The m⁶A modification is methylation on N6 of adenine A, which occurs on RNA. Regarded as most prevalent post-transcriptional modification on eukaryotic mRNA and lncRNA, m⁶A mediates more than 80% of RNA methylation (6, 11). According to recent studies, some m6A marks might be added co-transcriptionally, which could enable RNA modifications to take place at the same time as transcription and affect RNA processing. The early regulatory checkpoint that this co-transcriptional addition may offer influences on RNA maturation and function before transcription is completed (12). Studies have shown that m⁶A modification performs vital functions in RNA translation, splicing, transport, stabilisation and localisation (11).

The modification of m⁶A is a reversible process, which affects the post-transcriptional fate of modified RNA. The deposition, elimination, and recognition of m6A modifications on RNA molecules are controlled by RNA-modifying enzymes, which include writers, erasers, and readers ( Table 1 ).

Methyltransferases that catalyse the m6A modification of adenosine on mRNA include METTL3/14, WTAP, and KIAA1429 [17]. METTL3/14 and WTAP form a methyltransferase complex (MTC). Among them, METTL3 and METTL14 constitute the main components of MTC and perform the function of m6A deposition on nuclear RNAs. The WTAP mechanism guarantees the specific distribution of the METTL3-METTL14 complex to nuclear speckles and preserves its catalytic activity towards mRNA targets. Except for the canonical writers, MTC accessory subunits have a significant impact on the catalytic process include RBM15, ZC3H13, VIRMA, and HAKAI (12).

Demethylases, including two enzymes such as FTO and ALKBH5 (both belonging to the ALKB family), removed the methylated base that has undergone m⁶A modification. In detail, Previous studies such as Chuan He’s research group discovered FTO as the first m⁶A demethylase in 2011, and it primarily facilitates mRNA and snRNA demethylation at m⁶Am residues (13). But recent studies have explained that FTO is the first RNA demethylase that mediates the demethylation of m⁶A (!4). In 2013, ALKBH5 was discovered as another demethylase, which combines with specific m⁶A-modified single-stranded RNA to carry out the demethylation of m⁶A (14). In addition, Xie et al. have introduced flavin mononucleotide (FMN) as a new synthetic molecule derived from vitamin B2, plays a crucial role in enzymatic reactions (15).

The primary function of reader proteins is specifically recognised RNA that binds to the m⁶A modification, activating downstream regulatory pathways. YTH domain protein is well known as m⁶A recognition protein. YTHDF1 and YTHDF3 are mainly used to enhance the efficiency of mRNA translation, while YTHDF2 has been found to mediate RNA degradation (16, 17). Except for these direct readers, three heterogeneous nuclear ribonucleoproteins (hnRNPs) indirectly associate with m⁶A sites according to the “m⁶A switch” mechanism. Dysfunction of enzymes related to m⁶A modification may result in several autoimmunological diseases, such as rheumatoid arthritis and other musculoskeletal disorders (18). These m⁶A sites preferentially occur in the relatively conservative RRACH sequence (R represents A/G, H represents A/C/U), displays at the 3’-UTR, termination codon, and long exons of mRNAs (6).

Experiments have shown that m⁶A controls the dynamic equilibrium of the bone matrix and establishes an intricate network in bone metabolism (19). The m6A alteration at the cellular level modulates the expression of ALP, Runx2, Osterix, VEGF, and other associated genes, influencing physiological processes. m6A is involved in the delicate procedures of adipogenesis and osteogenesis from MSCs, which is important for bone homeostasis (19, 20).

m¹A methylation is another key post-transcriptional modification in RNA modification, identified by Dunn in the 1960s, which occurs on N1 adenosine (21). It is well-established that including this methyl group disrupts the specificity of base-pairing, revealing the regulatory roles of this alteration in RNA molecules. Multiple investigations have evidenced the extensive presence of m¹A in mRNA, tRNA, rRNA, and mitochondrial transcripts (22, 23).

