N6-methyladenosine (m6A) RNA modification in fibrosis and collagen-related diseases

Abstract

Fibrosis is an abnormal tissue healing process characterized by the excessive accumulation of ECM components, such as COL I and COL III, in response to tissue injury or chronic inflammation. Recent advances in epitranscriptomics have underscored the importance of m6A modification in fibrosis. m6A, the most prevalent modification in eukaryotic RNA, is catalyzed by methyltransferases (e.g., METTL3), removed by demethylases (e.g., FTO), and recognized by reader proteins (e.g., YTHDF1/2). These modifications are crucial in regulating collagen metabolism and associated diseases. Understanding the role of m6A modification in fibrosis and other collagen-related conditions holds promise for developing targeted therapies. This review highlights the latest progress in this area.

Introduction

Collagen is a crucial protein in the extracellular matrix (ECM) [1], providing essential structural support to various tissues and organs [2, 3]. Its metabolism involves a delicate balance between synthesis, assembly, and degradation. Dysregulation of collagen metabolism can lead to fibrosis, characterized by an excessive buildup of collagen and other matrix components in tissues. Fibrosis occurs as a response to tissue injury or chronic inflammation and can affect different organs in the body [4, 5].

In fibrosis, abnormal collagen synthesis and deposition play a central role. Activated fibroblasts and other cells increase collagen production, driven by signaling pathways like TGF-β [6] and CTGF [7]. Additionally, impaired collagen degradation contributes to fibrosis, resulting from an imbalanced ratio of matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs) [8].

Epitranscriptomics is an emerging and crucial field in recent years, encompassing over 170 distinct post-transcriptional RNA modifications or editing events, which play important roles in the regulation of fibroblasts and fibrosis [9, 10], such as the liver [11], lungs [12], kidneys, and heart [13]. Among these modifications, N6-adenosine methylation, known as m6A modification, stands out as the most prevalent modification in eukaryotic RNA and was first reported in 1974 [14]. The m6A modification is catalyzed by the methyltransferase complex (comprising METTL3, METTL14, and WTAP as co-factors), removed by demethylases (such as FTO and ALKBH5), and recognized by reader proteins (e.g., YTHDF1/2/3, YTHDC1/2, IGF2BP1/2/3), dynamically regulating gene expression at the post-transcriptional level and contributing to the development of various diseases [15]. Recent research has shed light on the roles of m6A regulatory factors in fibrosis and collagen-related diseases.

Understanding the role of m6A modification in fibrosis and other collagen-related conditions holds significant promise for the development of targeted therapies. This review highlights the latest advancements and progress in this area.

Collagen metabolism and fibrosis

The extracellular matrix is a vital three-dimensional macromolecular network consisting of collagen proteins, proteoglycans/glycosaminoglycans, elastin proteins, fibronectin, laminin, and other glycoproteins [16]. It plays a crucial role in tissue remodeling and the regulation of cell behavior. Collagen, a protein with a triple-helix structure [17], is the predominant constituent of the ECM, making up approximately 30% of the total protein content in the human body [1]. Its main functions include providing elasticity, stability, and support to tissues [2]. There are 28 different types of collagen identified so far, with COL I, COL III, and COL V mainly produced by fibroblasts, while COL IV is primarily expressed by epithelial cells and endothelial cells. In some cases, cancer cells and tumor-associated macrophages can also produce collagen [18].

Fibrosis is an abnormal tissue healing process characterized by excessive accumulation of ECM components such as COL I and COL III in response to tissue injury or chronic inflammation. It can affect various organs, such as the liver, lungs, kidneys, and heart. Fibrosis disrupts tissue architecture and function, leading to organ dysfunction and organ failure. Tissue healing involves three stages: inflammation, proliferation, and remodeling [19]. Fibroblasts play a significant role in this process, transforming into myofibroblasts with contractile force during the proliferation stage and driving wound contraction during the remodeling stage [20, 21]. The proper transformation of fibroblasts to myofibroblasts and their subsequent apoptosis are crucial for appropriate tissue healing. However, under pathological conditions, this normal wound healing process is disrupted, resulting in persistent myofibroblast presence and ECM remodeling [22].

In the context of fibrosis, abnormal collagen synthesis and deposition play a central role. Fibrotic tissues exhibit increased collagen production by activated fibroblasts and other cell types. This enhanced collagen synthesis is triggered by various signaling pathways, such as TGF-β [6] and CTGF [7]. These pathways promote the expression of collagen genes and drive fibroblast-to-myofibroblast transition, characterized by increased contractility and collagen production.

