Effect of siCDR1as and siDICER1 on SMSCs differentiation and the relationship between CDR1as and DICER1
DICER1 is central to miRNA-mediated silencing [40] and modulates miRNAs during myogenic differentiation [18]. Meanwhile, CDR1as plays a critical role in myogenesis by functioning as a molecular sponge for miR-7 [29]. To identify whether CDR1as and DICER1 are associated in myoblast differentiation, we in vitro cultured SMSCs isolated from Longissimus dorsi (LD) muscles of newborn Jiangzhou Big-Eared goats and knocked down the expression of CDR1as and DICER1 (Fig. 1a, d) as well as mouse myoblast C2C12 (Suppl. Figure 1, 2) by using their respective interfering RNAs. Similar to previous results, deficiency of CDR1as or DICER1 decreased the expression of MyoD mRNA, the master myogenic gene (Fig. 1b, e), and the formation of myotubes (Fig. 1c, h). Moreover, CDR1as was shown to be downregulated in SMSCs transfected with siDICER1 (Fig. 1f) and co-transfection of pCDNA3 (negative control for CDR1as) and siDICER1 (Fig. 1g), suggesting that the close relationship between DICER1 and CDR1as could help identify miRNAs and their targeted mRNAs regulated by CDR1as.
Figure 1 The Expression of Cerebellar Degeneration-Related protein 1 antisense (CDR1as) and DICER1 on skeletal muscle satellite cells (SMSCs) differentiation. (a and d) The expression level of CDR1as and DICER1 in SMSCs transfected with siCDR1as or siDICER1 was validated. (b and e) qPCR analysis of MyoD expression in SMSCs after transfection with siCDR1as or negative control (NC) and siDICER1 or NC. (c and h) A representative of microscopic images after the knockdown of CDR1as and DICER1. (f and g) SMSCs were transfected with NC or siDICER1, pCDNA3 + NC or pCDNA3 + siDICER1 to determine the relationship between CDR1as and DICER1. The p values were analyzed by Student’s test; p < 0.05.
Overview of mRNA and miRNA sequencing data associated with DICER1 and CDR1as
In order to predict mRNAs and miRNAs associated with DICER1 and CDR1as in SMSCs, we systematically cultured SMSCs and knocked down the expression of CDR1as and DICER1, with three biological replicates for each treatment. Cells were harvested at 48 h after transfection and the total RNA were extracted to construct the cDNA libraries individually for mRNA-seq and miRNA-seq using an Illumina HiSeq 2500 platform and 125 bp paired-end reads.
After removing low-quality sequences and adapters, considering the siCDR1as results, an average of 59,821,102 and 58,552,463 mRNAs were produced from raw and clean reads respectively (Table 1). In addition, an average of 15,350,087 raw reads and 14,614,375 clean reads miRNAs for siCDR1as were generated respectively (Table 2). On the other hand, siDICER1 has an average of 128,604,571 and 120,968,505 mRNAs were obtained from raw and clean reads respectively (Table 3). Also, an average of 14,260,428 raw reads and 14,028,682 clean reads miRNAs for siDICER1 were acquired accordingly (Table 4).
To explore the expression relationship of genes between samples, the Pearson’s correlation coefficient (PCC) of mRNAs and miRNAs expression levels of siCDR1as-1, 2 and 3, as well as, siDICER1-1, 2, and 3 in SMSCs were calculated and used to generate a correlation chart. As shown, the correlation coefficient of the siCDR1as-1, 2, and 3 as well as, siDICER1; 1, 2, and 3 in SMSCs ranged from 0.96 to 0.99 (average of 0.98), indicating that the samples replicate very well biologically (Fig. 2a, b, c, and d).
