miR-146a Deficiency Ablates β-Aminopropionitrile-induced Thoracic Aortic Aneurysmal Rupture in Mice

Thoracic aortic aneurysm (TAA) is a life-threatening disease with high mortality greater than 80% after rupture. Besides surgical interventions, no medical therapies are available to blunt TAA rupture.¹ miR-146a, a short non-coding microRNA (miRNA), is well known to regulate inflammatory and auto-immune processes under disease conditions including cardiovascular diseases,² by post-transcriptional repression of target genes. Increased miR-146a has been observed in plasma and dissected aorta of TAA patients.³ However, its role in TAA development and rupture is unknown.

To examine the potential role of miR-146a in TAA, we evaluated the miR-146a expression by in-situ hybridization in de-identified formalin-fixed human thoracic non-aneurysmal (NAA; n=5) and TAA sections (n=5, Source: One Health Biorepository, University of Missouri; protocol# 2101859 and BioLincc, NHLBI) using miR-146a or control detection probe. miR-146a positive nuclei (purplish red) were abundant in the aortic media of TAAs, whereas a weak expression observed in NAAs (Figure A, p<0.05, n=5). Next, we examined the miR-146a expression in experimental TAAs. Inhibition of lysyl oxidase (LOX), an enzyme that covalently cross-links elastin and collagen, by β-aminopropionitrile (BAPN) accelerates TAA formation and rupture in mice.⁴ To induce TAA, 3–4 weeks old C57BL/6J male and female mice were administered with either vehicle (water) or BAPN (0.5% wt/vol) in drinking water for 28 days. Consistent with human aortic findings, in situ hybridization revealed a significantly (p<0.05) increased miR-146a positive nuclei in BAPN-induced TAAs (n=4–5) compared to controls (NAA; n=4, Figure A). Immunohistochemical staining on serial human and mice aortic sections using antibodies against α-SMC actin and CD68 revealed that miR-146a positive nuclei localize to SMC-rich aortic medial tissue. qPCR analyses showed a significantly increased miR-146a in BAPN-induced TAAs compared to controls (Figure A, p<0.05, n=9). Next, to determine the role of miR-146a in TAA formation and rupture, miR-146a deficient (KO) mice in a C57BL/6J background were obtained from the Jackson laboratory (stock# 016239), and the genotype was confirmed as described in our previous publication.

Figure. miR-146a is a novel target for BAPN-induced TAA.

A, In situ hybridization staining of (i) miR-146a in human non-aneurysmal (NAA) and thoracic aortic aneurysmal (TAA) sections, and (ii) BAPN-induced mouse TAA sections (Purplish red color indicates positive miR-146a nuclei staining; N=5, *P<0.05 (P=0.008) TAA vs NAA, *P<0.05 (P=<0.001) BAPN vs vehicle. Immunohistochemical staining of α-SMC actin (SMA; 5μg/ml) and macrophages (CD68; 0.25μg/ml) in (i) human NAA and TAA sections, and (ii) BAPN-induced mouse TAA sections (Black arrows show positive area (brown color) of α-SMC actin, and CD68 staining). (iii) qPCR analyses of miR-146a abundance in mice administered with either Vehicle or BAPN (N=9, *P<0.001(P=0.0000155) BAPN vs vehicle. B, (i) Ex vivo ascending aortic width measurements of 28 days survived male and female miR-146a WT and KO mice administered with either vehicle (n=5) or 0.5% BAPN (n=12–18) for 4 weeks. *P<0.05 (Male P=0.013, Female P= 0.040) BAPN vs vehicle, (ii) Representative ex vivo thoracic aortic images of male and female miR-146a WT and KO mice administered with vehicle or BAPN. C, (i) TAA rupture percentage of male and female WT and miR-146a KO mice (n=12–18) administered with 0.5% BAPN for 4 weeks, (*P<0.05 (Male P=0.028, Female P=0.028) KO-BAPN vs WT-BAPN / vehicle); (ii) Representative ruptured thoracic aortic images of male and female miR-146a WT mice administered with BAPN. Survival curve of (iii) male and (iv) female WT and miR-146a KO mice (n=12–18) administered with 0.5% BAPN for 4 weeks (#P<0.05 (Male P=0.026, Female P= 0.014) WT-BAPN vs KO-BAPN). D, (i) Representative images of thoracic aortic sections from male and female WT and miR-146a KO mice administered with either vehicle or 0.5% BAPN stained for aortic medial elastin breakage (Verhoeff’s staining; black arrows highlighting the elastin breaks), and (ii) Quantification of medial elastin breaks (N=5, *P<0.05 (Male- P=0.014, Female- P=<0.001) BAPN vs vehicle). (iii) Representative images of thoracic aortic sections from male and female WT and miR-146a KO mice administered with either vehicle or 0.5% BAPN stained for collagen (Trichrome staining) and (iv) Quantification of Trichrome staining positive area (n=10–11, *P<0.05 (P=0.008) KO-BAPN vs KO-Vehicle; #P<0.05 (P=0.011) KO-BAPN vs WT-BAPN). E, (i) miR-146a target binding of Filamin A (FLNA) 3’UTR. (ii) Schematic of the putative binding sites or mutations of miR-146a in FLNA mRNA 3’UTR; Luciferase reporter assay validating FLNA as a target of miR-146a. HEK293T cells were co-transfected with the luciferase-3’-UTR construct of FLNA (wild type/mutant: 300ng) and control/miR-146a mimetic (50 nM), N=6, *P<0.001 WT-FLNA control vs miR-146a mimetics. (iii) Serial cross-sections from BAPN-induced TAA tissue sections were immunostained with anti-Filamin A (0.05 μg/ml). Red color (black arrows) indicates positive immunostaining. (iv) Quantification of Filamin A immunostaining, N=7–8, *P= <0.001, WT-BAPN vs WT-Veh; # P=0.009, WT-BAPN vs KO-BAPN. F, (i) Western blot analyses of FLNA protein in WT and miR-146a KO mice administered with either vehicle or BAPN (0.01 μg/ml; N=5). (ii) Quantification of Filamin A protein, *P<0.05 (P=0.005) BAPN vs vehicle; # P<0.05 (P=0.013) WT-BAPN vs KO-BAPN. (iii) Filamin A (0.5 μg/ml; red color,) and F-actin (green color, Phalloidin- Alexa Fluor 488) immunofluorescence staining of mouse aortic SMCs transfected with either control of miR-146a mimetics (100 nM). (iv) Quantification of SMC-Filamin A protein, N=13, *P= <0.001, miR-146a mimetics vs control mimetics. (v) qPCR analyses of pro-inflammatory genes in WT and miR-146a KO mice administered with either vehicle or BAPN (N=8–13), COX-2: *P=0.008 KO-BAPN vs KO-Veh; # P=0.032 KO-BAPN vs WT-BAPN. CCL2: *P<0.001 KO-BAPN vs KO-Veh; # P=0.002 KO-BAPN vs WT-BAPN. IRAK1: *P=0.010 KO-Veh vs WT-Veh. TNFα: *P<0.004 KO-BAPN vs KO-Veh. Number of samples are biological replicates for all data except F(iv) in which the number of samples are independent experiments. Statistics: P values were determined using the following tests: A(i) - Mann-Whitney Rank Sum test. A(iii), F(iv) – Student t test. A(ii), B(i), D(ii), D(iv), E(ii), E(iii), F(ii), F(v) -Two-Way Anova followed by Holm-Sidak post hoc test. C(i) -Fisher Exact test, C(iii) - Gehan-Breslow survival test followed by Holm-Sidak post hoc test. All data passed the normality test (Shapiro-Wilk) except for A(i), E(ii) and F(v) - COX and CCL2 mRNA. Equal variance test (Brown-Forsythe) was passed for all data except A(ii), F(v) – CCL2 and TNFα. Data are presented as mean ± SEM.