Furthermore, the degradation of collagen is finely regulated through a delicate balance between MMPs and TIMPs [23]. MMPs play a primary role in the breakdown of collagen [24], and their activity is controlled by TIMPs to prevent excessive degradation of the connective tissue. TIMPs counteract the effects of MMPs by forming complexes with them, impeding their interaction with substrates, and thus, slowing down the process of collagen degradation [25]. In fibrosis, this balance between MMPs and TIMPs is disrupted, resulting in reduced collagen breakdown and increased accumulation [8].

The regulation of collagen metabolism and fibrosis is a complex and dynamic process involving various factors. m6A modification may play a role in regulating collagen metabolism at multiple stages. Gaining insights into the molecular mechanisms of both collagen metabolism and m6A modification offers promising potential for developing targeted therapies for fibrosis and collagen-related diseases.

m6A regulatory proteins

“writers” of m6A methyltransferase

The m6A methyltransferase complex comprises METTL3, METTL14, and the co-factor WTAP [26]. METTL3, recognized as the catalytic core of the methyltransferase in 1997 [27], is the pioneering "writer" responsible for transferring the methyl group from S-adenosylmethionine (SAM) to the adenosine residues of RNA. METTL14 serves as an RNA-binding platform, facilitating RNA substrate binding and enhancing the complex's integrity [28, 29]. In human cells, METTL3 and METTL14 form a 1:1 stoichiometric complex [30], which localizes in the cytoplasm and then translocates to the nucleus through a nuclear localization signal within METTL3, where it associates with WTAP [31]. Although WTAP lacks methyltransferase activity, it interacts with the METTL3-14 complex and plays a regulatory role in recruiting the m6A methyltransferase complex to mRNA targets [32].

"erasers" of m6A demethylase

The "erasers" of m6A demethylase function akin to an eraser, removing m6A modifications from RNA. The first reported m6A demethylase in eukaryotic cells is the Fat mass and obesity-associated protein (FTO) [33]. The second identified m6A demethylase is ALKBH5, which has been shown to regulate mRNA output and RNA metabolism by reducing m6A levels in nuclear speckles [34].

"readers" of m6A modifications

"Readers" constitute a group of proteins that can recognize m6A modifications and regulate gene expression by influencing various biological processes, such as mRNA stability, splicing, structure, output, and translation efficiency [35]. Cytoplasmic m6A readers include YTHDF1/2/3, YTHDC2, and IGF2BP1/2/3. YTHDF1 enhances the translation of m6A methylated mRNA; while, YTHDF2 accelerates the degradation of m6A methylated mRNA. YTHDF3 collaborates with YTHDF1 and YTHDF2 to promote the metabolism of m6A methylated mRNA in the cytoplasm [36]. YTHDC2, located in nuclear speckles, preferentially binds to transcripts containing m6A modifications, leading to decreased mRNA abundance and increased translation efficiency through interactions with translation initiation and decay mechanisms [37]. Human insulin-like growth factor 2 mRNA-binding proteins (IGF2BPs) enhance mRNA stability by binding to target transcripts [38]. Nuclear m6A readers include YTHDC1, which interacts with splicing factors and nuclear export adapter protein SRSF3 to facilitate the transport of m6A-modified mRNA from the nucleus to the cytoplasm [39].

m6A modification in fibrotic diseases

Pulmonary fibrosis

Research related to pulmonary fibrosis is shown in Table 1. m6A levels increase in the lung tissues of patients with IPF and in mice with bleomycin (BLM)-induced fibrosis. This increase is attributed to elevated METTL3 expression. Silencing METTL3 reduces m6A levels and inhibits αSMA and COL I expression in TGF-β1-induced WI-38 cells. m6A modification, mediated by YTHDF1, regulates the fibroblast-to-myofibroblast transition (FMT) by modulating KCNH6 mRNA translation [41].

Table 1 Pulmonary fibrosis

Another study, through immunohistochemical analysis, observed a decrease in METTL3 expression in both pulmonary fibrosis patients and in a BLM-induced pulmonary fibrosis model in mice [45].

PM 2.5

PM2.5 exposure increases METTL3 expression, leading to heightened m6A modification of CDH1 mRNA. Moreover, enhanced recognition of CDH1 mRNA m6A modification by YTHDF2 inhibits its transcription and promotes its degradation, ultimately accelerating the progression of epithelial–mesenchymal transition (EMT) and pulmonary fibrosis after PM2.5 exposure [42].

Another study suggests that the upregulation of METTL3 plays a protective role in PM2.5 exposure. PM2.5 exposure-induced METTL3 expression promotes YTHDF1/IGF2BP1-mediated recognition of m6A sites on Nrf2 mRNA, leading to enhanced Nrf2 translation and activation of the Nrf2 antioxidant signaling pathway. Knockdown of METTL3 increases αSMA expression after PM2.5 exposure [44].

