All animal experiments were approved by The Laboratory Animal Welfare and Ethics Review Committee of Zhejiang University (Hangzhou, China) following The Guideline for the Care and Use of Laboratory Animals (approval no. 29060). C57BL/6J male mice were obtained from HFK Bioscience and housed with food and water ad libitum at 22±1°C and 40-60% humidity under a 12/12 h light-dark cycle (lights on at 7.00 am). Mice were randomly assigned to one of four groups: Sham, TAD, TAD + negative control (NC) or TAD + MEX3B KD. The TAD model was established by administration of BAPN (Shanghai Aladdin Biochemical Technology Co., Ltd.) and Ang II [GL Biochem (Shanghai) Ltd.] as previously described (18). Mice (age, 3 weeks, 9 g) were fed a normal diet and water containing 1 g/kg/day BAPN for 4 weeks. Osmotic mini pumps administering 1 μg/kg/min Ang II were implanted subcutaneously after anesthetization with isoflurane (induction, 3.0%; maintenance dose, 1.5%). At 24 h after implanting the micro-osmotic pumps, mice underwent echocardiography to measure the maximum diameter of the aorta under anesthesia with isoflurane. For sham group, mice were fed a normal diet and water for 4 weeks, followed by the infusion of physiological saline via micro-osmotic pumps. Mice were euthanized with 70% vol/min CO₂. Following euthanasia, cardiac and respiratory arrest, as well as fixed and dilated pupils, were observed in the mice for ~10 min to confirm death. Humane endpoints were as follows: a) sustained weight loss exceeding 20% of baseline; b) Impaired mobility or inability to access food or water autonomously; c) Signs of severe distress or pain unrelieved by analgesia; d) Clinical signs indicating severe systemic illness.

For MEX3B KD, short hairpin RNA (shRNA) targeting MEX3B or NC was synthesized by General Bio Co., Ltd. and inserted into adeno-associated virus subtype 9 (AAV9) vectors (Takara Bio Inc.). AAV-293 cells (Cellverse Bioscience Technology Co., Ltd.) were co-transfected with 7 μg of AAV transfer plasmid (MEX3B KD or NC), 14 μg pHelper (Takara Bio Inc.), and 7 μg pRC9 (Takara Bio Inc.) using Lipo3000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Following incubation for 72 h at 37°C, the cells were lysed to generate the AAV9 particles. The sequences of shRNA targeting MEX3B or NC were as follows: MEX3B, sense: 5'-GGGACAGACCACTATCCAAGT-3' and anti-sense: 5'-ACTTGGATAGTGGTCTGTCCC-3'; NC, sense: 5'-TTCTCCGAACGTGTCACGT-3' and anti-sense: 5'-ACGTGACACGTTCGGAGAA-3'. The AAV9 particles (8×10¹¹ viral genome/ml; 200 μl) were intravenously injected into the tail, followed by TAD induction after 3 days, as previously described (21). A double-blind method was used.

There were 88 mice: 12 mice for RNA-seq and 76 for subsequent experiments. For RNA-seq, in the sham group, aortic tissue from 5 mice were used for hematoxylin and eosin (H&E) staining. In the TAD group, 7 mice were initially enrolled; 2 died during the experiments and 2 were excluded because they did not develop TAD, indicating modeling failure. Thus, 3 mice were available for H&E staining and RNA-seq. For the subsequent experiments, the sham group consisted of 12 mice. The TAD and TAD + NC group each had 22 mice, while the TAD + MEX3B KD group had 20 mice. In the sham group, 6 mice were used for reverse transcription-quantitative (RT-q) PCR and western blot analysis and 6 were used for echocardiography followed by fixation for immunohistochemistry, immunofluorescence and pathological staining. In the TAD group, 4 mice died and 4 were excluded due to failure of TAD, leaving 14 mice; among these, 2 were used for pilot studies, 6 were used for RT-qPCR/western blotting and 6 were fixed for pathological examination. In the TAD + NC group, 5 died and 4 were excluded due to failure of TAD, leaving 13 mice, of which one was used for pilot studies, 6 for RT-qPCR/western blotting and 6 for pathological examination. In the TAD + KD group, 3 died and 5 were excluded due to failure of TAD, leaving 12 mice, with 6 for RT-qPCR/western blotting and 6 for pathological examination.

