Deletion of METTL14, a key methylation regulator, attenuates vascular ageing

Abstract

Background and Aims

Vascular ageing often accompanies inflammation, contributing to the onset of local or systemic vascular diseases. Nevertheless, limited research focuses on pivotal factors triggering chronic vascular inflammation and associated pathological changes. This study aimed to investigate the role of methyltransferase-like protein 14 (METTL14) in inflammation in the pathogenesis of vascular ageing.

Methods

The natural ageing mouse model, D-galactose induced ageing mouse model, and endothelial cell-specific METTL14 knockout mice were generated. The roles of METTL14 in vascular ageing were investigated in human, mice, and various endothelial cells.

Results

Up-regulation of METTL14 was observed in the aortic endothelial cells of aged mice, aged humans, and senescent human umbilical vein endothelial cells, human aortic endothelial cells, and mice aortic endothelial cells. Endothelium-specific knockdown or knockout of METTL14 notably inhibited arterial stiffness, arterial remodelling, and endothelial senescence, whereas endothelium-specific overexpression of METTL14 yielded opposing effects. At the cellular level, METTL14 knockdown ameliorated cellular senescence, inflammatory responses, and oxidative stress in senescent endothelial cells. Mechanistically, METTL14 facilitated m⁶A modification of Toll-like receptor 4 (TLR4) mRNA, thereby enhancing its stability. Knockdown of TLR4 reversed the detrimental effects of METTL14 on vascular ageing. Importantly, vascular ageing, along with related atherosclerosis and arteriosclerosis, positively correlated with blood METTL14 and TLR4 elevations in humans.

Conclusions

This study hints at the role of METTL14/TLR4 signalling in the pathogenesis of vascular ageing, and METTL14 knockdown emerges as a potential therapeutic strategy for mitigating vascular ageing and associated vascular diseases.

Structured Graphical Abstract

Structured Graphical Abstract.

Deletion of METTL14 attenuates vascular ageing.

Translational perspective.

Ageing is recognized as an independent risk factor for vascular dysfunction, yet the underlying mechanisms remain largely unknown. The data obtained from mice and humans demonstrate that elevated METTL14 level was associated with vascular ageing. Deletion of METTL14 in endothelial cells ameliorates ageing-associated arteriosclerosis via TLR4/MyD88/NF-κB signalling. These findings suggest that targeting METTL14 may offer a therapeutic strategy to alleviate vascular ageing and related diseases.

Introduction

Ageing significantly increases the susceptibility to cardiovascular diseases, contributing to elevated global mortality rates.¹ In the elderly, arteries undergo pathological remodelling, leading to age-related arterial diseases including hypertension, atherosclerosis, aneurysms, and others.² Enhancing vascular health by mitigating age-related signals not only promotes healthy ageing but also extends lifespan, making the concept of vascular rejuvenation promising in geroprotection.³ However, the precise mechanisms underlying age-induced vascular dysfunction remain elusive.

Endothelial cells are among the initial cells to undergo senescence with advancing age. This senescence results in impaired endothelium-dependent dilatation, and angiogenesis, and barrier dysfunction.⁴ Senescent endothelial cells are overactive and have characteristics of the senescence-associated secretory phenotypes (SASPs), comprising inflammatory chemokines, cytokines, and growth factors, along with a substantial production of reactive oxygen species (ROS).⁵ Evidence from endothelial cells and other cell types suggests that SASPs can induce senescence in neighbouring cells.⁶ The heightened inflammatory SASPs and ROS produced by senescent endothelial cells may have widespread deleterious effects on other cells in arteries and peripheral tissues, such as vascular smooth muscle cells (VSMCs), which in turn activates signals involved in arterial remodelling and contributes to arterial stiffening.⁷ Selective elimination of senescent cells and development of senolytic agents targeting these cells have highlighted the therapeutic potential in mitigating endothelial senescence.⁸ Yet, the origins and regulation of endothelial senescence require further in-depth investigation, both in vivo and ex vivo.

Epigenetic modifications, particularly N6-methyladenosine (m⁶A) as the most prevalent modification on mRNAs regulating RNA translation, stability, and degradation, exert a significant influence on ageing process. Among these modifications, methyltransferase-like protein 14 (METTL14), a key regulator promoting m⁶A methylation, is closely associated with cellular senescence. Its role in SASP production, regulation of p21 expression, and its impact on various cellular senescence pathways have been documented.⁹ However, its involvement in vascular ageing remains understudied. METTL14-mediated p21 m⁶A methylation enhances its translation level, leading to increased p21 expression in oxidative stress-induced cellular senescence.¹⁰ Additionally, METTL14 has been found to regulate intestinal cellular senescence through m⁶A modification of lamin B receptor (LBR).¹¹ Lamin A forms an interaction with METTL14 in nuclear speckles, and its deficiency compromises the nuclear speckle METTL14 reservoir, rendering these methyltransferases susceptible to proteasome-mediated degradation.¹² Furthermore, METTL14 enhances m⁶A modification of FOXO1 mRNA, promoting FOXO1 expression and inducing endothelial inflammation, contributing to the development of atherosclerosis.¹³ However, few studies have attempted to uncover the role of METTL14 in vascular ageing.

This study aims to elucidate the mechanisms through which METTL14 modulates vascular ageing by investigating its role in m⁶A modification. We employed diverse methods to analyse METTL14’s influence on vascular ageing in humans, mice, and cells. Our findings suggest that knockdown of METTL14 retards the progression of vascular ageing, by suppressing TLR4/Myd88/NF-κB signalling, hinting at its potential as a novel target for regulating ageing-related vascular diseases.

Methods

Supplementary data are available at European Heart Journal online.