Simultaneously, PM2.5 exposure downregulates ALKBH5 expression, which promotes m6A modification of Atg13 mRNA in BEAS-2B cells. This results in the upregulation of the ULK complex mediated by Atg13, promoting epithelial cell autophagy and inflammation under PM2.5 treatment. Consequently, the NF-κB/NLRP3 signaling pathway is activated, driving pulmonary fibrosis [46].

Silicosis

m6A-seq and RNA-seq analyses on silica-induced silicosis mice showed increased m6A levels and METTL3 expression; while, ALKBH5, FTO, YTHDF1, and YTHDF3 expression decreased. Furthermore, 307 genes showed high methylation; while, 52 genes exhibited hypomethylation, mainly enriched in pathways related to "phagosome," "antigen processing and presentation," and "apoptosis" [40].

In silicosis patients, SiO2-treated fibroblasts, and mice, METTL3 expression was found to increase. SiO2 induced m6A modification of hsa_circ_0000672 and hsa_circ_0005654 in lung fibroblasts through METTL3, and this process involved cooperation with eIF4A3. Consequently, lung fibroblast proliferation, migration, and activation were induced, ultimately leading to pulmonary fibrosis [43]

Additionally, upregulation of ALKBH5 in mice exposed to silica and TGF-β1-activated lung fibroblasts inhibited fibroblast activation. Mechanistically, ALKBH5 demethylated pri-miR-320a-3p, blocking its maturation process and preventing its regulation of fibrosis through FOXM1 mRNA 3'-UTR targeting. Furthermore, ALKBH5 could directly regulate FOXM1 in an m6A-dependent manner, promoting silica-induced pulmonary fibrosis [47].

Carbon black

In another study, the fibrosis-promoting factor, carbon black (CB), reduced the m6A modification of pri-miRNA-126 and its binding with the RNA-binding protein DiGeorge syndrome critical region gene 8 (DGCR8). This led to a decrease in mature miRNA-126 and activation of the PI3K/AKT/mTOR pathway, driving an increase in levels of pulmonary fibrosis markers, including αSMA, fibronectin, COL I, and hydroxyproline. [48]

Cardiac fibrosis

“writers” in cardiac fibrosis

Research related to cardiac fibrosis is shown in Table 2. In numerous studies, METTL3 has been consistently shown to play a promoting role in cardiac fibrosis. Upregulation of METTL3 was observed in human atrial fibrillation cardiac tissue [50,51,52], heart tissues of myocardial infarction mouse models [49, 51, 52], and TGFβ1-induced cardiac fibroblasts [49,50,51,52].

Table 2 Cardiac fibrosis

From a mechanistic perspective, silencing METTL3 alleviated TGF-β1-induced cell proliferation, FMT, and collagen production in CFs, and reduced the m6A modification levels of fibrosis-related genes [49]. Additionally, METTL3 downregulates AR expression through an m6A-YTHDF2 dependent mechanism, promoting glycolysis and cardiac fibroblast proliferation, which ultimately leads to cardiac fibrosis [50]. Furthermore, silencing METTL3 has been observed to downregulate the expression of IGFBP3, inhibiting the activation of CFs and reducing the degree of cardiac fibrosis [51]. Moreover, METTL3 increases m6A methylation of GAS5, leading to YTHDF2 binding to GAS5 and inhibiting its expression, which further promotes CF proliferation, migration, and mitochondrial fission [52]. In addition to its fibrotic effects, METTL3 is also involved in cardiac fibrosis and myocardial cell apoptosis by increasing the m6A level of TNC mRNA [53].

Furthermore, in the heart tissues of the myocardial infarction mouse model and TGF-β1-induced CFs, the expression of MetBil (METTL3 binding lncRNA) is significantly increased. MetBil overexpression enhances collagen deposition and CFs proliferation [54].

“erasers” in cardiac fibrosis

FTO plays a protective role against myocardial fibrosis. In heart failure mammalian hearts and hypoxic cardiomyocytes, FTO expression is reduced, leading to increased RNA m6A levels and impaired myocardial contractile function. Increased FTO expression in heart failure mice selectively demethylates contractile transcripts in the heart, preventing their degradation, thus mitigating the ischemia-induced increase in m6A and the decline in cardiac contractile function. This, in turn, reduces fibrosis and enhances angiogenesis [55]. In the diabetic cardiomyopathy mouse model, there is an increase in m6A levels and a downregulation of FTO. FTO overexpression can improve cardiac function in diabetic cardiomyopathy mice by reducing myocardial fibrosis and cardiomyocyte hypertrophy [56]. circCELF1 upregulates the expression of FTO, reducing m6A modification on DKK2 mRNA, inhibiting the binding of miR-636 to DKK2, and promoting DKK2 expression, thereby inhibiting the progression of myocardial fibrosis [57].

