CircRNA AIDA Regulates Development of Bovine Myoblast via Binding miR-29a

TL;DR

A circular RNA called circAIDA helps control muscle development in cattle by acting as a sponge for miR-29a, affecting cell growth and muscle regeneration.

Contribution

Discovery of a novel circAIDA/miR-29a regulatory axis in bovine skeletal muscle development.

Findings

circAIDA promotes myoblast proliferation and inhibits apoptosis and differentiation in vitro.

circAIDA attenuates skeletal muscle regeneration in vivo by sponging miR-29a.

circAIDA relieves miR-29a repression on AKT3 and CLCN2 genes.

Abstract

What are the main findings?
A novel circAIDA was identified as a miR-29a sponge that relieves the repression of AKT3 and CLCN2.circAIDA promotes myoblast proliferation and inhibits apoptosis and differentiation in vitro, while weakening skeletal muscle regeneration in vivo.

A novel circAIDA was identified as a miR-29a sponge that relieves the repression of AKT3 and CLCN2.

circAIDA promotes myoblast proliferation and inhibits apoptosis and differentiation in vitro, while weakening skeletal muscle regeneration in vivo.

What are the implications of the main findings?
This study reveals a novel circAIDA/miR-29a axis that regulates bovine skeletal muscle development.These findings provide new insights into the molecular mechanisms governing muscle growth and regeneration in bovine.

This study reveals a novel circAIDA/miR-29a axis that regulates bovine skeletal muscle development.

These findings provide new insights into the molecular mechanisms governing muscle growth and regeneration in bovine.

Circular RNAs (circRNAs) have emerged as critical regulators of skeletal muscle development, yet the functions of many muscle-derived circRNAs remain uncharacterized. In this study, we identified a novel circRNA, circAIDA, formed by exons 2 to 6 of the AIDA gene, based on bovine muscle sequencing data. Mechanistically, we demonstrate that circAIDA acts as a molecular sponge for miR-29a, thereby relieving the repression of its downstream targets, AKT3 and CLCN2. Functionally, circAIDA promotes bovine myoblast proliferation while inhibiting apoptosis and differentiation in vitro. Furthermore, experiments in vivo suggested circAIDA could attenuate regeneration of skeletal muscle in mice. In brief, we discovered a novel circAIDA/miR-29a interaction that regulates bovine myogenesis, providing new insights into the molecular networks controlling skeletal muscle development.

Figures

Full text

1. Introduction

Circular RNAs (circRNAs) are covalently closed, endogenous RNA molecules in eukaryotes with neither 5′ to 3′ polarity nor a polyadenylated tail [1,2], which originate in the back-splicing process and usually contain exons, introns or non-coding intergenic regions [3]. As high-throughput RNA sequencing and circRNA-specific bioinformatics have radically improved, there were thousands of circRNAs being identified in eukaryotes as well as viruses [4,5], and they showed specific expression patterns in tissues [6]. CircRNAs are highly stable molecules mainly found in cytosol, where they can function by binding with other molecules, such as proteins and microRNAs [7]. Some long non-coding RNAs (ncRNAs) can selectively sponge miRNAs to regulate their levels and/or activity. The initial observation that showed circRNAs possess many miRNA-binding sites led to the speculation that these molecules could work as miRNA sponges [8]. For instance, CDR1as, a well-known circRNA, contains more than 60 binding sites for miR-7. Its reduced expression decreases the abundance of mRNAs containing miR-7 binding sites, suggesting that CDR1as acts as a miR-7 sponge to participate in the gene expression network [9,10]. Crucially, in the context of skeletal muscle, CDR1as has been reported to induce myogenesis by sponging miR-7 to up-regulate key myogenic factors such as IGF1R and MyoD [11]. This highlights the specific regulatory potential of circRNAs in muscle development.