Figure 2 siCDR1as and siDICER1 Pearson correlation chart. (a-d) The abscissa and the ordinate were the respective samples, and the abscissa and the ordinate of each patch represented the correlation of siCDR1as and siDICER1 samples. Importantly, two completely related genomes had a value of 1. The closer to 1 the relative value is, the larger the Pearson correlation coefficient (PCC) for the siCDR1as and siDICER1 samples; conversely, the closer to 0 the relative value was, the smaller the PCC between the siCDR1as and siDICER1 samples.
To verify the RNA-sequencing data, in extension samples of SMSCs with siDICER1 or siCDR1as, five downregulated as well as five upregulated mRNA and miRNAs identified were randomly selected and quantified for their expression levels using qRT-PCR. As shown in (Fig. 3b; Suppl. Table 2a), (Fig. 4b; Suppl. Table 2b), (Fig. 5b; Suppl. Table 2c) and (Fig. 6b; Suppl. Table 2d), all the differential expression tendency were confirmed.
Differentially expressed (DE) mRNAs and their functional enrichment in SMSCs transfected with siDICER1
The mRNA expression profiles were detected in SMSCs transfected with siDICER1 and siNC, and hierarchical clustering (Fig. 3a) was performed to show differential mRNA expression patterns among samples. Considering the mRNA expression profiling data, a total of 1,113 mRNAs were differentially expressed in SMSCs transfected with siDICER1 (p < 0.05), of which 686 mRNAs (568 known transcript) were downregulated, and 427 mRNAs (365 known) were upregulated. Some myogenic genes including myogenin (MyoG), myocyte enhancer factor 2D (MEF2D) [41], bone morphogenetic protein 4 BMP4 [42], E2F transcription factor 2 (E2F2), insulin-like growth factor binding protein 5 (IGFBP5), cysteine-rich 61 (CCN1) and angiopoietin-1 (ANGPT1) were downregulated.
In addition, KEGG pathway analysis was conducted based on the differentially expressed mRNAs (Q value < 0.05). The results showed that among the top 20 most enriched KEGG pathways of the downregulatory genes include PI3K-AKT signaling pathway, Focal adhesion, and ECM receptor interaction (Fig. 3c), which are all associated with muscle development. Moreover, considering the upregulatory genes (Suppl. Figure 3), none of the 20 enriched KEGG pathways was associated with muscle development. These data show that the DICER1 gene is involved in the formation of SMSCs.
Figure 3 Expression profile of mRNAs in siDICER1 and NC (SMSCs). (a) Microarray analysis for mRNAs was performed with RNA extracted from siDICER1 (n = 3) and NC (n = 3) SMSCs. Hierarchical cluster analysis of significantly differentially expressed mRNAs: bright green, under-expression; gray, no change; bright red, over-expression. (b) Ten differentially expressed representative mRNAs were validated in SMSCs siDICER1 and NC by qPCR (n = 10 per group). GAPDH was used as an internal control. (c) KEGG of the downregulated mRNAs with the top 20 enrichment. Bubble color and size correspond to the Q value and gene number enriched in the pathway. The rich factor indicates the ratio of the number of DEGs mapped to a certain pathway to the total number of genes mapped to this pathway.
De Mirnas In Smscs Transfected With Sidicer1
A total of 542 miRNAs were detected in SMSCs samples, among which 22 miRNAs consisting of 7 downregulated and 15 upregulated miRNAs (FC > 1.25 and padj < 0.05) were significantly interfered by Dicer1 (si-DICER1) (Fig. 4a). Notably, myomiRNAs including miR-1, miR-206, and miR-133a/b [43], were upregulated by the deficiency of DICER1. However, miR-133 family contributes to myoblast proliferation, but also prevents differentiation by inhibiting Serum Response Factor (SRF) [13]. miR-1290 was not detected while the expression of miR-135a was almost undetectable in SMSCs treated with siDICER1/siCDR1as or not. Intriguingly, except for those DE miRNAs mentioned above, miR-7 was the most highly expressed one was insignificantly affected by siDICER1 (Table 5).
