CircRNA_0079586 and circRNA_RanGAP1 are involved in the pathogenesis of intracranial aneurysms rupture by regulating the expression of MPO

Several circRNAs have been reported to be dysregulated in human endothelial cells through sponging miRNAs. Previous reports demonstrated that MPO not only contributed to the formation and rupture of cerebral aneurysm but was also correlated with the degenerative remodeling predisposition to saccular intracranial aneurysm wall rupture, although its underlying mechanisms remain to be explored. Microarray screening was performed to compare the differential expression of circRNAs in the endothelial cells collected from UIAs and RIAs patients. Luciferase assays were used to explore the regulatory relationship between circRNAs and miRNAs, and between miRNAs and their target genes. Microarray screening analysis found a batch of up-regulated circRNAs in the endothelial cells harvested from RIAs patients, including circRNA-0079586 and circRNA-RanGAP1. Luciferase assays revealed the suppressive role of miR-183-5p/miR-877-3p in the expression of circRNA-0079586/circRNA-RanGAP1/MPO. And the expression of circRNA-0079586 and circRNA-RanGAP1 was respectively suppressed by the overexpression of miR-183-5p and miR-877-3p. And both the transfection of miR-183-5p and miR-877-3p mimics suppressed the relative expression level of MPO mRNA. The expression of circRNA-0079586, circRNA-RanGAP1 and MPO was significantly activated in the endothelial cells collected from RIAs patients when compared with UIAs patients, whereas the expression of miR-183-5p and miR-877-3p was remarkably suppressed in the endothelial cells collected from RIAs patients when compared with UIAs patients. We further altered the expression of circRNA-0079586 and circRNA-RanGAP1 using siRNA and overexpression in HUVECS, and the expression of circRNA-0079586 and circRNA-RanGAP1 was significantly and negatively correlated with the expression of miR-183-5p and miR-877-3p, but positively correlated with the expression of MPO under different conditions. In this study, we established two MPO-modulating signaling pathways of circRNA_0079586/miR-183-5p/MPO and circRNA_RanGAP1/miR-877-3p/MPO. These two signaling pathways are involved in the pathogenesis of intracranial aneurysms rupture.


Scientific Reports
| (2021) 11:19800 | https://doi.org/10.1038/s41598-021-99062-w www.nature.com/scientificreports/ consequences such as death as well as disability 4,5 . On the contrary, current IA treatments are associated with significant risk. Therefore, the accurate evaluation of IA rupture is vital for the clinician to consider surgery risk as well as the danger of natural IA rupture 6 . For many years, methods such as geometry, size, as well as location analyses of aneurysm have been used to study the risk factors of IA rupture 7,8 . While large size aneurysms have been thought to be associated with a higher risk of rupture, latest researches have actually shown that a lot of ruptured aneurysms are small in dimensions 8 . Shear stress has been considered a significant factor for platelet activation, while other factors like levels of collagen, ADP, as well as fibrinogen also play a role [9][10][11] . Under high shear stress, platelets can also be more easily activated for a brief period of time 12,13 . In terms of how wall shear stress (WSS) causes aneurysm rupture, it is strongly believed that the absolute value of WSS alone may not be used to reasonably forecast the onset of aneurysm rupture, because it does not involve any type of directional information. Physiological researches have also revealed that the endothelial cells of blood capillaries can respond to the changes in the direction as well as magnitude of WSS, and such responses may cause the onset of aneurysm re-modeling and additional growth and rupture of the aneurysm 7,14,15 .
Circular RNAs (circRNAs) belong to a subfamily of endogenous non-coding RNAs (ncRNAs) and are mainly consisted of exonic transcripts produced via back splicing 16 . Different from linear RNAs, circRNA is a covalently closed continuous loop without 5′-3′ polarity or a polyadenylated tail which commonly originates from protein-coding genes and complete exons 17 . And it was reported that eukaryotic circRNAs are mainly produced during splicing, catalyzed by spliceosomal machinery or by groups I and II ribozymes 18 . Moreover, circRNA is also known as miRNA sponge, which regulates the expression of miRNAs. Recent research has actually shown that circRNAs are richly expressed in cells to serve as a necessary regulator in transcriptional as well as posttranscriptional gene expression 19 . In colon cancer patients, circRNA CDR1as is significantly overexpressed in tumor tissues, while the silencing of CDR1as hinders the development of tumors by targeting miRNA-7 20 .
