MiR‐126a‐5p limits the formation of abdominal aortic aneurysm in mice and decreases ADAMTS‐4 expression

Abstract Abdominal aortic aneurysm (AAA) is a serious vascular disease featured by inflammatory infiltration in aortic wall, aortic dilatation and extracellular matrix (ECM) degradation. Dysregulation of microRNAs (miRNAs) is implicated in AAA progress. By profiling miRNA expression in mouse AAA tissues and control aortas, we noted that miR‐126a‐5p was down‐regulated by 18‐fold in AAA samples, which was further validated with real‐time qPCR. This study was performed to investigate miR‐126a‐5p's role in AAA formation. In vivo, a 28‐d infusion of 1 μg/kg/min Angiotensin (Ang) II was used to induce AAA formation in Apoe‐/‐ mice. MiR‐126a‐5p (20 mg/kg; MIMAT0000137) or negative control (NC) agomirs were intravenously injected to mice on days 0, 7, 14 and 21 post‐Ang II infusion. Our data showed that miR‐126a‐5p overexpression significantly improved the survival and reduced aortic dilatation in Ang II‐infused mice. Elastic fragment and ECM degradation induced by Ang II were also ameliorated by miR‐126a‐5p. A strong up‐regulation of ADAM metallopeptidase with thrombospondin type 1 motif 4 (ADAMTS‐4), a secreted proteinase that regulates matrix degradation, was observed in smooth muscle cells (SMCs) of aortic tunica media, which was inhibited by miR‐126a‐5p. Dual‐luciferase results demonstrated ADAMTS‐4 as a new and valid target for miR‐126a‐5p. In vitro, human aortic SMCs (hASMCs) were stimulated by Ang II. Gain‐ and loss‐of‐function experiments further confirmed that miR‐126‐5p prevented Ang II‐induced ECM degradation, and reduced ADAMTS‐4 expression in hASMCs. In summary, our work demonstrates that miR‐126a‐5p limits experimental AAA formation and reduces ADAMTS‐4 expression in abdominal aortas.


| INTRODUC TI ON
Abdominal aortic aneurysm (AAA) is a potential lethal vascular disease in the event of aortic rupture. 1 Currently, therapies are limited to open surgical and less invasive endovascular repair. 2 Effective drug-based intervention for this disease is still lacking, 3 highlighting an urgent need to better understand the underlying mechanisms involved in the formation and progression of AAA. The pathogenesis behind AAA is multifactorial, but it is well-known that extracellular matrix (ECM) degradation and loss of vascular smooth muscle cells (VSMCs) occur, which ultimately contribute to the weakening and of remodelling the aortic wall. 4,5 Members of ADAM metallopeptidase with thrombospondin type 1 motifs (ADAMTSs) are secreted enzymes sharing a common multidomain structure, and participate in the degradation of ECM proteins during development, morphogenesis and tissue remodelling. 6 ADAMTS-4 is a secreted metalloproteinase which belongs to the ADAMTS family, and it cleaves various ECM proteins including brevican, versican and aggrecan in the vasculature. 7 Increasing evidence has revealed an involvement of ADAMTS-4 in the pathogenesis of aneurysmal diseases. Ren and co-workers found that ADAMTS-4 expression was much higher in aortic tissues of patients with thoracic aortic aneurysm and dissection (TAA/D) than in control aortas. 8 This team further demonstrated that ADAMTS-4 deficiency attenuated angiotensin II (Ang II) infusion-induced sporadic TAA/D in Apoe -/mice. 9 Although these previous studies revealed that the elevation of ADAMTS-4 may contribute to arterial aneurysm formation, its role in AAA progression is not clear.
Interestingly, we prior established an Ang II-induced murine AAA model as described before, 10,11 and analysed the gene expression profile in the aneurysmal tissues and control aortas. The microarray profiling data showed a fourfold increase in ADAMTS-4 in the aortic tissues of AAA mice compared to control aortas (P < .01).
This finding inspired us to further elucidate ADAMTS-4's role in AAA formation.
MicroRNA-126a-5p (miR-126a-5p) is a 21-nt microRNA that has a high sequence conservation between different species. Like other miRNAs, it acts a negative regulator for genes by degrading mRNA and/or blocking the translation process. 12 Studies on miR-126a-5p in AAA formation are scarce; however, several literatures have revealed its involvement in artery diseases. Schober et al found that miR-126a-5p prevented endothelial cell function from hyperlipidaemia in an experimental atherosclerosis animal model. 13 Similar protective effects of miR-126a-5p on vascular endothelium have also be demonstrated by Esser et al. 14 In addition to analysing the mRNA profiles, our group also obtained the miRNA expression profiles in mouse aortic tissues. We found that miR-126a-5p decreased by 18-fold in AAA tissues. Interestingly, we noted potential binding sites on the 3' untranslated region (UTR) of both mus and homo ADAMTS-4 for miR-126a-5p.
Furthermore, miR-126a-5p expression was negatively correlated to ADAMTS-4 in the analysed aortic samples. These findings promoted us to study the role of the dysregulated miR-126a-5p-ADAMTS-4 axis in AAA formation.
In this study, experimental AAA formation was induced by Ang II infusion in Apoe -/mice. To explore how the down-regulated miR-126a-5p affects AAA development, its exclusive agomirs were given to Ang II-infused mice. Our work shows that re-expression of miR-126a-5p promotes the survival of mice injected with Ang II. On Ang II infusion, ADAMTS-4 expression increases, which is inhibited by miR-126a-5p. Dual-luciferase reporter results confirm that miR-126a-5p directly targets ADAMTS-4.