However, in another study, FTO played a contrasting role: HFpEF + Exercise training (EXT) mice showed higher m6A levels and downregulated FTO levels. FTO overexpression promoted myocardial cell apoptosis, myocardial fibrosis, and cardiomyocyte hypertrophy, thereby counteracting the benefits of exercise in HFpEF + EXT mice [58].

“readers” in cardiac fibrosis

YTHDF2 deficiency results in declined cardiac function in elderly mice, exacerbating the cardiac dysfunction and increasing fibrosis induced by the pressure overload from TAC surgery [59].

lncRNA Airn binds to IMP2, protecting it from degradation. The retained IMP2 recognizes m6A modifications on p53 mRNA, leading to increased stability and protein expression. This reduces α-SMA and COL I expression in high glucose-induced CFs, thereby reducing cardiac fibrosis in diabetic mice. Silencing METTL3 decreases m6A modification on p53, resulting in reduced stability and downregulation of p53 mRNA in CFs [60].

Pulmonary arterial hypertension

YTHDF1 interacts with Foxm1 mRNA and upregulates Foxm1 protein levels by enhancing translation efficiency through an m6A-dependent mechanism. This promotes the proliferation of hypoxic pulmonary arterial smooth muscle cells (PASMC) and the expression of proliferation markers. Silencing YTHDF1 alleviates pulmonary vascular changes and fibrosis [61].

Hepatic fibrosis

Research related to hepatic fibrosis is shown in Table 3. In the study of hepatic fibrosis progression and reversal, dynamic analysis of m6A methylation profiles revealed that during hepatic fibrosis, m6A methylation differences are primarily enriched in processes related to oxidative stress and cytochrome metabolism, while in hepatic fibrosis reversal, they are mainly associated with immune response and apoptosis [62].

Table 3 Hepatic fibrosis

“writers” in hepatic fibrosis

Regarding the role of METTL3 in hepatic fibrosis, it is upregulated in lipopolysaccharide (LPS)-activated THP-1 macrophages and plays a role in promoting the expression of fibrotic proteins, such as COL I, α-SMA, and fibronectin, through the Sp1/TGF-β1/Smad signaling pathway [63]. Additionally, METTL3 is upregulated in the CCl4-induced mouse liver fibrosis model and IFN-γ/LPS-activated M1 macrophages, where it promotes macrophage pyroptosis and inflammation via the PTBP1/USP8/TAK1 axis by increasing MALAT1 levels through m6A modification, thereby exacerbating liver fibrosis [64]. Silencing METTL3 in HSCs leads to inhibited HSC activation and reduced liver fibrosis. Mechanistically, silencing METTL3 increases the stability and protein expression of Lats2 mRNA, which leads to increased YAP phosphorylation, inhibiting YAP nuclear translocation and ultimately resulting in decreased expression of pro-fibrotic genes [65].

In CdCl2-exposed mouse liver tissue, METTL3 expression decreases over time and correlates with the severity of liver injury. Liver-specific overexpression of METTL3 in mice attenuates CdCl2-induced hepatic steatosis and fibrosis; while, METTL3 overexpression improves CdCl2-induced cytotoxicity and activation of HSCs [66]. Long-term exposure to chronic corticosterone (CORT) induces hepatic inflammation and fibrosis in chickens and increases the levels of various heat shock proteins (HSPs) mRNA and m6A methylation [67].

In non-alcoholic steatohepatitis (NASH) rats and LPS-treated Kupffer cells (KCs), METTL3/METTL14 is upregulated; while, FTO is downregulated. After LPS stimulation, NF-κB p65 directly activates METTL3 and METTL14, promoting cap-independent translation of TGF-β1 through m6A modification in the 5′UTR region. This upregulates TGF-β1 and exacerbates TGF-β1-mediated stellate cell activation, promoting the transition from NASH to liver fibrosis [68]. Acid-sensitive ion channel 1a (ASIC1a) regulates the processing of miR-350 through METTL3-dependent m6A modification. Mature miR-350 targets SPRY2 and further promotes liver fibrosis through the PI3K/KT and ERK pathways [69].