In a previous report, it was failed to show that MPO is linked to the clinical risk of IA rupture 21 . Therefore, it was proposed that the presence of MPO might indicate a more delicate wall of aneurysm that is prone to IA rupture. The level of MPO can be identified using enhanced MRI or blood samples collected from IA fundus. It was also suggested that MPO might be used as a biomarker to examine the risk of IA rupture 22,23 . Also, the measurement of MPO might improve the assessment of cardiovascular diseases. Moreover, it was suggested that myeloperoxidase (MPO) played a key role in inflammation as well as oxidative stress, and is involved in the onset and development of atherosclerosis. And circRNAs have been reported to take key regulatory roles in pathological biological processes such as inflammation and cell proliferation 24 . Besides, miR-183-5p was demonstrated to suppress inflammatory responses in intracerebral hemorrhage mice 25 , while miR-887-3p was found to lead to dysregulated inflammation and enhance lung injury by increasing the cytokine production in ICU patients with severe sepsis 26 .
Several circRNAs have been reported to be dysregulated in human endothelial cells 27,28 . Specially, cir-cRNA_0079586 was found to accelerate the progression of glioma via interacting with miR-183-5p, while circRNA_RanGAP1 was found to facilitate gastric cancer invasion and metastasis via interacting with miR-877-3p 29,30 . Moreover, previous reports demonstrated that MPO not only contributed to the formation and rupture of cerebral aneurysm, but was also correlated with the degenerative remodeling predisposition to saccular intracranial aneurysm wall rupture 22,31 . In this study, we aimed to study the molecular mechanism underlying the role of MPO and its potential signaling pathways.

Materials and methods
Human subjects sample collection. In this study, we recruited 34 IA patients and divided them into two groups based on their status of IA: RIAs (ruptured intracranial aneurysms, N = 18) and UIAs (unruptured intracranial aneurysms, N = 16). Intracranial aneurysm tissue samples were collected after removal surgery from patients in both groups following previous published methods 32,33 . Endothelial cells were isolated from surgically removed intracranial aneurysm. The clinical features including sex, age, hypertension, smoking history, lesion size were collected from the patients in the two groups and compared using Student's t tests. The Human Research Ethics Committees of the First Affiliated Hospital of Jinzhou Medical University has approved this research. All methods were performed in accordance with the last vision of the Declaration of Helsinki. Written informed consent was obtained from all patients before the study.
Primary culture of endothelial cells from human carotid artery. EC primary culture (Cell applications, San Diego, CA) was derived from human carotid arteries, and the characteristics of the EC primary culture were confirmed by immunohistochemistry.
