Abstract

   Vascular smooth muscle cell (VSMC) senescence is a pivotal driver of
   atherosclerosis (AS), but molecular links to ageing‐related dysfunction
   are unclear. It is aimed to identify regulators of VSMC senescence and
   develop clinical interventions for ageing‐related AS. Using single‐cell
   RNA sequencing of human atherosclerotic carotid arteries and
   immunofluorescence validation, activating transcription factor 3 (ATF3)
   is identified as central to VSMC senescence. Mechanistic studies employ
   SMC‐specific ATF3 knockout mice, CUT&Tag‐seq, RNA/protein interaction
   assays, and m6A epitranscriptomic analyses. To bridge discovery to
   therapy, high‐throughput virtual screening is performed for
   ATF3‐targeting compounds and functionally validated hits. ATF3
   deficiency in VSMCs accelerates ageing‐induced AS by promoting
   senescence. Multi‐omics showed ATF3 activates ATG7, triggering
   autophagy, while cytoplasmic ATG7 enhances ATF3 nuclear translocation,
   establishing a positive feedback loop. Ageing increases m6A methylation
   and decreases the stability of Atf3 mRNA. Terazosin (TZ) diminishes the
   interaction between YTH N6‐methyladenosine RNA binding protein F2
   (YTHDF2) and Atf3 mRNA, helping to preserve Atf3 mRNA stability. TZ is
   a promising therapeutic strategy for delaying VSMC senescence and
   preventing AS. ATF3 protects against VSMC senescence and AS by
   orchestrating autophagy via a novel ATF3‐ATG7 amplification loop.
   Repurposing TZ to stabilize ATF3 offers a translatable approach to
   combat ageing‐driven cardiovascular disease.

   Keywords: activated transcription factor 3, atherosclerosis, autophagy,
   terazosin, vascular smooth muscle cells senescence
     __________________________________________________________________

   Targeting ATF3 with terazosin to against VSMC Senescence and
   atherosclerosis. TZ reduces the binding affinity of YTHDF2 to Atf3
   mRNA, while enhancing mRNA stability. ATF3 promotes Atg7 transcription,
   enhancing autophagy. ATG7 binds to ATF3 in the cytoplasm, facilitating
   ATF3 translocation to the nucleus and establishing positive feedback.
   By modulating ATF3, VSMCs' senescence can be improved, which in turn
   inhibits atherosclerosis.

   graphic file with name ADVS-12-e02249-g008.jpg

1. Introduction

   Atherosclerosis (AS) is an age‐associated vascular disorder.^[ [46]^1 ,
   [47]^2 ^] Vascular smooth muscle cells (VSMCs) constitute a major
   cellular component of atherosclerotic plaques and are characterized by
   senescence and increased apoptosis.^[ [48]^3 ^] VSMC senescence
   accelerates VSMC phenotypic switching,^[ [49]^4 ^] resulting in
   pathological structural deterioration of vessels and increased vascular
   stiffness.^[ [50]^5 , [51]^6 , [52]^7 ^] Hence, protecting VSMCs from
   senescence has emerged as a promising therapeutic target for treating
   AS. However, hardly any pharmacological intervention has been proven to
   be safe and effective for reversing VSMC senescence and AS in clinical
   practice. Therefore, pharmacological treatment strategies for reversing
   VSMC senescence and AS are needed clinically.

   The Activator Protein 1/Activating Transcription Factor (AP1/ATF)
   family, which includes key members such as Proto‐oncogene c‐Fos,
   Proto‐oncogene c‐Jun, and ATF1 to ATF6, has been extensively studied
   for its critical roles in cardiovascular diseases.^[ [53]^8 ^] Our
   analysis of single‐cell RNA sequencing (scRNA‐seq) data from AS
   patients indicates that ATF3 may be a potential regulatory molecule for
   AS. ATF3 participates in extensive stress responses, and its proper
   activity is essential for sustaining normal physiological functions in
   cells.^[ [54]^9 ^] However, the role of ATF3 in senescence across
   different cell types and disease contexts is controversial. In certain
   models, ATF3 promotes senescence by activating senescence‐associated
   endogenous retroviruses and inducing an antiviral defense response.^[
   [55]^10 ^] In fibroblasts, ATF3 drives senescence‐associated gene
   expression by remodeling chromatin accessibility.^[ [56]^11 ^]
   Furthermore, ATF3 was upregulated in the VSMCs of murine aortic
   aneurysm and dissection models, where it suppressed apoptosis by
   inhibiting the DNA damage response.^[ [57]^12 ^] Further studies are
   essential to unravel its context‐dependent mechanisms in age‐related
   pathologies.

   Autophagy, a vital physiological process, promotes efficient cellular
   recycling by transporting dysfunctional cellular components to
   lysosomes for degradation and metabolic processing, supporting cell
   survival.^[ [58]^13 ^] The primary type of autophagy is autophagosome
   (AP) formation, which is facilitated by microtubule‐associated protein
   1 light chain 3 (MAP1LC3/LC3).^[ [59]^14 ^] Autophagy‐related proteins
   such as autophagy‐related 3 and autophagy‐related 7 (ATG7) facilitate
   this process.^[ [60]^15 , [61]^16 ^] Reduced autophagy is associated
   with cardiovascular functional decline and increased susceptibility to
   cardiovascular diseases in older adults.^[ [62]^17 ^] Growing evidence
   suggests that enhancing autophagy can mitigate vascular ageing.^[
   [63]^18 , [64]^19 ^] Enhancing autophagy to slow down vascular ageing
   is an emerging novel research field.

   Terazosin (TZ), a widely used medicine in clinical practice, reduces
   smooth muscle tension and is utilized for managing benign prostatic
   hyperplasia and hypertension. Recent studies have revealed that TZ has
   anti‐apoptotic properties independent of its ability to block
   α1‐adrenoceptors.^[ [65]^20 ^] TZ has also been shown to reduce the
   risk of age‐related neurodegenerative diseases. TZ has therapeutic
   potential for Parkinson's disease^[ [66]^21 ^] and reduces amyloid‐beta
   production in Alzheimer's disease by inhibiting glycogen synthase
   kinase 3 beta.^[ [67]^22 , [68]^23 ^] We found that the novel mechanism
   of TZ seems to involve altering ATF3 levels to exert anti‐ageing
   effects.

   This study aimed to explore the anti‐aging and anti‐AS effects of ATF3,
   as well as the underlying mechanisms.

2. Results

2.1. AS is Associated with Decreased ATF3 Expression and VSMC Senescence

   To identify potential AS‐related candidate genes within the AP1/ATF
   transcription factor family, single‐cell RNA sequencing analysis was
   conducted on twelve human carotid artery samples from the public GEO
   database ([69]GSE253903), including six cases with carotid
   atherosclerotic plaques. Based on the marker genes of each cluster,
   cells from carotid atherosclerotic plaques were primarily classified
   into nine cell types (Figure [70]1A; Figure [71]S1C, Supporting
   Information). Among the various cell types, the AP1 family exhibited
   the greatest fold change in ATF3 expression, especially in VSMCs
   (Figure [72]1B; Figure [73]S1D, Supporting Information).

Figure 1.