Results

METTL14 is up-regulated in aged vascular endothelium

To understand vascular ageing, we utilized a 21-month-old female C57BL/6J mouse model. Aged mouse aortas exhibited increased pulse wave velocity (PWV) and intima–media thickness (IMT), indicating significant vascular stiffness (see Supplementary data online, Figure S1A and B). Elevated levels of age-related markers, p53, p21, and p16, in ageing aortas further validated vascular ageing (see Supplementary data online, Figure S1C). Notably, senescence-associated β-galactosidase (SA-β-gal) staining revealed increased positivity, particularly within the endothelium layer, signifying endothelial senescence (Figure 1A). Focusing on m⁶A modifications in the vascular endothelium, RNA was extracted separately from the aortic endothelium layer (intima) and media and adventitia layers of young and aged mice. Consistent with our previous study,¹⁴ the high expression of CD31 in the endothelium component (Figure 1B), along with high expression of smooth muscle myosin heavy chain in the media and adventitia components (see Supplementary data online, Figure S1D), validated the successful isolation of pure RNA from the endothelium. A significant increase in m⁶A modification levels was observed in the aortic endothelium of aged mice compared to young mice (Figure 1C). Furthermore, the expression of genes related to m⁶A modification was assessed by qRT-PCR analysis, revealing a notable up-regulation of METTL14 mRNA in the aortic endothelium of aged mice (Figure 1D). In contrast, there was no alteration in the expression of METTL14 in the aortic media and adventitia layer of aged mice (see Supplementary data online, Figure S1E). To validate this notion across different species, an analysis was conducted on the human popliteal artery, which consistently demonstrated a specific elevation in METTL14 levels in the arterial endothelium of aged subjects (Figure 1E–G). Single-cell RNA-seq data from macaque aortas further supported the up-regulation of METTL14 in aged endothelium (see Supplementary data online, Figure S1F). At the cellular level, D-galactose (D-gal) was utilized to induce senescence in human umbilical vein endothelial cells (HUVECs), human aortic endothelial cells (HAECs), and mouse aortic endothelial cells (MAECs). Both qRT-PCR and western blot analyses showed an up-regulation of METTL14 expression following D-gal treatment in these endothelial cells (Figure 1H–J). Furthermore, we generated replicative senescence (RS) models in the same cell lines through prolonged serial passaging. Replicative senescence cells exhibited canonical senescence hallmarks, including increased SA-β-gal activity, reduced proliferative capacity evidenced by diminished Ki-67 positivity and 5-bromo-2-deoxyuridie (BrdU) incorporation, and elevated γ-H2AX foci indicative of persistent DNA double-strand breaks. Strikingly, METTL14 expression was also robustly elevated in all RS models (RS-HUVECs, RS-HAECs, and RS-MAECs), paralleling the expression pattern observed in D-gal-induced senescence (see Supplementary data online, Figure S2). These findings provide compelling evidence that METTL14 expression is up-regulated in vascular endothelium of aged mice, underscoring its potential role in endothelial cell senescence amidst vascular ageing.

Figure 1.

Up-regulation of METTL14 in aged vascular endothelium. (A) SA-β-gal staining of aorta sections revealed senescence-positive areas in aged mouse aorta (n = 6 per group), scale bar = 100 μm, the arrow points to the vascular endothelium. (B) qRT-PCR analysis of CD31 mRNA levels in mouse aortic endothelium, and in the media and adventitia (n = 6 per group). (C) m⁶A levels in young and aged mouse aortic endothelium measured by m⁶A colorimetric Epi Quik™m⁶A RNA Methylation Kit (n = 6 per group). (D) Expression levels of m⁶A modification related genes (METTL3, METTL14, WTAP, FTO, ALKBH5, YTHDF1, YTHDF2, YTHDC1) in mouse aortic endothelium assessed by qRT-PCR (n = 6 per group). (E) qRT-PCR analysis of METTL14 mRNA levels in human popliteal artery (n = 6 per group). (F and G) Representative images of immunofluorescence staining for the expression of METTL14 protein in human popliteal artery, bar graph summarizing the quantification of METTL14 fluorescence intensity co-localized with CD31 (n = 6 per group), scale bar = 50 μm. (H) qRT-PCR analysis of METTL14 mRNA levels in HUVECs, HAECs, and MAECs (n = 6 per group). (I and J) Western blot analysis of METTL14 protein levels in senescent HUVECs, HAECs, and MAECs (n = 6 per group). The data are expressed as the mean ± SEM. *P < .05, ***P < .001. SA-β-gal, SA-β-galactosidase; CTL, control; D-gal, D-galactose; HUVECs, human umbilical vein endothelial cells; HAECs, human aortic endothelial cells; MAECs, mice aortic endothelial cells

Endothelial cell-specific METTL14 knockdown or knockout mitigates ageing-induced arterial stiffness and vascular ageing in mice

The up-regulation of METTL14 expression within the endothelium of the aged aorta prompted us to investigate into its role in ageing-related arterial stiffness. To efficiently down-regulate endogenous METTL14 expression in the vascular endothelium, the AAVsig-sh-METTL14 vector that specifically targeting endothelial cells was transfected into female aged mice (18 months old) (METTL14 KD) (Figure 2A). Successful transfection was verified using qRT-PCR and immunofluorescence staining. The results indicated specific METTL14 down-regulation within the aortic endothelium without significant changes in the media and adventitia layer (Figure 2B–D, see Supplementary data online, Figure S3A). Under light microscopy, an increase in aortic wall thickness was observed in aged mice, yet this effect was mitigated in METTL14 KD mice (Figure 2E). Echocardiography-based assessment of vascular function revealed elevated levels of aortic PWV, carotid PWV, and carotid IMT in aged mice compared to their younger counterparts (Figure 2F–I). Additionally, the aged mice exhibited increased systolic blood pressure and pulse pressure (PP), but preserved diastolic blood pressure (see Supplementary data online, Figure S3B, Figure 2J). In contrast, these parameters were significantly reduced in METTL14 KD mice, suggesting that METTL14 knockdown mitigates age-related arterial stiffness. Exercise tolerance, which is closely correlated with vascular health as indicated by maximal run distance, maximal velocity, maximal run time, and workload, was reduced in aged mice but significantly improved after METTL14 knockdown (see Supplementary data online, Figure S3C). For further evaluation of METTL14’s effects on endothelium-dependent relaxation function, aortic dissociation for an ex vivo functional study was conducted. Acetylcholine-induced vascular relaxation was impaired in aged mice but were restored by METTL14 KD (Figure 2K and L). These findings suggest that METTL14 loss in endothelial cells improves age-induced vascular dysfunction, correlating with functional alterations in the endothelium. Anatomical and histological examinations were performed to determine if the improvement in vascular dysfunction in aged METTL14 KD mice was associated with vascular remodelling. Haematoxylin–eosin (HE) staining indicated increased aortic wall thickness in aged mice compared with young WT mice, a change mitigated by METTL14 KD (Figure 2M and N). Fibre breakage, a hallmark of increased arterial stiffness, assessed using Verhoeff–Van Gieson (VVG) staining, indicated that METTL14 KD inhibited elastin fibre breakage in the aortas of aged mice (Figure 2O and Q). Moreover, METTL14 knockdown mitigated collagen deposition and decreased the expression of col1a1, col3a1, and MMP2, which are all features associated with arterial stiffness (Figure 2P and R, see Supplementary data online, Figure S3D). Next, vascular ageing was examined, and the positive area of SA-β-gal staining and the expression levels of ageing-associated markers were assessed in mice aorta. Notably, METTL14 KD significantly alleviated vascular ageing in both the endothelium and media layers of the aorta in aged mice (Figure 2S and T, see Supplementary data online, Figure S3E). To assess potential sex differences in the effects of METTL14 in vascular ageing, we further examined the impact of METTL14 knockdown in male aged mice. The results revealed that endothelial cell-specific knockdown of METTL14 significantly improved vascular function in male aged mice, consistent with those observed in female aged mice (see Supplementary data online, Figure S4A–K).

Figure 2.