Loading shear stress. To measure the loading shear stress, the primary ECs were cultured on glass slides coated with gelatin overnight and then loaded with a shear stress of 0.05 and 3.0 Pa. After 24 h of culturing under the loaded shear stress, the primary ECs were harvested for other analyses. Additionally, the primary ECs were also subjected to the treatment with a turbulent flow as previously described 27 . Genotyping by direct sequencing. Total RNA of the samples was first extracted by using a Trizol reagent (Invitrogen, Carlsbad, CA) according to the directions of the reagent manufacturer. Then, the extracted total RNA was subsequently processed using DNase I (Takara, Dalian, China) according to the directions of the reagent manufacturer to get rid of genomic DNA contaminants. In the next step, equal volumes of total RNA extracted from each sample were subject to deep sequencing, which was carried out by BGI (Shenzhen, China) by making use of an Illumina Solexa platform. www.nature.com/scientificreports/ Microarray analysis. To examine the circRNA expression profile in collected tissue and cell samples, the total RNA, which also contained circRNAs and miRNAs, in each sample was isolated and studied using the Affymetrix GeneChip miRNA 2.0 microarray (Affymetrix, Santa Clam Clara, CA) according to the directions of the microarray manufacturer. Based on the results, circRNAs identified with changed expression were divided into 2 groups, i.e., a group of circRNAs with upregulated expression and a group of circRNAs with downregulated expression. The group of circRNAs with upregulated expression included 9 circRNAs: circRNA_0004543 (2.6fold increase in expression), circRNA_0079586 (4.1-fold increase in expression), circRNA_0000231 (3.5-fold increase in expression), circRNA_0003204 (2.7-fold increase in expression), circRNA_0454542 (3.3-fold increase in expression), circRNA_0091822 (3.5-fold increase in expression), circRNA_ RanGAP1 (3.9-fold increase in expression), circRNA_0943034 (3.1-fold increase in expression), and circRNA_320434 (1.9-fold increase in expression). The group of circRNAs with downregulated expression included 6 circRNAs: circRNA_0003492 (2.1-fold decrease in expression), circRNA_0011032 (2.5-fold decrease in expression), circRNA_0004264 (3.2fold decrease in expression), circRNA_0002331 (3.7-fold decrease in expression), circRNA_0004528 (2.8-fold decrease in expression), and circRNA_0000345 (2.6-fold decrease in expression).
RNA isolation and real-time PCR. RNA isolation and real-time PCR were carried out to study the expression of circRNA_0079586, circRNA_RanGAP1, miR-183-5p, miR-877-3p, and MPO mRNA in each sample. In brief, total RNA in each sample was extracted using a mirVana RNA extraction assay kit (Ambion, San Diego, CA) according to the directions of the assay kit manufacturer. Then, the integrity and content of extracted RNA was evaluated by using an Agilent 2100 BioAnalyzer (Agilent, San Jose, CA) according to the directions of the instrument manufacturer. In the next step, an NCode VILO cDNA synthesis assay kit (Qiagen, Germantown, MD) was used according to the directions of the assay kit manufacturer to create cDNA templates for real time quantitative PCR analysis, which was performed by using a QuantiFast SYBR Green master mix (Qiagen, Germantown, MD) according to the directions of the assay kit manufacturer to calculate the relative expression of circRNA_0079586, circRNA_RanGAP1, miR-183-5p, miR-877-3p, and MPO mRNA in each sample using the 2 −ΔΔCT approach. The real time quantitative PCR was carried out on a LightCycler 480 equipment (Roche, Basel, Switzerland) according to the directions of the equipment manufacturer.
Cell culture and transfection. HUVEC cells were acquired from ATCC and cultured in a Dulbecco's modified Eagle's medium (DMEM, Gibco, Thermo Fisher Scientific, Waltham, MA) containing 4.5 g/l of glucose, 10% fetal bovine serum as well as streptomycin and streptomycin (Sigma Aldrich, St. Louis, MO). The culture was done in a 37° C humidified environment of 5% carbon dioxide and 95% air. When HUVEC cells were confluent, they were used to establish 5 cellular models as below. In cellular model I, HUVEC cells were divided into 5 groups, i. Western blot analysis. Total protein was isolated from each sample by using a RIPA buffer (Sigma Aldrich, St. Louis, MO) according to the directions of the buffer manufacturer. Then, after 15 min of centrifugation at 13,400 g, the protein supernatant was quantified by using a BCA protein assay (Sigma Aldrich, St. Louis, MO) according to the directions of the assay kit manufacturer to determine protein concentration. In the next step, 10 µg of total protein from each sample was separated on a 10% SDS-PAGE gel and blotted onto a PVDF membrane (Roche, Indianapolis, IN), which was then blocked at room temperature for 2 h in a TBST buffer containing 5% goat milk, and then probed in sequence with primary anti-MPO monoclonal antibody and HRPconjugated IgG secondary antibody (Santa Cruz Biotechnology, Dallas, TX) according to the incubation conditions suggested by the antibody manufacturer. Finally, the target protein bands were identified by using an enhanced chemiluminescence reagent (ECL) (Sigma Aldrich, St. Louis, MO) according to the directions of the reagent manufacturer.