   Figure 1
   [74]Open in a new tab

   Vascular smooth muscle cells (VSMCs) in atherosclerosis (AS)
   simultaneously display signs of senescence and a decline in ATF3
   content. A) t‐SNE analysis of single cells from the carotid arteries of
   atherosclerosis patients, with 9 major cell types labeled in different
   colors. B) The violin plot illustrating the variation in levels of ATF3
   across different cell types. C) IF imaging of ATF3 (red), α‐SMA
   (green), CD31(orange), and nuclei (blue) in human carotid arteries
   (scale bars = 20 µm; n = 5). D) IF imaging of p53 (red), α‐SMA (green),
   and nuclei (blue) in human carotid arteries (scale bars = 20 µm;
   n = 5). E) Feature plot illustrating the expression distribution of
   ATF3 and p53 in VSMCs. F) Scatter plot illustrating the correlation
   between p53 and ATF3 expression levels. G) Experimental design:
   SMC‐specific knockdown of ATF3 in SAMP8 mice (n = 6). H) Pulse wave
   velocity (PWV) in SAMP8 mice (n = 6). I) SA‐β‐gal staining of mouse
   aortas (scale bars = 5 mm, n = 4). J) Western blot analysis of ATF3,
   p53, and p21 expression in SAMP8 mice aortas (n = 5). K) Experimental
   design: SMC‐specific knockdown of ATF3 in ApoE^‐/‐ mice. After 4 weeks,
   mice fed with high‐fat diet for 3 months (n = 8). L) PWV in ApoE^‐/‐
   mice (n = 6). M) SA‐β‐gal staining of mouse aortas (scale bars = 5 mm,
   n = 4). N) Western blot analysis of ATF3, p53, and p21 expression in
   ApoE^‐/‐ mice aortas (n = 5). O) Oil Red O staining and Hematoxylin and
   eosin staining of mouse aortas (scale bars = 100 µm, n = 5). Error bars
   represent the mean ± standard deviation. The unpaired t‐test was used
   to compare data; *p < 0.05, **p < 0.01, ***p < 0.001.

   Carotid atherosclerotic samples from patients with AS were collected.
   VSMCs were identified through immunofluorescence (IF) staining using
   the SMC‐specific marker alpha‐smooth muscle actin (α‐SMA). Endothelial
   cells (ECs) were labeled with platelet endothelial cell adhesion
   molecule‐1 (CD31).^[ [75]^24 ^] IF analysis further revealed that ATF3
   expression was downregulated in VSMCs but not in ECs (Figure [76]1C).
   On the other hand, VSMCs within atherosclerotic plaque areas exhibited
   pronounced senescence, as evidenced by upregulated expression of the
   senescence markers p53 and p21 (Figure [77]1D; Figure [78]S1E,
   Supporting Information). VSMC phenotypic modulation drives the
   progression of AS^[ [79]^25 ^] and represents a key hallmark of VSMC
   senescence.^[ [80]^26 ^] The phenotypic plasticity of VSMCs drives
   alterations in the architecture and physiology of the vessel wall,
   contributing to the progression of vascular stiffening.^[ [81]^27 ^]
   The IF assay revealed downregulation of a contractile phenotypic
   marker, smooth muscle protein 22α (SM22α), and upregulation of the
   synthetic marker osteopontin (OPN) compared with those in healthy
   controls (Figure [82]S1E, Supporting Information).

   To further confirm the above findings, we collected aortic specimens
   from high‐fat diet (HFD)‐fed ApoE^‐/‐ mice and found that the decrease
   in ATF3 was specific to VSMCs (Figure [83]S2A, Supporting Information)
   and that VSMCs in atherosclerotic plaques of the aorta underwent
   senescence and phenotypic transformation compared with those in the
   control group (Figure [84]S2B, Supporting Information). We identified
   in age‐stratified mouse cohorts that vascular senescence progressively
   intensifies with age, alongside elevated expression of p21 and p53
   proteins and a concomitant decline in ATF3 mRNA/protein levels (Figure
   [85]S2C,D, Supporting Information). We were interested in the
   association between ATF3 and VSMC senescence, and we observed that the
   expression distributions of ATF3 and p53 in the VSMCs trended in
   opposite directions (Figure [86]1E). Additionally, a negative
   correlation was observed between ATF3 and p53 levels (Figure [87]1F).
   Therefore, ATF3 might be an important molecule involved in the
   regulation of VSMC senescence and AS.

2.2. ATF3 Deficiency in VSMCs Exacerbates VSMC Senescence and AS

   To evaluate whether ATF3 deficiency contributes to VSMC senescence, we
   delivered shRNA‐ATF3 via recombinant adeno‐associated virus serotype 9
   vectors using the SM22α promoter (specific to SMCs) to knock down ATF3
   expression (Figures [88]S2E,F,S3A, Supporting Information).
   Senescence‐accelerated mouse prone 8 (SAMP8) mice were selected as a
   model to study age‐related changes in the vasculature (Figure [89]1G),
   and they displayed a disruption in vascular homeostasis and a markedly
   elevated incidence of atherosclerotic development compared with
   senescence‐accelerated mouse resistance 1 (SAMR1) mice starting at
   10–14 months.^[ [90]^28 ^] Pulse wave velocity (PWV) was used to assess
   vascular stiffness, as increased stiffness reflects aggravated VSMC
   senescence and vascular ageing.^[ [91]^29 ^] Compared with
   SAMP8+GFP^∆SMC mice, SAMP8+shATF3^∆SMC mice significantly increased
   vascular stiffness (Figure [92]1H). Knockdown of ATF3 exacerbated VSMC
   senescence, as evidenced by an increased SA‐β‐gal‐positive staining
   area and elevated levels of senescence markers (Figure [93]1I,J). To
   illustrate the effect of ATF3 on ageing‐induced AS, first, a SAMP8
   mouse model (vs SAMR1 controls) was used to explicitly evaluate the
   interplay between vascular ageing and AS susceptibility. Compared to
   age‐matched SAMR1 mice, SAMP8 mice exhibited no significant
   atherosclerotic features prior to 3 months of age (Figure [94]S3B,
   Supporting Information). However, with increasing age, distinct plaque
   formation was detected through Oil Red O staining of the aortic roots
   of 14‐month‐old SAMP8 mice (Figure [95]S3B, Supporting Information). In
   14‐month‐old SAMP8 mice, smooth muscle cells exhibited a significantly
   augmented inflammatory response (Figure [96]S3C, Supporting
   Information), as demonstrated by elevated expression of cluster of
   vascular cell adhesion molecule‐1, interleukin‐6, and tumor necrosis
   factor‐alpha. These molecular alterations represent a critical hallmark
   of the phenotypic transition from normal smooth muscle characteristics
   to an AS‐prone state.^[ [97]^30 ^] In line with existing research, the
   premature vascular ageing phenotype inherent in SAMP8 mice has been
   shown to be a predisposing factor for AS pathogenesis.^[ [98]^31 ^]
   ATF3 knockdown exacerbated AS progression in 14‐month‐old SAMP8 mice,
   as evidenced by increased plaque area compared with that of normal
   control littermates (Figure [99]S3B,C, Supporting Information).

   Next, we explored whether ATF3 deficiency exacerbates AS by promoting
   VSMC senescence in ApoE^‐/‐ mice. The experiment included two groups,
   ApoE^‐/‐+GFP^∆SMC mice and ApoE^‐/‐+shATF3^∆SMC mice, which were fed a
   HFD for 3 months (Figure [100]1K). We assessed PWV and plaque formation
   using SA‐β‐gal, hematoxylin‐eosin (HE) staining, Elastica van Gieson
   (EVG) staining, transmission electron microscope (TEM), and lipid
   deposition. As anticipated, ApoE^‐/‐+shATF3^∆SMC mice exhibited
   increased vascular stiffness (Figure [101]1L) and VSMC senescence
   (Figure [102]1M,N). The absence of ATF3 promoted the degradation of
   elastic fibers (Figure [103]S3D, Supporting Information). Western
   blotting (WB) and Immunofluorescence (IF) results demonstrated that
   ATF3 knockdown in VSMCs significantly increased OPN levels and reduced
   SM22α expression, suggesting enhanced VSMC phenotypic switching (Figure
   [104]S3D–F, Supporting Information). Furthermore, ApoE^‐/‐+shATF3^∆SMC
   mice exhibited lipid accumulation at the aortic valve and increased
   plaque development (Figure [105]1O).