In patients with biliary atresia, there is an increase in m6A levels, and the expression of METTL3, METTL14, and WTAP is upregulated; while, ALKBH5 is downregulated. The overexpression of METTL3 and METTL14 promotes the expression of COL1A1, MMP2, and THY1. THY1 may play a role in cholestatic fibrosis by interacting with the ITGAX/ITGB2 complex in bone marrow cells [70].

METTL16 is upregulated in the liver tissues of chronic hepatitis B (CHB) with severe fibrosis. Silencing METTL16 in HSCs downregulates the m6A modification level of HLA-DPB1 mRNA, and it is involved in the progression of fibrosis in CHB [71].

In Sorafenib, erastin, and RSL3-induced ferroptosis of HSCs, METTL4 expression is upregulated, and FTO is downregulated. YTHDF1 recognizes m6A binding sites and stabilizes BECN1 mRNA, triggering autophagy activation. Inhibition of m6A modification impairs erastin-induced ferroptosis in CCl4-induced liver fibrosis in mice and reverses the beneficial effect of erastin on liver fibrosis improvement [72].

The differentially expressed m6A genes in liver fibrosis mice are closely associated with processes such as the endoplasmic reticulum stress response, PPAR signaling pathway, and TGF-β signaling pathway. In liver fibrosis mice, the expression of WTAP, ALKBH5, and YTHDF1 is reduced. Decreased expression of WTAP leads to an increase in αSMA and COL I expression, promoting HSC activation and inducing the occurrence of liver fibrosis [73]. However, in another study, WTAP is highly expressed in liver fibrosis and it targets the 3'-UTR of Ptch1 mRNA to increase its stability. N-acetyl-seryl-aspartyl-lysyl-proline (AcSDKP) reduces the expression of WTAP and decreases the stability of Ptch1 mRNA, thereby exerting an anti-fibrotic effect [74].

“erasers” in hepatic fibrosis

FTO downregulation and consequent upregulation of m6A modification are essential for DHA-induced autophagy activation and HSC ferroptosis. YTHDF1 upregulation and FTO downregulation are involved in DHA-induced HSC ferroptosis by increasing the stability of BECN1 mRNA. Knocking down YTHDF1 can prevent this process, ultimately reducing the therapeutic effect of DHA on liver fibrosis [75].

In human fibrotic liver tissues and CCl4-induced mouse liver fibrosis, elevated m6A levels and decreased ALKBH5 expression are observed. ALKBH5 functions in a YTHDF1-dependent manner to inhibit mitochondrial fission, HSC proliferation, and migration by reducing Drp1 m6A modification. This regulatory process leads to a reduction in αSMA and COL I expression, contributing to improved liver fibrosis [76]. ALKBH5 is downregulated in both human and mouse liver fibrotic tissues. Its overexpression leads to reduced αSMA and COL I expression, decreased collagen protein accumulation, and interstitial fibrosis. ALKBH5's beneficial effects on liver fibrosis are achieved through m6A-dependent PTCH1 activation, inhibiting HSC activation [77].

ALKBH5 is upregulated in radiation-induced HSCs. It mediates m6A demethylation of toll-interleukin 1 receptor domain-containing adaptor protein (TIRAP) mRNA and regulates TIRAP expression in a YTHDF2-dependent manner, promoting HSC activation through the TIRAP/NF-κB pathway. ALKBH5 also regulates CCL5 secretion, facilitating monocyte recruitment and M2 polarization, further enhancing ALKBH5 expression and TIRAP/NF-κB pathway activation. Irradiated HSCs educate monocytes, leading to HSC activation and reduced HCC radiosensitivity through CCL20 secretion. Blocking the ALKBH5-CCR6 axis can alleviate radiation-induced liver fibrosis (RILF) and improve HCC radio sensitivity [78].

“readers” in hepatic fibrosis

A study suggests that DNA methylation (5mC) is essential for the initiation stage of HSC activation (myofibroblast transdifferentiation); while, m6A is crucial for the perpetuation stage of HSC activation (excessive ECM production). YTHDF1 enhances COL I A1 protein production by stabilizing its mRNA. Silencing YTHDF1 can alleviate CCl4-induced mouse liver fibrosis by inhibiting collagen synthesis [79].

YTHDF3 induces PRDX3 translation in an m6A-dependent manner, leading to the upregulation of PRDX3 expression. Through the mitochondrial reactive oxygen species (ROS)/TGF-β1/Smad2/3 pathway, it inhibits HSC activation, exerting a protective effect against liver fibrosis [80].