Endothelial-specific METTL14 knockdown mitigates ageing-induced arterial stiffness and vascular ageing. (A) Experiment timeline. C57BL/6J aged mice (18 months) were tail vein injected with AAVsig-sh-METTL14 vector three times, once every four weeks to establish endothelial-specific knockdown METTL14 mice (METTL14 KD). (B) Verification of endothelial-specific knockdown of METTL14 in aged mice by qRT-PCR (n = 6 per group). (C and D) Representative images of immunofluorescence staining for the expression of METTL14 protein in mouse aortic endothelium, bar graph summarizing the quantification of METTL14 fluorescence intensity co-localized with CD31 (n = 4 per group), scale bar = 50 μm, the arrow points to the merge area in vascular endothelium. (E) Light microscopy detected the morphological changes of mouse aorta (n = 3 per group). (F–I) Representative ultrasound images of the aortas, with aortic and carotid PWV parameters, and carotid IMT (n = 7 per group). d₀ represents the proximal aorta, d₁ represents the distal aorta, t₀ represents the time difference between the R wave of the ECG and the time when blood enters the proximal aorta, and t₁ represents the time difference between the R wave of the ECG and the time when blood enters the distal aorta. (J) PP in mice analysed using a tail non-invasive blood pressure instrument (n = 6 per group). (K and L) Aortic vasodilatory capacity measured in vitro using the DMT620 multichannel microvascular tension measurement system (n = 6 per group). (M and N) Representative HE staining images of mouse aortic sections to evaluate IMT thickness (n = 6 per group), scale bar = 150, 50 μm. (O and Q) VVG staining images showing number of elastin breaks in mouse aortic sections (n = 5 per group), scale bar = 150, 50 μm. (P and R) Representative Masson staining images of mouse aortic sections to evaluate collagen fibre deposition (n = 6 per group), scale bar = 150, 50 μm. (S and T) Representative SA-β-gal staining images showing senescence-positive areas in mice aortas (n = 6 per group), the arrow points to the SA-β-gal positive area. The data are expressed as the mean ± SEM. **P < .01, ***P < .001, ^##P < .01, ^###P < .001. METTL14 KD, METTL14 knockdown; PWV, pulse wave velocity; IMT, intima–media thickness; PP, pulse pressure; PE, phenylephrine; Ach, acetylcholine; HE, haematoxylin–eosin; VVG, Verhoeff–Van Gieson; SA-β-gal, SA-β-galactosidase

We then generated endothelial cell-specific METTL14 knockout mice (METTL14-CKO) to explore the role of METTL14 in vascular ageing. D-gal was used to induce ageing in METTL14-CKO mice and their WT littermates, as we reported previously¹⁵ (see Supplementary data online, Figure S5A). Similarly to naturally aged mice, D-gal treatment significantly elevated METTL14 expression; however, this increase was substantially reduced in METTL14-CKO mice (see Supplementary data online, Figure S5B–D). Arterial stiffness and vascular ageing were evaluated. Aortic PWV, carotid PWV, and carotid IMT were remarkably enhanced in WT mice following D-gal treatment. These detrimental effects were notably alleviated in METTL14-CKO mice (see Supplementary data online, Figure S5E–H). Additionally, arterial remodelling was evident in D-gal-treated WT mice, as indicated by HE, Masson, and VVG staining results, but was significantly reduced in METTL14-CKO mice (see Supplementary data online, Figure S5I–N). Meanwhile, the vascular ageing was also mitigated in METTL14-CKO mice (see Supplementary data online, Figure S5O–T). These findings further confirmed that METTL14 knockdown or deficiency in endothelial cells could effectively ameliorate arterial dysfunction in naturally aged and drug-induced ageing models.

Endothelial cell-specific overexpression of METTL14 induces arterial stiffness and vascular ageing in mice

In investigating the potential role of METTL14 in arterial stiffness and vascular ageing, we employed an endothelial cell-specific METTL14 overexpression plasmid (AAVsig-METTL14) administered to young WT mice (METTL14 OE) (Figure 3A). Remarkably, AAVsig-METTL14 markedly increased METTL14 levels in the aortic endothelium, with no significant alterations in the media and adventitia layers (Figure 3B–D, see Supplementary data online, Figure S3F). Microscopic examination revealed a substantial augmentation in aortic wall thickness in METTL14 OE mice (Figure 3E). Functional analysis further demonstrated that endothelium-overexpression of METTL14 resulted in vascular dysfunction, including enhanced arterial stiffness and vasorelaxation dysfunction (Figure 3F–L, see Supplementary data online, Figure S3G). Additionally, treadmill training experiments indicated decreased exercise capacity in METTL14 OE mice (see Supplementary data online, Figure S3H). Furthermore, METTL14 OE induced vascular remodelling, characterized by thickening of aortic wall, breakage of elastin fibres, excessive accumulation of collagen, and up-regulated col1a1, col3a1, and MMP2 (Figure 3M–O and Q–S, see Supplementary data online, Figure S3I). SA-β-gal staining and western blot results suggested that METTL14 OE mice exhibited remarkable vascular ageing (Figure 3P and T, see Supplementary data online, Figure S3J). These compelling findings underscore the deleterious action of METTL14 on arterial stiffness, thereby vascular ageing.

Figure 3.

Endothelial-specific overexpression of METTL14 induces arterial stiffness and vascular ageing. (A) Experiment timeline. C57BL/6J young mice (3 months) were tail vein injected with AAVsig-METTL14 vector three times, once every four weeks to establish endothelial-specific overexpression METTL14 mice (METTL14 OE). (B) Verification of METTL14 endothelial-specific overexpression in mice by qRT-PCR (n = 6 per group). (C and D) Representative images of immunofluorescence staining for the expression of METTL14 protein in mouse aortic endothelium, bar graph summarizing the quantification of METTL14 fluorescence intensity co-localized with CD31 (n = 4 per group), scale bar = 50 μm, the arrow points to the merge area in vascular endothelium. (E) Light microscopy detected the morphological changes of mouse aorta (n = 3 per group). (F–I) Representative ultrasound images of the aortas, with aortic and carotid PWV parameters, and carotid IMT (n = 7 per group). (J) PP in mice analysed using a tail non-invasive blood pressure instrument (n = 6 per group). (K and L) Aortic vasodilatory capacity measured in vitro using the DMT620 multichannel microvascular tension measurement system (n = 6 per group). (M and Q) Representative HE staining images of mouse aortic sections to evaluate aorta IMT thickness, scale bar = 150, 50 μm (n = 6 per group). (N and R) Representative VVG staining images of mouse aortic sections to evaluate the number of elastin breaks (n = 5 per group), scale bar = 150, 50 μm. (O and S) Representative Masson staining images of mouse aortic sections to evaluate the levels of collagen fibre deposition (n = 6 per group), scale bar = 150, 50 μm. (P and T) Representative SA-β-gal staining images to reveal senescence-positive areas in mouse arteries (n = 6 per group), the arrow points to the SA-β-gal positive area. The data are expressed as the mean ± SEM. **P < .01, ***P < .001; ^##P < .01, ^###P < .001. METTL14 OE, METTL14 overexpression; PWV, pulse wave velocity; IMT, intima–media thickness; PP, pulse pressure; PE, phenylephrine; Ach, acetylcholine; HE, haematoxylin–eosin; VVG, Verhoeff–Van Gieson; SA-β-gal, SA-β-galactosidase

Knockdown of METTL14 inhibits endothelial cell senescence and paracrine senescence