Nowadays, exploring the regulatory mechanism of myoblast proliferation and myogenesis is one of the research hotspots in developmental biology. As an important economic trait, the muscle development of livestock has increasingly attracted the attention of researchers [12]. A large number of studies have shown that circRNAs can further conduct the process of muscle development by combining with miRNAs to modulate the expression of genes associated with muscle development [13]. In pig skeletal muscle, there are 52,918 high-confidence circRNAs being detected across 27 developmental stages. CircFgfr2A, one of these circRNAs, was differentially expressed in various myogenesis systems and conserved across humans, mice and pigs. It was found to function as a sponge for miR-133 to regulate the Map3k20 gene and JNK/MAPK pathway thereby affecting skeletal muscle development and regeneration [14]. In bovine skeletal muscle, studies have analyzed circRNA high-throughput sequencing data and constructed a circRNA-miRNA-mRNA network, which also analyzed the presumed circHUWE1-miR-29b-AKT3 axis and confirmed its involvement in myogenesis [15]. In addition, many other circRNAs have been proven to bind miRNAs during bovine myoblast development such as circINSR [16], circRILPL1 [17], circCPE [18], circTTN [19] and so on. The above series of findings inspired us that circRNAs played important roles in the regulation of muscle development. Therefore, further research is needed to better explore the functions of circRNAs.

For this study, the previous sequencing data of circRNAs in bovine muscle from the GEO database of NCBI (ID: GSE87908) were analyzed, and then we focused on a novel circRNA, circAIDA, which was differentially expressed in fetal and adult bovine muscle tissue. Derived from exons 2 to 6 of the AIDA gene (axin interactor, dorsalization-associated) on chromosome 16, circAIDA has a length of 350 nt. Our findings showed that circAIDA could promote bovine myoblast proliferation and reduce cell apoptosis and differentiation by adsorbing miR-29a to regulate the expression of the AKT3 and CLCN2 genes. On balance, the data strongly demonstrated that circRNA played an important role in bovine myoblast development through sponging miRNAs.

2. Materials and Methods

2.1. Tissue Samples and Cell Lines

We collected muscle samples of Longissimus Dorsi Muscle from 90-day-gestation fetal calves and 24-month-old cattle (N = 3 each) at a local Qinchuan cattle slaughterhouse in Xi’an (China). Additional tissue samples, including heart, liver, spleen, lung, kidney, and muscle, were collected at the embryonic stage (90 days). All samples were snap-frozen in liquid nitrogen immediately after removal and stored in a −80 °C freezer until they were used for RNA extraction.

Bovine primary myoblasts were isolated from Longissimus Dorsi Muscle of 90-day-gestation fetal calves using collagenase digestion followed by differential adhesion, as described previously [20]. The cells were then cultured in high-glucose Dulbecco’s Modified Eagle Medium (DMEM, Biological Industries, Beit HaEmek, Israel) containing 20% Certified Fetal Bovine Serum (FBS, VivaCell, Shanghai, China) and 2% Penicillin–Streptomycin Liquid (Solarbio, Beijing China). For inducing myogenesis, DMEM with 2% Donor Equine Serum (HyClone, Logan, UT, USA) and 2% Penicillin–Streptomycin Liquid was used as differentiation medium for cells at around 90% confluence. The HEK293T cell line was purchased from the American Type Culture Collection (ATCC) and tested negative for mycoplasma contamination. These cells were cultured in DMEM with 10% FBS and 2% Penicillin–Streptomycin Liquid. All cells above were placed in a 37 °C incubator with 5% CO_2_ (Thermo Scientific CO_2_ Incubators).

2.2. RNA Extraction and Quantitative Real-Time PCR

Total RNA was extracted from bovine tissues and cultured cells using AG RNAex Pro Reagent (Accurate Biology, Changsha, China). Synthesis of cDNA from RNA reverse transcription used the Evo M-MLV RT Kit with gDNA Clean for qPCR II (Accurate Biology, China). Specific stem-loop primers were used for miRNA reverse transcription. The quantitative real-time polymerase chain reaction (qPCR) was performed using 2× Taq SYBRGreen^®^ qPCR PreMix (innovagene, Suzhou, China). The glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was applied as the internal control for both circRNA and mRNA detection. The expression of U6 was used for normalizing miRNAs expression. Primers in this study are shown in Table S1.

2.3. Treatment with RNase R and Actinomycin D

For RNase R assay, total RNA was incubated 5 U/μg RNase R (Epicenter, Madison, WI, USA) for 15 min at 37 °C. For Actinomycin D assay, myoblasts were cultured in growth medium supplemented with 2 μg/mL Actinomycin D (MilliporeSigma, Burlington, MA, USA) and samples were collected at 0, 4, 8, and 12 h.

