Angiotensin II inhibits apoptosis of mouse aortic smooth muscle cells through regulating the circNRG-1/miR-193b-5p/NRG-1 axis

Angiotensin II (Ang II) is known to promote proliferation of vascular smooth muscle cells (VSMCs) in vascular remodeling, but whether it has an anti-apoptotic effect needs to be explored. Neuregulin-1 (NRG-1) as a member of the epidermal growth factor family was reported to suppress the proliferation of VSMCs by activating ErbB receptors, and therefore we hypothesized that there might be a cross talk between the anti-apoptotic effect of Ang II and the anti-proliferative effect of NRG-1 in VSMCs. The aim of the present study was to observe the expression and role of NRG-1 underlying the inhibitory effect of Ang II on apoptosis of mouse aortic smooth muscle cells (MASMCs). It was found that NRG-1 expression was down-regulated via the circNRG-1/miR-193b-5p-mediated post-transcriptional mechanism in response to Ang II. In addition, NRG-1 overexpression reversed the inhibitory effect of Ang II on apoptosis in MASMCs. Our data may provide a molecular basis for further understanding the mechanism of Ang II in suppressing the apoptosis of MASMCs by decreasing NRG-1 expression at circular RNA and micro RNA levels. The circNRG-1/miR-193b-5p/NRG-1 axis may prove to be a potential target for Ang II to inhibit the apoptosis of VSMCs and lead to vascular remodeling.


Introduction
Vascular remodeling is a pathophysiological process in many cardiovascular diseases, such as atherosclerosis and hypertension 1 . Increasing evidence has demonstrated that proliferation and apoptosis of vascular smooth muscle cells (VSMCs) are key events in vascular remodeling 2 . Changes of the renin-angiotensin-aldosterone system (RAAS) may alter the balance between proliferation and apoptosis of VSMCs 3,4 . Angiotensin II (Ang II) is an effector peptide of the RAAS, and also a modulator of VSMC growth with proliferation/apoptosis effects mediated by activation of Ang II type 1 receptor (AT 1 R) or Ang II type 2 receptor (AT 2 R) 5 . In the early stages of vascular remodeling, Ang II promoted proliferation 6,7 but inhibited apoptosis 8,9 of VSMCs. A fundamental strategy for the treatment of these cardiovascular diseases is to accelerate apoptosis of VSMCs 10 . It is therefore important to gain insights into details of the molecular mechanism of Ang II in inhibiting apoptosis of VSMCs.
Neuregulin-1 (NRG-1) is a member of the epidermal growth factor (EGF) family, whose isoforms can be produced from the NRG-1 gene by alternative splicing 11 . Its transmembrane isoform includes an extracellular domain with an EGF-like sequence (NRG-1-ECD) and a highly conserved intracellular domain (NRG-1-ICD) 12 . NRG-1-ECD is a bioactive fragment, which can bind to ErbB family receptor tyrosine kinases to activate ErbB signaling in target cells 13 . NRG-1-ICD can translocate into the nucleus to regulate the gene expression, which has been confirmed by our previous and other studies 14,15 . Several lines of evidence have demonstrated that NRG-1 plays important roles in vascular physiopathology 16 . First, NRG-1 is expressed in vascular endothelial cells and smooth muscle cells, and its receptors are localized to the underlying smooth muscle cells 17 . Second, treatment of cultured VSMCs with NRG-1-ECD significantly decreases platelet-derived growth factor (PDGF)-BB-stimulated proliferation and migration 18 . Third, NRG-1-ICD is stimulated by transforming growth factor-β1 (TGF-β1), and translocates into the nucleus to regulate the α-SMA gene expression in human aortic smooth muscle cells (HASMCs) 15 . Previous studies showed that Ang II decreased NRG-1-ECD expression in endothelial cells 19 . However, the role of NRG-1 in the regulation of VSMC apoptosis in the context of Ang II signaling remains unclear.
Circular RNAs (circRNAs), which are shown to be a kind of critical gene regulator, are a novel class of noncoding RNAs with the characteristic of covalent bond linking the 3′ and 5′ ends generated by back splicing 20 . CircRNAs are known to play roles in the regulatory networks governing gene expression with multi-functions, such as cytoplasmic miRNA sponges 21 , RNA-binding protein participants 22 , and nuclear transcriptional regulators 23 . For example, one abundant circHIPK3, which regulates cell growth in cancerous tissues, acts as miRNA sponges to inhibit miR-124 activity by directly binding to miR-124 24 . The exon-intron circular RNAs circEIF3J and circPAIP2 interact with U1 snRNP and enhance transcription of their parental genes in the nucleus via specific RNA-RNA interaction 25 . However, the function of cir-cRNAs under the treatment of Ang II in VSMC apoptosis remains unknown.
In this study, we showed that Ang II inhibited apoptosis through decreasing the expression of NRG-1 in mouse aortic smooth muscle cells (MASMCs), and this inhibitory effect could be reversed by NRG-1. In addition, circNRG-1 targeted NRG-1 for degradation by binding to miR-193b-5p. The results of the present study revealed a novel anti-apoptosis mechanism of Ang II via the circNRG-1/ miR-193b-5p/NRG-1 axis, which may provide a potential therapeutic strategy for the prevention and treatment of vascular remodeling diseases.