In total, 14 mice died spontaneously. These mice were found dead in their cages without prior signs of prolonged distress. Upon autopsy, all mice presented with massive blood clots in the thoracic cavity, confirming rupture of the dissected aorta as the direct cause of death; aortic rupture is a well-documented and inherent complication of this model (22-25).

Human VSMCs isolated from thoracic aortic tissue were obtained from Procell Life Science & Technology Co., Ltd. (cat. no. CP-H081). Cells were cultured with iCell primary smooth muscle cell culture system (iCell Bioscience Inc.) under 5% CO₂ at 37°C. VSMCs were not contaminated by mycoplasma or other cell lines, demonstrated by STR profiling and mycoplasma detection and authentication.

A small interfering RNA (siRNA) pool with three siRNAs targeting MEX3B or versican (VCAN) was synthesized by JTS Scientific and transfected into VSMCs using Lipo3000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Briefly, cells were incubated with 75 pmol siRNA, 250 μl Opti-MEM and 7.5 μl lipo3000 at 37°C. After 24 h, VSMCs were treated with 100 nM Ang II for 48 h. The sequences of siRNAs targeting MEX3B or VCAN are listed in Table I. For the detection of VCAN mRNA expression, transfected VSMCs were treated with 10 μg/ml actinomycin D for 0, 4, 8 and 12 h at 37°C prior to RT-qPCR.

The data of MEX3A/MEX3B expression in aortic tissues of patients with TAD or normal samples was obtained from the National Center for Biotechnology Information (NBCI) Gene Expression Omnibus (GEO) database (accession no. GSE241265) (26,27). GSM7720285-GSM7720287 were regarded as the TAD group and GSM7720288-GSM7720290 were regarded as the control group.

TRIpure Reagent (BioTeke Corporation; cat. no. RP1001) was employed to obtain the total RNA of aortic tissue or VSMCs following the manufacturer's protocols. RNA-seq was performed by Novegene Bioinform Technology Co., Ltd. RNA integrity was assessed using the Agilent 5400 Fragment Analyzer system (Agilent Technologies). Libraries were sequenced on the Illumina Novaseq X Plus platform (Illumina, Inc.) in paired-end mode with a read length of 150 bp (PE150). Sequencing was performed using the Novaseq X series 25B Reagent Kit (300 cycles; cat. no. 20104706; Illumina, Inc.). The loading concentration of the final library was 4 pM, as determined by qPCR. Transcriptome reads from RNA-seq were mapped to the reference genome (hg38 and mm39) (28,29). Data were analyzed using R v4.3.2 (r-project.org/) and org.Hs.eg. db v3.18.0 (bioconductor.org/packages/3.18/data/annotation/html/org.Hs.eg.db.html). The significance was set at log₂ fold-change (FC) >1 and P<0.05. Differentially expressed genes were subjected to functional enrichment analysis by Gene Ontology (GO), Kyoto Encyclopedia of genes and genomes (KEGG) (30-32). Gene set enrichment analysis (GSEA) was performed using the GO gene sets and KEGG pathway gene sets (33). Principal component analysis (PCA) was performed to analyze the sample clustering. The RNA-seq data were uploaded to the NCBI GEO database with accession no. GSE315273.

Total mRNA was converted to cDNA using the All-in-One First-Strand SuperMix (Magen Biotechnology) according to the manufacturer's protocol. Amplification was performed using the 2X Fast Taq plus PCR Master Mix (Biosharp Life Sciences) and SYBR Green (Beijing Solarbio Science & Technology Co., Ltd.) following the manufacturers' protocols. Thermocycling conditions were as follows: Initial denaturation at 95°C for 5 min; 40 cycles at 95°C for 10 sec, 60°C for 10 sec and 72°C for 15 sec, then 72°Cfor 90 sec, 40°C for 1 min and a final melt at 60-90°C (every 1°C, 1 sec). Raw data were normalized to GAPDH, and the relative expression levels of targeted genes were calculated using 2^−ΔΔCq method (34). The sequences of primers are listed in Table II.