According to Dominissini et al., m¹A is abundant in the 5’-UTR of mRNA. It tends to present at the initiating codon upstream of the first splicing site, which may influence the process of mRNA translation (24). Besides, m¹A methylation may display the promising capability to negatively regulate reverse transcription and make changes according to relevant stimulation in this microenvironment (25). m¹A is inextricably linked with m⁶A modification because m1A can not only rearrange to m⁶A (Dimroth rearrangement) under alkaline conditions but also share some regulators with this modification (26). Like m⁶A methylation, these enzymes of m¹A, including “writer” (TRMT10C/61B, TRMT6/61A), “eraser” (ALKBH1, ALKBH3), and “reader” (YTHDF1-3, and YTHDC1), exert their effects in the post-transcriptional stage of mRNA and ncRNAs. YTHDF1-3 and YTHDC1 are regulatory factors that have participated in the detailed procedures for m¹A and m⁶A (27). Respectively, the deeply advanced research about the regulation of m⁶A may be supported to investigate the mechanism of m¹A further (28).

Moreover, recent studies have illustrated that m¹A affects RNA base pairing, thereby influencing the structure and function of target RNA molecules. Human rRNAs and tRNAs contain numerous diverse m¹A modification sites, such as position 1322 of 28S rRNA, position 947, and position 58 of tRNA (26). In mRNA, m¹A occurs in each segment, including CDS, 5’-UTR, 3’-UTR performs effective capabilities according to the different regions or subcellular locations (28, 29).

m⁷G, the RNA methylation of guanine at the N7 position, accounts for about 0.4% of all guanosines. As a well-known 5’ cap (m7GPPPN) modification of mRNA, m⁷G regulates mRNA splicing and nuclear export and contributes to mRNA translation (30). m⁷G modification is like m¹A in modification level and widely exists in tRNA (at position 46), 18s rRNA, mature and even pre-miRNAs (31), which is accumulated in transcript fragments of mRNA 5’-UTR, CDS, and 3’-UTR, pre-mRNAs (32).

In m⁷G, METTL1, RNA (guanine-7-) methyltransferase (RNMT), WD repeat domain 4 (WDR4) complex, and Williams-Beuren syndrome chromosomal region 22 protein (WBSCR22) are recognised as components involved in its modification (33). Among them, RNMT is necessary for effective cap methylation and is activated by the mini protein (RAM). For most non-coding RNAs, the demethylation process of the m⁷G cap can be completed by cleavage or further modification to m^2,2,7G trimethyl guanosine. The recognition of m7G cap is mostly performed through eIF4E and cap-binding complex (CBC), therefore, regulating RNA maturation, nuclear export, and translation (34, 35). Interestingly, the working mechanism of METTL1 and WDR4-modified tRNAs has been further explored by two research teams and reached consistent and complementary conclusions, revealing that in addition to the N-terminal of METTL1, the C-terminal of WDR4 is also very important for normal activity of METTL1, and found that the N-terminal of METTL1 participates in the construction of active pockets, and inhibits its methyltransferase activity through the steric hindrance effect triggered by phosphorylation (36).

m⁷G is present in mRNA and is crucial in orchestrating translation procedures. Moreover, it exerts considerable influence on stem cell proliferation and differentiation (37).

An alteration known as RNA m⁵C was initially discovered in 1958, involving the methylation of the fifth carbon atom of RNA cytosine. m⁵C methylation occurs on every RNA molecule, such as tRNA, rRNA, mRNA, and ncRNA, which are key molecules to maintain the regular physiological functions of eukaryotic cells (29, 38).

It enhances RNA stability and regulates protein synthesis and translation. According to Amort et al., m⁵C is primarily located at mRNA’s 5’ and 3’-UTRs, particularly close to the translation start codon. It increases the stability of RNA and controls the processes of protein synthesis and translation (39). m5C can stabilise tRNAs by enhancing the thermal stability of hydrogen bonds with guanine, providing protection against inappropriate degradation. Meanwhile, the m⁵C residue of rRNA mostly regulates the quality of ribosome biogenesis (40).