2.3. ATF3 Promotes Autophagy by Increasing Atg7 Transcription

   To investigate how ATF3, as a transcription factor, exerts its
   transcriptional regulatory functions, CUT&Tag and RNA‐seq were
   performed to map the genome‐wide DNA‐binding sites and likely
   transcriptional targets of ATF3. According to the CUT&Tag
   identification, compared with the control, ATF3 displayed binding peaks
   around the transcription start site, 2 kb away (Figure [106]2A). As
   shown in Figure [107]2B, after negative selection of VSMCs in the
   vasculature was performed, RNA‐seq analysis was conducted, and the
   heatmap displays the differentially expressed genes (DEGs) between the
   SAMP8+GFP^∆SMC and SAMP8+shATF3^∆SMC mice. Cross‐analysis of the
   CUT&Tag and RNA‐seq data led to the identification of 65 candidate
   transcriptional targets directly regulated by ATF3 (Figure [108]2C).
   These genes were primarily enriched in signaling pathways related to
   autophagy and collagen synthesis (Figure [109]2D). Autophagy is a
   pivotal juncture for cellular processes that are disrupted in
   age‐related diseases.^[ [110]^32 ^] real‐time quantitative PCR (qPCR)
   experiments revealed that Atg7 was significantly downregulated after
   ATF3 knockdown in SAMP8 mice (Figure [111]2E). The CUT&Tag results
   revealed significant binding peaks of ATF3 in the promoter region of
   Atg7 (Figure [112]2F). Specific primers targeting the Atg7 promoter
   region were designed for Chromatin Immunoprecipitation (ChIP)‐qPCR, and
   the results revealed significant enrichment of ATF3 binding signals in
   the Atg7 promoter region (Figure [113]2F). We propose that Atg7 could
   be a potential target gene of ATF3. The JASPAR database was used to
   predict the binding region of ATF3 to the Atg7 promoter. Atg7 promoter
   mutations 1 (−32 to −25 bp) and 2 (−1538 to −1531 bp) were designed,
   and a dual‐luciferase reporter assay was conducted for further
   validation in HEK‐293T cells. The results revealed that ATF3
   significantly increased luciferase activity in the wild‐type Atg7
   3’‐untranslated region (3'UTR) reporter, whereas ATF3 did not affect
   the luciferase activity of Atg7 promoter mutations (Figure [114]2G).
   These findings suggest direct binding of ATF3 to the Atg7 promoter
   region. scRNA‐seq of AS lesions demonstrated co‐expression patterns
   between ATF3 and ATG7 in VSMCs (Figure [115]S4A, Supporting
   Information), which was supported by a statistically significant
   positive correlation (Figure [116]S4B, Supporting Information). IF
   labeling demonstrated the concurrent downregulation of ATG7 in response
   to ATF3 knockdown in SAMP8 mice (Figure [117]2H,I). ATG7 acts as an
   E1‐like enzyme, facilitating the conjugation of LC3 to
   phosphatidylethanolamine.^[ [118]^33 ^] As expected, the decrease in
   ATG7 levels coincided with a reduction in autophagy (Figure [119]2J,K).
   In SAMP8+shATF3^∆SMC mice, the LC3B‐II/LC3B‐I ratio decreased resulting
   in an increase in the level of the autophagy substrate sequestosome1
   (SQSTM1).^[ [120]^34 ^] Fewer APs were observed in the VSMCs with ATF3
   deficiency via TEM (Figure [121]3L).

Figure 2.

   Figure 2
   [122]Open in a new tab

   ATF3 facilitates autophagy through the upregulation of Atg7
   transcription. A) The CUT&Tag calling peaks are enriched in the
   transcription start site (TSS) area by the heatmap. B) Clustering
   heatmap of differentially expressed genes between SAMP8+GFP^∆SMC and
   SAMP8+shATF3^∆SMC mice. C) Veen diagram of CUT&Tag and RNA‐seq
   intersecting targets. D) Dot plot of the GO pathway enrichment analysis
   of intersecting targets. E) Variations in the expression of genes
   enriched in the top three signaling pathways. F) ATF3 exhibits a
   significant binding peak in the promoter region of the Atg7 gene in
   CUT&Tag. And ChIP‐qPCR of ATF3 binding to the Atg7 promoter region. G)
   Luciferase assay of PGL3‐basic vector carrying wild‐type or mutant Atg7
   3′‐UTR co‐transfected with ATF3 or negative control in HEK‐293 T cells
   (n = 3). TK: pRL‐TK, Renilla Luciferase and Thymidine Kinase promoter.
   H,I) Immunofluorescent staining of ATF3 (red), ATG7 (green), α‐SMA
   (pink), and nuclei (blue) in mouse aortas (scale bars = 20 µm; n = 5).
   J,K) Western blot analysis of ATF3, LC3B‐II/LC3B‐I, SQSTM1 and ATG7
   expression in mouse aortas (n = 5). L) TEM observation of autophagic
   phenomena in VSMCs in mouse aortas (n = 5), arrows indicate
   autophagosomes or autolysosomes (scale bars, 1 µm and 500 nm). M)
   Fluorescence analysis of VSMCs transfected with the mCherry‐GFP‐LC3
   reporter (n = 3); red: autophagosomes, yellow: autolysosomes (scale
   bars, 10 µm). N) TEM observation of autophagic phenomena in VSMCs (n =
   3), arrows indicate autophagosomes or autolysosomes (scale bars, 2 µm
   and 500 nm). Error bars represent the mean ± standard deviation. The
   unpaired t‐test (I,K), one‐way ANOVA (G), and Mann–Whitney U‐test (M)
   were used to compare data; *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 3.

   Figure 3
   [123]Open in a new tab

   ATG7 promotes the translocation of ATF3 into the nucleus. A,B)
   Pull‐down proteomics analysis of ATF3 and IgG antibodies: differential
   protein Venn diagram (A) and KEGG pathway classification analysis (B).
   C) Fluorescence analysis of the cellular distribution of ATF3 and ATG7
   (scale bars = 100 µm). D) Immunoprecipitation of ATG7 from VSMCs using
   ATF3 and IgG antibodies, respectively. E) Immunoprecipitation of ATF3
   from VSMCs using ATG7 and IgG antibodies, respectively. F)
   GST‐pull‐down of His‐ATG7 from HEK‐293T cells with a GST‐ATF3 fusion
   protein. G) Visualization of the interaction between ATF3 and ATG7;
   blue: ATF3, purple: ATG7. The spatial structure is presented through a
   cartoon, and the surface is presented as a combination pocket. H)
   Different protein complex components from WT, Mutant1‐ (ATF3–A93G),
   Mutant2‐ (ATF3–L97G), or Mutant3‐ (ATF3–A104G) transfected HEK‐293T
   cells immunoprecipitated using the Flag tag antibody. I) 2D structure
   diagram of ATF3. J) Western blot analysis of ATF3 expression in the
   nucleus and cytoplasm following transfection of VSMCs with Si‐NC or
   Si‐ATG7 (n = 3). K) Fluorescence analysis of ATF3 distribution in the
   nucleus and cytoplasm (scale bars, 20 µm; n = 3). Input: whole cell
   lysates, IP: immunoprecipitants, Mut: Mutant. Error bars represent the
   mean ± standard deviation. The Mann–Whitney U‐test was performed to
   compare data; **p < 0.01.