The effect of METTL14 expressed in vascular endothelial cells on vascular ageing led us to investigate its mechanisms within these cells. Senescence was induced in endothelial cells using D-gal, and METTL14 small interfering RNA (si-METTL14) was employed to silence endogenous METTL14 expression (Figure 4A, see Supplementary data online, Figure S6A). METTL14 expression, up-regulated in both the cytoplasm and nucleus of D-gal-induced HUVECs, was effectively attenuated by si-METTL14 (Figure 4B). Notably, si-METTL14 significantly alleviated senescence phenotypes in HUVECs. Molecular analyses revealed that si-METTL14 down-regulated key senescence-associated proteins, including p53, p21, and p16 (Figure 4C). These molecular changes were corroborated by functional assays: Flow cytometry demonstrated a marked reduction in senescent cell populations (Figure 4D and E), while cell cycle analysis showed partial rescue of G0/G1 phase arrest (Figure 4F and G), a hallmark of senescence. Consistently, si-METTL14-treated cells exhibited attenuated senescence markers (reduced SA-β-gal-positive cells and ROS levels, see Supplementary data online, Figure S6B–D) alongside restored proliferative capacity, as evidenced by elevated Ki-67 positivity and BrdU incorporation and diminished γ-H2AX foci indicative of persistent DNA double-strand breaks (see Supplementary data online, Figure S6C and E–G). Furthermore, the anti-senescence effects of si-METTL14 were consistently observed in replicative senescent HUVECs (see Supplementary data online, Figure S6H–J) and D-gal-treated HAECs and MAECs, as evidenced by a remarkable reduction in SA-β-Gal-stained cells following si-METTL14 treatment (see Supplementary data online, Figure S6K–O).

Figure 4.

Knockdown of METTL14 inhibits ECs senescence and paracrine senescence. (A) HUVECs were treated with D-gal and transfected with METLL14 siRNA. qRT-PCR analysis of METTL14 mRNA knockdown efficiency was performed (n = 6 per group). (B) Immunofluorescence staining was performed to analyse the localization and expression of METTL14 in HUVECs (n = 7 per group), scale bar = 20 μm. (C) Protein levels of METTL14, p53, p21, and p16 detected by western blot (n = 6 per group). (D and E) Flow cytometry and SA-β-Gal staining assessed the number of senescent HUVECs (n = 6 per group). (F and G) Statistical results of cell cycle changes of senescent HUVECs detected by flow cytometry (n = 6 per group). (H and I) Telomere detection kit detects the length of telomere and activity of telomerase in HUVECs (n = 6 per group). (J) Migration ability of HUVECs was detected by wound healing experiment (n = 8 per group), scale bar = 200 μm. (K–M) Co-culture of HUVECs and VSMCs to test SA-β-gal-positive cells (scale bar = 100 μm) and ROS accumulation levels (scale bar = 100 μm) in VSMCs (n = 6 per group). (N and O) qRT-PCR was used to analyse the levels of col1a1 and col3a1 mRNA in VSMCs (n = 6 per group). The data are expressed as the mean ± SEM. **P < .01, ***P < .001; ^##P < .01, ^###P < .001. CTL, control; D-gal, D-galactose; si-METTL14, METTL14 siRNA; si-NC, siRNA negative control; SA-β-gal, SA-β-galactosidase; ROS, reactive oxygen species; VSMCs, vascular smooth muscle cells

Given the vulnerability of telomeres to ROS, we further investigated the impact of si-METTL14 on telomere length and activity and found that si-METTL14 restored both telomere length and telomere activity in senescent HUVECs (Figure 4H and I). Meanwhile, si-METTL14 also enhanced the capacity of growth and migration of senescent HUVECs (Figure 4J). Considering the influence of EC-METTL14 on vascular remodelling and ageing in the medial layer, we proposed that these effects may be attributed to paracrine senescence. As illustrated in Figure 4K–O, co-culture with senescent HUVECs induced cell senescence and stimulated ROS production in normal VSMCs. Additionally, it up-regulated the activated fibroblast markers col1a1 and col3a1. These findings reveal that silencing METTL14 inhibits senescence in endothelial cells, ameliorating vascular ageing by blocking paracrine senescence.

Overexpression of METTL14 promotes endothelial cell senescence and paracrine senescence

We then turned to investigate the impact of METTL14 overexpression on endothelial cell senescence. The METTL14 overexpression plasmid (pcDNA3.1-METTL14) was transfected into senescent HUVECs, resulting in a significant elevation of METTL14 expression levels (see Supplementary data online, Figure S7A and B). Notably, compared to the D-gal-treated group, METTL14 overexpression enhanced senescence in HUVECs. This was evident from the substantial up-regulation of p53, p21, and p16 protein levels (see Supplementary data online, Figure S7C), validated by flow cytometry analysis (see Supplementary data online, Figure S7D and E). METTL14 overexpression also triggered cell cycle arrest in the G0/G1 phase and hindered HUVECs’ growth (see Supplementary data online, Figure S7F and G). Additionally, METTL14 overexpression resulted in a significant increase in SA-β-gal-positive cells and elevated ROS levels. This pro-senescence effect was further supported by diminished proliferative capacity, evidenced through reduced Ki-67 expression and impaired BrdU incorporation, and elevated γ-H2AX foci indicative of persistent DNA double-strand breaks (see Supplementary data online, Figure S6P-U). Furthermore, it exacerbated telomere shortening and inhibited telomere activity (see Supplementary data online, Figure S7H and I). Moreover, METTL14 overexpression limited the capacity of HUVECs to grow and migrate (see Supplementary data online, Figure S7J). Co-culture experiments further demonstrated that METTL14 significantly accelerated paracrine senescence in VSMCs (see Supplementary data online, Figure S7K–O). These results highlight the role of METTL14 overexpression in driving EC senescence and fostering paracrine senescence.

METTL14 promotes TLR4 m⁶A modification leading to SASP overproduction

The accumulation of SASPs plays a pivotal role in ageing-related vascular diseases and paracrine senescence.^16,17 To identify SASPs influenced by METTL14, we utilized the Cytokine Array Panel A kit to screen differentially secreted factors (Figure 5A). Aged mouse serum showed elevated levels of inflammatory factors such as ICAM-1, IL-27, CXCL12, IL-23, CXCL13, CXCL1, INF-γ, and IL-2, which decreased upon METTL14 knockdown (Figure 5B, see Supplementary data online, Figure S8A). To further elucidate the regulatory signalling of these SASPs in senescent endothelial cells, we conducted RNA-seq on normal and METTL14-overexpressing endothelial cells. METTL14-overexpressing endothelial cells exhibited up-regulation of 1461 genes and down-regulation of 569 genes (Figure 5C). Gene ontology (GO) enrichment analysis of biological process gene signatures demonstrated a significant enrichment of inflammatory processes, influenced by METTL14, which specifically involved the Toll-like receptor (TLR) signalling pathway (Figure 5D). Consistently, bioinformatics analyses on normal and senescent endothelial cells (GSE13712) demonstrated an up-regulation of TLR signalling pathway in senescent endothelial cells (see Supplementary data online, Figure S8B).

