Hsa_circ_0079474 is upregulated in IUA patients
Since EMT is a pivotal event in the progression of fibrosis, the EMT hallmarks including epithelial cell markers (E-cadherin and CK-18) and mesenchymal cell markers (α-SMA and VIM) were determined using qRT-PCR (Fig. 1A), Western blot (Fig. 1B) and immunohistochemistry (IHC) (Fig. 1C) techniques in both IUA tissues and normal tissues. The results, as depicted in Fig. 1a-1c, demonstrated a reduction of the levels of E-cadherin and CK-18 whereas the levels of α-SMA and VIM were obviously enhanced in IUA tissues compared to normal tissues, suggesting involvement of EMT in IUA progression. To identify the circRNAs expression profiles involved in IUA progression, a circRNA microarray analysis on 3 paired IUA samples and adjacent normal samples under hysteroscopy was performed. The criteria for selecting differentially expressed circRNAs was absolute fold change > 2.0 and p value less than 0.05. Among them, 348 differentially expressed circRNAs were reported in circBase (www.circbase.org/) and 235 differentially expressed circRNAs were detected in all samples including 72 upregulated circRNAs and 163 downregulated circRNAs. A heatmap was constructed to show different expressed circRNAs (Fig. 1D). Furthermore, top 10 circRNAs were selected for further verification by qRT-PCR by using designed divergent primers and found that only hsa_circ_0132015 and hsa_circ_0079474 were confirmed. We further found that hsa_circ_0079474 had more significant upregulation in IUA tissues and more abundance in IUA and normal tissues than hsa_circ_0132015 which was selected for further analysis (Fig. 1E). Furthermore, the expressions of E-cadherin and CK-18 were decreased while α-SMA and VIM were increased in TGF-β1-induced EECs than control EECs using qRT-PCR (Fig. 1F), Western blot (Fig. 1G) and immunocytochemistry (ICC) (Fig. 1H). The expression of hsa_circ_0079474 was also higher in TGF-β1-induced EECs than control EECs (Fig. 1I).
Characteristics of hsa_circ_0079474
Analysis of circBase showed that hsa_circ_0079474 is originated from back splicing of exons 6-13 of human gene DGKB (also known as hsa_circ_DGKB-009) which was located on chromosome 7, with the length 812 nt (Fig. 2A). To confirm the results of the expression of hsa_circ_0079474 in IUA tissues, Sanger sequencing on the PCR products of hsa_circ_0079474 was performed. Sanger sequencing verified the back-splicing junction with the expected size and predicted splicing site in the PCR products (Fig. 2B). Agarose gel electrophoresis of reverse transcribed RNA (cDNA) and genomic DNA (gDNA) were performed by convergent and divergent primers for hsa_circ_0079474. The results demonstrated that the divergent primers could amplify products from cDNA but not gDNA (Fig. 2C). As a confirmation of the stability of hsa_circ_0079474, the total RNA samples were treated with RNase R. The results of qRT-PCR displayed that the levels of linear DGKB (DGKB mRNA) was decreased significantly under RNase R treatment, whereas circDGKB-009 was resistant to the digestion of RNase R (Fig. 2D). The stability of hsa_circ_0079474 was further confirmed by treating EECs with Actinomycin D (a transcription inhibitor) for different periods of time. Results disclosed that the linear DGKB mRNA level decreased gradually with time but the level of circDGKB-009 was more stable and resistant to Actinomycin D administration (Fig. 2E). In addition, hsa_circ_0079474 was mainly located in cytoplasm measured by cytoplasmic and nuclear fractionation assay, with the positive control U6 in nucleus and GAPDH in cytoplasm, respectively (Fig. 2F). In brief, these findings indicated that hsa_circ_0079474 is a circRNA which is localized in the cytoplasm, has good stability, and may involve in the occurrence and development of IUA.
Hsa_circ_0079474 promoted the progression of EMT
To explore the functional roles of hsa_circ_0079474, gain-of-function and loss-of-function assays were performed, respectively. Firstly, overexpression plasmid of hsa_circ_0079474 was constructed and transfected to Ishikawa cells. The level of hsa_circ_0079474 was increased more than 60,000-fold in hsa_circ_0079474-overexpression of Ishikawa cells compared with control by qRT-PCR (Fig. 3A). Secondly, three specific siRNAs targeting hsa_circ_0079474 were transfected into Ishikawa cells to silence hsa_circ_0079474 expression. The results demonstrated that si-hsa_circ_0079474-3 had the highest efficiency and was selected for further functional trials (Fig. 3B).