2.4. Vector Construction and Cell Transfection

The second to sixth exon sequences of the bovine AIDA gene were cloned to construct the pcDNA2.1-circAIDA overexpression vector (Geneseed, Guangzhou, China). The small interfering RNA (siRNA)-targeting circAIDA, miR-29a mimics, miR-29a inhibitors and corresponding negative control (NC) were synthesized by General Biology (Hefei, China). The wild-type and mutant full-length sequences of circAIDA were separately inserted into the psiCHECK-2 vector (Promega, Madison, WI, USA). Two copies of miR-29a reverse complementary sequences were synthesized to construct the miR-29a biosensor (psiCHECK-2-miR-29a 2×). The predicted 3′UTR fragment containing the miR-29a binding site in AKT3 was also cloned into the psiCHECK-2 vector as the wild-type. The sequences were replaced as indicated to mutate the binding site. All transfections were conducted with LipoGene^TM^ 2000 Star Transfection Reagent (UE, Shanghai, China) according to the manufacturer’s protocols.

2.5. Dual-Luciferase Reporter Assay

HEK293T cells were cultured in 96-well plates and transfected with previously constructed psiCHECK-2 reporter plasmids, mimics NC, mimics miR-29a, pcDNA2.1-circAIDA or their combination. After 24 h, Firefly and Renilla luciferase activities were detected by Dualucif^TM^ Firefly & Renilla Assay Kit (UE, Shanghai, China) in accordance with the manufacturer’s instructions.

2.6. Cell Counting Kit-8 (CCK-8) and 5-Ethynyl-2′-Deoxyuridine (EdU) Assay

For cell proliferation assay, Cell Counting Kit-8 (CCK-8, UE, Shanghai, China) and BeyoClick™ EdU Cell Proliferation Kit with Alexa Fluor 594 (Beyotime, Shanghai, China) were respectively used according to their manufacturer’s protocols. The transfected myoblasts were cultured in 96-well plates with growth medium.

2.7. Immunofluorescence Staining

Myoblasts were cultured in differentiation medium to induce differentiation for 4 days. Then, they were fixed by 4% paraformaldehyde for 20 min. After washing 3 times with PBS, cells were permeabilized by 0.5% TritonX-100 for 15 min. After blocking with 5% BSA, they were incubated overnight with antibody-MyHC (1:250; GTX20015, GeneTex, Irvine, CA, USA) at 4 °C. The secondary antibody we used was goat anti-mouse IgG (H&L)-Alexa Fluor 594 (1:500; RS3608; Immunoway Biotechnology, Plano, TX, USA), which was incubated with the cells at room temperature for 2 h. The nuclei were stained with Hochest 33342 (UE, Shanghai, China). Finally, a fluorescent microscope (DM5000B; Leica, Wetzlar, Germany) was used for observation.

2.8. Western Blot

Extraction of total proteins from cultured cells with different treatments used RIPA buffer (high) containing 1mM PMSF (Solarbio, Beijing, China). Then, they were loaded and separated by 12% SDS-PAGE gel for 2 h and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, Darmstadt, Germany) for 2.5 h. After blocking with 5% non-fat milk, the membranes were incubated at 4 °C overnight with specific primary antibodies: anti-GAPDH (1:5000, #M20006, Abmart, Shanghai, China), anti-CyclinE (1:750, #WL01072, Wanleibio, Shenyang China), anti-PCNA (1:500, #WL02208, Wanleibio, Shenyang, China), anti-CyclinD1 (1:500, #WL01435a, Wanleibio, Shenyang, China), anti-CDK2 (1:500, #WL02028, Wanleibio, China), anti-p53 (1:500, #WL01919, Wanleibio, Shenyang, China), anti-caspase9 (1:1000, # CY5049, Abways Technology, Shanghai, China), anti-Bax (1:1000, # CY5059, Abways Technology, China), anti-MyoD1 (1:1000, #CY6736, Abways Technology, Shanghai, China), anti-MyoG (1:1000, ab103924, Abcam, Cambridge England), anti-AKT3 (1:500, #WL0005a, Wanleibio, China). After incubation with appropriate secondary antibodies, the membranes were detected using the ChampChemi™ Top 610 (SAGECREATION, Beijing, China).