Total protein from aortic tissues or VSMCs was obtained using the RIPA lysis solution and phenylmethanesulfonyl fluoride (100:1, v/v; Beyotime Biotechnology). The concentration of protein was quantified using a BCA protein assay kit (Beyotime Biotechnology) according to the manufacturer's protocols. The proteins (20 μg/lane) were separated with 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred to polyvinylidene fluoride membranes (Abcam). Following blocking with 1xQuickBlock Blocking Buffer (Beyotime Biotechnology) for 60 min at room temperature, proteins were incubated with the following primary antibodies: Anti-MEX3B (1:300; Santa Cruz Biotechnology, Inc.; cat. no. sc-515833), anti-VCAN (cat. no. A19655), anti-OPN (cat. no. A1361), anti-myosin heavy chain 11 (SM-MHC; all ABclonal Biotech; cat. no. A10827) and anti-GAPDH (all 1:1,000; Proteintech Group, Inc.; cat. no. 60004-1-Ig) at 4°C overnight. Membranes were incubated with the secondary antibodies including horseradish peroxidase-conjugated goat anti-rabbit (cat. no. A0208) or anti-mouse IgG (both 1:5,000; both Beyotime Biotechnology; A0216) at 37°C for 45 min. The bands were visualized using a BeyoECL Plus Enhanced Chemiluminescence Reagent (Beyotime Biotechnology). Strip analysis was performed using Tanon Image software version 1.10 (Tanon Science & Technology Co., Ltd.).

For H&E staining, murine aortic tissue was fixed with 4% paraformaldehyde at 4°C for 24 h, followed by embedding with paraffin and sliced into sections (5 μm). Next, the sections were deparaffinized in xylene and rehydrated in an ethanol gradient with distilled water. The sections were stained with hematoxylin for 5 min and eosin for 3 min at room temperature. The sections were observed with a BX53 light microscope (Olympus Corporation) in brightfield mode.

Elastic Van Gieson (EVG) staining was performed to visualize ELN fibers. Briefly, sections of aortic tissue were obtained as aforementioned. Then, sections were stained with Verhoeff solution and eosin at room temperature. A BX53 light microscope was used to capture the images in brightfield mode.

For immunofluorescent staining, murine aortic tissue was fixed with 4% paraformaldehyde at 4°C for 24 h, followed by embedding with paraffin and sliced into sections at a thickness of 5 μm. Next, the sections were deparaffinized in xylene and rehydrated in an ethanol gradient with distilled water. The sections were placed in boiling antigen retrieval solution and retrieved in a microwave oven at low power for 10 min. Following blocking with 1% bovine serum albumin at room temperature for 15 min, sections were incubated with primary antibodies including anti-MEX3B (1:50; Santa Cruz Biotechnology, Inc.; cat. no. Sc-515833) or anti-α-smooth muscle actin (SMA; 1:100; Cell Signaling Technology, Inc.; cat. no. 19245) at 4°C overnight, followed by incubation with Alexa Fluor 555-conjugated anti-rabbit IgG (green) or Alexa Fluor 488-conjugated anti-mouse IgG (red; both 1:200; both Cell Signaling Technology, Inc.; cat. no. 4413 and 4408) at room temperature for 90 min away from the light. The sections were counterstained with DAPI. A BX53 fluorescence microscope was used to capture the images.

For immunohistochemical staining, murine aortic tissue was fixed with 4% paraformaldehyde at 4°C for 24 h, followed by embedding with paraffin and then sliced into sections at a thickness of 5 μm thickness. Next, the sections were deparaffinized in xylene and rehydrated in an ethanol gradient with distilled water. The sections were placed in boiling antigen retrieval solution and retrieved in a microwave oven at low power for 10 min. After washed with phosphate buffered saline, sections were treated with 3% H₂O₂ for 15 min at room temperature and blocked with 1% bovine serum albumin at room temperature for 15 min. Then, sections were incubated with primary antibody anti-proliferating cell nuclear antigen (PCNA; 1:1,500; Proteintech Group, Inc.; cat. no. 10205-2-AP) at 4°C overnight and secondary antibody horseradish peroxidase-conjugated goat anti-rabbit IgG (1:200; Sangon Biotech Co., Ltd.; cat. no. D110058) at room temperature for 60 min. Sections were treated with 100 μl DAB (Beijing Solarbio Science & Technology Co., Ltd.) and stained with hematoxylin (Beijing Solarbio Science & Technology Co., Ltd.; cat. no. H8070) at room temperature for 3 min. Images were captured with a BX53 light microscope in brightfield mode (magnification, ×400).

