative control (mimic NC) treatment, suggesting that it inhibits proliferation ( Figure 5B). After 48 h of transfection, the apoptosis pathway was activated, as evidenced by an increase in the relative expression level of caspase 8, caspase 9, and caspase 3 ( Figure 5C,D). EdU staining further supported the fact that gga-miR-12207-5P overexpression significantly repressed the viability of CPMs ( Figure 5E,F). Overall, the viability of CPMs may be downregulated by overexpression of gga-miR-12207-5P.

gga-miR-12207-5P Inhibits Cell Proliferation and Promotes Apoptosis
To explore the biological function of gga-miR-12207-5P, we designed the overexpression vector of gga-miR-12207-5P and termed it the gga-miR-12207-5P mimic. After 48h of transfection, the relative expression level of gga-miR-12207-5P was significantly higher than the control group, indicating reliable overexpression efficiency. ( Figure 5A). After transferring gga-miR-12207-5P mimic into CPMs 48h, cyclin D1, cyclin D2, and cyclin B2 were all significantly downregulated compared to the mimic negative control (mimic NC) treatment, suggesting that it inhibits proliferation ( Figure 5B). After 48 h of transfection, the apoptosis pathway was activated, as evidenced by an increase in the relative expression level of caspase 8, caspase 9, and caspase 3 ( Figure 5C,D). EdU staining further supported the fact that gga-miR-12207-5P overexpression significantly repressed the viability of CPMs ( Figure 5E,F). Overall, the viability of CPMs may be downregulated by overexpression of gga-miR-12207-5P.

MDM4 Promotes Cell Proliferation and Inhibits Cell Apoptosis through the P53 Signaling Pathway
To confirm MDM4′s biological function in the myogenesis process, we created an overexpression plasmid of MDM4 to detect it and transfected this plasmid into CPMs, which showed a significant repression of the relative mRNA expression of p53, a critical signaling pathway regulating cell apoptosis, indicating that MDM4 may inhibit myogenesis via apoptosis. ( Figure 6A). As such, we detected the relative expression level of cyclin D1, cyclin D2, and cyclin D3 when MDM4 was overexpressed in CPMs. The results revealed that these genes were significantly promoting expression at the mRNA level. Additionally, we examined the potential interaction between gga-miR-12207-5P, MDM4, and circPLXNA2. Overexpression of gga-miR-12207-5P decreased MDM4 and circPLXNA2 mRNA abundance, suggesting that it inhibits their transcriptional activity ( Figure 5G).

MDM4 Promotes Cell Proliferation and Inhibits Cell Apoptosis through the P53 Signaling Pathway
To confirm MDM4 s biological function in the myogenesis process, we created an overexpression plasmid of MDM4 to detect it and transfected this plasmid into CPMs, which showed a significant repression of the relative mRNA expression of p53, a critical signaling pathway regulating cell apoptosis, indicating that MDM4 may inhibit myogenesis via apoptosis. ( Figure 6A). As such, we detected the relative expression level of cyclin D1, cyclin D2, and cyclin D3 when MDM4 was overexpressed in CPMs. The results revealed that these genes were significantly promoting expression at the mRNA level.  EdU staining assays also suggested that MDM4 overexpression induced CPM proliferation ( Figure 6B-D). On the contrary, silencing the MDM4 gene repressed the relative mRNA expression of cyclin D1, cyclin D2, and cyclin D3, as well as the total number of stained cells (Figure 6E-H). Furthermore, we detected the relative expression level of apoptosis-related genes and found that caspase 3, caspase 8, and caspase 9 were significantly downregulated 48 h after transfecting an overexpression MDM4 plasmid ( Figure 6I). Western blotting analysis also showed a decreasing trend of caspase 3, caspase 8, and caspase 9 after MDM4 overexpression ( Figure 6J). Conversely, we observed the opposite effect of caspase 3, caspase 8, and caspase 9 as the expression of MDM4 was interfered with ( Figure 6K). Congruously, these results supported the biological role of MDM4 in promoting cell proliferation and inhibiting apoptosis. We used the RNAhybrid tool to predict the target binding region of the 3 UTR of MDM4 and the gga-miR-12207-5P seed sequence and identified the best-matched region possessing the least free energy ( Figure 6L) (https://bibiserv.cebitec.uni-bielefeld.de/rnahybrid). Subsequently, dual-luciferase reporter vectors (i.e., Pmir-GLO-MDM4-WT and Pmir-GLO-MDM4-MUT) were constructed and transferred into DF-1 cells to further explore the potential interaction of MDM4 and gga-miR-12207-5P. RT-qPCR results demonstrated the targeted binding relationship between gga-miR-12207-5P and MDM4 ( Figure 6M), which was further confirmed by a biotin-coupled miRNA pulldown assay ( Figure 6N,O).