Statistical analysis.
All experiments were done ≥ 3 times. The results were ex-shown as mean ± SEM. Statistical analysis was done in SPSS 17.0 (SPSS, Chicago, IL) by utilizing the Student's t-tests for inter-group comparisons. P < 0.05 was deemed statistically significant.

Results
Clinical features of the patients. We recruited IA (intracranial aneurysm) patients and divided them into two groups based on the status of IA: RIAs (ruptured intracranial aneurysms) and UIAs (unruptured intracranial aneurysms). Tissue samples were collected after surgery from patients in both groups. The clinical features including sex, age, hypertension, smoking history, lesion size were compared between the two groups. No obvious difference was found between patients in the two groups (Table 1).

Shear stress reshaped the expression of circRNAs in the endothelial cells of IA patients.
Previous reports revealed that circRNAs played crucial roles in the pathogenesis of IA. We collected the endothelial cells from IA tissue samples and subjected them to microarray screening to examine the differentially expressed circRNAs between RIAs and UIAs. The expression of circRNA_0004543, circRNA_0079586, circRNA_0000231, circRNA_0003204, circRNA_0454542, circRNA_0091822, circRNA_RanGAP1, circRNA_0943034 and cir-cRNA_320434 was remarkably elevated in RIAs when compared with UIAs. On the contrary, the expression of circRNA_0003492, circRNA_0011032, circRNA_0004264, circRNA_0002331, circRNA_0004528 and cir-cRNA_0000345 was notably suppressed in RIAs (Fig. 1A). Furthermore, we treated HUVECs with WSS (wall shear stress) and TF (turbulent flow), followed by circRNAs screening to find the differentially expressed cir-cRNAs in HUVECs treated under the two conditions. The expression of circRNA_0008748, circRNA_0032813, circRNA_0000221, circRNA_0079586, circRNA_0019832, circRNA_RanGAP1, circRNA_0000911 and cir-cRNA_00009321 was significantly enhanced in TF-treated HUVECs when compared with WSS-treated www.nature.com/scientificreports/ HUVECs, while the expression of circRNA_0009035, circRNA_0000732, circRNA_0009841, circRNA_0000948 and circRNA_0000322 was apparently repressed in TF-treated HUVECs when compared with WSS-treated HUVECs (Fig. 1B).
Luciferase assays revealed the regulatory relationship between circRNA_0079586/miR-183-5p, circRNA_RanGAP1/miR-877-3p, miR-183-5p/MPO, and miR-877-3p/MPO. Sequence analysis showed that miR-183-5p could potentially target circRNA-0079586, so luciferase vectors containing wild type and mutant circRNA-0079586 were established and transfected into HUVEC cells with miR-183-5p. The luciferase activity of wild type circRNA-0079586 was significantly suppressed by miR-183-5p, while the luciferase activity of mutant circRNA-0079586 remained unchanged when compared with the control (Fig. 2A). Sequence analysis showed that miR-877-3p could potentially target circRNA-RanGAP1, so luciferase vectors containing wild type and mutant circRNA-RanGAP1 were established and transfected into HUVEC cells with miR-877-3p. The luciferase activity of wild type circRNA-RanGAP1 was significantly suppressed by miR-877-3p, while the luciferase activity of mutant circRNA-RanGAP1 remained unchanged when compared with the control (Fig. 2B). Sequence analysis showed that miR-183-5p could potentially target MPO, so luciferase vectors containing wild type and mutant MPO were established and transfected into HUVEC cells with miR-183-5p. The luciferase activity of wild type MPO was significantly suppressed by miR-183-5p, while the luciferase activ-  (Fig. 2C). Sequence analysis showed that miR-877-3p could potentially target MPO, so luciferase vectors containing wild type and mutant MPO were established and transfected into HUVEC cells with miR-877-3p. The luciferase activity of wild type MPO was significantly suppressed by miR-877-3p, while the luciferase activity of mutant MPO remained unchanged when compared with the control (Fig. 2D). And to further