   Small‐interfering RNA (siRNA) was used to investigate the effects of
   ATF3 in vitro (Figure [124]S4C–F, Supporting Information), ATF3
   knockdown led to a decrease in ATG7 at both the mRNA and protein levels
   (Figure [125]S4G–I, Supporting Information). Bafilomycin A1 (BafA1)
   prevents the autophagosome‐lysosome fusion and subsequently results in
   the accumulation of LC3‐II.^[ [126]^35 ^] A lake of ATF3 further
   decreased the LC3B‐II/LC3B‐I ratio in the presence of BafA1 in mVSMCs
   (Figure [127]S4J,K, Supporting Information). mVSMCs were infected with
   Adenovirus (AdV) carrying GFP‐mCherry‐LC3 fusion protein. GFP signal
   quenching occurred in the highly acidic environment of lysosomes but
   not by mCherry (red fluorescent protein). We found that the numbers of
   mCherry‐single positive (red dot) and mCherry‐GFP double positive
   puncta (yellow dot) were decreased in mVSMCs with ATF3 deficiency
   (Figure [128]2M). Loss of ATF3 decreased the level of APs and
   autolysosomes in mVSMCs by TEM (Figure [129]2N). These findings suggest
   that ATF3 facilitates autophagy through the stimulation of ATG7
   transcription.

2.4. ATG7 Promotes the Nuclear Entry of ATF3

   Co‐immunoprecipitation (Co‐IP) assay and mass spectrometry analysis
   were used to identify the proteins binding to ATF3. Proteomic analysis
   revealed that 217 proteins were pulled down by ATF3 (Figure [130]3A).
   Notably, KEGG enrichment analysis highlighted significant enrichment of
   the autophagy pathway (Figure [131]3B). There were 16 proteins enriched
   in the autophagy pathway. Among these, ATG7 was highly ranked in both
   intensity‐based absolute quantification and confidence (Q‐score) within
   the autophagy pathway (Table [132]S2, Supporting Information). IF
   co‐localization analysis revealed significant co‐localization of ATF3
   and ATG7 in the cytoplasm (Figure [133]3C). Interactions between ATF3
   and ATG7 were detected using Co‐IP and WB. The ATF3‐specific antibody
   immunoprecipitated endogenous ATG7 from VSMC lysates (Figure [134]3D).
   Conversely, the ATG7 antibody also immunoprecipitated endogenous ATF3
   (Figure [135]3E). The results of the Glutathione S‐transferase
   (GST) pull‐down assay confirmed direct binding between ATF3 and ATG7
   (Figure [136]3F). ATF3 immobilized on the CM5 sensor chip can bind ATG7
   with an affinity constant of 963 nm as determined via a surface plasmon
   resonance assay (Figure [137]S4L, Supporting Information). These
   findings suggest that ATF3 directly binds to ATG7

   The binding pattern of ATG7 to ATF3 was analyzed using the molecular
   docking method. The strongest binding complex was selected for
   visualization (Figure [138]3G), with a central energy of
   −542 kcal mol^−1. To further clarify the actual binding site between
   ATG7 and ATF3, the amino acid binding site was predicted (Table
   [139]S3, Supporting Information). The corresponding mutant ATF3
   plasmids were constructed with three variants: ATF3‐A93G, ATF3‐L97G,
   and ATF3‐A104G. Compared with those of wild‐type ATF3, Co‐IP
   experiments demonstrated a reduced interaction of all ATF3 variants
   with ATG7 (Figure [140]3H), suggesting that a stable complex formed
   between the compounds ATG7 and ATF3 at amino acids 93, 97, and 104.

   According to the forecasts, these three binding sites are located in
   the basic structural domain of the ATF3 (amino acids 88–110), which is
   responsible for the nuclear translocation of the ATF3 (Figure [141]3I).
   To confirm whether ATG7 acts as an auxiliary factor for ATF3, we
   knocked down ATG7 in mVSMCs (Figure [142]S5A–D, Supporting Information)
   and assessed the nuclear translocation of ATF3. The
   nuclear‐to‐cytoplasmic ratio of ATF3 was significantly lower in the
   ATG7‐knockdown group than in the control group (Figure [143]3J). IF
   analysis further confirmed disrupted ATF3 concentrations in the nucleus
   after ATG7 knockdown, in contrast with the results of the control group
   (Figure [144]3K). Moreover, we demonstrated that ATG7 promotes ATF3
   nuclear translocation without concomitant translocation into the
   nucleus. WB and IF analyses confirmed ATF3 enrichment in nuclei upon
   ATG7 overexpression in VSMCs, whereas ATG7 exhibited cytoplasmic
   retention without nuclear entry (Figure [145]S5E–G, Supporting
   Information).

   These results indicate that a positive feedback loop formed between
   ATF3 and ATG7. ATG7 bound to ATF3 in the cytoplasm, increasing the
   nuclear translocation of ATF3, while ATF3 bound to the promoter region
   of Atg7 and increased its transcription in the nucleus, increasing the
   expression of ATG7 mRNA and protein and initiating the downstream
   autophagy process.

2.5. TZ Enhances Atf3 mRNA Stability to Inhibit VSMC Senescence and Phenotype
Switching

   We explored a clinical drug that specifically targets ATF3. ATF3 mRNA
   levels were lower in SAMP8 mice than in SAMR1 mice (Figure [146]4A). We
   assessed the half‐life (t[1/2] ) of Atf3 mRNA in VSMCs using qPCR.
   Compared with that in the control group, the half‐life of Atf3 mRNA in
   senescent VSMCs was shorter (Figure [147]4B). The most abundant
   internal modification in mRNAs is N6‐methyladenosine (m6A), which
   affects the stability and translation efficiency of mRNAs.^[ [148]^36
   ^] We also investigated whether the decreased expression of ATF3 mRNA
   and protein arises from a m6A‐mediated epitranscriptomic mechanism. We
   analyzed publicly available m6A sequencing data using RMBase v2.0 and
   SRAMP.^[ [149]^37 , [150]^38 ^] For detection purposes, we selected the
   three methylation sites (777, 1434, and 1627) that received the highest
   scores in both prediction databases. Data from the single‐base
   elongation‐ and ligation‐based qPCR amplification (SELECT) assay
   revealed that the m6A level at adenosine residue 1627 in Atf3 mRNA was
   increased in senescent VSMCs (Figure [151]4C). The m6A reader protein,
   YTH N6‐methyladenosine RNA binding protein F2 (YTHDF2) binds to
   m6A‐modified Atf3 mRNA, triggering its decay.^[ [152]^39 ^] Thus,
   blocking the binding of YTHDF2 to Atf3 mRNA could serve as a potential
   approach to target ATF3. We conducted a virtual screening of 16 000 000
   compounds from the FDA‐approved drug database, traditional Chinese
   medicine (TCM) database, and PubChem database to identify therapeutic
   agents that attenuate the binding between Atf3 mRNA and YTHDF2
   (Figure [153]4D). The top 10 molecules with the highest affinity are
   listed in Table [154]S4 (Supporting Information). The stability of the
   complex was assessed using molecular dynamics (MD) simulations.
   Ultimately, the compound Terazosin (TZ), which had the highest binding
   affinity, was prioritized for further mechanistic validation. The
   binding modes of TZ, Atf3 mRNA, and YTHDF2 are depicted in
   Figure [155]4E. The root‐mean‐square deviation (RMSD), radius of
   gyration (Rg), and root mean square fluctuation (RMSF) were monitored
   throughout the 100‐ns MD simulations to assess the structural stability
   of the systems (Figure [156]4F; Figure [157]S6A,B, Supporting
   Information). As a result of TZ addition, Atf3 mRNA exhibited enhanced
   structural fluctuations and a destabilized binding interface with
   YTHDF2. The increase in total energy (ΔEtot) upon TZ addition further
   supports the weakened binding affinity between Atf3 mRNA and YTHDF2
   (Figure [158]4F; Figure [159]S6C,D, Supporting Information). An RNA
   immunoprecipitation (RIP) assay was then performed to assess the
   binding of YTHDF2 with Atf3 mRNA. Notably, the presence of TZ
   significantly diminished the affinity of YTHDF2 for m6A‐modified Atf3
   mRNA (Figure [160]4G). TZ extended the Atf3 mRNA half‐life in ageing
   VSMCs (Figure [161]4H).