VSMCs were fixed with 4% paraformaldehyde at 4°C for 15 min and incubated with 0.1% TritonX-100 for 30 min at room temperature. Cells were washed with PBS and incubated with 1% bovine serum albumin at room temperature for 15 min. VSMCs were incubated with anti-α-SMA (1:100) at 4°C overnight and Alexa Fluor 555 conjugated anti-rabbit IgG (1:200; both Cell Signaling Technology, Inc.; cat. no. 19245 and 4413) at room temperature for 60 min, followed by counterstaining with DAPI. Images were with captured with a BX53 fluorescence microscope (magnification, ×200).

VSMCs were centrifuged at 300 g for 10 min at 4°C, and the supernatant was collected. The levels of MMP2, MMP9 and ELN in the supernatant of VSMCs were detected using the Human MMP2 and MMP9 (both MultiSciences Biotech; cat. nos. EK1M02 and EK1M09) and ELN ELISA kits (Wuhan Fine Biotech Co., Ltd.; cat. no. EH1505) according to the manufacturers' instructions.

A total of 5×10³ VSMCs/well were seeded into the 96-well plates. Cells were treated with 10 μl CCK-8 solution (Beijing Solarbio Science & Technology Co., Ltd.) for 2 h. Absorbance was measured at 450 nm with an 800TS microplate reader (BioTek; Agilent Technologies, Inc.).

The proliferative ability of VSMCs was assessed using the E-Click EdU Cell Proliferation Flow Cytometry Assay kit (Elabscience Bionovation Inc.; cat. no. E-CK-A370) according to the manufacturer's instructions. Briefly, VSMCs were incubated with 10 μM EdU solution at 37°C for 2 h, followed by fixation with 4% paraformaldehyde at room temperature for 15 min. Cells were washed and treated with Click solution at room temperature for 30 min, followed by detection using a NovoCyte flow cytometer (ACEA Biosciences, Inc.; Agilent Technologies, Inc.), and data analysis with NovoExpress software version 1.4.1 (ACEA Biosciences, Inc.).

VSMCs were seeded into a 6-well plate and treated with 10 μg/ml mitomycin C (Millipore Sigma) for 1 h at 37°C. When the cells reached confluence (80%, a 200 μl pipette tip was used to create a scratch. Cells were washed with a serum-free iCell Primary Smooth Muscle Cell Culture Medium (Cellverse Bioscience Technology Co., Ltd.,) and cultured for another 24 h. The images at 0 and 24 h were captured using an IX53 light microscope. Wound closure was calculated as: (de-d)/de ×100%, where de is the initial wound width at 0 h and d is the residual width at 24 h.

A total of 200 μl cell suspension (within 6×10³ cells) resuspended in a serum-free iCell Primary Smooth Muscle Cell Culture Medium was planted into the upper chamber of a 24-well plate. VSMCs were allowed to migrate to the lower chamber, which contained 800 μl iCell Primary Smooth Muscle Cell Culture Medium with 10% fetal bovine serum (Cellverse Bioscience Technology Co., Ltd.) at 37°C. After 24 h, VSMCs that migrated to the lower chamber were fixed with 4% paraformaldehyde at room temperature for 20 min and stained with 0.5% crystal violet (Amresco, LLC) at room temperature. The stained cells were observed with an IX53 light microscope and counted with Image J software version 1.51 j8 (National Institutes of Health).

A RIP assay was performed using the EZ-Magna RIP kit (MilliporeSigma (cat. no. 17-701) according to the manufacturer's protocol. In brief, cells were collected and washed with pre-cooled PBS following by centrifuging at 720 × g for 10 min at 4°C. Then, the cells were lysed using the RIP lysis buffer. The magnetic beads were conjugated with 5 μg anti-MEX3B antibody (Santa Cruz Biotechnology, Inc.; cat. no. sc-515833) or normal mouse IgG (1:500; Beyotime Biotechnology; cat. no. A7028) at room temperature for 30 min. The cell lysates were centrifuged at 13,618 g for 10 min at 4°C and 100 μl of supernatant was incubated with the prepped magnetic beads at 4°C overnight. The magnetic beads were separated using a magnetic stand. Beads were digested with 18 μl of proteinase K and 15 μl of 10% sodium dodecyl sulfate at 55°C for 30 min. The immunoprecipitated RNA was analyzed by RT-qPCR employing specific primers targeting VCAN, as aforementioned. The primers were synthesized by Beijing Solarbio Science & Technology Co., Ltd., and the sequences were as follows. Forward 5'-TGGCTTTCATAACACTA-3'; Reverse 5'-CATTCTATCCGTCACAT-3'.