CircPLXNA2 Affects the Proliferation and Apoptosis of Myoblasts by Regulating MDM4 through the Competitive Sponge of gga-miR-12207-5P
To further characterize the potential interaction relationship between circPLXNA2, gga-miR-12207-5P, and MDM4, we measured MDM4 expression levels after overexpressing both gga-miR-12207-5P mimic and circPLXNA2. To be more specific, compared to the gga-miR-12207-5P mimic and circPLXNA2 co-transfection groups, the mRNA expression level of MDM4 was significantly upregulated in the gga-miR-12207-5P mimic and PCD25 co-transfection groups, as well as the mimic NC and circPLXNA2 co-transfection groups. Furthermore, the relative mRNA expression level of MDM4 was lowest in the group that had both gga-miR-12207-5P mimic and PCD25 co-transfected. In contrast, the expression level of MDM4 was highest when circPLXNA2 was overexpressed, which suggests that circPLXNA2 may reverse the effect of gga-miR-12207-5P on MDM4 ( Figure 7A).
We further quantified the cyclin B2, cyclin D1, and cyclin D2 genes after overexpressing gga-miR-12207-5P and circPLXNA2 in CPMs, individually or in combination, to confirm the potential proliferation effect of both molecules. As shown, the RT-qPCR results indicated that as circPLXNA2 was overexpressed independently, the mRNA relative expression of the three genes was drastically higher than other groups (p < 0.05), whereas gga-miR-12207-5P independent overexpression reduced the mRNA relative expression of the three genes. We, therefore, co-transfected circPLXNA2 and gga-miR-12207-5P mimic and discovered that the activity levels of these three genes were partially restored ( Figure 7B). Furthermore, we measured the abundance of the apoptosis markers caspase 3, caspase 8, and caspase 9 mRNAs after overexpression of gga-miR-12207-5P and circPLXNA2, respectively, or co-overexpression of both. RT-qPCR measurements showed that when circPLXNA2 was overexpressed alone, the three apoptosis markers were drastically downregulated compared to the other three groups. However, when we transfected the gga-miR-12207-5P mimic alone, the mRNA abundance of the three markers increased significantly (p < 0.05). Subsequently, we co-transfected both gga-miR-12207-5P mimic and circPLXNA2, and caspase 3, caspase 8, and caspase 9 were reverted between transfections. (Figure 7C). These findings led us to hypothesize that circPLXNA2 could free MDM4 by competing with it for binding to gga-miR-12207-5P, restoring MDM4 s ability to promote cell proliferation while simultaneously inhibiting apoptosis. In order to measure the levels of circPLXNA2, we cotransfected gga-miR-12207-5P mimic and MDM4. We discovered a similar tendency when compared to the MDM4 expression levels as we transfected gga-miR-12207-5P mimic and circPLXNA2 separately or both synchronously. We also measured the relative levels of proliferation and apoptosis markers after gga-miR-12207-5P and circPLXNA2 co-transfection or independent transfection, which showed a close variable trend ( Figure 7D-F). Overall, these results back our hypothesis, that is, circPLXNA2 recovered the function of MDM4 by competitively binding to gga-miR-12207-5P, thereby recovering the promotion of cell proliferation and inhibition of apoptosis by MDM4.