2.9. Animal Studies

All animal procedures were performed in accordance with the guidelines of the Animal Care and Use Committee of Northwest A&F University. A total of 24 female C57BL/6 mice (6 weeks old) were maintained in the laboratory animal facility under a 12 h light/dark cycle with ad libitum access to food and water. Mice were randomly assigned to experimental groups. To monitor muscle regeneration, 50 μL of 10mM cardiotoxin (CTX) in PBS was injected to make muscle injury in the right tibialis anterior (TA) muscle of 6-week-old mice. For experimental overexpression of circAIDA in vivo, a total of 6.25 μg pcDNA2.1-circAIDA plasmid was prepared by preincubating with Entranster™-in vivo (Engreen, Beijing, China) for 15 min and mixed with 10% glucose solution to a final volume of 50 μL. The same amount of pcDNA2.1 vector was used as the control group. They were injected three times: at 12 h, 2 d, and 4 d after CTX treatment. Mice were sacrificed by cervical dislocation at 3 d and 6 d to harvest TA muscles for RNA extraction and histological analysis.

2.10. Statistical Analysis

All results were shown as the mean ± SEM of at least three independent biological replicates. The normality of data distribution was assessed using the Shapiro–Wilk test. Comparisons between two groups were analyzed by Student’s t test using GraphPad Prism 8.0.2 software. Statistical differences were considered significant if p < 0.05 and indicated with one asterisk, while two asterisks indicated p < 0.01.

3. Results

3.1. Identification and Expression Pattern of Bovine circAIDA

We characterized the constitution of circAIDA, which is derived from exons 2 to 6 of the AIDA gene (350 nt). Divergent primers were used to amplify the back-splice junction of circAIDA, and products were verified by Sanger sequencing (Figure 1A). To confirm the circular nature of circAIDA, divergent and convergent primers were designed. PCR analysis demonstrated that divergent primers were able to amplify the circular isoform of circAIDA using cDNA as a template, but not genomic DNA (gDNA), while convergent primers were able to generate the linear product not only from cDNA but also from gDNA (Figure 1B). We used bovine myoblasts after RNase R or Actinomycin D treatment to detect the stability of circAIDA. The results revealed that circAIDA was resistant to RNase R digestion compared to the linear host gene (Figure 1C). Furthermore, circAIDA exhibited higher stability than AIDA mRNA following Actinomycin D treatment (Figure 1D). Next, we detected the expression of circAIDA in the nucleus and cytoplasm, which showed that circAIDA was mainly located in the cytoplasm (Figure 1E). Analysis of expression patterns in tissues showed that the expression of circAIDA in embryonic muscle was significantly higher than the adult stage, which was consistent with the previous sequencing data. We also measured the expression of circAIDA in various fetal bovine tissues. It is generally expressed in the heart, liver, spleen, lung, kidney and more in muscle (Figure 1F). We analyzed the abundance of circAIDA and AIDA after overexpression or reduction with circAIDA. The results showed that expression of the circular type altered, but the linear AIDA mRNA did not change (Figure 1G,H).

3.2. Effect of circAIDA on Bovine Myoblast Proliferation

To investigate the physiological function of circAIDA, bovine myoblasts were transfected with either the pcDNA2.1-circAIDA overexpression vector or specific small interfering RNA (siRNA) to overexpress or knockdown circAIDA, respectively. The results showed that the overexpression of circAIDA could increase the expression of proliferation-related genes such as Cyclin E, PCNA, Cyclin D1 and CDK2 at both mRNA (Figure 2A) and protein (Figure 2C) levels. Conversely, circAIDA knockdown decreased their expression at both mRNA (Figure 2B) and protein (Figure 2D) levels. The Cell Counting Kit-8 (CCK-8) assay revealed that cell viability increased after overexpressing circAIDA (Figure 2E) and decreased after silencing circAIDA (Figure 2F). Finally, the analysis of 5-Ethynyl-2′-deoxyuridine (EdU) assay showed overexpression of circAIDA significantly accelerated cell proliferation (Figure 2G,H). In contrast, silencing circAIDA exerted the opposite effect (Figure 2G,I). Briefly, these data suggested that circAIDA promotes the proliferation of bovine myoblasts.