Subsequently, CCK-8 assays and Edu staining assays were conducted to determine the impact of hsa_circ_0079474 on the proliferation of Ishikawa cells. The findings disclosed that hsa_circ_0079474 overexpression led to improvement of the proliferation of Ishikawa cells whereas silence of hsa_circ_0079474 suppressed the proliferation of Ishikawa cells (Fig. 3C, D). Flow cytometry analysis elucidated that the up-regulation of hsa_circ_0079474 resulted in a decrease in the proportion of the G0/G1 phase but an increase in the percentage of the S phase of the cell cycle in Ishikawa cells. Conversely, down-regulation of hsa_circ_0079474 resulted in elevation in the percentage of the G0/G1 phase and a reduction in the proportion of the S phase of the cell cycle in Ishikawa cells, demonstrating that hsa_circ_0079474-overexpressing promoted cell cycle progression while down-regulation of hsa_circ_0079474 led to cell cycle arrest (Fig. 3E). Nevertheless, both up-regulation and down-regulation of hsa_circ_0079474 were found to have no significant impact on the proportion of G2/M phase of the cell cycle in Ishikawa cells. To further investigate the effects of hsa_circ_0079474 on EMT, qRT-PCR, Western blot and ICC were employed. The obtained results validated that overexpression of hsa_circ_0079474 led to a reduction in the levels of epithelial cell markers E-cadherin and CK-18 while increasing the levels of mesenchymal cell markers α-SMA and VIM. Conversely, knockdown of hsa_circ_0079474 yielded the opposite outcomes (Fig. 3F-H). In conclusion, hsa_circ_0079474 showed effects of promotion of cell proliferation, acceleration of cell cycle progression and improvement of EMT.
Hsa_circ_0079474 served as a sponge for miR-630
It was postulated that hsa_circ_0079474 exerts its function through the mechanism of miRNA sponges, which is a prevalent mechanism for circRNAs. Circular RNA Interactome and CircBank datasets were employed to predict the potential binding miRNAs of hsa_circ_0079474. As exhibited in Venn diagram in Fig. 4A, three candidate miRNAs (hsa-miR-630, hsa-miR-892a and hsa-miR-935) were identified to have putative binding sites on hsa_circ_0079474. The miRanda dataset was performed to suggest that miR-630, miR-892a and miR-935 could potentially bind to hsa_circ_0079474, with binding scores of 162, 156 and 151, and a total free energy of -14.39 kCal/Mol, -14.32 kCal/Mol and -16.67 kCal/Mol, respectively. These findings indicate that miR-630 is the most likely to bind hsa_circ_0079474. The expression levels of the aforementioned three miRNAs were subsequently assessed in both IUA and normal tissues. It was observed that miR-630 exhibited a significant decrease in IUA tissues compared to normal tissues whereas no significant differences were found in miR-892a and miR-935 (Fig. 4B). Additionally, the levels of miR-630 were determined in Ishikawa cells with overexpression of hsa_circ_0079474. The results indicated a significant decrease in miR-630 expression (Fig. 4C). Consequently, miR-630 was selected for further experimentation. The predicted binding sites of hsa_circ_0079474 to miR-630 were also displayed (Fig. 4D) and dual-luciferase reporter assay was performed to confirm the interaction between hsa_circ_0079474 and miR-630. The wild-type hsa_circ_0079474 sequences, which included potential miR-630 binding sites, as well as a mutation version, were inserted into the psiCHECK2 luciferase reporter. Results manifested that miR-630 remarkably inhibited the luciferase activity while miR-630 inhibitor notably enhanced luciferase activity in hsa_circ_0079474-wt reporter while not in the hsa_circ_0079474-mut reporter. This suggested that hsa_circ_0079474 has the ability to bind to miR-630 (Fig. 4E). Additionally, to understand the molecular mechanisms of hsa_circ_0079474 in IUA progression, subcellular localization was conducted in EECs by FISH. As presented in Fig. 4F. FISH results exhibited co-localization of miR-630 and hsa_circ_0079474 in the cytoplasm of EECs. RIP assays demonstrated that hsa_circ_0079474 was remarkably immunoprecipitated by Ago2 antibody in miR-630-transfected EECs compared to miR-NC-transfected EECs (Fig. 4G). These results confirmed that hsa_circ_0079474 served as a sponge for miR-630 in EECs.
MiR-630 was downregulated in IUA and alleviated EMT process
To further explore the involvement of miR-630 during IUA development, EECs and TGF-β1-induced EECs were collected and subjected to qRT-PCR to evaluate the expression of miR-630. Results manifested that the expression of miR-630 were lower in both IUA tissues and TGF-β1-induced EECs compared to normal tissues and EECs, respectively. (Fig. 5A). In addition, transfection of miR-630 mimics and miR-630 inhibitors into EECs confirmed the upregulation and downregulation of miR-630 (n=3) (Fig. 5B). Subsequently, CCK-8 and Edu staining assays were employed to observe the impacts of miR-630 on the property of proliferation of EECs. Results revealed that miR-630 inhibited the proliferation of EECs whereas suppression of miR-630 promoted the proliferation of EECs (Fig. 5C, D). Flow cytometry analysis further exhibited that miR-630 was found to have a restraining effect on the cell cycle while inhibition of miR-630 induced cell cycle progression (Fig. 5E). The results of expression of EMT markers proved that miR-630 enhanced the levels of epithelial cell markers (E-cadherin and CK-18) while decreasing the expression of mesenchymal cell markers (α-SMA and VIM) (Fig. 5F-H). In summary, miR-630 yielded properties of inhibition of cell proliferation, cell cycle progression and EMT.