Discussion
CircRNAs are new members of the non-coding RNA family that emerged from the backbone of pre-mRNA back splicing and were previously classified as "transcriptional noise" in mammals [22,23]. Furthermore, the conventional mode of study is inapplicable due to the circRNA's low abundance and expressive quantity [24]. However, with the progress of the genome sequencing approach, circRNAs are increasingly recognized to be implicated in regulating disease phenotypes. Depending on their subcellular location, circRNAs are regulated in two distinct ways [25]. In the nucleus, circRNAs combine with RNA polymerase II to regulate the transcriptional activity [26]. Alternatively, in the cytoplasm, circRNAs act as molecular sponges to bind with specific miRNA to promote the function of target genes [27][28][29][30][31][32]. For example, it has been demonstrated that circPTPN4 can stimulate NAMPT expression by sponging miR-499-3P in the cytoplasm and acting as a competing endogenous RNA [13]. Hence, here we favor a model to identify the circRNA-miRNA-mRNA interaction network in which the axes may be implicated in the progress of CPMs myogenesis in a combination of multi-omics (i.e., circRNA-seq and riboseq). To be more specific, we first used circRNA-seq to identify a total of 360 circRNAs with differential expression during the GM compared to the DM. We further presented the 3342 ribo-seq-identified differentially-transcribed genes, thereby selecting the genes most likely to play a role in myogenesis during the GM as opposed to the DM. Combining

Discussion
CircRNAs are new members of the non-coding RNA family that emerged from the backbone of pre-mRNA back splicing and were previously classified as "transcriptional noise" in mammals [22,23]. Furthermore, the conventional mode of study is inapplicable due to the circRNA's low abundance and expressive quantity [24]. However, with the progress of the genome sequencing approach, circRNAs are increasingly recognized to be implicated in regulating disease phenotypes. Depending on their subcellular location, circRNAs are regulated in two distinct ways [25]. In the nucleus, circRNAs combine with RNA polymerase II to regulate the transcriptional activity [26]. Alternatively, in the cytoplasm, circRNAs act as molecular sponges to bind with specific miRNA to promote the function of target genes [27][28][29][30][31][32]. For example, it has been demonstrated that circPTPN4 can stimulate NAMPT expression by sponging miR-499-3P in the cytoplasm and acting as a competing endogenous RNA [13]. Hence, here we favor a model to identify the circRNA-miRNA-mRNA interaction network in which the axes may be implicated in the progress of CPMs myogenesis in a combination of multi-omics (i.e., circRNA-seq and ribo-seq). To be more specific, we first used circRNA-seq to identify a total of 360 circRNAs with differential expression during the GM compared to the DM. We further presented the 3342 ribo-seq-identified differentially-transcribed genes, thereby selecting the genes most likely to play a role in myogenesis during the GM as opposed to the DM. Combining these, we characterized a circRNA-miRNA-mRNA network that may participate in regulating myogenesis.
In our study, we identified a novel circRNA with heterogeneous expression in GM relative to DM and termed circPLXNA2, which is generated from the 3-4 exon of the PLXNA2 gene. Overexpression, real-time quantitative PCR, a dual-luciferase reporter assay, RNA pulldown, and the addition of EdU were the methods that we used to demonstrate the potential biological role of circPLXNA2. We found that circPLXNA2 could promote the relative mRNA abundance of cyclin D1, cyclin D2, and cyclin D3 and inhibit mRNA abundance of caspase 3, caspase 8, and caspase 9, facilitating cell proliferation and repressing apoptosis. In addition, the interaction between gga-miR-12207-5P and circPLXNA2 was investigated further and found to be consistent with the results of our in-silico analysis.
MDM4 is constituted as a p53 inhibitor to regulate p53 activity, thereby inhibiting cell apoptosis and promoting proliferation [20]. P53 is a sensor of genotoxic stress that could protect cells from DNA damage by inducing cell-cycle arrest [21]. Numerous studies have shown that p53 induces apoptosis via both transcriptional activation of pro-apoptotic and repression of anti-apoptotic genes, as well as non-transcriptional mechanisms [14,15]. As exemplified by Cai et al., miR-16-5p could inhibit proliferation and promote apoptosis using the p53 signaling pathway by regulating SESN1 mRNA relative expression [33]. Here, we showed that gga-miR-12207-5P has the opposite effect on circPLXNA2, namely, that it can suppress MDM4, which controls myoblast proliferation and apoptosis via the p53 signaling pathway (Figure 8). By simultaneously overexpressing circPLXNA2 and gga-miR-12207-5P, we restored the relative abundance of MDM4, which reversed proliferation and inhibited apoptosis. these, we characterized a circRNA-miRNA-mRNA network that may participate in regulating myogenesis. In our study, we identified a novel circRNA with heterogeneous expression in GM relative to DM and termed circPLXNA2, which is generated from the 3-4 exon of the PLXNA2 gene. Overexpression, real-time quantitative PCR, a dual-luciferase reporter assay, RNA pulldown, and the addition of EdU were the methods that we used to demonstrate the potential biological role of circPLXNA2. We found that circPLXNA2 could promote the relative mRNA abundance of cyclin D1, cyclin D2, and cyclin D3 and inhibit mRNA abundance of caspase 3, caspase 8, and caspase 9, facilitating cell proliferation and repressing apoptosis. In addition, the interaction between gga-miR-12207-5P and circPLXNA2 was investigated further and found to be consistent with the results of our in-silico analysis.
MDM4 is constituted as a p53 inhibitor to regulate p53 activity, thereby inhibiting cell apoptosis and promoting proliferation [20]. P53 is a sensor of genotoxic stress that could protect cells from DNA damage by inducing cell-cycle arrest [21]. Numerous studies have shown that p53 induces apoptosis via both transcriptional activation of pro-apoptotic and repression of anti-apoptotic genes, as well as non-transcriptional mechanisms [14,15]. As exemplified by Cai et al., miR-16-5p could inhibit proliferation and promote apoptosis using the p53 signaling pathway by regulating SESN1 mRNA relative expression [33]. Here, we showed that gga-miR-12207-5P has the opposite effect on circPLXNA2, namely, that it can suppress MDM4, which controls myoblast proliferation and apoptosis via the p53 signaling pathway (Figure 8). By simultaneously overexpressing circPLXNA2 and gga-miR-12207-5P, we restored the relative abundance of MDM4, which reversed proliferation and inhibited apoptosis.