RNA m⁵C was also monitored by “writers”, “erasers”, and “readers” as presented in Table 1 . m⁵C methyltransferases (“writers”), including NSUN1-7 and DNA methyltransferase-like 2 (DNMT2) (41, 42). Specific m⁵C writers catalyse different RNA subsets. TET proteins catalyse the removal process (TET1-3) and ALKBH1, mainly oxidising m⁵C to 5-hydroxymethylcytidine (hm5C) (38). In all cases, the formation of hm5C will reduce the modification of m⁵C, so TETs and ALKBH1 are identified as erasers (40). The Aly/REF nuclear export factor (ALYREF) is the first identified reader, which can bind to m⁵C sites to perform biological functions (43). Dominissini et al. confirmed that ALYREF facilitates the nuclear export of mRNAs. Subsequently, ELAVL1 and FMRP were also identified, which could directly bind to m⁵C to modulate the distribution of m⁵C in mRNA and affect the maturation of rRNA by regulating the m⁵C levels (44). The level of m⁵C is closely related to immune diseases.

ac4C is the RNA acetylation of cytosine at N4 position, which occurs in cytidine conservatively. It is the only acetylation modification that has been studied in eukaryotic RNA. Like many RNA modifications, ac4C was first detected in tRNA, rRNA, and mRNA (45). So far, only n-acetyltransferase 10 (NAT10) is considered the writer of ac4C, an indispensable ATP-dependent RNA acetyltransferase (46, 47). It can catalyse ac4C modification in multiple RNAs, for example, 18S rRNA, tRNA and mRNA. During the formation of ac4C, two additional proteins, including C/D snoRNA U13 and THUMPD1, are required to act together in the catalytic process (48).

As a newly discovered RNA modification, ac4C is still unknown, especially its regulator and molecular function. Only one writer, eraser, and reader has yet to be found. The presence of ac4C in the mRNA CDS region can significantly improve mRNA stability and promote protein translation, especially in the translation process, which may influence the interaction with homologous tRNAs (42). In 18sRNAs, ac4C plays a crucial role in pre-rRNA processing and ribosome production, directly impacting translation capability. On the other hand, ac4C can mainly promote the stability of tRNA and serve as a monitoring indicator for the maturation of eukaryotic tRNA (47–49).

Pyrimidine nucleoside is formed by linking Ψ, a trace nucleic acid base, located at the fifth position of uracil and ribose. Ψ is connected to ribose base with covalent bonding. One of the post-transcriptional modifications, Ψ, is widespread in RNAs, particularly in long non-coding RNA (lncRNA), where it enhances the activity of tRNA and rRNA by fortifying the structure of the RNA chains (50). The studies have reported about pseudouridine synthase 1 (DKC1) count and RNA-independent single pseudouridine synthetases (PUSs) comprising PUS1/3/7/10, PUSL1/7, TRUB1/2, and RPUSD1-4 that led to distinct cellular localisation and RNA targets (51).

Researchers have demonstrated that Ψ enormously influences RNA’s biogenesis, structure, stability, and function and regulates gene expression. There are many pseudouridine sites in tRNA, which are crucial for the stability of the structure and mediate the base pairing of tRNA codon-anticodon, thus affecting the translation process (50). Moreover, Ψ regulates the pre-mRNA stage through various functional modes, such as mRNA/rRNA/snRNA structure, stability, translation, and termination (49). The chemical structure of all six modifications are showed in Figure 1 .

FLS are highly specialised mesenchymal cells in the synovium of joints, mainly in the intimal lining layer. FLS is the core target cell of RA, which could switch from early immunosuppressive to stimulatory in response to relevant factors (55). During normal physiological conditions, the intimal lining forms a thin and permeable barrier at the boundary between the sublining and the synovial fluid canal. Synovial fluid synthesis (FLS) regulates the composition of the extracellular matrix (ECM) and synovial fluid, providing lubrication and nourishment to the cartilage surface (56, 57). Therefore, as significant effector cells, FLS have special aggressive behaviours and function actively in the pathogenesis and progression of RA. Figure 3 illustrates the main RNA modification in FLS.