verify the regulatory association between the miRNA and its possible target gene, we observed the relative expression of potential target genes in HUVEC cells overexpressing miR-183-5p or miR-887-3p. Accordingly, the relative expression of circRNA-0079586 ( and circRNA-RanGAP1 siRNAs were transfected into TF-treated HUVECs to evaluate its effect on the expression of circRNA-0079586, circRNA-RanGAP1, miR-183-5p, miR-877-3p and MPO. The expression of cir-cRNA-0079586 was dramatically suppressed in TF-treated HUVECs by circRNA-0079586 siRNA (Fig. 4A). The expression of circRNA-RanGAP1 was dramatically suppressed in TF-treated HUVECs by circRNA-RanGAP1 siRNA (Fig. 4B). TF treatment remarkably down-regulated the expression of miR-183-5p and miR-877-5p when compared with WSS treatment. The expression of miR-183-5p was significantly activated in TF-treated HUVECs by circRNA-0079586 siRNA (Fig. 4C). The expression of miR-877-3p was significantly activated in TFtreated HUVECs by circRNA-RanGAP1 siRNA (Fig. 4D). The expression of MPO mRNA and protein was notably up-regulated in TF-treated HUVECs when compared with WSS-treated HUVECs. CircRNA-0079586 and circRNA-RanGAP1 siRNAs obviously decreased the expression of MPO mRNA (Fig. 4E) and protein (Fig. 4F) in TF-treated HUVECs.

Discussion
In this study, we recruited UIAs and RIAs patients and collected their tissue samples to compare the different expression of circRNAs using microarray screening. We found a batch of up-regulated and down-regulated circRNAs between RIAs and UIAs patients. In addition, we performed luciferase assay to explore the regulatory relationship between circRNAs and miRNAs, as well as between the miRNAs and their target genes. The luciferase activity of circRNA-0079586 was suppressed by miR-183-5p and the luciferase activity of circRNA-RanGAP1 was suppressed by miR-877-3p. The luciferase activity of MPO was inhibited by miR-183-5p and miR-877-3p.
High WSS values may be detected at possible locations of IA rupture as well as the neck section of IA lesions 34,35 . Also, experimental IA models with IA rupture induced by high WSS gradients showed that IA is extremely sensitive to the diseases of intracranial arteries induced by high WSS 36,37 . The effect of WSS on IA rupture remains unclear, although WSS with excessively reduced focal hemodynamics may lead to structural www.nature.com/scientificreports/ fragility as well as reduced resistibility of the aneurysmal wall 38 . A reduced magnitude of WSS can also enhance chronic inflammation as well as atherosclerotic changes induced by macrophages, which in turn cause the thinning of the aneurysmal wall to cause further rupture 39 . In this study, we performed qPCR to compare the different expression of circRNA-0079586, circRNA-RanGAP1, miR-183-5p, miR-877-3p and MPO in the endothelial cells collected from UIAs and RIAs patients. The expression of circRNA-0079586 and circRNA-RanGAP1 was significantly enhanced in RIAs patients when compared with UIAs patients. The expression of miR-183-5p and miR-877-3p was remarkably repressed in RIAs patients when compared with UIAs patients. The expression of MPO was significantly activated in RIAs patients when compared with UIAs patients. Wang et al. used high throughput sequencing to screen the effects of circRNAs on GM and paired noncancer tissues. They found that circ_0079586 was up-regulated in GM specimens 40 . Located at chr7: 23353140-23383472 and containing 586 base pairs, circ_0079586 is generated by the splicing of IGF2BP3. In this study, The expression of miR-877-3p was apparently suppressed in HUVECs transfected with p-circRNA-RanGAP1 and p-circRNA_0079586 + p-circRNA_RanGAP1 (*P < 0.05 vs. NC). (E) The expression of MPO mRNA was significantly increased in HUVECs transfected with p-circRNA-0079586 or p-circRNA_RanGAP1 alone, and to a higher extent in HUVECs transfected with p-circRNA_0079586 + p-circRNA_RanGAP1 (*P < 0.05 vs. NC; **P < 0.05 vs. p-circRNA_0079586).