Animals and Cell
Thigh muscles of 11 embryo-aged chickens were isolated on sterile forceps using autoclaved surgical equipment and then completely minced. The crushed muscle tissue was

Animals and Cell
Thigh muscles of 11 embryo-aged chickens were isolated on sterile forceps using autoclaved surgical equipment and then completely minced. The crushed muscle tissue was digested in a constant temperature incubator at 37 • C for 15 min using 0.25% trypsin-EDTA (Gibco, Carlsbad, CA, USA). Cells were treated with high glucose Dulbecco's modified Eagle medium (Gibico) containing fetal bovine serum (FBS)10% (v/v) and 0.2% penicillin/streptomycin) to terminate digestion. The digested suspension was then filtered through a 70 µm strainer and centrifuged at 1000× g for 5 min. The precipitated cells were seeded uniformly in cell culture dishes by blowing with complete medium, and adherent purification was performed twice for 40 min at 37 • C in an incubator with 5% CO 2 to eliminate the interference of chick embryo fibroblasts. Finally, the medium containing CPM cells was uniformly seeded in cell culture dishes for subsequent experiments.
We incorporated different mediums to prompt cell to reach specific life cycles. For myogenic proliferation, cells were treated with growth medium, which is composed of high glucose Dulbecco's modified Eagle medium (Gibico) containing FBS 10% (v/v) and 0.2% penicillin/streptomycin. For myogenic differentiation, the growth medium was replaced with differentiation medium (RPMI-1640 medium (Gibico) with 2% horse serum) when myoblasts reached 90% cell confluence.