| Ethic statement
We performed the animal experiments according to the Guide for the Care and Use of Laboratory Animals (National Institutes of Health). The Ethic Committee of Dalian Medical University has approved our study.

| Ang II infusion AAA model and miRNA agomir administration
AAA was induced in Apoe -/mice according to the methods reported by Daugherty et al. 10 In short, male Apoe -/mice (15 weeks; C57BL/6J background) were infused with 1 μg/kg/min Ang II (GL BioChem) or normal saline (Dubang Pharmaceutical Co., Ltd) with Model 2004 Alzet Osmotic minipumps (Alzet) that were implanted subcutaneously. Mice were anaesthetized via 2%-3% isoflurane before surgical procedure. The formation of arterial aneurysm was defined by a 50% or greater dilation in the external diameter of suprarenal aorta.
For survival test, a total of 36 mice were included (n = 6 in sham group, n = 12 in AAA + miR-126a-5p agomirs, n = 18 in AAA + NC agomirs). The mortality was recorded for 28 d. Aortic tissues from survived mice were used in analysis of morphological changes.
Additional 12 mice (n = 4 per group) were subjected to analyse protein and mRNA alterations of targeted genes.

| Dual-luciferase reporter assay
Measurement of normalized firefly luciferase activity was performed by using the pmirGLO Dual-Luciferase miRNA Target Expression Vector (Promega Corporation) as per manufacturer's recommendations. Dual-luciferase reporter constructs containing the 3'UTR of ADAMTS-4 with miR-126a-5p binding sites were cotransfected with NC or miR-126a-5p agomirs into cells.
Corresponding mutant 3'UTR fragments were also inserted into pmirGLO plasmid. Forty-eight hours post the transfection, cells were analysed for luciferase. For each transfection, the firefly luciferase activity was normalized to renilla luciferase activity, and the luciferase activity was averaged from three replicates.

| Aortic diameter measurement
Mouse aortic diameters were measured via GE volusonE8 ultrasound system at baseline and day 28 post-aneurysm induction.

| Human aortic smooth muscle cells (hASMCs) and treatment
The hASMCs were obtained from Z.q.x.z Cell Research and kept in ScienCell smooth muscle culture medium. For some experiment, hASMCs were stimulated with different doses of Ang II (0.5, 1 or 5 μM) for 12 hours or with 1 μM Ang II for 6, 12 or 24 hours.

| Western blot
Protein samples were collected from aortic tissues and hASMCs by

| Statistics
GraphPad Prism software version 8.0 (GraphPad Software) was used to analyse all data. Values were presented as means ± SD or SEM. Kaplan-Meier curve followed by log-rank (Mantel-Cox) test was used to analyse the survival data. Data from two groups were analysed with Student's t test, while data from multiple groups were analysed with one-way ANOVA followed by Tukey's multiple comparisons test. Cell counts in IF staining of aortic tissues were analysed with Kruskal-Wallis test followed by Dunn's multiple comparison test. P < .05 was considered statistically significant.