Figure 5.

METTL14 promotes TLR4 m⁶A modification to induce SASP production. (A and B) Mouse Cytokine Array Panel A kit was used to detect the levels of inflammatory cytokines in mouse serum, and the values were measured by ImageJ software. (C) RNA-seq analysis of differentially expressed genes in control and METTL14 overexpressing HUVECs. (D) The enrichment analysis of GO biological process gene in METTL14 overexpressing HUVECs. (E) TLRs mRNA levels in HUVECs were tested by qRT-PCR (n = 6 per group). (F) Interaction network of altered SASPs and TLR signalling, visualized using the STRING database. (G) qRT-PCR analysis of METTL14 mRNA levels in human popliteal artery (n = 6 per group). (H) Representative images of immunofluorescence staining for the expression of TLR4 protein in human popliteal artery, bar graph summarizing the quantification of TLR4 fluorescence intensity co-localized with CD31 (n = 6 per group), scale bar = 50 μm. (I) Analysis of the m⁶A modification sites on human TLR4 mRNA using SRAMP. (J and K) MeRIP experiments assessed the adenine methylation of TLR4 mRNA in HUVECs. (n = 3 per group). (L) RIP analysis assessed the interaction between METTL14 protein and TLR4 mRNA (n = 3 per group). (M) Detection of the TLR4 mRNA stability following METTL14 overexpression (OE-METTL14) or knockdown (si-METTL14) (n = 5 per group). (N) Western blot tested the protein levels of TLR4, MyD88, NF-κB and p-NF-κB in HUVECs (n = 6 per group). The data are expressed as the mean ± SEM. *P < .05, **P < .01, ***P < .001; ^#P < .05, ^##P < .01, ^###P < .001. CTL, control; D-gal, D-galactose; GO, gene ontology, SASP, senescent associated secretory phenotype; OE-METTL14, METTL14 overexpression; si-METTL14, METTL14 siRNA; si-NC, siRNA negative control

To validate this finding, we assessed the expression of nine TLR subtypes, observing a remarkable inhibition of TLR4 expression upon METTL14 knockdown in HUVECs (Figure 5E). These regulatory effects of METTL14 on TLR4 expression were consistently observed in HAECs and MAECs (see Supplementary data online, Figure S8C and D). Further analysis using the String database uncovered stronger interaction of TLR4 with dysregulated SASPs compared to TLR7 and TLR9 (Figure 5F). Additionally, we detected the expression of TLR4 in the human popliteal artery samples obtained from young and elderly patients, which confirmed the increased expression of TLR4 in aged arteries (Figure 5G and H). Consequently, TLR4 was prioritized for further in-depth investigations. Exploring the potential link between METTL14 and TLR4, we predicted four high-confidence m⁶A modification sites (837, 6831, 7813, and 8819 nt) on the 3′UTR of human TLR4 gene and two sites on the 3′UTR of mouse TLR4 gene through a dedicated website (http://www.cuilab.cn/sramp) (Figure 5I, see Supplementary data online, Figure S8E). Subsequent methylated RNA immunoprecipitation (MeRIP) experiments demonstrated that the adenine at position 8819 of TLR4 exhibited the highest level of methylation in HUVECs, and was decreased upon METTL14 knockdown (Figure 5J and K). RNA immunoprecipitation (RIP) analysis further verified the interaction between METTL14 and TLR4 mRNA (Figure 5L). In MAECs, the adenine at position 6168 in the TLR4 3′UTR displayed the highest degree of methylation (see Supplementary data online, Figure S8F). METTL3 always participate the regulatory effects of METTL14 on m⁶A modification. Generally, METTL3 and METTL14 form a stable 1:1 heterodimer structure and act together for m⁶A modification.¹⁸ Our results consistently demonstrated an increased interaction between METTL3 and METTL14 in senescent HUVECs compared to the control group, which is attributed to the elevated expression of METTL14 during ageing (see Supplementary data online, Figure S8G and H). Furthermore, we conducted pulldown analysis to identify potential ‘readers’ that serve as the final executor of m⁶A modification on TLR4 mRNA. Our results indicated a pronounced binding of YTHDF1 and YTHDC1 to TLR4 mRNA (see Supplementary data online, Figure S8I), with both readers demonstrating an ability to enhance mRNA stability.^{19  20} Consistent with these observations, we also noted an increase in TLR4 mRNA stability upon overexpression of METTL14 and reduction in its stability following METTL14 knockdown (Figure 5M). These findings suggest that METTL14 regulates TLR4 through m⁶A modification, thus influencing SASP production in senescent endothelial cells.

TLR4/MyD88/NF-κB signalling mediates METTL14’s role in vascular ageing

The TLR4/MyD88/NF-κB signalling pathway is well recognized for its involvement in inflammatory responses. Our study revealed that silencing METTL14 significantly blocked the activation of the TLR4/MyD88/NF-κB signalling pathway in senescent HUVECs (Figure 5N). In contrast, overexpressing METTL14 in senescent HUVECs significantly promoted the activation of this signalling pathway (see Supplementary data online, Figure S9A). Immunofluorescence staining results suggested that METTL14 KD significantly inhibited the up-regulation of p-NF-ĸB and TLR4 in aortas of aged mice, whereas METTL14-OE exerted opposite effects (see Supplementary data online, Figure S9B–E). These findings strongly imply that METTL14 activates TLR4/MyD88/NF-κB signalling in vascular ageing. To further elucidate the role of TLR4 in METTL14-mediated EC senescence, co-transfection experiments using TLR4 overexpression or knockdown plasmids in conjunction with si-METTL14 were conducted. Overexpression of TLR4 counteracted the beneficial effects of si-METTL14 on inflammation signalling and cellular senescence (Figure 6A). Additionally, concurrent knockdown of both METTL14 and TLR4 exerted synergistic effects, resulting in a further reduction in TLR4/MyD88/NF-κB signalling activation and cellular senescence (see Supplementary data online, Figure S10A). Furthermore, the involvement of TLR4 in vascular ageing was investigated in vivo. Endothelial cell-specific knockdown of TLR4 (TLR4-KD) remarkably improved arterial function in aged mice, as demonstrated by reductions in aortic PWV, carotid PWV, and carotid IMT (Figure 6H–K). Arterial remodelling, as indicated by HE, Masson, and VVG staining, was also notably reduced following TLR4-KD intervention (Figure 6L–Q). As expected, TLR4 expression in aortic endothelium was remarkably decreased flowing TLR4-KD, which was reversed by METTL14-OE (Figure 6R). These findings provide compelling evidence supporting the role of TLR4 in mediating the detrimental effects of METTL14 on vascular ageing.

Figure 6.