Parallel Generation of circRNA and riboRNA Libraries
We prepared the CPM samples at growth medium (GM) and differentiation medium (DM) for three biological replicates with good cell viability, respectively. Each sample was individually pooled for subsequent RNA extraction and library construction. Total RNA was extracted using TRI-zol ® Reagent (Invitrogen, CA, USA) according to the manufacturers' instruction. Subsequently, the total RNA was treated with RNase R to degrade the linear RNAs and purified using a RNeasy MinElute Cleanup Kit (Qiagen, Hilden, Germany). Genomic DNA was then removed using DNase I (Takara, Dalian, China). CircRNA libraries were constructed using the TruSeqTM Stranded Total RNA Sample Preparation Kit from Illumina (Illumina, Los Angles, USA) after the removal of host ribosomal RNA using Illumina Ribo-ZeroTM rRNA Removal Kits (Qiagen) and RNase R treatment. Next, a strand-specific library was constructed using a VAHTS Total RNA-seq (H/M/R) Library Prep Kit (Vazyme, Nanjing, China) from Illumina, following the manufacturer's instructions. Briefly, ribosome RNAs were removed to retain circRNAs. The enriched circRNAs were fragmented into short fragments by using fragmentation buffer and reverse transcribed into cDNA with random primers. Second-strand cDNA were synthesized by DNA polymerase I, RNase H, dNTP (dUTP instead of dTTP), and buffer. Next, the cDNA fragments were purified with VAHTSTM DNA Clean Beads (Vazyme), end repaired, poly(A) was added, and then they were ligated to Illumina sequencing adapters. Then, UNG (Uracil-N-Glycosylase) was used to digest the second-strand cDNA. The digested products were purified with VAHTSTM DNA Clean Beads, PCR amplified, and sequenced using Illumina Novaseq 6000 (Illumina).

Ribosome Footprinting
Analogously, we prepared the CPMs during GM and DM three times as biological replicates. The ribosome footprints were generated after immunoprecipitation of cardiomyocyte-specific monosomes with anti-HA magnetic beads after treating the lysate with RNase I. Then, the libraries were constructed according to the manufacturer's instruction using a mammalian ribo-seq kit (Illumina) with each replicate. Then, the libraries were sequenced using Illumina Novaseq 6000 (Illumina).

CircRNA Validation
We first validated the biological structure of circPLXNA2 using convergent and divergent primers. Then, the junction sequence of the circPLXNA2 was sequenced by Tsingke Biotechnology Co., Ltd. (Beijing, China). Furthermore, we detected the RNase R sensitivity using RT-qPCR.

Primers
All primers for this study were designed in Premier Primer v5.0 (Premier Biosoft International, San Francisco, CA, USA), and synthesized by Tsingke Biotechnology Co., Ltd. (Beijing, China). The detailed information on all primers is listed in Table 1. Table 1. Primers used in this study.

RNA Oligonucleotides and Plasmids Construction
Gga-miR-12207-5P mimic, mimic NC; si-MDM4 and siRNA negative control were designed and synthesized by RiboBio (Guangzhou, China). The detailed information of sequences is listed in Table 2. Table 2. Oligonucleotide sequences in this study.

Name Sequences (5 -3 )
si-MDM4 AGACTAACCAGGACATAGA gga-miR-12207-5P mimic ACGCAGCACGCAGGAGAGCCGAG Construction of the MDM4 overexpression vector was as follows. The full-length sequence of MDM4 was amplified by conventional PCR primers, and the amplified sequence fragment was cloned into the commercial vector pcDNA3.1 (Promega, Madison, WI, USA) through the restriction site of HindIII and XhoII.
Construction of the circPLXNA2 overexpression vector was as follows. First, the full-length linear sequence of circPLXNA2 was synthesized according to the sequencing results. Subsequently, the EcoRI restriction site, forward looping mediate sequence, and AG receptor were added into the 5 end of the forward primer. Similarly, the BamHI restriction site, reverse cyclization mediating sequence, and GT donor were added into the 5 end of the reverse primer. A homology arm was then added to the full-length linear sequence of circPLXNA2 by PCR reaction, and the fragment was cloned into the circular vector pCD25-ciR by homologous recombinase (Geneseed Biotech, Guangzhou, China).
Construction of PmirGLO dual-luciferase target reporter vector was as follows. The DNA templates of circPLXNA2, MDM4 WT, and circPLXNA2,MDM4 MUT were synthesized by Tsingke Bio-technology Co., Ltd. (Beijing, China). The circPLXNA2 and MDM4 3'UTR segments containing the putative gga-miR-12207-5P binding sequence were then subcloned into the XhoI and SalI restriction sites of the pmirGLO dual-luciferase reporter vector. (Promega, USA) (Table S3).