Materials and methods
Cell culture and treatment MASMCs (ATCC, No.CRL-2797TM) were cultured in low-glucose Dulbecco's modified Eagle's medium (DMEM) containing penicillin, streptomycin and 10% fetal bovine serum (FBS). Human embryonic kidney 293A cells were obtained from ATCC (Manassas, VA) and maintained in high-glucose DMEM supplemented with 10% FBS. The cells were all in a humidified incubator at 37°C with 5% CO 2 . Before stimuli and infection with adenovirus vectors or plasmids, MASMCs were incubated in serum-free medium for 24 h. Angiotensin II (A9525), Losartan (6188) and PD123319 (P186) were purchased from Sigma-Aldrich.
Cell apoptosis assay MASMC apoptosis assays were performed with the Annexin V-FITC Apoptosis Detection Kit (KeyGEN BioTECH) according to the manufacturer's recommendations. Cells were trypsinized, washed with cold PBS twice, re-suspended in 500 μl binding buffer containing 5 μl annexin V-FITC and 5 μl PI, and cultured at room temperature in the dark for 15 min. Finally, apoptosis was detected with a flow cytometer. All groups were evaluated in a minimum of three separate wells per experiment.

Microarray analysis
Circular RNA expression profiling was performed using Arraystar Mouse circRNA Array V2 analysis (Arraystar, USA). Circular RNAs of MASMCs were extracted using QIAzol Lysis Reagent (QIAGEN, Catalog no.79306) according to the manufacturer's instructions. The sample preparation and microarray hybridization were performed based on the Arraystar's standard protocols. Briefly, total RNAs were digested with Rnase R (Epicentre, Inc.) to remove linear RNAs and enrich circular RNAs. Then, the enriched circular RNAs were amplified and transcribed into fluorescent cRNA utilizing a random priming method (Arraystar Super RNA Labeling Kit; Arraystar). The labeled cRNAs were hybridized onto the Arraystar Mouse circRNA Array V2 (8 × 15K, Arraystar). After washing slides, the arrays were scanned by the Agilent Scanner G2505C. Agilent Feature Extraction software (version 11.0.1.1) was used to analyze acquired array images.

Fluorescence in situ hybridization
For circRNA FISH, cells were fixed in 4% paraformaldehyde for 5 min at room temperature, then permeabilized with 0.5% Triton X-100 and washed with PBS. The process used a Ribo TM Fluorescent In Situ Hybridization Kit (RiboBio, China). In brief, MASMCs were blocked with pre-hybridization buffer for 30 min at 37°C and then incubated with a circNRG-1 FISH Probe Mix and hybridization buffer overnight at 37°C. Cell slides were washed with hybridization washing buffer I, II, and III for 15 min at 42°C and 0.2 × SSC at 53°C.
The double FISH assay was performed as described previously 15 . The procedures were conducted with a miR-193b-5p double-fluorescein (both 5′-and 3′-were labeled with FITC) FISH probe (Genepharma, China) and a cy3-circNRG-1 FISH probe. Hybridization was performed using fluorescence-labeled probes in hybridization buffer by incubation at 55°C for 1.5 h. After stringent washing with SSC buffer, then DAPI (157574, MB biomedical) was used for nuclear counter staining. Images were captured by confocal microscopy and processed by LAS AF software.

