CircAP2A2 acts as a ceRNA to participate in infantile hemangiomas progression by sponging miR‐382‐5p via regulating the expression of VEGFA

Abstract Background Increasing evidences reveal that circular RNAs (circRNAs) play crucial functions in cancer development. However, the expression pattern and roles of circRNAs in infantile hemangiomas (IH) remain unclear. Methods In this study, qRT‐PCR was performed to determine the expression of circAP2A2, miR‐382‐5p, and VEGFA in IH tissues and cell lines. Moreover, MTT assay, colony formation, transwell assay, and Western blot analysis were conducted to assess the function of circAP2A2 or miR‐382‐5p on cell proliferation, and migration in vitro, respectively. Also, dual luciferase assay was used to confirm the interactions among circAP2A2, miR‐382‐5p, and VEGFA. Results CircAP2A2 was confirmed to be highly expressed in IH. CircAP2A2 knockdown or miR‐382‐5p overexpression decreased the proliferation, colony formation, migration, and invasion of HemECs and HUVEC cells. Conclusion CircAP2A2 could promote proliferation and invasion of IH by regulating miR‐382‐5p/VEGFA axis.


| INTRODUC TI ON
Infantile hemangiomas (IH) are the most common soft-tissue tumors of infancy. The prevalence of infantile hemangioma in mature neonates is around 4.5%. 1 Infantile hemangiomas have been well characterized by its distinctive presentation, namely rapid growth during infancy followed by subsequent regression in early childhood. 2 This lesion is usually harmless; however, ~10% of lesions (those that compromise vision or vital systems like the airways) require treatment during the proliferative stage. Periocular lesions are of special concern as they pose a threat to the development of normal visual sensory function, leading to amblyopia in 43%-60% of cases. 3,4 Approximately 80% of lesions affect the head and neck. However, no available clinical biomarker is used to sensitively and specifically diagnose IH. 5 Thus, the pathogenesis and possible therapeutic targets of IH should be investigated.
In the past, circular RNAs (circRNAs) are a class of long, non-coding RNAs molecules. However, recent researchers have discovered that circRNAs can be translated. 6 circRNAs are characterized by covalently closed continuous loop which have no 5′-3′ polarity and contain no polyA tail. 7 Emerging evidences have indicated that thousands of endogenous circRNAs are present in mammalian cells. 8 One representative function of circRNA is that it could act as a mi-croRNA (miRNA) sponge to regulate the stability and/or translation efficiency of other RNAs, and such a circRNA is called a competing endogenous RNA (ceRNA). 9 Furthermore, circRNAs might play important roles in colon cancer, 10 pancreatic ductal adenocarcinoma, 11 gastric cancer 12 and breast cancer 13 and serve as diagnostic or predictive biomarkers of cancer.
Infantile hemangiomas is the result of dysregulation of both vasculogenesis and angiogenesis. 2 Despite extensive literature, the pathogenesis is still not clear. Recently, there is a most likely hypothesis that can explain the pathogenesis, involving hypoxic stress as the triggering signal, 14 by inducing overexpression of angiogenic factors (such as VEGF) via the HIFα pathway. 15 In response to VEGF overexpression, stem cells (expressing CD133), naturally present or recruited in fetal skin, proliferate and differentiate into immature endothelial cells (expressing CD31), but also pericytes (expressing SMA), dendritic cells (expressing factor XIIIa), and mesenchymal cells have the potential adipogenic activities. 16 Therefore, VEGFA (vascular endothelial growth factor A) plays a key role in pathological angiogenesis and angiogenesis of IH. 17 Furthermore, some researchers have demonstrated that circRNA could directly bind to miRNA and regulates VEGF expression in bladder carcinoma. 18 Considering the critical role of circRNAs and VEGF in cancer, it has been proposed that whether circRNAs could directly bind to miRNA by regulation of VEGFA expression in IH. Based on the previous microarray assay of circRNA in infantile hemangiomas, 19 we screened circAP2A2 and verified its high expression in IH by qRT-PCR. By using the bioinformatics software, miR-382-5p was predicted to be a target of circAP2A2, and VEGFA was suggested to be a putative target gene of miR-382-5p. Therefore, we speculated that the circAP2A2/miR-382-5p/VEGFA regulatory feedback circuit could affect the proliferation, migration, and invasion of IH cells. By exploring the hypothesis, we may gain insights into the pathogenesis of IH and provide new therapeutic targets for the treatment of IH.

| Human tissue specimens
Infantile hemangiomas samples were collected at the Ningbo Women and Children's Hospital from 2015 to 2019. The excised samples were immersed overnight in RNAstore (CWBIO) at 4°C and preserved at −80°C.

| RNA extraction and reverse transcription
Total RNAs were extracted from tissues or cultured cells using TRIzol reagent (Invitrogen), according to the manufacturer's instructions. The extracted RNAs were reverse transcribed into cDNAs using ReverTra Ace qPCR RT Kit (TOYOBO).