3.3. Effect of circAIDA on Bovine Myoblast Apoptosis and Differentiation

To explore the role of circAIDA in bovine myoblast apoptosis, we measured the mRNA and protein levels of key apoptosis-related genes. The results indicated up-regulation of circAIDA could increase the mRNA expression of Bcl2 while inhibiting p53, caspase9 and Bax (Figure 3A). Conversely, circAIDA knockdown yielded the opposite effects (Figure 3B). For protein expression, circAIDA overexpression also decreased the levels of p53, caspase9 and Bax (Figure 3C), whereas circAIDA knockdown increased them (Figure 3D). Next, in order to determine whether circAIDA regulates myoblast differentiation, we collected cells 3 days after inducing myogenesis by differentiation medium for qPCR and Western blot. The analysis of qPCR showed the mRNA expression of MyHC, MyoD and MyoG was down-regulated by circAIDA overexpression (Figure 3E) and up-regulated by the interference of circAIDA (Figure 3F). Western blot results also explained that overexpression of circAIDA decreased the protein abundance of these differentiation-related genes including MyoD and MyoG at the protein level (Figure 3G), while interference of circAIDA increased their protein expression levels (Figure 3H). As shown by immunofluorescence staining results (Figure 3I), overexpression of circAIDA not only suppressed the expression of MyHC but also inhibited the formation of myotubes, whereas interference had the opposite effect. Collectively, these findings demonstrated that circAIDA suppressed the apoptosis and differentiation of bovine myoblasts.

3.4. CircAIDA Acts as a miR-29a Sponge

As circAIDA is primarily localized in the cytoplasm (Figure 1E), we hypothesized that it functions as an miRNA sponge. To identify potential targets, we used RNAhybrid to predict interactions between circAIDA and key myogenesis-related miRNAs. Bioinformatics analysis showed that there was a potential binding site for circAIDA to interact with miR-29a (Figure 4A). Expression analysis showed that circAIDA overexpression decreased miR-29a levels, whereas circAIDA knockdown increased them (Figure 4B,C). Next, we respectively transfected miR-29a mimics and inhibitors into bovine myoblasts to confirm significant overexpression or silencing of miR-29a (Figure 4D,E). To confirm the direct interaction between circAIDA and miR-29a, we performed dual-luciferase reporter assays. We constructed psiCHECK-2 reporter plasmids containing either wild-type (WT) or mutant (MUT) circAIDA sequences (Figure 4F). The results of the dual-luciferase reporter assay in HEK293T cells showed that the overexpression of the miR-29a significantly inhibited relative luciferase activity of the circAIDA-WT plasmid but not the mutant (Figure 4G). The miR-29a biosensor was constructed to further determine the complementarity. We found that miR-29a overexpression reduced the relative luciferase activity of the biosensor, whereas co-transfection of circAIDA relieved the repression of miR-29a (Figure 4H). Given all of that, the adsorption of miR-29a by circAIDA was confirmed.

3.5. Effect of miR-29a on Bovine Myoblast Proliferation, Apoptosis and Differentiation

To investigate whether circAIDA functions by sponging miR-29a, we performed rescue experiments using co-transfection strategies. Cells were transfected with mimics NC, miR-29a mimics, or miR-29a mimics + circAIDA. Parallel experiments were conducted using inhibitors and si-circAIDA. Regarding proliferation, qPCR and Western blot results showed that miR-29a overexpression significantly suppressed proliferation markers at both mRNA (Figure 5A) and protein (Figure 5C) levels. Importantly, co-transfection with circAIDA alleviated this inhibition. Conversely, miR-29a inhibition yielded the opposite effects (Figure 5B,D). Consistent results were obtained using CCK-8 (Figure 5E,F) and EdU (Figure 5G–I) assays, confirming that circAIDA rescues the proliferation arrest caused by miR-29a. Regarding apoptosis, miR-29a overexpression and inhibition significantly altered the expression of apoptosis markers at both mRNA (Figure 6A,B) and protein (Figure 6C,D) levels. Notably, circAIDA effectively reversed these effects. Regarding differentiation, the expression of differentiation-related genes was similarly altered by miR-29a modulation (Figure 6E–H), and these changes were prevented by circAIDA. Immunofluorescence staining further illustrated that miR-29a influenced MyHC expression and myotube formation, while circAIDA attenuated this impact (Figure 6I,J). Ultimately, these data indicated miR-29a inhibited the proliferation of bovine myoblasts while promoting apoptosis and differentiation. But these effects could be reversed by circAIDA, which also proved the interaction between circAIDA and miR-29a.