Luciferase assay
Human embryonic kidney 293A cells were maintained as previously described 27 . For luciferase assays, HEK 293A cells were transfected with miR-193b-5p mimic (Gene pharma; Shanghai) and NC mimic combined with luciferase reporter plasmid of circNRG-1, circNRG-1 mut, NRG-1 3′-UTR, NRG-1 3′-UTR mut or an empty vector. Cells were harvested and luciferase activity was measured using a dual luciferase assay kit (Promega) after 24 h. The specific target activity was expressed as the relative activity ratio of firefly luciferase to renilla luciferase. All constructs were evaluated in a minimum of three separate wells per experiment.

Statistical analysis
All the data are presented as the means ± S.E.M. Differences between two groups were assessed using analysis of variance followed by a Student's t-test. A value of P < 0.05 was considered statistically significant. Statistical analysis was performed using Graphpad Prism 5 software (GraphPad Software, San Diego, CA, USA).

Ang II decreases NRG-1 protein expression in MASMCs
To examine the effect of Ang II on NRG-1, we firstly analyzed the expression of NRG-1 in MASMCs treated with Ang II. Knowing that NRG-1 included an extracellular domain (ECD) and an intracellular domain (ICD), we first detected the expression levels of full-length NRG-1, NRG-1-ICD, and NRG-1-ECD, respectively. Western blot showed that Ang II significantly decreased both fulllength NRG-1 and NRG-1-ICD protein levels in a doseand time-dependent manner (Fig. 1a, c). Furthermore, there was a reduction in NRG-1 from cell membrane lysates and accumulation of NRG-1-ECD in the media from MASMCs in response to Ang II (Fig. 1b, d). Next, we sought to determine whether Ang II decreased the transcription of NRG-1 gene. MASMCs were treated with Ang II 10 −7 M for different times and then NRG-1 mRNA was detected by qRT-PCR. Interestingly, we found that NRG-1 mRNA level was not affected by Ang II treatment (Fig. 1e). Based on these findings, we hypothesized that Ang II reduced protein expression of NRG-1 through post-transcriptional regulation.

Ang II suppresses circNRG-1 expression in MASMCs
The finding that Ang II downregulated NRG-1 protein level without affecting mRNA level suggested that Ang II regulated NRG-1 expression post-transcriptionally via miRNAs or circRNAs. An Arraystar mouse circRNA Microarray was accomplished with MASMCs to detect whether circRNAs were involved in the regulation of NRG-1. The differential expression levels of circRNAs between Ang II and control groups were identified through scatter plots (Fig. 2a). In the scatter plots, cir-cRNAs above the top green line and below the bottom Hierarchical clustering analysis showed the differentially expressed circRNAs over 2.0-fold change. Red color indicates high expression level, and blue color indicates low expression level. c Divergent and convergent primers were used to verify whether circNRG-1 was a circRNA. Convergent primers were used to detect NRG-1 mRNA. Divergent primers amplified circNRG-1 in cDNA but not gDNA. GAPDH served as linear control and size marker in base pairs. d Sanger sequencing confirmed head-to-tail junction of circNRG-1. e RNA fluorescence in situ hybridization for circNRG-1 was detected. Nuclei were stained with DAPI. Scale bars = 50 μm. f qRT-PCR detected circNRG-1 expression in MASMCs treated with Ang II (10 −7 M) for the different times. Data represent the means ± SEM of three independent experiments. *P < 0.05, ***P < 0.001 vs. Ang II for 0 h green line showed >1.5-fold change between the two compared groups. A heat map of 382 circRNAs was differentially expressed between Ang II and control groups. Among the 382 circRNAs, 235 were upregulated and 147 were downregulated by >2-fold change in Ang II group. The results showed that mmu-circRNA-42742, whose parental gene is NRG-1, was significantly down-regulated in the MASMCs treated with Ang II as compared with the control group (Fig. 2b). To further validate the results, we used convergent and divergent primers, respectively, to amplify total RNA and circular RNA transcripts derived from the NRG-1 gene by RT-PCR (Fig. 2c). Its PCR products were confirmed by DNA sequencing (Fig. 2d). The result of fluorescence in situ hybridization (FISH) showed that circNRG-1 mostly located in the cytoplasm of MASMCs (Fig. 2e). A general consistency was shown between the results of qRT-PCR and microarray analysis. The results showed that circNRG-1 was decreased in a time-dependent manner in the group treated with Ang II in MASMCs (Fig. 2f). These data suggested that Ang II reduced circNRG-1 expression.