Jiang and colleagues established the m6A transcriptional map in human rheumatoid arthritis fibroblast-like synovial cell line MH7A cells using m⁶A seq and RNA seq. The potential relationship between m⁶A and FLS in RA is described for the first time (58). Compared with the healthy controls (HCs), single-cell analysis and machine learning efficiently predicted that IGFBP2 was elevated and METTL3 was downregulated in RA-FLS. By searching the GEO database, four different data sets showed the increased expression of the IGF2BP3 gene in RA synovial cells. Moreover, molecular docking and in vitro experiments suggested that IGF2BP3 is also a target gene of triptolide (TP), which has been proven to have therapeutic potential for RA.

METTL3, a crucial component of the m6A methyltransferase complex, has been confirmed to promote activation and inflammation of FLS through the NF-κB signalling pathway in RA (59). Furthermore, Shi et al. demonstrated that METTL3 reduced autophagy activity by weakening the stability of autophagy-related 7 (ATG7) mRNA in a manner dependent on m6A-YTHDF2. This, in turn, facilitated the ageing of FLS and the progression of arthritis (60). In addition, Lin et al. identified an m⁶A methylation-mediated gene TGM2 and portrayed the regulatory mechanism in controlling the immune microenvironment of RA-FLS (10). This cross-validated experiment verified the role of m⁶A RNA methylation in synovium and identified a novel target for RA treatment.

Although little is known about other RNA modifications that regulate FLS cells in the synovium microenvironment, previous literature has confirmed that RNA modification is closely related to RA-FLS biology. Figure 3 illustrates the main RNA modification in FLS.

The synovial lining and sublining layers contain numerous HLA-DR+ macrophages in the synovium, which are activated through antigen presentation (61). Generally, synovium macrophages are composed of two types, tissue-resident macrophages and circulating macrophages. New evidence suggests that the cellular function of macrophages varies from the different subsets of macrophages in synovial homeostasis (62). These macrophages are potent cytokine producers with apparent activated status and professional antigen-presenting cells (APCs) that stimulate an adaptive immune response. After that, T lymphocytes were activated and largely promoted joint destruction. T lymphocytes at synovium and pan membranous cartilage junction increased significantly. Abundant-activated macrophages are an early hallmark of arthropathies, and many macrophages have become a significant feature of inflammatory lesions (63). The degree of synovial macrophage infiltration is correlated with the degree of joint erosion and depletion of these macrophages from inflamed tissue (64). This phenomenon also has a profound therapeutic benefit. Importantly, there is bidirectional intercellular communication between macrophages and FLS in the arthritis synovium, which is beneficial in governing the maintenance or disruption of joint homeostasis ( Figure 4 ).

Macrophages M1/M2 polarisation is mainly mediated immune/inflammatory response and maintained the dynamic balance in synovium (65). Classical theory presumed that M1 (pro-inflammatory) macrophages are activated by interferon-γ, lipopolysaccharide (LPS), or tumour necrosis factor-alpha (TNF-α). Meanwhile, it secretes large amounts of proinflammatory cytokines and mediators, such as TNF-α, interleukin-1/6/12 (IL-1/6/12), and cyclooxygenase-2 (COX-2). M2 (anti-inflammatory) macrophages termed as wound-healing macrophages, are polarised by the Th2 cytokines, IL-4 and IL-13. These are further classified into specific subtypes. M2 macrophages shows an anti-inflammatory phenotype, which is helpful for tissue repair and remodelling (66). However, M1/M2 can’t be regarded as the extreme of the activation state spectrum of macrophages, which is just a point reflecting the complex environment of the synovium. While the imbalance of M1/M2 macrophages, correlated significantly positively with synovial homeostasis (67). A complex cytokine and chemokine network is involved in the inflammatory environment.