293T cells (Cellverse Bioscience Technology Co., Ltd.) were planted into 12-well plates and cultured at 37°C for 24 h. The cells were transfected with 1 μg pmirGLO (Fenghuishengwu Co., Ltd.)-VCAN 3'untranslated region (UTR) and 1 μg MEX3B overexpression or MEX3B mutKH vector using the Lipo3000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocols. MEX3B mutKH vector expresses an RNA-binding-deficient MEX3B mutant, in which key residues associated with RNA binding, including G83 and G177 in the KH domains, were replaced with D (15). After 48 h transfection at 37°C, a dual-luciferase reporter assay was performed using Dual Luciferase Reporter Gene Assay kit (Biosharp Life Sciences) according to the manufacturer's protocols. Firefly/Renilla luminescence ratios were normalized to control.

The data were analyzed by GraphPad Prism software version 8.0 (Dotmatics) and presented as mean ± standard deviation (n=3 for the in vitro experiments; n=6 for the in vivo experiments) as previously described (21). A post hoc power analysis was conducted using G*Power 3.1.9.7 (Heinrich Heine University Düsseldorf, Düsseldorf, Germany) to evaluate the sufficiency of the sample size. Given α=0.05, n=6/group and MMP2 concentration was 4.60±0.57 ng/ml (Sham) vs. 9.57±1.56 ng/ml (TAD) vs. 10.16±1.51 ng/ml (TAD + NC) vs. 6.02±1.06 ng/ml (TAD + MEX3B KD), the power was 1.00, which exceeded the conventional threshold of 0.9. RNA-seq. The data were tested for normality test and homogeneity of variance. Unpaired Student's t-test or one-way ANOVA with post hoc Bonferroni's correction were utilized to analyze the differences. Effect size was calculated to determine the power of the analysis, as previously described (35). Student's t-tests are suitable when sample size is small and effect size is large (36,37). P<0.05 was considered to indicate a statistically significant difference.

To understand the molecular mechanism of TAD pathogenesis, RNA-seq was performed to analyze the differentially expressed genes in thoracic aortic tissue from TAD mice and controls. The murine TAD model was established using BAPN and Ang II (Fig. 1A). H&E staining demonstrated tearing of the aortic wall as well as formation of thrombi in the false lumens, indicating TAD formation (Fig. 1B). Total RNA from the thoracic aortic tissues was subjected to sequencing (Fig. 1C). PCA demonstrated that the samples in the TAD and sham groups were well clustered, indicating each group had a distinctive expression (Fig. 1D). There were 1,642 up- and 1,199 downregulated genes (|log₂FC|>1; P<0.05; Fig. 1E). Next, GSEA was performed based on GO and KEGG analysis. GO functional enrichment analysis indicated that these genes were mainly enriched in 'vascular wall composition', 'elastin catabolic process'), 'protein-RNA binding' ('protein-RNA complex organization' and 'RNA binding'), 'cell population proliferation' ('positive regulation of smooth muscle cell proliferation') and 'cell migration' (Fig. 1F). KEGG analysis by GSEA suggested that TAD was associated with 'vascular smooth muscle contraction', 'PI3K-Akt signaling pathway' and 'cytokine-cytokine receptor interaction' (Fig. 1F). These enriched terms demonstrated the potential pathogenesis of TAD.