circNRG-1 upregulates NRG-1 expression by acting as miR-193b-5p sponge
Knowing that Ang II decreased the expression levels of NRG-1 and circNRG-1, we sought to determine the relationship between them by overexpressing or knocking down circNRG-1 in MASMCs. Knowing that the exogenous circNRG-1 was effectively overexpressed in pLVX-circNRG-1-transfected MASMCs, we designed a siRNA targeting the backsplice sequence for circNRG-1 (si-circNRG-1) as shown by qRT-PCR (Fig. 3a). Western blot analyses revealed that circNRG-1 overexpression or knockdown markedly increased or decreased NRG-1 protein level, respectively (Fig. 3b). Knowing that cir-cRNAs can function as miRNA sponges, we predicted miRNAs, whose nucleotide sequences are known to be complementary to a region of circNRG-1, and found out their interactions. The circRNA/miRNA interaction was predicted with Arraystar's home-made miRNA target prediction software based on TargetScan and miRanda. The result showed that circNRG-1 contained sequences complementary to miR-193b-5p seed sequence (Fig. 3c). In addition, Ang II significantly increased the miR-193b-5p level (Fig. 3d). Localization of miR-193b-5p and circNRG-1 was showed by RNA in situ hybridization in MASMCs, and their co-localization was predominantly in cytoplasm (Fig. 3e). To further clarify whether the sponge effect of circNRG-1 on miR-193b-5p affected gene expression, circNRG-1 sequence was inserted immediately in downstream of the luciferase reporter gene. Luciferase assay revealed that transfection with miR-193b-5p mimic significantly decreased the luciferase activity via mediation of wild-type circNRG-1 sequence but not via its mutant (Fig. 3f). These findings suggested that circNRG-1 regulated NRG-1 expression and bound miR-193b-5p to prevent it from binding to the target genes.
It has been shown that endocardial endothelia and small subendocardial vessels are the primary cell source of NRG-1, and NRG-1 and its receptor ErbB play crucial roles in cardiac development, structural maintenance, and functional integrity of the heart 28 . Although our previous study found that TGF-β1 and PDGF-BB increased and decreased NRG-1 expression in HASMCs 15 , respectively, whether and how Ang II regulated NRG-1 in MASMCs remained unknown. The present data support the idea that Ang II decreases the NRG-1 protein expression in MASMCs and down-regulates it post-transcriptionally. NRG-1/ErbB signaling is an important potential mechanism for regulating vascular remodeling 29 . Hedhli  30 . Interestingly, we found that full-length NRG-1, NRG-1-ECD and NRG-1-ICD were all downregulated by Ang II, p-ErbB 2 and p-ErbB 4 expression levels were also decreased in a time-dependent manner after stimulation by Ang II (Supplementary Fig. S1). This means that reduced phosphorylation of ErbB receptors was due to the low expression of NRG-1-ECD. b HEK 293A cells were co-transfected with miR-193b-5p mimic and the wild-type pmirGLO-NRG-1 3′-UTR or its mutant pmirGLO-NRG-1 3′-UTR mut. After 24 h, luciferase activities were measured. Data are the means ± SEM of three independent experiments. **P < 0.01 vs. pmirGLO or pmirGLO-NRG-1 3′-UTR mut. c qRT-PCR detection of miR-193b-5p expression in MASMCs transfected with miR-193b-5p mimic or anti-miR-193b-5p and their corresponding controls. Data represent the means ± SEM of three independent experiments. ***P < 0.001 vs. miR-Ctl, ###P < 0.001 vs. anti-miR-Ctl. d MASMCs were transfected with miR-193b-5p mimic or anti-miR-193b-5p, as well as with their corresponding controls. NRG-1 expression was analyzed by Western blot. The right panel shows densitometric analyses from three independent experiments. **P < 0.01 vs. miR-Ctl, ##P < 0.01 vs. Anti-miR-Ctl. e Western blot detected NRG-1 expression in MASMCs transfected with pLVX-circNRG-1, anti-miR-193b-5p, or both of them. NRG-1 expression was quantitated as described above (Down panel). ***P < 0.001 vs. pLVX-Ctl+Anti-miR-193b-5p-untreated group. f MASMCs were transfected with si-circNRG-1, miR-193b-5p, or both of them for 24 h. NRG-1 expression was analyzed by Western blot and was quantitated as described above (Down panel). **P < 0.01, ***P < 0.001 vs. si-Con+miR-193b-5p-untreated group Cleaved-caspase-3, caspase-3, bax