Figure 4.

   Figure 4
   [162]Open in a new tab

   Terazosin (TZ) improves age‐related phenotypes in SAMP8 mice via ATF3.
   A) qPCR analysis of Atf3 in SAMP8 mice carotid arteries (n = 5). B)
   Senescence decreases the half‐life of Atf3 mRNA in VSMCs (n = 3). C)
   Relative m6A methylation abundances of Atf3 mRNA at the 1627 site
   measured by the SELECT method in VSMCs (n = 3). D) Experimental design:
   screening small molecular drugs targeting ATF3. E) Visualization of the
   interaction between Atf3 mRNA, YTHDF2, and TZ. blue: YTHDF2, yellow:
   Atf3 mRNA with m6A modifications (blue), red: TZ. F) Structural
   backbone RMSD variations from molecular dynamics simulations and
   MM‐PBSA‐derived binding free energy values. G) Carry out YTHDF2‐RNA
   immunoprecipitation (RIP) in TZ‐treated VSMCs, then use qPCR to
   evaluate Atf3 mRNA expression. Analyze the influence of TZ on YTHDF2's
   affinity for Atf3 mRNA (n = 3). H) TZ increases the half‐life of Atf3
   mRNA in VSMCs (n = 3). I) Experimental design: Administered TZ via
   gavage to SAMP8 mice for three months, with SAMR1 as the control group
   (TZ1: 20 µg kg^−1, TZ2: 60 µg kg^−1, TZ3: 180 µg kg^−1; n = 6). J) PWV
   in SAMR1 mice and SAMP8 mice (n = 6). K) SA‐β‐gal staining of mouse
   aortas (scale bars = 5 mm, n = 5). L) Western blot analysis of ATF3,
   p53, and p21 expression in mouse aortas (n = 5). Error bars represent
   mean ± standard deviation. The one‐way ANOVA (G, J, L) and unpaired
   t‐test (A–C) were used to compare data; *p < 0.05, **p < 0.01,
   ***p < 0.001.

   We further investigated the potential regulatory effects of TZ on ATF3
   in ECs. In contrast to that in VSMCs, ATF3 mRNA in ECs was inherently
   highly stable and was not further extended by TZ (Figure [163]S7A,B,
   Supporting Information). A plausible explanation is that m6A
   modification of ATF3 mRNA in ECs remains unaltered during ageing
   (Figure [164]S7C, Supporting Information).

   To exclude the possibility that TZ may target other members of the ATF
   family, we conducted RNA‐seq on TZ‐treated VSMCs from SAMP8 mice.
   Enrichment analysis of the DEGs revealed that the ATF3‐CHOP signaling
   axis was the most prominently altered pathway (Figure [165]S7D,
   Supporting Information). Notably, qPCR validation demonstrated no
   significant alterations in the mRNA levels of other ATF family members,
   including ATF2 and ATF4, compared with those in vehicle controls
   (Figure [166]S7E, Supporting Information).

   Further validation of the therapeutic effects of TZ was carried out
   through in vivo experiments. SAMP8 and SAMR1 mice were randomly divided
   into five groups and treated with three different TZ doses
   (Figure [167]4I). After 3 months of treatment, the basic
   characteristics of the mice were recorded (Table [168]S5, Supporting
   Information). No significant differences were observed among the five
   groups in terms of body weight, serum glucose level, or blood lipid
   level at multiple time points during drug administration.

   Additionally, HE staining of liver and kidney histology revealed no
   apparent structural disarrangement, indicating rare hepatotoxicity and
   renal toxicity (Figure [169]S8A, Supporting Information). Mice treated
   with TZ showed a decrease in blood pressure (BP), but the difference
   was not statistically significant (Figure [170]S8B,C, Supporting
   Information).

   The PWV of SAMP8 mice was greater than that of SAMR1 mice. TZ treatment
   decreased vascular stiffness in SAMP8 mice (Figure [171]4J). The
   senescence markers SA‐β‐gal activity, p53, and p21 were assessed. TZ
   treatment not only increased ATF3 mRNA and protein levels but also
   decreased SA‐β‐gal activity, p53, and p21 mRNA and protein levels in
   SAMP8 mice (Figure [172]4K,L; Figure [173]S8D–F, Supporting
   Information). The vessels of prematurely aged SAMP8 mice expressed more
   synthetic phenotypic markers, while those of TZ‐treated SAMP8 mice
   showed more contractile phenotypic markers (Figure [174]S8G–L; Figure
   [175]S9A, Supporting Information). TZ treatment reduced the broken ends
   of elastic fibers and restored the arrangement of collagen fibers, as
   shown by HE and EVG staining and TEM analysis (Figure [176]S9B,
   Supporting Information). These findings suggest that TZ slows the
   vascular ageing process in SAMP8 mice. The lowest effective dose that
   reduced vascular stiffness (60 µg Kg^−1) was selected as the
   therapeutic dose for further experimental exploration based on the
   above results.

   Mouse primary VSMCs were utilized to illustrate the protective effect
   of TZ in vitro. Cell viability assays showed enhanced cell activity at
   concentrations ranging 1–20 nm (Figure [177]5A). Consequently, the
   concentration ranges of TZ that do not decrease cell viability were
   categorized into three groups: TZ1 (1 nm), TZ2 (10 nm), and TZ3
   (20 nm). Doxorubicin was used to induce VSMC senescence, as previously
   reported.^[ [178]^40 , [179]^41 ^] Rapamycin (RAPA), known for its
   ability to inhibit cell senescence, was selected as a positive control
   in this study.^[ [180]^42 , [181]^43 ^] The anti‐ageing effects of TZ
   were assessed, and a significant reduction in SA‐β‐gal activity was
   observed after TZ administration (Figure [182]5B,C). Similar to RAPA,
   TZ2 and TZ3 effectively reduced the degree of senescence in VSMCs
   (Figure [183]5D), leading to the use of a 10 nm TZ concentration as the
   experimental therapeutic dose for further validation. In addition, we
   investigated whether the anti‐ageing effect of TZ depends on α1
   blockade. In experiments including SA‐β‐gal and WB, the anti‐ageing
   effect of TZ was not altered by prior exposure to the irreversible
   α‐blocker phenoxybenzamine (phen). Doxazosin (DZ) failed to protect
   VSMCs from senescence, indicating that TZ does not exert its
   anti‐ageing effects by acting as an α1‐adrenoceptor blocker
   (Figure [184]5E–H). WB (Figure [185]5I,J) and IF assays
   (Figure [186]5K,L) subsequently revealed that TZ inhibited the
   phenotypic transformation of VSMCs.

Figure 5.

   Figure 5
   [187]Open in a new tab

   TZ reverses VSMC senescence and phenotype conversion. A) Analysis of
   cell viability in VSMCs treated with varying TZ concentrations for 48 h
   (n = 3). B,C) SA‐β‐gal staining of VSMCs (scale bars = 50 µm; n = 3).
   D) Western blot analysis of ATF3, p53, and p21 expression in VSMCs
   (n = 3). E,F) Prior to adding TZ to the senescent VSMCs, pre‐treat with
   Phen (phenoxybenzamine) for 4 h, using DZ (Doxazosin) as a negative
   control. E) SA‐β‐gal staining of VSMCs (scale bars = 50 µm; n = 3). F)
   Western blot analysis of ATF3, p53, and p21 expression in VSMCs
   (n = 3). G–I) Western blot analysis of SM22α and OPN expression in
   VSMCs (n = 3). J–L) Immunofluorescent staining of SM22α (red), OPN
   (red), and nuclei (blue) in VSMCs (scale bars = 20 µm; n = 3). Error
   bars represent mean ± deviation. The Kruskal–Wallis test was performed
   for data comparison. *p < 0.05, **p < 0.01, ***p < 0.001.