CCl4-induced liver fibrosis and primary HSCs exhibit elevated levels of methylation, increased expression of ZC3H13, and decreased expression of FTO. Lowering m6A levels can reduce the protein levels of αSMA and COL I, thus improving liver fibrosis. YTHDC1 is upregulated in CCl4-induced liver fibrosis and primary HSCs, promoting the degradation of nuclear receptor subfamily 1 group d member 1 (NR1D1) mRNA. The absence of NR1D1 inhibits phosphorylation of DRP1S616, leading to weakened mitochondrial fission function, increased mtDNA release, activation of the cGAS pathway, and promotion of liver fibrosis progression. DHA alleviates liver fibrosis by promoting the proteasomal degradation of YTHDC1 in activated HSCs, restoring NR1D1 expression [81].

Renal fibrosis

“writers” in renal fibrosis

Research related to renal fibrosis is shown in Table 4. In renal fibrosis, the UUO mouse model shows decreased m6A levels and reduced METTL3/METTL14 expression; while, FTO is upregulated. Differentially methylated genes are mainly associated with the TGFβ signaling pathway (downregulated genes) and the axon signaling pathway (upregulated genes) [82]. However, another study using the UUO model found that METTL3 upregulation increased pri-miR-21 m6A modification, promoting miRNA-21-5p maturation. This triggered the SPRY1/ERK/NF-κB pathway, driving inflammation and the development of obstructive renal fibrosis. And HNRNPA2B1 may be involved in recognizing m6A modifications in pri-miR-21 and facilitating the maturation of miR-21-5p [83].

Table 4 Renal fibrosis

Long noncoding RNA AI662270 promotes the transcriptional stage of CTGF expression by recruiting METTL3 to the CTGF promoter and depositing m6A modifications on nascent mRNA. This activation of CTGF drives the activation of interstitial fibroblasts and promotes renal fibrosis [84]. TGF-β1 treatment upregulates METTL3, METTL14, and WTAP in HK2 cells. Inhibiting METTL3 reduces MALAT1 expression and contributes to DHA's anti-fibrotic effect against TGF-β1-induced renal fibrosis through the MALAT1/miR-145/FAK axis [85]. High glucose treatment in mouse mesangial cells (SV40-MES-13) results in decreased m6A levels and downregulation of METTL3 expression. Overexpression of METTL3 enhances the stability of Nuclear receptor-binding SET domain protein 2 (NSD2) mRNA through YTHDF1, promoting its expression. Consequently, this alleviates kidney impairment and renal fibrosis in diabetic nephropathy [86].

METTL14 reduces the stability of TUG1 mRNA by increasing its m6A modification, thereby inhibiting TUG1 expression. TUG1, in turn, interacts with LIN28B, leading to the inactivation of the MAPK1/ERK signaling pathway. Knockdown of METTL14 or overexpression of TUG1 protects diabetic kidney disease (DKD) mice from renal damage and renal fibrosis induced by streptozotocin (STZ) [87].

“erasers” in renal fibrosis

Canagliflozin increases m6A levels in HK2 cells while reducing FTO expression. FTO overexpression weakens the effect of canagliflozin on autophagy induction, leading to decreased stability of SQSTM1 mRNA. Deletion of SQSTM1 abolishes the protective effect of canagliflozin against renal fibrosis. Therefore, canagliflozin combats renal lipotoxicity and interstitial fibrosis through the m6A-modified SQSTM1/autophagy/STAT6 axis [88]. In UUO kidneys, the expression of FTO, METTL3, and METTL14 increases, while ALKBH5 expression decreases. In kidneys subjected to unilateral ischemia–reperfusion (UIR) and TGF-β1-treated HK-2 cells, FTO expression is elevated. In vivo and in vitro blocking of FTO can reduce the upregulation of Kcnk5, encoding TWIK-related acid-sensitive K+ channel-2 (TASK-2), cell cycle arrest, and renal fibrosis. TASK-2 is upregulated through FTO-mediated Kcnk5 demethylation and is activated by intracellular alkalization, leading to reduced intracellular K+ concentration, G2/M cell cycle arrest, and exacerbation of renal fibrosis [89]. FTO expression increases in TGF-β1-treated HK-2 and HKC-8 cells, as well as in UUO mouse kidney tissues. FTO suppresses the expression of lncRNA GAS5 by reducing its m6A modification. Knockdown of FTO inhibits TGF-β1 and UUO-induced EMT and inflammatory response, resulting in reduced expression of αSMA and COL I [90].