Furthermore, using TargetScan, RNAhybrid, and miRanda, a total of 7,194 targets mRNAs were predicted as targets of those upregulated miRNA genes whilst 1,686 target mRNAs of the downregulatory miRNAs were screened. All the mRNAs for the upregulated-miRNAs were enriched in the KEGG pathway such as, Focal adhesion, FoxO signaling pathway, MAPK signaling pathway, PI3K-AKT signaling pathway, Rap1 signaling pathway, and mTOR signaling pathway which are related to muscle development (Fig. 4c; Suppl. Figure 4). Moreover, this result is consistent with the functional enrichment for DE mRNAs mentioned before.
Figure 4 Differentially expressed miRNAs in siDICER1 and NC (SMSCs). (a) Sequencing analysis for miRNAs was performed from siDICER1 (n = 3) and NC (n = 3) of SMSCs. Hierarchical cluster analysis of significantly differentially expressed miRNAs: bright red, overexpression; white, no change; bright blue, under-expression. (b) Differential expression of ten representative miRNAs was validated in siDICER1 and NC of SMSCs by qPCR (n = 10 per group). (c) KEGG analysis of the upregulated miRNA-mRNA network. Bubble color and size correspond to the Q value and gene number enriched in the pathway. The rich factor indicates the ratio of the number of DEGs mapped to a certain pathway to the total number of genes mapped to this pathway.
De Mrnas In Smscs Transfected With Sicdr1as
From mRNA expression profiling data, a total of 789 mRNAs (Fig. 5a) were differentially expressed in SMSCs transfected with siCDR1as (p < 0.05), of which 401 mRNAs were downregulated (316 known protein-coding genes, 85 novel transcripts), and 389 mRNAs were upregulated (277 protein-coding genes, 112 novel transcripts). Some myogenic genes including signal transducer and activator of transcription 2 (STAT2) [44], ANGPT1 [45], intercellular adhesion molecule-1 (ICAM1), E2F2 [46], CCN1 [47], fibroblast growth factor receptor 1 (FGFR1) [48], and MEF2C [49] were downregulated according to the sequencing data.
Further, we performed KEGG pathway analysis with KOBAS software based on the differentially expressed mRNAs genes (Q value < 0.05). The results showed that among the top 20 most enriched pathways of the downregulated genes in siCDR1as samples, some well-known muscle-related pathways including PI3K-AKT signaling pathway signaling pathway, Focal adhesion, Rap1 signaling pathway, and MAPK signaling pathway (Fig. 5c) were identified. While those genes up-regulated by siCDR1as were mainly enriched in fatty acid biogenesis and metabolism (Suppl. Figure 5). These results indicate that just like DICER1, the downregulated mRNAs caused by the deficit of CDR1as are closely related to muscle development.
Figure 5 Expression profile of mRNAs in siCDR1as and NC (SMSCs). (a) Microarray analysis for mRNAs was performed with RNA extracted from siCDR1as (n = 3) and NC (n = 3) SMSCs. Hierarchical cluster analysis of significantly differentially expressed mRNAs: bright green, under-expression; gray, no change; bright red, over-expression. (b) Ten differentially expressed representative mRNAs were validated in SMSCs siCDR1as and NC by qPCR (n = 10 per group). GAPDH was used as an internal control. (c) KEGG of the downregulated mRNAs with the top 20 enrichment. Bubble color and size correspond to the Q value and gene number enriched in the pathway. The rich factor indicates the ratio of the number of DEGs mapped to a certain pathway to the total number of genes mapped to this pathway.
De Mirnas In Smscs Transfected With Sicdr1as
From mRNA expression profiling data, a total of 789 mRNAs (Fig. 5a) were differentially expressed in SMSCs transfected with siCDR1as (p < 0.05), of which 401 mRNAs were downregulated (316 known protein-coding genes, 85 novel transcripts), and 389 mRNAs were upregulated (277 protein-coding genes, 112 novel transcripts). Some myogenic genes including signal transducer and activator of transcription 2 (STAT2) [44], ANGPT1 [45], intercellular adhesion molecule-1 (ICAM1), E2F2 [46], CCN1 [47], fibroblast growth factor receptor 1 (FGFR1) [48], and MEF2C [49] were downregulated according to the sequencing data.