2.6. Silencing ATF3 in VSMCs Abolishes the Anti‐Ageing Effects of TZ

   To clarify whether the observed vascular results are dependent on the
   effects of TZ in ECs or VSMCs, we investigated the changes in the
   expression of ATF3 in ECs and VSMCs. According to the IF assay results,
   the level of VSMC‐labeled ATF3 protein decreased in SAMP8 mice and
   recovered after TZ intervention (Figure [188]S10A, Supporting
   Information). Moreover, ATF3 expression in ECs was not affected by
   ageing or TZ intervention (Figure [189]S10A, Supporting Information),
   indicating that VSMC‐derived ATF3 acts as a key regulator contributing
   to the protective effects of TZ. We administered TZ to SAMP8+GFP^∆SMC
   and SAMP8+shATF3^∆SMC mice for 3 months and divided the experimental
   groups into three categories (Figure [190]6A).

Figure 6.

   Figure 6
   [191]Open in a new tab

   TZ improves VSMC senescence by promoting the autophagy signaling
   pathway via ATF3. A) Experimental design: SAMP8 mice were injected tail
   vein with adenoassociated virus. After 4 weeks, mice received TZ
   (60 µg kg^−1) by oral gavage for 3 months (n = 6). B) PWV in SAMP8 mice
   (n = 6). C–E) SA‐β‐gal staining, Hematoxylin and eosin staining, EVG
   staining, and TEM of mouse aortas (scale bars, 5 mm, 50 µm, 100 µm,
   1 µm, and 500 nm). EF: elastic fiber, CF: collagen fiber; arrows
   indicate areas of collagen fiber disarray (n = 5). F–H) Western blot
   analysis of p53 and p21 expression in mouse aortas (n = 5). I–L)
   Western blot analysis of LC3B‐II/LC3B‐I, SQSTM1, and ATG7 expression in
   mouse aortas (n = 5). M) TEM observation of autophagic phenomena in
   VSMCs in mouse aortas (n = 5); arrows indicate autophagosomes or
   autolysosomes (scale bars, 1 µm and 500 nm). Error bars represent
   mean ± standard deviation. The one‐way ANOVA was performed to compare
   data; *p < 0.05, **p < 0.01, ***p < 0.001.

   SAMP8 mice treated with TZ rarely presented significant alterations in
   body weight, blood glucose levels, blood lipids, or abnormalities in
   liver and kidney histology (Table [192]S6, Figure [193]S10B, Supporting
   Information).

   A comparison of the SAMP8+GFP^∆SMC +TZ and SAMP8+shATF3^∆SMC+TZ groups
   revealed that decreased ATF3 expression significantly inhibited the
   regulatory effects of TZ and accelerated VSMC senescence and vascular
   ageing in SAMP8 mice. Compared with the SAMP8+GFP^∆SMC+TZ group, the
   SAMP8+shATF3^∆SMC+TZ group exhibited higher PWV (Figure [194]6B). ATF3
   knockdown inhibited the reparative effect of TZ on vascular ageing and
   structure, as shown by SA‐β‐gal and EVG staining and TEM
   (Figure [195]6C–E). WB revealed a significant increase in senescence
   markers and synthetic phenotypic markers in the SAMP8+shATF3^∆SMC+TZ
   group compared with those in the SAMP8+GFP^∆SMC+TZ group
   (Figure [196]6F,H), with similar trends observed in the qPCR and IF
   assays (Figure [197]S10C–E, Supporting Information).

   In vitro studies also demonstrated this effect. ATF3‐knockdown siRNA
   blocked the beneficial effects of TZ on anti‐ageing and anti‐phenotype
   switching in senescent VSMCs (Figure [198]S11, Supporting Information).

2.7. TZ Reduced VSMC Senescence and Vascular Ageing through the Autophagy
Signaling Pathway

   We examined the variations in the autophagy pathway after the
   administration of TZ. TZ increased ATG7 levels, promoted the conversion
   of LC3‐II to LC3‐I, and enhanced SQSTM1 degradation in SAMP8 mice, as
   shown by WB (Figure [199]6I–L). This trend was reversed by the absence
   of ATF3 (Figure [200]6I–L). Compared with SAMR1 mice, SAMP8 mice showed
   fewer APs in the vascular media under TEM. When ATF3 was expressed, TZ
   led to an increased number of APs (Figure [201]3M). These results
   showed that TZ administration activated autophagy in an ATF3‐dependent
   manner.

   In vitro experiments were conducted to confirm the above findings. TZ
   elevated ATG7 mRNA and protein levels in senescent VSMCs, facilitating
   the degradation of SQSTM1 (Figure [202]S12A–D, Supporting Information).
   TZ further increased the LC3B‐II/LC3B‐I ratio in the presence of BafA1.
   However, ATF3 silencing prevented TZ from promoting LC3‐II formation
   (Figure [203]S12E,F, Supporting Information). Both TEM and LC3 virus
   transfection revealed that TZ increased the AP content, whereas ATF3
   knockdown reversed this effect (Figure [204]S12G,H, Supporting
   Information). Taken together, these results suggest that TZ treatment
   activates autophagic flux via ATF3.

   To investigate whether TZ played a role in regulating cellular ageing
   and phenotype switching through the ATG7/LC3 signaling pathway, ATG7
   and LC3 were inhibited in VSMCs. In ATG7‐depleted senescent VSMCs, TZ
   failed to decrease the expression of ageing markers or significantly
   reverse the contraction‐secretion phenotype transition (Figure
   [205]S13, Supporting Information). DC‐LC3in‐D5, an LC3 lipidation
   inhibitor^[ [206]^44 ^] suppresses cellular autophagy and exacerbates
   cellular senescence. The inhibition of autophagy led to VSMC senescence
   and secretory phenotype transition, a change that could not be reversed
   by TZ treatment, as shown by the results of the qPCR, WB, IF assays
   (Figure [207]S14, Supporting Information). These findings indicate that
   the inhibition effect of TZ on VSMC senescence depended on the
   activation of autophagic flux.

2.8. TZ Improves AS by Ameliorating VSMC Senescence

   We verified the ATG7/LC3 signaling pathway in the atherosclerotic
   vasculature. IF analysis revealed that the abundance of LC3 spots in
   atherosclerotic plaques was reduced, indicating decreased autophagy
   activity in AS human samples (Figure [208]7A).

Figure 7.

   Figure 7
   [209]Open in a new tab

   TZ improves age‐related phenotypes in ApoE^‐/‐ mice. A)
   Immunofluorescent staining of LC3B (red) and nuclei (blue) in human
   carotid arteries (scale bars, 20 µm; n = 5). B,C) Immunofluorescent
   staining of ATG7 (red), α‐SMA (green), and nuclei (blue) in human
   carotid arteries (scale bars, 20 µm; n = 5). D) Experimental design:
   Administered TZ via gavage to in APOE^‐/‐ mice fed with a High‐fat diet
   for three months (TZ: 60 µg kg^−1, n = 8–10). E) Western blot analysis
   of ATF3, p53, and p21 expression in mouse aortas (n = 5). F) PWV in
   ApoE^‐/‐ mice (n = 8–10). G–I) SA‐β‐gal staining, Hematoxylin and eosin
   staining, Oil Red O staining, and TEM of mouse aortas (scale bars,
   5 mm, 100 µm, 1 µm, and 500 nm). EF: elastic fiber, CF: collagen fiber;
   arrows indicate areas of collagen fiber disarray (n = 5). J,K) Western
   blot analysis of LC3B‐II/LC3B‐I, SQSTM1, and ATG7 expression in mouse
   aortas (n = 5). Error bars represent mean ± standard deviation. The
   unpaired t‐test (A,B) and one‐way ANOVA (E,F,G,I,K) were performed to
   compare data; *p < 0.05, **p < 0.01, ***p < 0.001.