In the UUO model, the total m6A level increases; while, ALKBH5 expression decreases. Knocking down ALKBH5 suppresses E-cadherin expression and promotes αSMA and Snail levels. Genistein improves renal fibrosis by restoring ALKBH5 expression and regulating EMT [91]. Another study found that inhibiting ALKBH5 increases the m6A modification of CCL28 mRNA, leading to enhanced stability of CCL28 through recognition by IGF2BP2. This upregulates CCL28 levels, recruiting Tregs (regulatory T cells), which protect the kidneys from inflammation and immune cell infiltration. As a result, inhibiting ALKBH5 has a protective effect against ischemia–reperfusion-induced acute kidney injury (AKI) and fibrosis [92].

“readers” in renal fibrosis

YTHDF1 is highly expressed in human fibrotic kidneys and upregulated in fibrotic mouse kidneys induced by UUO, high-dose folic acid administration, or the unilateral ischemia–reperfusion injury (IRI). Knocking down YTHDF1 in cultured cells induced by TGF-β treatment and UUO mouse models alleviates the progression of renal fibrosis. This effect is likely mediated by YTHDF1's regulation of Yes-associated protein (YAP) [93].

Retinal

During the process of laser-induced choroidal neovascularization and subretinal fibrosis in mice, METTL3 is upregulated in retinal pigment epithelial (RPE) cells. METTL3 enhances the stability of HMGA2 mRNA through m6A modification, leading to an increase in HMGA2 protein expression. This activation of HMGA2 induces the transcription factor SNAIL, promoting EMT. However, silencing METTL3 effectively reduces subretinal fibrosis in the retina [94].

In patients with proliferative vitreoretinopathy (PVR), the expression of METTL3 is reduced in retinal pigment epithelial cells. The expression of METTL3 is downregulated in ARPE-19 cells after EMT. Overexpression of METTL3 inhibits cell proliferation and weakens the ability of TGFβ1 to induce EMT by modulating the Wnt/β-catenin pathway. Intravitreal injection of cells overexpressing METTL3 delays the occurrence of PVR [95].

High glucose upregulates the m6A modification level of PARP1 mRNA in human retinal microvascular endothelial cells (hRMECs) and downregulates YTHDF2. Overexpression of YTHDF2 reduces the expression of Poly (ADP-ribose) polymerase 1 (PARP1) in hRMECs in an m6A-dependent manner, enhances hRMEC viability, and prevents glucose-induced inflammation, fibrosis, and angiogenesis [96].

Oral

In Oral submucous fibrosis (OSF) tissues, there is an increase in m6A modification levels. Arecoline promotes the expression of METTL3 and METTL14 through TGFβ signaling. Silencing METTL14 reverses the effects of arecoline on Hacat cell proliferation and apoptosis by inhibiting MYC m6A modification and reducing TIMP1 expression [97].

m6A modification in non-fibrotic collagen-related diseases

Osteoarthrosis

m6A modification promotes intervertebral disc degeneration [98] and osteoarthritis [99, 100], and enhances chondrocyte differentiation [101] and osteoblast differentiation [102].

In degenerative human endplate cartilage tissue, m6A levels are increased. Mechanical tension stimulation increases METTL3-mediated m6A levels in human endplate chondrocytes. METTL3 mediates m6A modification of SOX9 mRNA and disrupts the stability of SOX9 mRNA, leading to the inhibition of downstream COL II α1 expression. Suppression of METTL3 expression in endplate cartilage can alleviate mechanical imbalance-induced intervertebral disc degeneration [98].

In ATDC5 chondroprogenitor cells treated with IL-1β, m6A levels and METTL3 expression increase. Silencing METTL3 reduces IL-1β-induced cell apoptosis, levels of inflammatory cytokines, and NF-κB signaling in chondrocytes. METTL3 silencing promotes extracellular matrix degradation by reducing MMP13 and COL X expression, and increasing aggrecan and COL II expression [99]. Similarly, IL-1β stimulation in C28/I2 chondrocyte cell line results in increased m6A levels and METTL3 expression, along with decreased ALKBH5 expression. Overexpression of ALKBH5 downregulates IL-1β-induced MMP13 and COL X expression, while upregulating COL II and aggrecan expression [100].

METTL3, METTL14, and m6A modification levels are increased in synovium-derived mesenchymal stem cells (SMSCs) during chondrogenic differentiation. Knockdown of METTL3 inhibits chondrogenic differentiation, downregulates SOX9, ACAN, and COL II α1, and increases MMP3, MMP13, and GATA3 expression [101]. METTL3-mediated m6A methylation of LncRNA MIR99AHG increases the expression of Osterix, COL I α1, bone sialoprotein, and RUNX2 by targeting miR-4660, enhancing the osteogenic differentiation of bone marrow mesenchymal stem cells (BMSCs) [102].