(F) The expression of MPO protein was significantly increased in HUVECs transfected with p-circRNA-0079586 or p-circRNA_RanGAP1 alone, and to a higher extent in HUVECs transfected with p-circRNA_0079586 + p-circRNA_RanGAP1 (*P < 0.05 vs. NC; **P < 0.05 vs. p-circRNA_0079586).   www.nature.com/scientificreports/ we altered the expression of circRNA-0079586 and circRNA-RanGAP1 in HUVECs. The expression of cir-cRNA-0079586 and circRNA-RanGAP1 was negatively correlated with the expression of miR-183-5p and miR-877-3p, respectively, but positively correlated with the expression of MPO mRNA and protein. Located on chromosome 7, miR-183 has been shown to play a role as prospective oncogene in lung cancer, prostate cancer, breast cancer, colorectal cancer as well as hepatocellular cancer 41 . In general, miR-183-5p is either abnormally upregulated or downregulated in different types of cancers 42 . As an example, miR-183-5p is upregulated in hepatocellular cancer to inhibit cancer cell apoptosis by suppressing PDCD4 expression 43 . However, miR-183-5p also inhibited the invasion of gastric cancer by targeting Ezrin 44 . In one study, circ-RanGAP1 expression was found to be considerably upregulated in GC tissues, especially in the later stage of GC 45 . MiR-877-3p was shown to promote the differentiation and proliferation of lung and bladder cancer cells 46,47 . Also, miR-877-3p was shown to inhibit the differentiation of mesenchymal stem cells in the lungs to myofibroblasts to alleviate lung fibrosis caused by bleomycin 46 . It was also suggested that circ-RanGAP1 can promote the invasion and metastasis of GC by sponging the expression of miR-877-3p, a tumor suppressor 40 . Another research signified that miR-877 might serve as a tumor suppressor to block the expansion of hepatocellular carcinoma cells 48 .
In the presence of H2O2, MPO is generated by neutrophilic respiratory burst to promote the production of chlorinating agents such as hypochlorous acid. Besides its prominent role in protection against microbes, MPO has also been implicated in the destabilization of atherosclerotic plaques 49 . It was actually found that a short treatment using a low dose of Ox-LDL noticeably boosted the number EC-binding monocytes as well as the expression of ICAM-1 on the surface of EC cells. Furthermore, a prolonged treatment using a high dose of Ox-LDL caused increased cytotoxicity as well as EC monolayer detachment, indicating that Ox-LDL stimulation may lead to atherosclerosis by promoting the expression of adhesion molecules on EC surface. In cerebral aneurysm patients, the concentrations of circulating MPO are locally increased in CA sacs to enhance the number of MPO-positive cells such as neutrophils in the aneurysm. In a CA mouse model, the deficiency of MPO reduced the levels of pro-inflammatory molecules in cerebral arteries while lowering the number of leukocytes to alleviate the rupture of cerebral aneurysm 50 . Since MPO shortage is commonly seen in human populations, it is important to analyze the rate of cerebral aneurysm in MPO deficiency 31 . In UIA patients, MPO showed the association with the danger of aneurysm rupture, indicating that MPO may be used as a potential biomarker for aneurysm rupture 51 .

Conclusion
In this study, we established two MPO-modulating signaling pathways of circRNA_0079586/miR-183-5p/MPO and circRNA_RanGAP1/ miR-877-3p/MPO. These two signaling pathways are involved in the pathogenesis of intracranial aneurysms rupture. In patients with ruptured intracranial aneurysms, the expression of MPO was up-regulated with promoted expression of circRNA_0079586 and circRNA_RanGAP1.

Data availability
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.