   ATG7 expression was lower in atherosclerotic plaque samples than in
   healthy control samples (Figure [210]7B,C). Additionally, to elucidate
   the impact of TZ on senescent VSMCs within the context of AS, we
   adopted an AS model in which ApoE^‐/‐ mice were fed a HFD
   (Figure [211]7D). As anticipated, the AS model mice exhibited overt
   signs of vascular senescence, as evidenced by elevated expression of
   VSMC senescence markers (Figure [212]7E; Figure [213]S15A–C, Supporting
   Information) and increases in PWV (Figure [214]7F) and the senescent
   cell fraction (Figure [215]7G,H). Treatment with TZ increased the ATF3
   content in AS model mice and ameliorated VSMC senescence
   (Figure [216]7E). We also evaluated the vascular morphology of ApoE^‐/‐
   mice via HE and EVG staining, TEM analysis, and lipid deposition. TZ
   significantly decreased elastic fiber rupture at the plaque site and
   the Oil Red O‐positive staining area in the aortic valve
   (Figure [217]7H,I; Figure [218]S15D,E, Supporting Information).
   Remarkably, TZ ameliorated the extent of vascular ageing without
   significantly lowering the BP of ApoE^‐/‐ mice (Figure [219]S15F,
   Supporting Information). mRNA and protein analyses collectively
   revealed that TZ countered VSMC phenotypic switching during AS
   development (Figure [220]S15G–L, Supporting Information). TZ
   demonstrated robust safety and therapeutic efficacy in AS mouse models.

   TZ restored the diminished ATG7 levels associated with AS progression
   in ApoE^‐/‐ mice (Figure [221]7J,K). Similar changes were observed by
   qPCR and IF (Figure [222]S16A,B, Supporting Information). Furthermore,
   TEM revealed diminished AP formation within the VSMCs of AS model mice.
   However, this effect was reversed upon TZ administration (Figure
   [223]S16C, Supporting Information). WB analysis confirmed that
   atherosclerotic conditions were associated with reduced formation of
   LC3B‐II and increased SQSTM1 in VSMCs, and these effects were reversed
   by TZ treatment (Figure [224]7J,K). Taken together, these data indicate
   that TZ activates autophagy through the ATF3/ATG7/LC3 signaling
   pathway, ultimately delaying VSMC senescence and reducing AS.

3. Discussion

   This study demonstrated that the lack of ATF3 in VSMCs contributes to
   AS and VSMC senescence. ATF3 activates the transcription of ATG7,
   promoting autophagy to improve VSMC senescence. A new target of TZ is
   to maintain the stability of Atf3 mRNA to exert anti‐AS effects. ATF3
   knockdown or autophagy inhibition negates the vascular protective
   effects of TZ.

   VSMC senescence has been implicated as a contributing factor in the
   development of diverse vascular pathologies.^[ [225]^45 ^] AS is
   characterized by premature vascular ageing,^[ [226]^46 ^] particularly
   in atherosclerotic plaque areas in the aorta. Senescent VSMCs
   exacerbate plaque growth and increase the risk of rupture.^[ [227]^47
   ^] Anti‐senescence strategies, such as histone deacetylase‐modulating
   agents^[ [228]^48 ^] and vitamin D,^[ [229]^49 ^] have been explored to
   ameliorate vascular dysfunction in AS.

   ATF3 is vital in cellular functions and decreases in expression with
   age. Its activity and expression are influenced by transcriptional
   regulation, post‐transcriptional control, post‐translational
   modification, and the cellular milieu.^[ [230]^37 ^] Wang et al.
   demonstrated that inactivation of ATF3 results in VSMC
   dedifferentiation, causing loss of the contractile phenotype.^[
   [231]^50 ^] Other studies have reported the beneficial actions of ATF3
   in the context of AS. Hepatocyte‐derived ATF3 exerts an inhibitory
   effect on AS development through its ability to modulate lipid and bile
   acid metabolism.^[ [232]^51 ^] We observed the greatest changes in ATF3
   expression in VSMCs from AS patients compared with those from baseline
   controls, while ECs from AS patients showed no significant alterations
   (in ATF3 expression) compared with their baseline counterparts. In
   addition, ATF3 levels were significantly decreased in the VSMCs of aged
   mice. The expression distributions of ATF3 and p53 in VSMCs were
   negatively correlated. Moreover, additional knockdown of ATF3 in VSMCs
   exacerbated AS in APOE^‐/‐ mice through accelerated VSMC senescence.

   Autophagic activity decreases in the vascular tissues of elderly mice
   and humans,^[ [233]^32 ^] increasing the risk of arterial stiffness and
   vascular calcification. Impaired VSMC autophagy in APOE^‐/‐ mice with
   SMC‐specific ATG7 deficiency accelerates plaque formation.^[ [234]^52
   ^] Previous studies have shown that inducing autophagy can suppress
   VSMC phenotypic transformation and migration induced by oxidized
   low‐density lipoprotein.^[ [235]^53 ^] The modulatory effect on VSMC
   phenotypic transformation induced by high uric acid can be reversed,
   delaying AS.^[ [236]^54 , [237]^55 ^] However, the dual role of
   autophagy in organisms^[ [238]^56 ^] requires great caution in the
   administration of autophagy‐modulating drugs. Our findings align with
   those of other studies, suggesting that ATF3 acts as a potential novel
   pathway for autophagy regulation. ATF3 activation enhances autophagy
   upon bacterial infection, increasing the defense capabilities of the
   organism.^[ [239]^57 ^] ATF3 overexpression in mouse lung tissues
   following Mycobacterium infection facilitates the expression of
   autophagy‐related proteins while suppressing inflammation.^[ [240]^57
   ^] However, impaired autophagy hinders the nuclear translocation of
   ATF3, compromising its anti‐inflammatory effects.^[ [241]^58 ^] Tumor
   cells lacking ATG7 or ATG5 exhibit downregulated ATF3 expression due to
   impaired autophagy.^[ [242]^59 ^] In our study, ATF3 interacted with
   ATG7 as determined by MS, Co‐IP, GST pull‐down, and co‐localization
   assays. ATG7 binds to the nuclear translocation region of ATF3 in the
   cytoplasm, and three key amino acids essential for binding in this
   region were confirmed. The role of ATG7 in modulating the nuclear entry
   of ATF3 was assessed through supplementary experiments. ATG7‐deficient
   VSMCs showed reduced nuclear ATF3 content with a more dispersed
   distribution, emphasizing the role of ATG7 in regulating ATF3 nuclear
   localization and distribution. Once in the nucleus, ATF3 binds to the
   ATG7 promoter area, facilitating its transcription and leading to the
   activation of autophagy. ATG7 promotes ATF3 nuclear translocation
   without concomitant nuclear translocation. This mechanism mirrors
   established regulators of ATF transcription factors, such as 14‐3‐3
   proteins that dissociate prior to nuclear entry.^[ [243]^60 ^]

   Depletion of ATF3 leads to a decline in autophagic flow. In VSMCs
   transfected with AdV carrying the GFP‐mCherry‐LC3 fusion protein, APs
   that do not fuse with lysosomes exhibit yellow fluorescence, whereas
   autolysosomes display red fluorescence. A deficiency reduces the levels
   of both fluorescence types, with a notable decrease in yellow
   fluorescence.