To induce TAA, 3–4-week-old male and female miR-146a KO and wild type (WT) littermate control mice were administered with either vehicle (n=5) or BAPN (0.5% wt/vol; n=12–18) for 4 weeks in drinking water. BAPN administration significantly but equivalently developed TAA, as measured by ex vivo maximal external ascending aortic width, in both male (1.6 ± 0.2 mm) and female (1.87 ± 0.2 mm) mice compared to controls (1.1 ± 0.02 mm) (Figure B). However, miR-146a deficiency had no influence on BAPN-induced TAAs in both males and females compared to WT-BAPN (Male: KO-1.9 ± 0.2 vs WT-1.6 ±0.2 mm; Female: KO-1.4 ±0.1 vs WT-1.9 ± 0.2 mm; Figure B). BAPN administration resulted in a 33% rupture in both male (6/18) and female (4/12) mice compared to controls (0%). Interestingly, miR-146a deficiency completely and significantly protected both male (0/13) and female (0/15) mice from TAA rupture (Figure C). Furthermore, miR-146a deficiency significantly improved mice survival rate from BAPN-induced aortic rupture (Figure C). Verhoeff’s staining analysis revealed that LOX inhibition accelerated elastin breaks equivalently in both WT and miR-146a KO mice (Figure D). Interestingly, Trichrome staining showed a significantly increased collagen-positive area in the adventitia of KO-BAPN mice compared to WT-BAPN (Figure D).

Based on the observed preserved extracellular matrix collagen in the miR-146a KO mice, next we examined the cytoskeletal structural protein Filamin A (FLNA) organization in the aortic section. Cytoskeletal structural proteins bind with the extracellular matrix to maintain structural integrity and function. TargetScan analysis revealed miR-146a binding sites in the 3’-UTR of FLNA (Figure E), a major cytoskeletal protein, downregulated in human thoracic dissected aortas.⁵ Luciferase reporter assay using control or miR-146a mimetics on HEK293T cells expressing WT or mutant 3’-FLNA UTR revealed significantly decreased activity with WT-3’-UTR/miR-mimic compared to WT-3’-UTR/control. In contrast, no significant change in luciferase activity was observed in cells transfected with mutant-3’UTR/miR-mimic compared to mutant-3’UTR/control, which further confirms that miR-146a directly targets FLNA via 3’-UTR (Figure E). Immunostaining revealed a significant reduction of FLNA in the aortic media of WT-BAPN mice whereas miR-146a deficiency showed significantly preserved FLNA (Figure E). Western Blot analyses revealed a significant reduction of FLNA in thoracic aortic lysates of BAPN-WT mice compared to controls (Figure F). Interestingly, miR-146a deficiency protected the loss of FLNA with BAPN administration (Figure F). Immunofluorescence staining of mouse aortic SMCs transfected with miR-146a mimetics showed a significant suppression of FLNA along with disorganized filamentous actin (Figure F). qPCR analyses revealed significantly increased pro-inflammatory genes- COX-2 and CCL2 in thoracic aorta of BAPN-KO mice compared to WT-BAPN group. Furthermore, miR-146a deficiency resulted in significantly increased IRAK-1 and TNFα compared to vehicles which was not further influenced with BAPN administration (Figure F). In conclusion, our study results demonstrate that miR-146a deficiency significantly protects mice from BAPN-induced TAA rupture, which is associated with preserved aortic medial cytoskeletal structural protein FLNA and adventitial collagen despite increased pro-inflammatory genes in the aorta.