| Ang II infusion alters the expression miR-126-5p and ADAMTS-4 in mouse aorta tissues
Ang II was infused into Apoe -/mice for continuous 28 d via osmotic minipumps. On day 28, the survival mice were killed, and type III-IV arterial aneurysms ( Figure 1A) were collected for microarray analysis (n = 3). Aorta tissues from four sham-operated mice ( Figure 1A) were collected as the control for microarray assay (n = 4). Real-time qPCR data illustrated that miR-126a-5p expression was significantly lower in AAA tissues than in control aortas ( Figure 1B). ADAMTS-4 showed opposite expression pattern to miR-126a-5p ( Figure 1B).
These data demonstrate that Ang II infusion alters the expression of miR-126a-5p and ADAMTS-4 to opposite directions in mouse abdominal aortas. In addition, genes potentially targeted by and negatively expressed (Pearson correlation coefficient <−0.7, and P value <.05) with miR-126a-5p in mouse aortic tissues (n = 4 for sham, n = 3 for AAA) were listed in Table S1.
Data from real-time qPCR showed that miR-126a-5p overexpression revealed that Ang II-induced increase in aorta diameter was attenuated by miR-126a-5p agomirs ( Figure 4B). Abdominal aortas were then stained with H&E ( Figure 4C), and EVG ( Figure 4D). The staining images clearly indicated a degradation of elastic laminas in Ang II-infused aortas. As compared to AAA mice injected with NC agomirs, mice administrated with miR-126a-5p agomirs had increased layers of elastic laminas. The integrity of elastic laminas was also impaired by Ang II, which was protected by miR-126a-5p agomirs as well. Sirius red could stain collagens in yellow and red.
As indicated in Figure 4E, miR-126a-5p agomirs attenuated the degradation of collagens induced by Ang II injection. These data together reveal a protective role of miR-126a-5p in experimental AAA formation.

| ADAMTS-4 increases in response to Ang II infusion, whereas decreases when miR-126a-5p is overexpressed
ADAMTS-4 mRNA and protein levels were up-regulated by Ang II infusion and down-regulated by miR-126a-5p overexpression in mouse aortic tissues ( Figure 5A-B). To determine what type of cells expresses ADAMT-4, SMCs were marked via SM-22α and macrophages were labelled with CD68. Immunofluorescent images showed that, in response to Ang II infusion, a strong up-regulation of ADAMTS-4 presented in SMCs of aortic media ( Figure 5C). An elevation of ADAMT-4 was also observed in macrophages infiltrated in the arterial aneurysm ( Figure S1). Immunofluorescent data validate the down-regulation in ADAMTS-4 can be induced by miR-126a-5p agomirs.

F I G U R E 2
MiR-126a-5p agomirs improve the survival of mice treated with Ang II. Kaplan-Meier curve followed by log-rank (Mantel-Cox) test was performed to assess the survival data (A). P < .05 was considered statistically significant. The expression levels of miR-126a-5p in mouse aortas were determined with realtime qPCR on day 28 (B). Values were presented as means ± SEM (n = 4). † † † P < .001

| Leniviruses expressing pre-mir-126 reduces Ang II-induced elevation in ADAMTS-4 in primary hASMCs in vitro
In keeping with in vivo data, Ang II decreased miR-126-5p expression in a dose-and time-dependent way in hASMCs ( Figure 6A