| Treatment with RNase R and RNA localization experiment
RNase R treatment was carried out for 15 minutes at 37°C using RNase R (Epicenter) 3 U/mg. For RT-PCR, the treated RNA was directly reverse transcribed into cDNA. Cytoplasmic and nuclear RNA isolations were performed using a PARIS Kit (Invitrogen), following the manufacturer's instructions. The extracted RNAs were reverse transcribed into cDNAs.
β-actin was 5′-CACCAGCATCTTTTCCAACC-3′ (forward primer) and 5′-AAGGCCGACTCTCCTACACA-3′ (reverse primer). The relative expression was shown as ∆C t value by subtracting the β-actin C t value from the circRNA C t value. Higher gene expression was indicated by a smaller ∆C t value.

| siRNAs, miRNA mimic, and transfection experiments
The circAP2A2 siRNAs and the mimic of miR-382-5p, as well as the non-targeting negative control, were purchased from GenePharma. Lipofectamine 2000 (Invitrogen) was used for transfection. Two circAP2A2 siRNAs with sequences of si-circAP2A2-1,

| MTT Assay
Cell proliferation was determined by Thiazolyl Blue (MCE) assay. In brief, 20 μL MTT (5 mg/mL in PBS) was added to the medium and incubated at 37°C for 4 hours. The supernatants were carefully aspirated, and 100 μL DMSO was added to each well. Absorbance values at 490 nm were measured.

| Colony formation assay
For the cell colony formation assay, stably transfected cells (1000 per well) were plated in six-well plates. After culturing for 7-9 days, cells were fixed with 70% ethanol and stained with 0.5% crystal violet solution. Colonies with more than 50 cells per colony were counted. All experiments were conducted three times in triplicate.

| Transwell migration and matrigel invasion assays
The migration and matrigel invasion assays were conducted by using transwell chamber (for migration assay) or transwell pre-coated matrigel chamber (for invasion assay) according to the manufacturer's protocol (Corning). The homogeneous single cell suspensions (4 × 10 4 cells/well for migration, 1 × 10 5 /well for invasion) were added to the upper chambers and incubated for 24 hours. The migration and invasion rates were quantified by counting the migration and invaded cells at least three random fields.

| Dual luciferase assay
Dual luciferase assay was conducted according to the protocol from the manufacturer (Promega). WT and mutant circAP2A2 and VEGFA were amplified and cloned into the luciferase reporter vector pGL3 (Promega). 293T cells were co-transfected with luciferase plasmids and miR-382-5p mimics. After 48 hours, the activities of firefly luciferase were normalized to that of Renilla luciferase.

| Western blot analysis
Total protein was separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE). The proteins in SDS-PAGE were transferred onto nitrocellulose membranes (GE Healthcare). The membrane was incubated with rabbit anti-VEGFA (ab46154, 1:1000 dilution; Abcam) primary antibodies overnight at 4°C and then with secondary antibody at room temperature for 1 hour. Proteins of interest were visualized using ECL Plus Western blotting Detection Reagents (Millipore).

| Statistical analysis
Unless otherwise stated, all data are shown as mean ± standard error of the mean (SEM). Statistical analysis was performed with twotailed t tests to compare mean values (Prism; GraphPad Software). A P-value less than .05 was considered statistically significant. *P < .05, **P < .01, ***P < .001.

| Screening of candidate circRNAs
We selected two aberrantly expressed circAP2A2 and circFAM53P as circRNAs. 19 In the hemangiomas specimens, circAP2A2 is expressed in hemangiomas by qRT-PCR, which is consistent with the microarray analysis. Also, circAP2A2 consists of two exons, derived from exons 15 and 16 of AP2A2 gen, with a length of 250 nt ( Figure 1A).

| CircAP2A2 is highly expressed in hemangiomas and is mainly localized in cytoplasm
We designed a reverse primer spanning the cyclization site and confirmed the single and reverse splicing target fragment of the amplified product by qRT-PCR and Sanger sequencing based on the circAP2A2 sequence ( Figure 1A).
In the second step, we used RNase R digestion to confirm the presence of circAP2A2, and circAP2A2 can tolerate RNase R digestion because RNase R can digest linear RNA and it has no polyA tail. The results of qRT-PCR showed that there was no significant change in the expression level of circAP2A2 before or after RNase R treatment in HUVEC and HemECs cell lines, but the expression of AP2A2 mRNA was significantly decreased after treatment with RNase R (P < .001) ( Figure 1B). Then, we used qRT-PCR to detect the expression of circAP2A2 in normal tissues of 20 pairs of hemangiomas. The expression level of circAP2A2 was more highly in hemangiomas compared with normal tissues (P < .05, Figure 1C).
Therefore, we confirmed the presence of circAP2A2 in hemangiomas and cell lines. Moreover, it has been found that circAP2A2 was mainly localized in the cytoplasm of HemECs and HUVEC cells by nuclear separation experiments, and by using both U6 (localized in the nucleus) and β-actin as the control (localized in the cytoplasm) ( Figure 1D).