3.6. Overexpression of circAIDA Attenuates Injury-Induced Mouse Muscle Regeneration In Vivo

We established a cardiotoxin (CTX)-induced muscle regeneration model in mice to validate the above discovery, avoiding the challenges of in vivo experiments in big animals. Bioinformatics analysis using miRBase, MEGA 7.0.14, and WebLogo 2.8.2 revealed that miR-29a is highly conserved among vertebrates such as humans, mice, cows and so on (Figure 7A). The tibialis anterior (TA) muscles were collected at 0, 3, 6 and 9 days after CTX injection (Figure 7B) to perform Hematoxylin and Eosin (H&E) staining. Histological examination revealed that the muscle fibers in the 3d-CTX group were extensively degenerated and vacuolated at the injury site, accompanied by a significant infiltration of inflammatory cells (primarily neutrophils and macrophages) and scattered nuclei. Newly formed muscle fibers containing the central nucleus replaced part of the injury muscle fibers at 6d-CTX, while the 9d-CTX muscle showed advanced regeneration (Figure 7C). These meant the above model was made successfully. Then, we injected the pcDNA2.1-circAIDA overexpression vector into the TA muscle following the outlined protocol in Figure 7D. In terms of phenotype, overexpression of circAIDA at 3 d after CTX injection could find a significant decline in nascent muscle fibers, while there were also fewer newly formed muscle fibers containing the central nucleus than the control group at 6 d (Figure 7E). qPCR analysis confirmed that circAIDA was successfully overexpressed and the abundance of miR-29a and PAX7 were significantly decreased in the overexpression group at both 3 d (Figure 7F) and 6 d (Figure 7G). In summary, these data represented that circAIDA could attenuate injury-induced mouse muscle regeneration in vivo by delaying the maturation of myofibers.

3.7. CircAIDA Regulates AKT3 and CLCN2 by Sponging miR-29a

Bioinformatics analysis using TargetScan 7.2 predicted that AKT3 is a potential target of miR-29a (Figure 8A). It has been reported that miR-29a played a role in skeletal muscle development by targeting AKT3 [21]. Consistent with this, we found that the mRNA (Figure 8B,C) and protein (Figure 8D,E) expression levels of AKT3 was increased after circAIDA overexpression, while after interference with circAIDA they were decreased. We further investigated whether circAIDA modulates AKT3 via miR-29a. As expected, up-regulation of miR-29a inhibited AKT3 expression, which could be abolished by circAIDA (Figure 8F,H). Interference with miR-29a had the opposite effect (Figure 8G,I). We constructed psiCHECK-2 reporter vectors including AKT3-3′UTR-WT and AKT3-3′UTR-MUT as represented in Figure 8J and performed dual-luciferase reporter assays to confirm this interaction between miR-29a and AKT3. The results showed mimics miR-29a significantly decreased the luciferase activity of the AKT3-3′UTR-WT vector, but it was rescued by circAIDA overexpression (Figure 8K). To sum up, all research supported the conclusion that circAIDA regulates AKT3 expression by sponging miR-29a in the development of bovine myoblasts.