Cell Transfection
All cell transient transfections in this study were performed after cells grew to 60-80% confluence according to the manufacturer's protocol of Lipofectamine 3000 reagent (Invitrogen). The nucleic acids were also diluted using OPTI-MEM medium (Gibco). For miRNA mimic, mimic NC, and siRNA transfection, the final transfection concentration was 100 nM.

5-Ethynyl-20-Deoxyuridine (EdU) Assays
To investigate the proliferation of the cells, 50 µM 5-ethynyl-20-deoxyuridine was first added into cells to incubate for 2 h after 36 h of transfection (EdU) (RiboBio, China), with fixing with 4% paraformaldehyde (PFA) for 30 min and neutralizing with 4% PFA with 2 mg/mL glycine; then, they were permeabilized with 0.5% Triton X-100. Cells were subsequently stained using the C10310 EdU Apollo In Vitro Imaging Kit (RiboBio). Finally, fluorescence microscopy (DMi8; Leica Microsystems, Wetzlar, Germany) was used to obtain cell images. At least three random fields were selected for each well to photograph, and Image J was used to count the cells; hence, the cell proliferation rate was calculated.

Biotin-Coupled miRNA Pull Down Assay
When myoblasts in two 10 cm dishes reached 60-80% confluence, 3 end biotin-labelled gga-miR-12207-5P mimic or mimic NC (100 nM, p < 0.05) were transfected into the cells alternatively. Cells were collected and washed with PBS after 36 h transfection. Using a miRNA pulldown kit (BersinBio, Guangzhou, China), the lysed cells were incubated with the sealed magnetic beads for 4 h at 4 • C with mild rotation (20 rpm/min) to pulldown the biotin-coupled RNA complex. The RNA was eluted and precipitated as required to obtain the RNA for subsequent RT-qPCR that specifically interacted with gga-miR-12207-5P. 4.11. RNA Extraction, cDNA Synthesis and Quantitative Real-Time PCR As described, total RNA was extracted from tissues or cells using TRIzol reagent ® (Invitrogen) according to the manufacturer's requirements. The concentration and integrity of extracted RNA were tested using NanoDrop One and 1% gel electrophoresis. Then, the cDNA was synthesized using PrimeScript RT Reagent Kit with gDNA Eraser (perfect real-time) (TaKaRa). Real-time quantitative PCR was performed as previously described [13,33,34].

Statistical Analysis
In this study, all results were showed as mean ± S.E.M with 3-6 independent replications. An independent sample t-test was used to test the statistically significant difference between groups. We considered p < 0.05 to be statistically significant. Here, * p < 0.05; ** p < 0.01; *** p < 0.001.

Conclusions
In conclusion, our study provided a comprehensive profiling of interaction axes which may regulate the process of myogenesis involving circRNA, miRNA, and mRNA. With these, we proved that a novel circRNA, circPLXNA2, is highly heterogeneously expressed in DM relative to GM. Mechanistically, circPLXNA2 could function as ceRNA to sponge gga-miR-12207-5P to release MDM4, a negative regulator of the p53 signaling pathway, thereby exerting myoblast proliferation and repressing apoptosis. Overall, this study showed the first evidence of circPLXNA2 of regulating skeletal myogenesis, and the circPLXNA2-gga-miR-12207-5p-MDM4 axis may become a potential therapeutic target for muscle development.

Data Availability Statement:
The circRNA-seq and the ribo-seq profiles have been deposited in the China National GeneBank DataBase (CNGBdb) under the accession number CNP0003882. Any remaining data could be acquired in the supplement files or requested to the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.