| Knocking down circAP2A2 can significantly inhibit the growth, proliferation, invasion, and migration of Hemangioma cells
Previously, we have demonstrated that circAP2A2 is highly expressed in hemangiomas, suggesting that it may have potential carcinogenic functions. Therefore, we transfected circAP2A2 siRNA into HUVEC and HemECs. qRT-PCR results showed that the expression of circAP2A2 was significantly decreased in the si-circAP2A2 group transfected with circAP2A2 siRNA compared with the NC group (P < .05 and P < .01, Figure 2A). The effect of circAP2A2 on cell proliferation was examined by MTT assay, which showed that the cell proliferative ability of si-circAP2A2 group was significantly decreased compared with NC group (P < .01, showed that the cell growth ability of the si-circAP2A2 group was significantly decreased compared with the NC group (P < .05 and P < .01, Figure 2C), indicating that si-circAP2A2 can inhibit HUVEC and growth of HemECs cells. Moreover, the Transwell assay was used to detect cell migration and invasion. The results showed that the ability of cell migration and invasion was inhibited in the si-circAP2A2 group compared with the NC group (P < .05 and P < .01, Figure 2D), indicating that si-circAP2A2 can inhibit migration and invasion of HUVEC and HemECs cells.
Detection of circAP2A2 expression after transfection of miRNA mimics, respectively, it was found that only miR-382-5p was significantly affected by the negative regulation of circAP2A2 by other miRNAs ( Figure 3B). To further investigate whether circAP2A2 has a regulatory effect on miR-382-5p, we first upregulated

| CircAP2A2 is a "molecular sponge" of miR-382-5p to play a regulatory role
We used software to search for the potential binding site of miR-382-5p on the circAP2A2 base sequence, and also constructed a wild-type luciferase vector (PGL3-circAP2A2-WT) and a mutant luciferase vector (PGL3-circAP2A2-MUT) containing a partial sequence of circAP2A2 ( Figure 3D). Next, we transfected miR-382-5p mimics and circAP2A2 wild-type luciferase vectors simultaneously into HUVEC and HemECs cells, respectively. It has been found that the relative activity of dual luciferase was significantly downregulated, while the relative activity of the dual luciferase was not changed, followed by miR-382-5p mimics and circAP2A2 mutant luciferase vectors transfection. These experimental results indicate that circAP2A2 binds to circAP2A2, and circAP2A2 regulates the expression of miR-382-5p by "miRNA sponge action"( Figure 3E). Moreover, the correlation analysis of the specimens was performed. We detected the expression of miR-382-5p by qRT-PCR in 20 ( Figure 3F) pairs of fresh hemangiomas and corresponding adjacent tissues ( Figure 3G), and found that there were significantly low expression levels of miR-382-5p in hemangiomas. Statistical analysis showed that miR-382-5p is negatively correlated to circAP2A2 in this 20 pairs of tissues ( Figure 3G).

| VEGFA is a direct target of miR-382-5p
We found that miR-382-5p has potential interaction with VEGFA  breast cancer, 32 and prostate cancer, 33 and is related to the chemoresistance of cancer cells. 34 Furthermore, miR-1307 contributes to prostate cancer proliferation by targeting FOXO3A. 35 In our study, miR-382-5p expression was confirmed to be regulated in HUVEC cells. Moreover, we confirmed the presence of a binding site for circAP2A2 and miR-382-5p by luciferase assay, suggesting that cir-cAP2A2 can act as a miR-382-5p sponge to regulate downstream mRNA of miR-382-5p.
VEGFA levels are significantly elevated in the proliferative phase of infantile hemangiomas, and gradually decrease in the regression phase, which is highly correlated with the development of infantile hemangiomas. 36 Also, it is well known that the VEGFA-mediated VEGFA/VEGFR pathway is an important signaling pathway involved in the development of hemangiomas, and also VEGFA interacts with HIF-1α to regulate angiogenesis. 37 In our research, bioinformatics analysis revealed that VEGFA is a miR-382-5p target. Also, it has been confirmed that the inhibition of circAP2A2 could downregulate VEGFA expression in HemECs and HUVEC cells in vitro. In the meanwhile, the changed expression of VEGFA caused by circAP2A2 could be reversed by miR-382-5p inhibitor, which is in accordance with the conversion of enhanced cell proliferation in HemECs and HUVEC cell lines.
Therefore, our studies indicated that circAP2A2 can regulate the occurrence of hemangiomas through the circAP2A2/miR-382-5p/VEGFA axis.
In conclusion, our study confirms upregulation of circAP2A2 in HC, and also proposes a key role of circAP2A2 in IH, functioned as ceRNA by regulating the expression of VEGFA via sponge miR-382-5p and participating in IH progression. The study may provide a new potential target for IH diagnosis and treatment.

ACK N OWLED G M ENTS
Project support was provided by the Medical Health Science and