Recently, accumulated evidence has revealed the potential links between m⁶A modification and macrophage polarisation. METTL3, the writer of m⁶A, facilitates M1 macrophage polarisation through the methylation of STAT1 mRNA directly and attenuates M2 macrophage polarisation (68). While the reader of m⁶A (69), FTO strengthened the polarisation of M1 and M2 macrophages, FTO activated the NF- κB signal path, and increased the mRNA stability of STAT1 and PPAR-γ through YTHDF2 involvement, thereby accelerating macrophage activation (70). Combined with methylated RNA immunoprecipitation (MeRIP seq) and RNA sequencing results, Wan et al. evaluated the whole transcriptome m6A modification in the synovium of RA patients. They showed that M1 macrophages in RA patients were significantly increased (71).

Wu et al. found that YTHDF3 (m¹A reader) may participate in the polarisation of M1 macrophages by regulating a series of target genes (72). Utilising the Lasso regression analysis, researchers constructed a novel m⁷G risk signature or m⁵C prognostic model in both hepatocellular carcinoma and prostate cancer (73–75). They found a significant correlation between the expression level of m⁵C regulator genes (NSUN2/6, TET1/3) and immune cells, including M1/M2 macrophage infiltration characteristics (76, 77). Similarly, consistent with the previous results, Aly/REF and NUSN2/3 (m⁵C modulators) interacted with lncRNA to trigger the occurrence of disease, which shed new light on the role of m⁵C modification within the process of immune infiltration of macrophages (74). NAT10, as a critical gene regulating the formation of N4-acetylcytidine in RNA, has a significant positive correlation between immune infiltration (75), including B cells, CD8⁺T cells, CD4⁺T cells, neutrophils, macrophages, and other immune cells (69). Seven writers of the pseudouridine (Ψ) modification were significantly enriched in the immune-related signal pathway of macrophages (78). Unfortunately, these studies were carried out in other diseases, and there is no research report on RNA modification except m⁶A in synovium.

Specific T cells originate from pluripotent stem cells in the bone marrow, mature in the thymus, and then trafficked to the inflammatory sites, developing into the pathophysiology of synovitis (79, 80). T cells are transferred from the bloodstream to the synovial tissue through interactions with the endothelial cells in the posterior venules of synovial capillaries (73). Then, many immune cells infiltrate, including macrophages, granulocytes and B cells, especially CD4⁺ T and CD8⁺ T cells, increasing the expression of adhesion molecules and chemokines in the endothelium, thus activating a systemic immune response and forming a local synovium. It has been proved that CD4⁺ T cells, rather than CD8⁺ T cells, initiated the synovial inflammation process, the hyperactivation of the immune response, and the presence of autoantibodies in the synovial microenvironment is sufficient to trigger the development of rheumatic diseases (74, 75). Figure 5 presents the machinery of RNA modifications and their molecular function in T cells ( Figure 6 ).

Among them, CD4⁺ T cells infiltrated in synovial joints could be subdivided into several subsets based on their differential effector functions. For instance, TH1 subsets defended against intracellular pathogens and activated macrophages. TH1/Tfh/T helper-like can sustain the B cell-Ig switch, while the primary function of TH17 was to recruit neutrophils and induce proinflammatory factors. Cytotoxic CD4⁺ T cells mediated cell death and participated in the hypercitrullination of Neutrophil Extracellular Traps through a perforin-dependent mechanism (81). Parallel to the characterisation and classification of T cell subsets, the contribution of various effector T cells to synovial homeostasis has been investigated in numerous studies, although sometimes with contradictory results. Reduced functions and frequencies of Treg cells contribute to the pathogenesis of RA (82). m⁶A plays a significant role in maintaining T cells’ homeostasis and differentiation stability. Zhao group established a nomogram based on the immune cell model employing GEO databases and cluster analysis (83). Finally, CLP1 was highly correlated with Tfh infiltration, which may play a non-negligible role in the synovium via altering immune cell infiltration. It could be a novel target for clinical diagnosis and therapy of joint disorders (84).