   VSMC senescence is accompanied by increased m6A methylation of Atf3
   mRNA and decreased mRNA stability. By employing YTHDF2 to recognize the
   binding patterns of Atf3 mRNA m6A sites, we performed high‐throughput
   virtual screening and discovered that TZ could potentially change their
   affinity. Through RIP assay and mRNA stability analysis, TZ was shown
   to increase the expression of ATF3 by reducing the binding affinity of
   YTHDF2 for m6A‐containing Atf3 mRNAs. Molecular dynamics simulations
   further confirmed that adding TZ destabilizes the binding between
   YTHDF2 and ATF3 mRNA. Additionally, sequencing analysis of vascular
   cells from TZ‐treated SAMP8 mice revealed that DEGs were predominantly
   enriched in the ATF3‐CHOP signaling pathway. qPCR demonstrated that TZ
   does not alter the mRNA levels of other ATF family members.
   Collectively, these findings suggest the specificity of TZ in targeting
   ATF3 mRNA. Additionally, ATF3 expression in ECs remained unaltered by
   ageing or TZ treatment. In contrast to that in VSMCs, ATF3 mRNA in ECs
   is inherently highly stable and is not further extended by the TZ. A
   plausible explanation is that the m6A modification of ATF3 mRNA in ECs
   remains unaltered during ageing.

   The clinical safety of TZ, a widely used drug, is well‐accepted. TZ has
   demonstrated significant benefits in various diseases, particularly
   neurodegenerative disorders, such as amyotrophic lateral sclerosis and
   Parkinson's disease.^[ [244]^61 ^] However, there is a relative dearth
   of data regarding the benefits of TZ for cardiovascular diseases. These
   findings suggest that the ATF3‐autophagy signaling axis may serve as a
   potential molecular target for TZ in AS intervention. In our previous
   pilot study, low‐dose TZ significantly improved arterial stiffness, as
   evidenced by reduced brachial‐ankle pulse wave velocity, with the
   beneficial effects persisting throughout a one‐year follow‐up period.^[
   [245]^62 ^] On this basis, larger randomized controlled trials
   evaluating the efficacy of TZ in treating AS are still needed. By
   integrating these preliminary clinical observations with our
   experimental findings, this work provides foundational data and a
   feasibility assessment for the design of future large‐scale, in‐depth
   clinical trials. Knocking down ATF3 weakens the anti‐ageing effect of
   TZ. Moreover, TZ reduces VSMC senescence and phenotypic transformation
   through an autophagy‐dependent pathway, and autophagy inhibitors block
   the anti‐aging effects of TZ.

   This study provides strong evidence and valuable insights into the
   relationship between VMSC senescence and the ATF3/ATG7/LC3 signaling
   pathway. While the findings showed a trend toward reduced BP with
   low‐dose TZ, they did not reach statistical significance. In
   particular, the TZ dose used in our rodent experiments, when translated
   into human equivalent doses, is far below the recommended dose for
   clinical use to lower BP according to the expert consensus‐based
   clinically equivalent dose estimates and dosing recommendations.^[
   [246]^63 ^] Therefore, the downward tendency in blood pressure in
   ageing‐reversed mice might be attributed to improved vascular
   compliance due to reduced vascular stiffness. Our data demonstrate that
   TZ exerts anti‐ageing effects distinct from its α1‐AR antagonism.
   Pre‐treatment with phen fully abrogated TZ's α1‐AR antagonism, yet TZ
   retained its ability to rescue senescence markers in VSMCs. In
   contrast, DZ—a structurally similar α1‐AR antagonist—failed to mitigate
   cellular senescence at equimolar doses, despite exhibiting comparable
   receptor binding affinity. Future studies are still needed to examine
   the anti‐ageing effects of hypotensors across various tissues and cell
   types. The current data primarily provide theoretical foundations for
   clinical translation, and further translational medical studies are
   needed to validate TZ's clinical efficacy. Future research could focus
   on the following directions. Cross‐Species Validation: Expanding
   preclinical evaluation by assessing the efficacy of TZ in
   ageing‐related AS using large animal models (e.g., non‐human primates),
   with a focus on defining its dose‐response relationships and ATF3
   dependency. Clinical correlation analysis: Retrospective cohort studies
   to compare plaque progression in AS patients treated with TZ with that
   in patients treated with other antihypertensive agents, and
   histological analyses of vascular tissues to quantify ATF3‐autophagy
   pathway activity and explore its clinical relevance.

4. Experimental Section

Human Samples

   A case‐control study involving human biological sample collection was
   performed. Five carotid artery samples were procured from the sample
   repository maintained by the Department of Cardiovascular Surgery at
   Tongji Hospital from 2020 to 2022. Control samples were acquired from
   three healthy donors who underwent heart transplantation. Table [247]S1
   (Supporting Information) summarizes the characteristics of the
   patients. The Ethics Committee of Tongji Hospital, Huazhong University
   of Science and Technology (approval number: TJ‐IRB20191215) approved
   the study. All experiments were performed in accordance with the
   principles expressed in the Declaration of Helsinki, and informed
   consent was obtained from all participants.

Animals

   The animal experiments were approved by the Experimental Animal Center
   of Huazhong University of Science and Technology (20 220 981). All
   animal procedures conformed to the current National Institution of
   Health guidelines.

   Eight‐month‐old male SAMR1 mice and SAMP8 mice weighing 28—32 g were
   obtained from the Experimental Animal Science Department, Peking
   University School of Medicine, Beijing, China. Additionally, 8‐week‐old
   ApoE^‐/‐ male mice weighing 25–28 g and 18‐month‐old C57BL/6 male mice
   weighing 30–35 g were purchased from Guangdong Yaokang Biotechnology
   Company.

   ApoE^‐/‐ mice were fed a HFD comprising 19.97% protein and 39.85% fat
   (Special Diet Services, XTZ08C, Xietong Bio‐engineering, Jiangsu,
   China) for 12 weeks to induce advanced plaques. The mice were housed
   under conventional and specific pathogen‐free conditions. Isoflurane, a
   volatile anesthetic agent, was used to induce general anesthesia in the
   mice once a month. The mice were euthanized via intraperitoneal
   injection of pentobarbital sodium (1%, 0.2 mL), and the hearts were
   subsequently perfused with phosphate‐buffered saline. The detailed
   methods are provided in the Supporting Information.

TZ Administration

   TZ (T4680, Sigma‐Aldrich, USA,
   sigmaaldrich.cn/CN/zh/product/sigma/t4680), with a purity of 99.7%, was
   used in this experiment. The molecular structure
   ([248]www.chemblink.com) and chromatographic identification of TZ are
   in Figure [249]S1A,B (Supporting Information). TZ was dissolved in
   distilled water and administered to mice by oral gavage, with the
   control group receiving an equivalent volume of distilled water. The TZ
   dosages were divided into three groups based on previous reports and
   preliminary designs: 20 µg kg^−1 (TZ1 group), 60 µg kg^−1 (TZ2 group),
   and 180 µg kg^−1 (TZ3 group).

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   H.N. and T.J. contributed equally to this work and shared first
   authorship. H.N. and T.J. was dealt with conceptualization,
   investigation, validation, and writing original draft. J.Y. and C.Z.
   dealt with funding acquisition, writing review and editing, conceived
   the study, and designed the experiments. Y.H., J.H., H.L., L.Z., and
   Z.Y. dealt with data analysis. W.L., Y.G., Y.L., L.Z., and L.R. dealt
   with supervision.

Supporting information

   Supporting Information
   [250]ADVS-12-e02249-s002.docx^ (18.3MB, docx)

   Supporting Information
   [251]ADVS-12-e02249-s003.docx^ (43.5KB, docx)

   Supporting Information
   [252]ADVS-12-e02249-s004.docx^ (62.2KB, docx)

   Supporting Information
   [253]ADVS-12-e02249-s001.docx^ (22.5KB, docx)

Acknowledgements