RNA-binding proteins are key components of post-transcriptional regulatory mechanism, which are involved in numerous steps of RNA metabolism including RNA shearing, RNA stability and RNA translation (38). RNA binding proteins are associated with the cardiovascular system (13). RNA-binding proteins serve a key role in cardiovascular diseases including TAD (39,40). As RNA-seq demonstrated genes enriched in protein-RNA binding, the role of RNA binding proteins in TAD was assessed. There were 92 differently expressed genes enriched in 'protein-RNA complex organization and RNA-binding (Fig. 1G). RNA recognition motif (RRM) and the K homology (KH) domain are the most common motifs in RNA binding proteins (41). There were 15 factors with RRM domains and two with KH domains. Notably, all these factors with RRM or KH domains were upregulated in aortic tissue of TAD mice compared with the Sham group (Fig. 1H). MEX3A and MEX3B were the only two members containing KH domain. MEX3B was markedly upregulated in aortic tissue of patients with TAD according to GSE241265, whereas MEX3A was not (Fig. 1I), suggesting MEX3B may serve a role in the progression of TAD.

To investigate the effects of MEX3B on TAD, MEX3B KD was performed via AAV9 vectors intravenously injected into mice prior to TAD induction (Fig. 2A), leading to the reversal of the MEX3B upregulation in thoracic aortic tissue of TAD mice (Fig. 2B and C). Immunofluorescence staining demonstrated that MEX3B colocalized with α-SMA in the aortic wall of mice with TAD, and the colocalization between MEX3B and α-SMA was decreased in the medial layer following AAV9 vector injection (Fig. 2D). Echocardiography demonstrated that the thoracic aorta was markedly dilated in TAD mice; this effect was reduced by MEX3B KD (Fig. 2E). The aortas of TAD mice were enlarged from the root to the thoracic segment. Hematomas were observed in the lesions, suggesting thrombosis. These pathological changes were alleviated by MEX3B KD (Fig. 2F). Next, H&E staining displayed that interfering with MEX3B expression preserved the integrity of the aortic structure and prevented the thrombi in the false lumens (Fig. 2G). EVG staining confirmed the depletion and fragmentation of elastic fibers in TAD mice, which was mitigated by MEX3B KD (Fig. 2H). The data suggested that MEX3B KD alleviates TAD formation in vivo.

The function of MEX3B in TAD-induced VSMC phenotype transformation was explored in vivo. MEX3B KD reversed the TAD-induced downregulation of contractile markers myocardin and calponin 1 and upregulation of synthetic marker OPN (Fig. 3A). Loss of MEX3B upregulated the protein expression of myocardin and calponin 1 and downregulated the protein levels of OPN (Fig. 3B). MEX3B KD also upregulated the protein expression of contractile marker SM-MHC (Fig. 3B). Immunohistochemical staining demonstrated that TAD induced the upregulation of PCNA of VSMCs in the aortic medial layer, indicating the induction of the synthetic phenotype of VSMCs, while the loss of MEX3B decreased the expression of PCNA (Fig. 3C).

To explore the function of MEX3B, a siRNA pool targeting MEX3B was synthesized and transfected into VSMCs, followed by the establishment of the in vitro model with Ang II (Fig. 4A). Ang II-induced upregulation of MEX3B was decreased in VSMCs transfected with MEX3B siRNA pools (Fig. 4B and C). The KD effects of individual siRNAs are shown in Fig. S1A: MEX3B was downregulated in VSMCs transfected with MEX3B siRNA1-3 (Fig. S1A). The mRNA expression of myocardin and calponin 1 was upregulated and OPN was downregulated in MEX3B-depleted VSMCs (Fig. 4D). Similar changes were observed in the protein expression of these factors (Fig. 4E). SM-MHC was upregulated by MEX3B KD (Fig. 4E). Immunofluorescence demonstrated that MEX3B KD reversed the downregulation of α-SMA induced by Ang II (Fig. 4F). The arterial medial layer consists of VSMCs and extracellular matrix (42). ELN and collagen are the most abundant components of extracellular matrix in the aorta and responsible for the maintenance of aortic mechanical properties (8). Changes in ELN and the degradation factors MMP2 and MMP9 were measured to investigate the degradation of extracellular matrix. The loss of MEX3B elevated ELN and decreased MMP2 and MMP9 levels (Fig. 4G). These results suggested that MEX3B KD blocks Ang II-induced VSMC phenotype switching and ELN degradation in vitro.