TLR4 mediates the regulatory effects of METTL14 on vascular ageing. (A–G) Western blot analysis of TLR4, MyD88, NF-κB, p-NF-κB, p53, p21, and p16 in senescent HUVECs, transfected with si-METTL14 or OE-TLR4 (n = 6 per group). (H–K) Representative ultrasound images of aortas, aortic and carotid PWV parameters, and carotid IMT (n = 6 per group). (L and O) Representative HE staining images of mouse aortic sections to evaluate the levels of aorta IMT thickness (n = 6 per group), scale bar = 150, 50 μm. (M and P) Representative Masson staining images of mouse aortic sections to evaluate the levels of collagen fibre deposition (n = 6 per group), scale bar = 150, 50 μm. (N and Q) Representative VVG staining images of mouse aortic sections to evaluate the quantities of elastic fibre lamellae (n = 5 per group), scale bar = 150, 50 μm. (R) TLR4 mRNA levels in mouse vascular endothelium measured by qRT-PCR (n = 6 per group). The data are expressed as the mean ± SEM. ***P < .001; ^###P < .001; ^&&P < .01; ^&&&P < .001. D-gal, D-galactose; si-METTL14, METTL14 siRNA; si-NC, siRNA negative control; OE-TLR4, TLR4 overexpression; OE-NC, OE-TLR4 negative control; TLR4-KD, TLR4 knockdown; METTL14-OE, METTL14 overexpression; PWV, pulse wave velocity; IMT, intima–media thickness; HE, haematoxylin–eosin; VVG, Verhoeff–Van Gieson

METTL14 and TLR4 are positively correlated with vascular ageing

The information concerning 62 subjects detailed in Supplementary data online, Table S1 highlighted distinct differences between younger (ages 28–59) and older (ages 60–75) individuals. Carotid ultrasound examinations conducted across different age groups showed that aged subjects (ages 60–75) had visibly thicker carotid arteries and a higher incidence of carotid plaque formation compared to their younger counterparts (ages 28–59) (Figure 7A–D). Moreover, aged subjects displayed a significant increase in brachial-ankle pulse wave velocity (baPWV) (Figure 7E). Subsequent analysis of the expression of METTL14 and TLR4 mRNA in the whole blood revealed its up-regulation in aged subjects (Figure 7F and G), suggesting that circulating METTL14 and TLR4 levels were associated with vascular ageing. However, the methylated form of TLR4 mRNA remains constant between the young and aged subjects (see Supplementary data online, Figure S11A).

Figure 7.

Increased levels of METTL14 and TLR4 are correlated with vascular ageing in humans. (A and B) Doppler ultrasound performed to analysis subjects of varying ages, younger subjects (ages 28–59, n = 31), older subjects (ages 60–75, n = 31), scale bar = 10 mm, 3 mm. (C and D) Doppler ultrasound analysis of carotid IMT and plaque formation in varying ages subjects (n = 31 per group). (E) baPWV was detected in younger and older subjects using PWV measurement device (n = 31 per group). (F and G) Detection the mRNA levels of METTL14 and TLR4 in the human whole blood by qRT-PCR (n = 31 per group). (H and I) ROC curve analysis of the predictive capabilities of the blood METTL14 (n = 31 per group). (J and K) ROC curve analysis of the predictive capabilities of the blood TLR4 (n = 31 per group). Data were expressed by mean ± SEM. ***P < .001. IMT, intima–media thickness; baPWV, brachial-ankle pulse wave velocity; ROC, receiver operating characteristic

Association of METTL14 and TLR4 with human vascular ageing

To further describe the more accurate relationship between blood METTL14 and TLR4 and human vascular ageing, we performed univariate and multivariate linear regression analyses (Table 1). Univariate analysis showed that age, IMT, PWV, and the presence of plaques were significantly positively correlated with blood METTL14, and E/A ratio was significantly negatively correlated with blood METTL14 levels. Multivariate linear regression analysis showed that age (P = .002) and plaques (P = .048) were independently associated with METTL14 (adjusted r² = 0.273, F = 5.497, P < .001). We also found that age, systolic blood pressure, diastolic blood pressure, IMT, PWV, and plaques were significantly positively correlated with blood TLR4, and E/A ratio was significantly negatively correlated with blood TLR4 level. Multiple linear regression analysis showed that age (P = .048) was independently correlated with TLR4 (adjusted r² = 0.442, F = 7.776, P＜.001). Receiver operating characteristic (ROC) curve analysis was employed to evaluate the predictive power of blood METTL14 and TLR4 for arteriosclerosis and atherosclerosis. The results were promising, showing that METTL14 and TLR4 levels had predictive significance for these conditions, with high area under the curve values of 0.7823, 0.8120, 0.8208, and 0.7724, respectively (Figure 7H–K). To figure out the resources of METTL14 and TLR4 mRNAs in circulating, we detected their mRNA levels in blood exosomes and white blood cells (WBCs) of young and aged individuals. Specifically, the METTL14 mRNA levels in blood exosomes of aged individuals were significantly elevated, whereas no such elevation was noted in WBCs (see Supplementary data online, Figure S11B and C). Consistently, exosomes derived from D-gal-treated HUVECs and RS-HUVECs exhibited a parallel up-regulation of METTL14 transcripts (see Supplementary data online, Figure S11D and E). In contrast, levels of TLR4 mRNA increased in both blood exosomes and WBCs (see Supplementary data online, Figure S11F and G). Furthermore, the enrichment of METTL14 and TLR4 mRNAs in exosomes was also validated by the browse results from exoRBase website (see Supplementary data online, Figure S11H). Therefore, we hypothesize that in the ageing process, endothelial cells release exosomes into the blood that contain a significant amount of METTL14 mRNA, which is linked to vascular ageing. These findings underscore the role of circulating METTL14 and TLR4 as biomarkers for human vascular ageing and age-related vascular diseases. Importantly, the observed associations in humans align with earlier observations made in mice, strengthening the link between METTL14, TLR4, and vascular ageing across species.

Table 1.