Skin

In the mouse model of bleomycin-induced scleroderma, the differentially m6A-hypermethylated mRNAs were most significantly associated with growth hormone synthesis, secretion, and action, insulin secretion, and amphetamine addiction. On the other hand, the differentially m6A-hypomethylated mRNAs were most significantly associated with rheumatoid arthritis, Toll-like receptor signaling pathway, and amoebiasis [103]. In keloid tissue, m6A modification was decreased, and the expression of m6A demethylase FTO was increased. FTO overexpression in skin fibroblasts stimulated fibroblast migration and increased the expression of COL I α1 and α-SMA. FTO upregulates COL I α1 expression by regulating its m6A modification and stabilizing mRNA, thus promoting keloid formation. [104]. m6A sequencing and RNA sequencing revealed that differentially methylated m6A-related genes were associated with fibrosis-related pathways in hyperplastic scars compared to normal skin. Highly methylated genes were mainly related to the P13K-Akt signaling pathway, focal adhesion, and ECM-receptor interaction. On the other hand, lowly methylated genes were mainly associated with the MAPK signaling pathway and the NF-κB signaling pathway [105].

Cancer

The role of m6A varies in different types of tumors. In U87 and U251 cells, METTL3 reduces the methylation level of COL IV α1, upregulates its expression, and stimulates the malignant development of glioblastoma [106]. In lung cancer, cancer-associated fibroblasts (CAFs) derived from lung squamous cell carcinoma (LUSC) upregulate the m6A modification of COL X α1 by increasing METTL3 expression, stabilizing COL X α1 expression, promoting LUSC cell proliferation, and inhibiting apoptosis-induced oxidative stress [107]. Silencing METTL3 can upregulate the expression of COL III α1 chain by increasing m6A levels, ultimately promoting the metastasis of triple-negative breast cancer tumor cells [108]. lncRNA NIFK-AS1 is highly expressed in HCC tissues and cells, and this upregulation is dependent on METTL3-mediated m6A methylation. NIFK-AS1 affects HCC progression through the NIFK-AS1/miR-637/AKT1 axis, regulating MMP7 and MMP9 expression. Knockdown of NIFK-AS1 inhibits HCC cell proliferation, colony formation, migration, and invasion [109]. In prostate cancer tissues, however, METTL3 is highly expressed and can regulate the expression of integrin β1 (ITGB1) through m6A modification, thereby affecting the binding of ITGB1 to COL I and promoting prostate cancer bone metastasis。 [110]

Cerebrovascular

Downregulation of m6A reader protein proline-rich coiled-coil 2B (PRRC2B) mediates selective splicing of COL XIIα1 chain in an m6A-dependent manner and regulates the decay of MMP14 and ADAM metallopeptidase domain 19 (ADAM19) mRNA in an m6A-independent manner, promoting hypoxia-induced endothelial cell migration. Conditional knockout of PRRC2B in endothelial cells enhances hypoxia-induced vascular remodeling and cerebral blood flow redistribution, thereby alleviating hypoxia-induced cognitive decline [111].

Conclusion and perspectives

In recent years, RNA epigenetics, particularly m6A modification, has emerged as a prominent research area. Among more than 100 different RNA modifications, m6A stands out as the most abundant in eukaryotic cells. This dynamic and reversible modification is meticulously controlled by "writers" and "erasers," while "readers" play a crucial role in its recognition and functionality. The significance of m6A modification in regulating collagen metabolism across various diseases cannot be overstated. This comprehensive review aims to provide an overview of the functions and mechanisms of m6A modification in organ fibrotic diseases and non-fibrotic collagen-related conditions. While the majority of research has focused on the core methyltransferase METTL3, there have been some investigations into other methyltransferases and demethylases, albeit with fewer studies dedicated to m6A readers.

In summary, the field of m6A regulation in collagen metabolism holds tremendous potential for further exploration. Recent advances have been made in elucidating the role of m6A in collagen regulation; however, many aspects of m6A modulators in collagen-related diseases remain unexplored, necessitating further inquiry.

Future research should prioritize the following areas: 1. Investigating the roles of other m6A methyltransferases and demethylases in collagen metabolism and their impact on collagen-related diseases. 2. Exploring the functions of various m6A readers and their contributions to collagen regulation. 3. Unraveling the intricate molecular mechanisms by which m6A modification regulates collagen synthesis, deposition, and degradation. 4. Developing potential therapeutic interventions targeting m6A modification for treating collagen-related diseases.

In conclusion, ongoing investigations into m6A modification in collagen metabolism offer promising directions for future research. Sustained efforts in this area will undoubtedly deepen our understanding of the regulatory mechanisms of m6A in collagen-related diseases and open up new possibilities for therapeutic applications.