Further, we performed KEGG pathway analysis with KOBAS software based on the differentially expressed mRNAs genes (Q value < 0.05). The results showed that among the top 20 most enriched pathways of the downregulated genes in siCDR1as samples, some well-known muscle-related pathways including PI3K-AKT signaling pathway signaling pathway, Focal adhesion, Rap1 signaling pathway, and MAPK signaling pathway (Fig. 5c) were identified. While those genes up-regulated by siCDR1as were mainly enriched in fatty acid biogenesis and metabolism (Suppl. Figure 5). These results indicate that just like DICER1, the downregulated mRNAs caused by the deficit of CDR1as are closely related to muscle development.
Fig. 5 Expression profile of mRNAs in siCDR1as and NC (SMSCs). (a) Microarray analysis for mRNAs was performed with RNA extracted from siCDR1as (n=3) and NC (n=3) SMSCs. Hierarchical cluster analysis of significantly differentially expressed mRNAs: bright green, under-expression; gray, no change; bright red, over-expression. (b) Ten differentially expressed representative mRNAs were validated in SMSCs siCDR1as and NC by qPCR (n=10 per group). GAPDH was used as an internal control. (c) KEGG of the downregulated mRNAs with the top 20 enrichment. Bubble color and size correspond to the Q value and gene number enriched in the pathway. The rich factor indicates the ratio of the number of DEGs mapped to a certain pathway to the total number of genes mapped to this pathway.
Anchoring The Novel Core Mirnas Regulated By Cdr1as
Considering the major function of DICER1 on modulating miRNAs during myogenesis, and its closely positive effect on CDR1as as shown by expression of CDR1as as well as the enrichment results. First of all, we investigated the critical miRNAs that mediate the function of DICER1, using online tool MSigDB (http://www.broadinstitute.org/gsea/msigdb/index.jsp) and the DE mRNAs dataset caused by siDICER1. We found out 100 miRNAs were potentially targeted by these DE mRNA genes altered by siDICER1 (FDR q-value < 0.004), among which 11 miRNAs including miR-1, miR-199b-5p, miR-206, miR-27a-5p, miR-19b-3p, miR-30b-5p, miR-129-5p, miR-128-3p, miR-30e-5p, miR-27b-5p, and miR-424-5p were overlapped with those differentially expressed miRNAs detected by using miRNA-seq in SMSCs transfected with siDICER1. This indicates that these miRNAs may play important roles in mediating DICER’s function in myogenic differentiation of SMSCs.
Furthermore, six miRNAs including miR-1, miR-206, miR-424-5p, miR-30b-5p, miR-128-3p, and miR-19b-3p were enriched in targeting the upregulated mRNA genes caused by interfering CDR1as (siCDR1as) while only miR-199a was enriched in downregulated genes too (Fig. 8a; Suppl. Table 3). Additionally, miR-146a, miR-19a/b, miR-27a-5p, and miR-30e-5p were overlapped in DE miRNAs caused by si-DICER1 and si-CDR1as, suggesting that the following 11 miRNAs including miR-1, miR-206, miR-424-5p, miR-30b-5p, miR-128-3p, miR-19b-3p, miR-199a, miR-146a, miR-19a/b, miR-27a-5p, and miR-30e-5p, are much more likely to be novel target for CDR1as. Further, RNAhybrid was used to predict the complementary interactions between CDR1as and these miRNAs. The results indicate that CDR1as has potential binding sites for these miRNAs ( Fig. 8b).
Figure 8 Prediction of CDR1as-related miRNAs binding sites. (a) Predicted target sites of CDR1as-related miRNAs on mRNAs using Targetscan. (b) Binding sites of miRNAs found on CDR1as with the use of RNAhybrid.