Clinical and ultrasonic testing variables associated with METTL14 or TLR4

	METTL14				TLR4
	Univariable		Multivariable		Univariable		Multivariable
	Sβ (95% CI)	P value	Sβ (95% CI)	P value	Sβ (95% CI)	P value	Sβ (95% CI)	P value
Gender (male vs female)	−0.14 (−0.221 to 0.199)	.916			−0.039 (−0.166 to 0.122)	.762
Age (years, 28–59 vs 60–75)	0.500 (0.219 to 0.574)	<.001	0.599 (0.043 to 0.911)	.002	0.606 (0.218 to 0.441)	<.001	0.506 (0.003 to 0.550)	.048
BMI (kg/m2)	0.141 (−0.018 to 0.061)	.273			0.243 (−0.001 to 0.052)	.057
SBP (mmHg)	0.194 (−0.003 to 0.022)	.130			0.379 (0.005 to 0.021)	.002	0.077 (−0.006 to 0.011)	.538
DBP (mmHg)	0.012 (−0.017 to 0.018)	.924			0.313 (0.003 to 0.026)	.013	0.205 (−0.001 to 0.020)	.077
PP (mmHg)	0.209 (−0.002 to 0.025)	.104			0.176 (−0.003 to 0.016)	.172
TC (mmol/L)	0.127 (−0.065 to 0.193)	.325			0.007 (−0.087 to 0.092)	.957
LDL cholesterol (mmol/L)	0.028 (−0.023 to 0.029)	.827			0.076 (−0.012 to 0.023)	.558
Triglycerides (mmol/L)	0.140 (−0.043 to 0.148)	.277			0.221 (−0.008 to 0.121)	.084
Leucocytes (109/L)	−0.101 (−0.082 to 0.036)	.436			0.098 (−0.025 to 0.056)	.446
Eosinophils (109/L)	−0.043 (−0.848 to 0.607)	.741			−0.219 (−0.910 to 0.064)	.087
Monocytes (109/L)	−0.013 (−0.618 to 0.556)	.917			0.082 (−0.274 to 0.529)	.528
Lymphocytes (109/L)	0.073 (−0.144 to 0.258)	.571			0.006 (−0.135 to 0.141)	.964
Ureophil (mmol/L)	0.228 (−0.008 to 0.156)	.075			0.241 (−0.002 to 0.109)	.059
IMT (mm)	0.346 (0.199 to 1.125)	.006	−0.085 (−0.087 to 0.484)	.618	0.492 (0.349 to 0.939)	<.001	0.079 (−0.299 to 0.505)	.609
BaPWV (m/s)	0.391 (0.021 to 0.86)	.002	−0.201 (−0.089 to 0.034)	.377	0.470 (0.023 to 0.065)	<.001	−0.153 (−0.053 to 0.024)	.452
Plaque	0.407 (0.137 to 0.516)	.001	0.247 (0.002 to 0.396)	.048	0.356 (0.063 to 0.329)	.005	0.092 (−0.073 to 0.174)	.412
E/A	−0.326 (−0.438 to −0.061)	.01	−0.110 (−0.283 to 0.114)	.399	−0.464 (−0.365 to −0.123)	<.001	−0.224 (−0.238 to 0.003)	.056

Univariable linear regression models were constructed to identify the clinical and ultrasonic testing variables associated with blood METTL14 or TLR4 levels. Factors with a significance level of P < .05 were selected as independent variables for the multivariable analysis. 95% confidence intervals were provided. Bold values indicate values less than .05. Other abbreviations are shown in Supplementary data online, Table S1.

Sβ, standardized B.

Collectively, this study provides compelling evidence for blood METTL14 and TLR4 levels as a predictive marker for vascular ageing and age-related vascular diseases in humans, reinforcing its significance in this context.

Discussion

The current study presents a compelling insight into the role of METTL14, an m⁶A modification factor, in the context of vascular ageing across both mouse and human models. The findings indicate a pivotal association between METTL14 expression and various facets of vascular ageing, shedding light on its mechanistic involvement. There are several main and novel findings. First, METTL14 showed significant up-regulation in the aortic endothelium of aged mice, popliteal arterial endothelium of aged individuals and senescent HUVECs, HAECs, and MAECs, suggesting its potential involvement in the ageing process. Second, the endothelium-specific knockdown/knockout of METTL14 in aged mice demonstrated positive effects: ameliorating arterial stiffness, vascular ageing, cell senescence, oxidative stress, and inflammatory response. Third, conversely, the overexpressing METTL14 exacerbated these ageing-promoting effects, indicating a direct influence of METTL14 levels on vascular ageing pathways. Fourth, human studies also indicated a correlation between elevated METTL14/TLR4 levels and ageing, as well as with vascular diseases associated with ageing, adding translational significance to the findings. Collectively, this study proposes a mechanistic understanding of how METTL14 may influence vascular ageing. Specifically, it governs the production of SASPs by methylating and up-regulating TLR4, thereby regulating cell senescence and contributing to vascular ageing. This mechanistic understanding links METTL14’s epigenetic regulatory role to the modulation of cellular senescence and inflammation in the context of vascular ageing (Structured Graphical Abstract).

Ageing is recognized as an independent risk factor for vascular dysfunction, yet the underlying mechanisms remain largely unexplored. Endothelial cells are acknowledged as one of the initial cell types to undergo senescence with advancing age.⁴ Despite the investigation of m⁶A modification roles in vascular diseases in young mice, specific roles in vascular ageing, particularly in endothelial ageing, remain elusive. The functions of m⁶A modification in diseases significantly hinge on the expression levels of regulatory genes. Analysis of m⁶A gene expression patterns revealed a noteworthy up-regulation of METTL14 in the aortic endothelium of aged mice, popliteal arterial endothelium of elderly individuals, and senescent HUVECs, HAECs, and MAECs. Interestingly, Zhang et al.¹¹ reported a decrease in METTL14 in the intestine of aged mice and Drosophila melanogaster. Qian et al.²¹ recently demonstrated the up-regulation of METTL14 in aged ovary. These studies, along with ours, illustrate the diverse expression patterns of METTL14 in ageing. Prior studies examining the roles of METTL14 in blood vessels have predominantly concentrated on vascular diseases in young individuals. For example, METTL14 has been implicated in the methylation of vascular osteogenic transcripts, thereby regulating vascular calcification induced by indoxyl sulfate.²² Additionally, the METTL3–METTL14 complex induces necroptosis and inflammation of VSMCs by promoting m⁶A modification of receptor-interacting protein 3 (RIP3) in abdominal aortic aneurysms.²³ METTL14 increases the m⁶A modification of pri-miR-19a and promotes the processing of mature miR-19a, thereby fostering the proliferation and invasion of atherosclerotic vascular endothelial cells.²⁴ However, the roles of endogenous METTL14 in vascular ageing remain undiscovered.

The development of biological or chemical drugs facilitating the selective ablation of senescent cells (senolytics) has presented additional evidence supporting senescent endothelial cells as a viable therapeutic target.²⁵ The AAVsig vector synthesized for the mouse study demonstrated a high affinity for endothelial cells. Including the endothelial cell-targeting peptide SIGYPLP, this vector has become a widely used genetic tool for its specific targeting capabilities and shows promise as a candidate for endothelium-related senolytics.^26,27 Our findings in aged animals provided compelling evidence that endothelium-specific knockdown of METTL14 effectively alleviates ageing-induced arterial stiffness and vascular ageing, while the overexpression of METTL14 produces opposite effects. Additionally, we prompted accelerated ageing in mice using D-gal, a pharmacologic agent that mimics natural ageing processes and has been extensively utilized in studies concerning organ ageing.^28–30 In line with the present study, various previously published reports have demonstrated that D-gal administration in rodents leads to vascular remodelling, impaired endothelial function, and elevated blood pressure.^31,32 METTL14 deficiency in vascular endothelium ameliorated above detrimental effects, thereby underscoring the potential therapeutic application of METTL14 in alleviating drug-induced ageing. Nevertheless, further studies are needed to assess the role of METTL14 in other cell types during the process of vascular ageing.