and bcl-2 expression levels were detected by Western blot. The right panel shows densitometric analysis from three independent experiments. ***P < 0.001 vs. 0 h. b MASMCs were treated with different times of Ang II (10 −7 M) and then the apoptotic rate was determined by flow cytometry analysis using annexin V/propidium iodide double staining. The right panel shows the apoptotic ratio from three independent experiments. ***P < 0.001 vs. 0 h. c MASMCs were infected with Ad-GFP-NRG-1 or Ad-GFP for 24 h and then treated with or without Ang II (10 −7 M) for an additional 12 h, Western blot detected NRG-1, Cleaved-caspase-3, caspase-3, bax and bcl-2 levels. Data represent the means ± SEM of three independent experiments. ***P < 0.001 vs. Ad-GFP+Ang II-untreated group. d The apoptotic rate was determined and quantitative analysis for the flow cytometry experiment is presented on the right panel. Data represent the means ± SEM, n = 3. ***P < 0.001 vs. Ad-GFP+Ang IIuntreated group Notably, we found that Ang II down-regulated NRG-1 protein expression without affecting the mRNA level in MASMCs, which urged us to make clear whether some miRNAs or circRNAs regulated the NRG-1 expression at the post-transcriptional level. According to the circRNA microarray, we found that circNRG-1 was also downregulated by Ang II, and this circRNA was not on the list of circBase, meaning that circNRG-1 is a new circRNA, so we used RT-PCR and Sanger sequencing to confirm whether circNRG-1 was derived from NRG-1 gene. In addition, we found that circNRG-1 overexpression or knockdown could increase or decrease the NRG-1 expression, so we wondered whether this mechanism of circNRG-1 was involved in the regulation of NRG-1 protein level. To simplify this problem, we tried to use the ceRNA theory 31 to explain how circNRG-1 exerted its effect through bioinformatic analysis by establishing a circRNA-miRNA-mRNA regulatory model based on the ceRNA theory. As expected, we found that circNRG-1 acted as a miR-193b-5p sponge and interacted with miR-193b-5p to target the 3'-UTR of NRG-1 mRNA, thus increasing the NRG-1 expression in MASMCs.
In general, Ang II has been shown to induce or prevent apoptosis in several cell types 32,33 but the role of Ang II in regulating apoptosis of VSMCs remains controversial. Here, we firstly demonstrated a novel mechanism of Ang II in inducing MASMCs to exhibit apoptosis resistance. The results showed that Ang II (10 −7 M) decreased the apoptosis gene Cleaved-caspase-3 and bax, and increased the anti-apoptosis gene bcl-2 in MASMCs. In addition, the apoptotic rate of MASMCs was reduced by about 50%. This is similar to the finding in a previous study that Ang II inhibited VSMC apoptosis 8 , but different from the result of other studies that Ang II promoted apoptosis of VSMCs 34 . This may be due to different concentrations of Ang II used, or the different receptors activated by Ang II. In the early stages of vascular remodeling, a low Ang II environment may inhibit VSMC apoptosis, but with the phase advancing, the high Ang II environment stimulated VSMC apoptosis. In addition, Ang II exerted its biological effect through AT 1 R and AT 2 R, which are known to be associated with proliferation and apoptosis 35,36 . Several studies demonstrated that AT 1 R activation resulted in proliferation and anti-apoptosis whereas AT 2 R activation induced apoptosis in VSMCs 37 . In the present study, we blocked AT 1 R or AT 2 R to observe the apoptotic rate of MASMCs, and found that AT 1 R blockage could reverse the apoptotic rate of Ang II, but not the AT 2 R inhibitor ( Supplementary Fig. S2). Finally, Ang II-induced antiapoptosis was shown to be dose-dependent, and AT 1 R signaling was found to be a decisive response.
To further investigate the role of NRG-1 in Ang IIinduced apoptosis resistance, we overexpressed NRG-1 and found that higher expression of NRG-1 reversed the inhibitory effect of Ang II on apoptosis in MASMCs. Together, these data suggest NRG-1 as a critical factor for mediating the resistance to apoptosis conferred by Ang II in MASMCs, and NRG-1 may be a potential therapeutic target for vascular remodeling associated with RAAS overactivation in atherosclerosis, hypertension and vascular diseases.