To investigate phenotype transformation, a CCK-8 assay and EdU staining were used to assess the proliferative ability of VSMCs. Ang II-induced increased cell viability was suppressed in MEX3B KD cells (Fig. 5A). Similar proliferation inhibition was observed by EdU staining assay (Fig. 5B). Next, wound healing and Transwell assays were used to investigate the migratory ability of VSMCs. MEX3B KD weakened the Ang II-induced migratory ability (Fig. 5C-F). These data indicated that MEX3B KD dampens Ang II-induced phenotype switching of VSMCs.

RNA-seq was performed to analyze the differentially expressed genes in VSMCs with MEX3B KD in the presence of Ang II (Fig. 6A). PCA revealed that NC and MEX3B KD groups were well clustered in space, indicating each group had distinctive expression (Fig. 6B). There were 135 up- and 107 downregulated genes (|log₂FC|>1; P<0.05; Fig. 6C). GO functional analysis was performed based on biological process, cellular component and molecular function. The differentially expressed genes were primarily enriched in items related to cardiovascular function. The GO items associated with the pathogenesis of TAD were screened, including 'extracellular matrix structural constituent conferring compression resistance', 'collagen-containing extracellular matrix', 'regulation of smooth muscle cell migration' and 'negative regulation of smooth muscle cell proliferation', suggesting the potential function of downstream factors of MEX3B (Fig. 6D). There were nine genes enriched in >1 pathway (Fig. 6E). VCAN is an aggregating proteoglycan, accumulation of which aggravates the TAD and the synthetic phenotype switching (43,44). In the present study, VCAN was upregulated in aortic tissue of TAD mice and downregulated in VSMCs with MEX3B KD (Fig. 6F), prompting investigation of the regulation between MEX3B and VCAN.

Consistent with the results of RNA-seq, both mRNA and protein expression of VCAN were downregulated by the loss of MEX3B (Fig. 7A). Similar results were also observed in aortic tissue of TAD mice with MEX3B KD (Fig. 7B and C). As an RNA binding protein, MEX3B regulates the mRNA stability of targets by binding to their 3'UTR. A similar mechanism between MEX3B and VCAN was confirmed by RIP-PCR assay. The RNA fragment bound to MEX3B antibody (Fig. 7D). Next, MEX3B-overexpression vectors were transfected into VSMCs, leading to the upregulation of MEX3B (Fig. 7E). Dual-luciferase reporter assay was performed using a vector expressing an RNA-binding deficient MEX3B mutant (MEX3B mutKH), in which key residues that participate in RNA binding, including G177 and G83 in KH domains, are replaced with D. MEX3B overexpression-induced elevated relative luciferase activity was decreased by MEX3B mutKH (Fig. 7F). Whether MEX3B control the mRNA stability of VCAN was investigated. Following actinomycin D treatment, VCAN mRNA levels declined in a time-dependent manner, with a faster decay upon MEX3B knockdown and a slower decay following MEX3B overexpression (Fig. 7G). These data suggested that MEX3B potentiates VCAN mRNA stability.

The present study assessed whether the effects induced by MEX3B were dependent on its RNA-binding functions. MEX3B overexpression induced the upregulation of VCAN, whereas the expression of VCAN was not changed by MEX3B mutKH (Fig. 8A). The upregulation of MEX3B induced the synthetic phenotype of VSMCs in the presence of Ang II, as evidenced by the downregulation of SM-MHC and the upregulation of OPN (Fig. 8B and C), as well as the enhanced cell viability and migration (Fig. 8D and E). The ELN levels were decreased in VSMCs with MEX3B overexpression (Fig. 8F). These phenotypes were not significantly changed by MEX3B mutKH (Fig. 8B-F). The data indicated that MEX3B regulates the phenotype transformation of VSMCs in an RNA-binding-dependent manner.

As loss of MEX3B blocked the abnormal VSMC phenotype switching, VCAN might serve a similar role as the downstream RNA substrates of MEX3B. siRNAs targeting VCAN were synthesized and transfected into VSMCs, resulting in the downregulation of VCAN (Figs. 9A and B and S1B). VCAN KD inhibited the Ang II-induced VSMC phenotype transformation, as evidenced by the upregulation of SM-MHC and downregulation of OPN (Fig. 9C). The concentration of ELN was increased by loss of VCAN (Fig. 9D). The proliferative and migratory abilities were suppressed by VCAN depletion (Fig. 9E and F).