MiR-630 could directly target YAP1
To assess the potential target genes of miR-630, various datasets including miRDB, miRWalk, miRTarbase, RNACentral and TargetScan were utilized to predict. As a result, only one candidate gene (YAP1) was filtrated for further analysis (Fig. 6A). The two predictive binding sites on YAP1-3’UTR for miR-630 by miRanda were presented in Fig. 6B. To validate the binding between miR-630 and YAP1, dual-luciferase report assays was further conducted. Wild-type (Wt) YAP1-3’UTR and miR-630 binding site Mut luciferase reporter plasmids were transfected along with miR-630 mimic and miR-630 inhibitor. Results indicated that the relative luciferase activity of wt YAP1-3’UTR group was marked declined when transfected miR-630 mimic but remarkably increased when transfected miR-630 inhibitor. Conversely, no significantly differences of the relative luciferase activity were observed in mut YAP1-3’UTR groups (Fig. 6C). In addition, the expression of YAP1 was evaluated using qRT-PCR, Western blot and ICC following transfection with miR-630 mimic and miR-630 inhibitor and indicated that elevation of miR-630 dramatically inhibited the expression of YAP1 whereas downregulation of miR-630 distinctly improved the expression of YAP1 (Fig. 6D-F). These results confirmed that miR-630 could target YAP1 directly.
YAP1 was increased in IUA and promoted EMT
To investigate the underlying role of YAP1 in IUA development, samples of IUA tissues, normal tissues, EECs and TGF-β1-induced EECs were collected and subjected to qRT-PCR, Western blot and IHC/ICC analysis. Results concluded that the level of YAP1 in IUA tissues and TGF-β1-induced EECs was significantly higher than adjacent normal tissues and EECs respectively (Fig. 7A-C). To further explore the role of YAP1 in IUA development, YAP1-overexpression plasmid and siRNAs were transfected into EECs. The efficiency of transfection was confirmed using qRT-PCR. The transfection of YAP1-overexpression plasmid into EECs resulted in an increased level of YAP compared to the control group. (Fig. 7D). Next, three specific siRNAs targeting YAP1 were designed and transfected into EECs to silence YAP1 expression. It was illustrated that si-YAP1-3 exhibited the highest efficiency and was selected for further experiments (Fig. 7E). Further functional trials demonstrated that YAP1 overexpression resulted in a dramatically improvement of cell proliferation, cell cycle progression and EMT while knockdown of YAP1 disclosed opposite effects (Fig. 7F-K).
Hsa_circ_0079474 promotes IUA progression via miR-630/YAP1 axis
Furthermore, our study proposed to investigate the involvement of miR-630 in the regulation of hsa_circ_0079474 in Ishikawa cells. It was observed that the presence of hsa_circ_0079474 significantly enhanced cell proliferation, facilitated cell cycle progression, and promoted EMT. However, these effects were effectively counteracted by the restoration of miR-630 expression (Fig. 8A-F). Conversely, the depletion of hsa_circ_0079474 in Ishikawa cells was rescued by the administration of miR-630 inhibitor (Fig. 8G-L). The findings elucidated that hsa_circ_0079474 exerted its effects on Ishikawa cells, at least in part, by acting as a sponge for miR-630.
Hsa_circ_0079474 promoted IUA via miR-630 in vivo
The expression of E-cadherin and CK-18 were decreased while the expression of α-SMA and VIM increased compared with sham group by qRT-PCR and immunohistochemical staining, confirming that EMT was involved in the progression of IUA in rat (Fig.9A, B). Masson staining was applied to determine the degree of fibrosis. In IUA group, more fibrotic area was found than that of control group, suggesting that the IUA rat model was established successfully (Fig. 9B). It was further validated that IUA+AAV-hsa_circ_0079474 group notably decreased levels of E-cadherin and CK-18, accompanied by increased levels of α-SMA and VIM, as well as a greater presence of collagen fibers compared to the IUA group by qRT-PCR, IHC and Masson staining (Fig.9c-9d). Conversely, injection of si-hsa_circ_0079474 yielded contrasting results of AAV- hsa_circ_0079474 (Fig. 9C, D). In IUA+miR-630 agomir group, intrauterine injection of miR-630 agomir demonstrated higher levels of E-cadherin and CK-18 whereas lower levels of α-SMA and VIM, as well as lower rate of fibrotic area, which disclosed contrasting effects of hsa_circ_0079474 using qRT-PCR, IHC and Masson staining (Fig. 9E-F). Furthermore, administration of miR-630 antagomir exhibited opposite effects (Fig. 9E-F). However, these effects of AAV- hsa_circ_0079474 were at least partly reversed by administration of miR-630 agomir (Fig. 9G, H). Furthermore, the effects of hsa_circ_0079474-siRNA-chol were reversed at least partially by miR-630 antagomir (Fig. 9G, H). These findings displayed that hsa_circ_0079474 improved IUA via miR-630 in vivo.