In addition to AKT3, the seed sequence of miR-29a could also target the 3′UTR of the CLCN2 gene mRNA (Figure 9A). Since the CLCN2 gene was rarely studied in bovine muscle, its function was probed next. We transfected the si-CLCN2 to bovine myoblasts, after which there was a remarkable reduction in CLCN2 expression by qPCR detection (Figure S1A). Functionally, qPCR and Western blot analyses demonstrated that CLCN2 knockdown significantly decreased the expression of proliferation markers, including Cyclin E, PCNA, Cyclin D1, and CDK2 (Figure S1B,C). The data of CCK-8 (Figure S1D) and EdU (Figure S1E,F) also suggested that the viability of cell proliferation was attenuated compared with the control after transfection with si-CLCN2. As for the effect on apoptosis and differentiation of bovine myoblasts, qPCR (Figure S1G,I) and Western blot (Figure S1H,J) assay showed that the knockdown of CLCN2 did alter the expression of marker genes including Bcl2, p53, caspase9, Bax, MyHC, MyoD and MyoG. Imaging of immunofluorescence staining (Figure S1K) indicated myotube formation was better after disruption of CLCN2 than in controls. These results indicate that CLCN2 promotes bovine myoblast proliferation while inhibiting apoptosis and differentiation. The function of CLCN2 is consistent with the function of circAIDA on bovine myoblasts, which provided support for our speculation that circAIDA conformed to the ceRNA mechanism.

Finally, we validated the regulation of CLCN2 by the circAIDA/miR-29a axis. The analysis of qPCR showed the overexpression of circAIDA significantly increased the expression level of CLCN2 (Figure 9B), while after silencing circAIDA it was reduced (Figure 9C). And up-regulation of miR-29a decreased CLCN2 expression, which could be rescued by circAIDA (Figure 9D). Knockdown of miR-29a did the opposite (Figure 9E). Next, we constituted psiCHECK-2 reporter vectors including CLCN2-3′UTR-WT and CLCN2-3′UTR-MUT (Figure 9F). The dual-luciferase reporter assay suggested that miR-29a brought significant reduction in the activity of CLCN2-3′UTR-WT, but there was no impact on CLCN2-3′UTR-MUT (Figure 9G). Crucially, co-transfection with circAIDA recovered the luciferase activity (Figure 9G). In conclusion, circAIDA acts as a sponge for miR-29a, impairing its repression of AKT3 and CLCN2.

4. Discussion

Although circular RNAs (circRNAs) were identified approximately 40 years ago, they have only recently gained significant attention in the mainstream science [22]. Accumulating evidence shows that circRNAs specifically manage development or disease progression by acting as miRNA sponges and many others [23]. Several circRNAs have been suggested to function together with lncRNAs and miRNAs as novel regulators of skeletal muscle growth and differentiation [24]. The circNfix/miR-204/MEF2C axis was speculated to regulate myogenesis in mice [25]. The most typical circRNA, CDR1as, has been reported to induce myogenesis in goats via the miR-7/IGF1R regulatory pathway [11]. However, the functions of many other muscle-related circRNAs warrant further investigation.

In this study, we focused on a novel circRNA, circAIDA, which is generated from exons 2 to 6 of the AIDA gene. Sanger sequencing, RNase R treatment, Actinomycin D assay and other experiments were conducted to identify circAIDA. Overexpression and interference experiments showed that circAIDA function was indeed related to cell proliferation, apoptosis and differentiation. The differential expression of circAIDA between fetal and adult bovine muscle, combined with its cytoplasmic localization, suggested a potential ceRNA mechanism. Therefore, we searched for miRNAs that bind to circAIDA and finally identified miR-29a as a candidate through bioinformatics analysis and experimental validation.

miR-29a is a member of the miR-29 family and is commonly implicated in various diseases. It plays a major role in pathophysiological processes, involving cardiorenal disease [26,27], liver disease [28], immune disease [29], tendon disease [30] and others [31]. Meanwhile, studies have supported that miR-29a could function as the potential therapeutic target and promising biomarker for kinds of tumors [32]. The function of miR-29a in cancer indirectly illustrated the importance of miR-29a in the regulation of cell growth and development. miR-29a may up-regulate p53 to regulate cell growth, senescence and apoptosis by targeting the regulatory subunit of PI3 kinase (p85α) and a Rho family GTPase (CDC42) [33]. In the context of muscle biology, miR-29a is characterized as an enhancer of myoblast differentiation and an inhibitor of rhabdomyosarcoma (RMS) through NF-κB and YY1 regulatory circuits [34]. In keeping with previous work, we found that miR-29a inhibited proliferation, and promoted apoptosis and differentiation in bovine myoblasts. Moreover, co-transfection with circAIDA reversed these effects, supporting our hypothesis that circAIDA functions as a molecular sponge for miR-29a.