The roles of METTL14 in senescent endothelial cells have not yet been thoroughly investigated. Several studies have demonstrated its regulatory effects on senescence in various cell types. For instance, Zhang et al.¹¹ proposed that the knockdown of METTL14 reduced the m⁶A level of LBR, resulting in LBR mRNA instability and thereby inducing cellular senescence. Additionally, overexpression of METTL14 promoted the expression of SASP in induced pluripotent stem cells (iPSCs) in an NF-κB-dependent manner.³³ METTL14-mediated m⁶A methylation was shown to promote p21 translation, resulting in increased p21 expression during cellular senescence induced by oxidative stress.¹⁰ METTL14 enhances the stability and production of PTCHD4 mRNA through m⁶A modification, thereby protecting senescent cells from senolysis and apoptosis.³⁴ METTL14 is also involved in TNF-α-induced miR-34a-5p m⁶A modification, promoting senescence in neural progenitors and nucleus pulposus cells in patients with intervertebral disc degeneration.³⁵ Our findings in the present study illustrate that METTL14 regulates cell senescence in endothelial cells (HUVECs, HAECs, and MAECs) across both D-gal-induced and RS cell models. This suggests that targeted inhibition of intracellular METTL14 using siRNA or other specific inhibitors could hold promise as a potential anti-ageing therapy for preventing vascular ageing and related vascular diseases in humans. Numerous studies have showcased the substantial therapeutic potential of METTL3 inhibitors across various diseases,^36,37 however, pharmacological therapies targeting METTL14 are still largely unexplored. Isoliquiritigenin (ISO) was found to up-regulate METTL14 mRNA by activating the transcription factor FOXO3a.³⁸ Metformin increased the expression of METTL14, subsequently promoted methylation levels in pancreatic β-cells.³⁹ Further investigations focused on METTL14 inhibitors are necessary to evaluate the application of METTL14 as a therapeutic target for treating ageing-related diseases.

Our findings demonstrate that the modulation of METTL14 influences cell senescence in neighbouring VSMCs in a paracrine manner. Consistent with in vitro studies, altering METTL14 specifically in the endothelium also influences arterial remodelling and contractility, with mechanistic correlations to VSMC damage. Moreover, these gene therapies also lead to systemic impairments, such as decreased exercise tolerance. Further research is necessary to elucidate whether the observed influence on overall health is primarily attributed to modified functions in blood vessels or skeletal muscle. Since the transcriptome data of senescent HUVECs and METTL14-overexpressed HUVECs revealed the reprogramming of inflammatory pathways, we focused on inflammation, which serves as a signal regulator of oxidative stress and cell senescence. Furthermore, our results provide support for the association between METTL14-mediated cell senescence and SASP production. TLR4 signalling was identified as a contributor to inflammation, and inhibition of this signalling exerted beneficial effects on vascular diseases and ageing. For example, activation of TLR4 signalling induces the transcription of proinflammatory cytokines, thereby exacerbating pathological retinal angiogenesis.⁴⁰ Moreover, Salmonella typhimurium triggers inflammation in murine intestinal tissue through the activation of TLR4-mediated molecular ageing, subsequently impairing the host defense against intestinal inflammation.⁴¹ Rosiglitazone attenuates high-fat diet-induced atherosclerotic plaque formation in ApoE^−/− mice through decreasing TLR4-mediated production of proinflammatory cytokines.⁴² The m⁶A modification of TLR4 has been identified in various biological processes, including neutrophil activation, diabetic osteoporosis, and multiple myeloma.^43–45 Here, we revealed that METTL14 increased the mRNA stability of TLR4 by enhancing its m⁶A modification, which is conserved between humans and mice. Consequently, the MyD88/NF-κB signalling pathway, which is activated by TLR4, was also influenced by METTL14. Notably, TLR4 knockdown exerted protective effects against vascular ageing and endothelial cell senescence, which were counteracted by METTL14 overexpression. These findings emphasize the role of TLR4 as a key regulator in the METTL14-SASPs-cell senescence axis, governing the development of vascular ageing. Exploring TLR4 inhibitors may offer promising therapeutic strategies for addressing vascular ageing and related diseases. However, SASP production is complex. The observed METTL14-induced SASP overload may not be specific to TLR4 alone, as we also revealed an influence of METTL14 on TLR7 and TLR9. In addition, Liu et al.⁹ reported that nuclear METTL14 binds to the enhancer of NF-κB and drives the production of SASPs in an m⁶A-independent manner. Therefore, it would be interesting to study the role of other signalling pathways in METTL14-expressed tissues and cells during vascular ageing.

Furthermore, we demonstrated that the METTL14 and TLR4 mRNA levels in whole blood are correlated with vascular ageing in humans. As evidenced by others, circulating exosomes released from endothelial cells contain potential valuable biological information for biomarker discovery in various diseases.⁴⁶ The elevated levels of METTL14 mRNA were observed in blood exosomes but not WBCs, suggesting that vascular endothelium might secrete exosomes containing METTL14 mRNAs into blood. However, levels of TLR4 mRNA increased in both blood exosomes and WBCs. The enrichment of METTL14 and TLR4 mRNA in exosomes was also validated by the browse results from exoRBase website. The elevated TLR4 mRNA levels in WBCs may be attributed to the abundance of SASPs in the blood, which frequently promote systemic inflammation associated with ageing.⁴⁷ Additionally, our results indicate that circulating METTL14 and TLR4 mRNA levels are sensitive to predict age-related aortic diseases, including atherosclerosis and arteriosclerosis. The Asp299Gly SNP frequency of TLR4 was shown to correlated with risk of atherosclerosis, acute coronary events, and longevity.⁴⁸ Additionally, the increased circulating TLR4 level was negatively related to lifespan and healthspan in elderly.^49,50 These findings along with ours highlighted the possibility that increase of circulating METTL14 and TLR4 may serve as biomarkers for biological ageing and a prognostic factor for ageing-related vascular diseases. However, additional studies are still required to investigate whether circulating METTL14 or TLR4 can modulate other age-related diseases.

This study had several limitations. First, while our findings demonstrate that the TLR4/MyD88/NF-κB pathway partially contributes to the role of METTL14 in vascular ageing, it is important to note that METTL14 also affects other Toll-like receptors. Therefore, additional mechanisms may contribute to the effects of METTL14 on vascular ageing, warranting further investigation. Second, we found that the knockdown of METTL14 in endothelium attenuated vascular ageing, but we did not test its long-term effect on the lifespan, which requires further evaluation.

Supplementary data

Supplementary data are available at European Heart Journal online.

Declarations

Disclosure of Interest

All authors declare no disclosure of interest for this contribution.

Data Availability

All data are available in the manuscript or the supplementary materials.

Funding

This study was supported by the Noncommunicable Chronic Diseases-National Science and Technology Major Project (2024ZD0537908), the National Natural Science Foundation of China (82273919, 82270396, U24A20813 and U21A20339), the HMU Marshal Initiative Funding (HMUMIF-21022), and the Science Foundation for the Excellent Youth Scholars of Heilongjiang Province (JJ2023YX0509).

Ethical Approval

The Ethic Committees of Harbin Medical University approved study protocols in human blood samples and animals (approval number: IRB1001719) and human popliteal arteries samples (approval number: IRB5065724). The study protocols in human blood and human popliteal arteries samples were performed in accordance with the ethics principles in the Declaration of Helsinki. Written informed consent was given by each patient before participation in the study. Use of animals was conformed to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996).

Pre-registered Clinical Trial Number

None supplied.