| D ISCUSS I ON
In the present study, miR-126a-5p or NC agomirs were given to mice infused with Ang II intravenously. As compared to sham-op- Both mature miR-126a-5p (previous name: miR-126*) and miR-126a-3p (miR-126) are generated from precursor mir-126a (pre-mir-126a) via dicer. 15 Hua et al found that Ang II reduced miR-126-3p in vascular SMCs in vitro. 16 Such findings together with ours suggest that Ang II may interfere the processing of primary-mir-126 (pri- As our microarray results suggested that ADAMTS-4 expression was negatively correlated to miR-126a-5p in aneurysms and control aortas in mice, we then investigated whether miR-126a-5p could down-regulate ADAMTS-4. Ren and co-workers found that ADAMTS-4 overexpressed in the both aortic media and adventitia in mice challenged with Ang II. 9 Our data in consistent with their study demonstrated an up-regulation of ADAMTS-4 aortic wall in AAA mice. Degeneration and disorganization of the elastic media is the key histopathological feature for AAA development. 25,26 Dysfunction of ASMCs contributes to the aortic dilatation and eventual rupture in progressed AAA. 27 Elastic fibres fail to form normal elastic lamina-SMC connections due to a local elevation of Ang II signalling. 28 Labelling cells with SM-22α antibody enabled us to observe an elevation of ADAMTS-4 in aortic SMCs in mice. By staining the elastic fibres with EVG, we observed that F I G U R E 5 ADAMTS-4 expression up-regulates in aortic SMCs upon Ang II infusion, whereas is decreased by miR-126a-5p agomirs in vivo. The mRNA and protein expression of ADAMTS-4 or versican degradation products (VDP) were analysed with real-time qPCR (A) and Western blot (B). Values were presented as means ± SEM (n = 4). The protein expression levels of ADAMTS-4 (green) and SM-22α (red, a marker for vascular SMCs) were also probed with immunofluorescent assay and quantified (C). Bar, 50 μm. Values were presented as means ± SEM (n = 6 for sham, n = 8 for AAA + NC agomirs, n = 9 for AAA + miR-126a-5p agomirs). † † † P < .001 up-regulation of miR-126a-5p prevented elastin from degradation.
We next treated hASMCs with Ang II at different doses. Ang II treat- . Values were presented as means ± SD (n = 3). † † P < .01, † † † P < .001 up-regulation with inflammation has been reported in other human diseases associated with inflammation. 29,30 However, a previous study also reported an anti-inflammatory role of ADAMTS-4 in the central nervous system. 31 Although we performed the present work to investigate the role of miR-126a-5p-ADAMTS-4 axis in regulation ECM homeostasis, to understand its role in AAA-associated inflammation is needed.
MMP-9 and MMP-2 are the two most widely studied MMPs because of their ability to degrade elastin and collagen. 19,32 Interestingly, we noted that the expression of these two MMPs decreased following administration of miR-126-5p agomirs. However, several online databases (TargetScan, miRDB or Starbase) predict MMP-13 and MMP-16, but not MMP-9 and MMP-2, as potential targets for miR-126a-5p. These findings suggest that miR-126-5p may not regulate their expression directly. Reduction of MMP-9 and MMP-2 may occur when AAA was attenuated by miR-126-5p.
We also determined the expression of MMP-13 and MMP-16, and observed that only MMP-13 expression decreased following miR-126a-5p overexpression. Such data suggest that, aside from ADAMTS-4, miR-126a-5p may also attenuate ECM degradation by directly targeting MMPs. In addition, Ventricular Zone Expressed PH Domain Containing 1 (VEPH1) was identified as a novel regulator of phenotypic switch of vascular SMCs by our group (data not shown), and it was a predicted target of miR-126a-5p. To further investigate whether miR-126a-5p/VEPH1 axis plays a role in phenotypic transition of aortic SMCs will help to fully evaluate the anti-AAA effects of miR-126a-5p.
We understand that ADAMTS-1, ADAMTS-4 and ADAMTS-5 are the major proteases responsible for the cleavage and degradation of hyaluronan-binding proteoglycans, 6 but not collagen I or elastin. However, a previous study from Hong-Brown et al showed that silencing of ADAMTS-1 blocked alcohol-induced degradation of elastin in myocytes. 33 Considering the important function of collagen I and elastin in maintaining the integrity of aortic wall, we F I G U R E 8 MiR-126-5p inhibits Ang II-induced ECM loss and reduces ADAMTS-4 in hASMCs. Primary hASMCs were infected with pre-mir-126 or anti-miR-126-5p sponge lentiviruses before Ang II stimulation (A-D). Re-expression of ADAMTS-4 was also mediated with lentiviruses (E-F). The expression levels of versican degradation products, collagen I, elastin and ADAMTS-4 were determined with Western blot (A-F). Values were presented as means ± SD (n = 3). † P < .05, † † P < .01, † † † P < .001 further analysed their expression in Ang II-treated hASMCs. Our data illustrated that miR-126-5p protected collagen I and elastin from Ang II-induced degradation, and interestingly, ADAMTS-4 inhibition was involved such protective effects. It has been suggested that ADAMTS-4 is not able to directly cleave collagen I or elastin. 34 We then asked why these two ECM components were reduced post-ADAMTS-4 overexpression. ADAMTS-4 neutralizing antibody has been demonstrated to block pro-inflammatory interleukin 1 (IL-1) signalling transduction in chondrocytes, 35,36 suggesting a potential pro-inflammatory role of ADAMTS-4.
Overgeneration of pro-inflammatory cytokines, such as IL-1, is also one major cause for ECM degradation in aneurysm formation. 37 These findings together with our work imply that ADAMTS-4 may indirectly contribute to collagen I and elastin degradation by inducing inflammation.
Collectively, our work demonstrates miR-126a-5p down-regulation as a pathogenic factor in Ang II-induced AAA formation.
Overexpression of miR-126a-5p limits murine AAA formation, in which ADAMTS-4 suppression is involved. This study suggests miR-126a-5p-ADAMTS-4 as a potential therapeutic target for AAA development.

ACK N OWLED G EM ENTS
Dr RA Khalil was in part funded by National Heart, Lung and Blood Institute grants (HL111775 and 1R56HL147889-01).

CO N FLI C T O F I NTE R E S T
We have no conflict of interest.

AUTH O R CO NTR I B UTI O N S
LL wrote the paper. LL, WM and SP performed the research. LL, YL, HW and BW analysed the data. LL and WM designed the research study. RK helped to design the research study and revised the manuscript. All authors approved the submission.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data are presented in this manuscript.