To further explore the role of circAIDA on myogenesis in vivo, we adopted a well-established muscle regeneration model that induced muscle injury by intramuscular injection of cardiotoxin (CTX) in mice, leading to regeneration. Bioinformatics analysis confirmed that the miR-29a seed regions are highly conserved across species (Figure 8A), justifying the use of a mouse model. We proposed that exogenous circAIDA could adsorb miR-29a in mouse muscle, thereby weakening muscle regeneration. This is likely because muscle regeneration requires a precise temporal transition from proliferation to differentiation [35]. We speculate that sustained overexpression of circAIDA locks myoblasts in the proliferative phase, hindering the differentiation steps necessary for mature muscle fiber formation. Next, our study determined AKT3 and CLCN2 as targets of miR-29a and examined the crosstalk including circAIDA-miR-29a-AKT3 and circAIDA-miR-29a-CLCN2 by dual-luciferase reporter assays.

It is widely known that there are three closely related members of the AKT family including AKT1 (PKBα), AKT2 (PKBβ) and AKT3 (PKBγ) [36]. The primary function of AKT isoforms is to promote cell survival, proliferation, and metabolic changes through downstream targets such as glycogen synthase kinase 3 (GSK3) [37,38], forkhead box O (FoxO) family transcription factors [39,40], tuberous sclerosis complex 2 (TSC2) [41] and the mechanistic target of rapamycin complex 1 (mTORC1) [42,43]. In papillary thyroid carcinoma (PTC), miR-29a has been reported to inhibit proliferation and metastasis by targeting AKT3 [44]. Similarly, miR-29a has been considered to mediate AKT3 silencing to retard cell proliferation and promote differentiation in the C2C12 cell line [21]. Our experiments verified the existence of this miR-29a/AKT3 axis in bovine muscle cells and confirmed that circAIDA could alleviate the inhibitory effect of miR-29a on AKT3.

The CLCN2 gene encodes a voltage-gated chloride channel (CLC-2), which is expressed in many different tissues; its physiological functions are quite diverse, including electrogenesis, homeostatic control of cell volume, and maintenance of ion gradients [45,46]. Previous studies have linked CLCN2 to cell proliferation and differentiation. For example, CLCN2 played roles in IGF-1-induced regulation of vascular smooth muscle cell proliferation [47] and also in differentiation and migration of human conjunctival fibroblasts through the PI3K/AKT signaling pathway [48]. Based on these findings, we investigated CLCN2 function in bovine myoblasts using siRNA knockdown. The results showed that CLCN2 promotes proliferation, inhibiting differentiation while inhibiting apoptosis. Therefore, we propose that CLCN2 acts as a complementary downstream effector to AKT3 within the circAIDA/miR-29a axis. Together, they may synergistically coordinate myoblast proliferation and survival. Further experiments confirmed that circAIDA could relieve the inhibitory effect of miR-29a on the CLCN2 gene in bovine myoblasts.

To our knowledge, this is the first study to validate the function of circAIDA in bovine myoblast development and consider that it involves the miR-29a/AKT3 axis. Of course, there are certain limitations. The efficiency of ceRNA-based regulation depends critically on the relative abundance and competitive dynamics of the specific RNAs involved [49]. Additionally, whether circAIDA regulates other miRNAs or acts through alternative mechanisms remains to be explored. Finally, the technical limitation regarding the cross-species design (bovine and mouse) should be noted, and further in vivo rescue experiments are needed to fully validate the conservation of this mechanism.

Overall, this study identified circAIDA as a novel molecular sponge for miR-29a. circAIDA relieves the repression of AKT3 and CLCN2 genes, thereby promoting myoblast proliferation and suppressing its apoptosis and differentiation in bovines. Meanwhile, experiments in vivo suggested circAIDA could weaken regeneration of skeletal muscle in mice. For animal husbandry, we discovered a novel circAIDA/miR-29a interaction that could regulate bovine muscle cell development and be applied to molecular marker-assisted breeding. As for medical prospects, given the high stability of circRNAs and the inhibitory role of circAIDA in myogenesis, it may serve as a potential diagnostic marker or therapeutic target for muscle-related diseases, where silencing its expression could potentially enhance muscle repair.