Luo et al., 2021 has described the role of m6A as a prevalent internal mRNA modification that regulates various physiological functions and disease progression, specifically emphasising its involvement in chronic inflammation, which is a hallmark of rheumatoid arthritis. They argue that emerging evidence points to m6A methylation as a significant player in inflammatory responses, suggesting that understanding this modification could provide insights into the pathogenesis of inflammation-related diseases (85).

Xu et al., have highlighted that m6A modifications not only regulate post-transcriptional processes but also influence cellular responses, including necrosis, apoptosis, and autophagy, which are critical in the context of inflammatory diseases. Their findings indicate that m6A is integral to the regulation of inflammation, thereby reinforcing the notion that m6A dysregulation may contribute to the development of conditions such as rheumatoid arthritis (86).

Huang et al., have described the myth about m6A methylation that is intertwined with fibrosis and inflammation regulation, particularly through the functional dynamics of m6A methyltransferases, demethylases, and binding proteins. The authors posit that disruptions in this network can lead to abnormal m6A levels, resulting in altered RNA expression and the onset of autoimmune disorders. They also note that m6A modification is crucial for innate immune responses, suggesting that m6A dysfunction could be a key factor in the pathophysiology of autoimmune diseases like rheumatoid arthritis (87). The crosstalk of RNA modification regulators and the potential roles in T cell infiltration and tumor microenvironment rheumatoid arthritis have been widely researched. Zhou Jing’s team has confirmed that METTL3 produces a marked effect on T cells (88). It is investigated that in condition of rheumatoid arthritis METTL3 governs the homeostasis and differentiation of T cells by targeting the IL-7/STAT5/SOCS signal axis and regulates the stability of Treg cells via targeting the IL-2/STAT5/SOCS pathway (89). Likewise, Yao et al. showed conditional deletion of METTL3 in CD4⁺ T cells impaired Tfh differentiation and germinal center responses in a cell-intrinsic manner in mice (90). Subsequently, their research group confirmed that m6A eraser ALKBH5, but not FTO, maintains the ability of naïve CD4⁺ T cells to induce adoptive transfer colitis, which resulted in attenuated CD4⁺ T cells responses and diminished recruitment of neutrophils during T cell-mediated inflammation and autoimmunity (88). RNMT (m⁷Gcap methyltransferase), under the action of m⁷Gcap binding protein LARP1, promotes T cell activation by increasing ribosome abundance (91).

Guo et al. identified 455 m⁵c and 26 ac4C methylation transcripts by RIP-Seq and mass spectrometry in CD4⁺T cells (84). The methylation expression profile analysis found that NSUN2, as the major m⁵C methyltransferase, significantly decreased its expression in CD4⁺T cells. These ac4C-related dysregulated transcripts mainly monitor mRNA’s catabolic process and translation initiation (92). Besides, m⁷G modification level is closely related to the immune infiltration characteristics in patients with COVID-19. The expression of m⁷G-related genes in COVID-19 patients was higher than in non-COVID-19 patients (93).

Both synovial and intra-articular B cells are recruited from the peripheral blood to inflammation sites, driven by CC and CXC chemokines that express receptors on B lymphocytes (94). Therefore, synovial tissue becomes the active area for B cell aggregation, proliferation, plasma cell differentiation and autoantibody production. Synovial B cells are usually located in the T-B cell region and maintained under sustained stimulation by cytokines, including proliferation-inducing ligand (APRIL) and lymphocyte stimulator (BLyS) (95). Currently, studies have found that in blood and synovium of patients with RA, memory (CD27+), B lymphocytes (including swIg and LM B lymphocytes) increased, accompanied by an activated (CD95+CD21 low) phenotype (96). Figure 6 explains the molecular mechanism and RNA modification with their functions in B cells.

B cells secreted substantial proinflammatory cytokines under the control of transcription factor NF-kB. In the early and late stages of arthritis, activated NF-kB is generally present in B cells and other cells (including macrophages, T cells, and fibroblast synoviocytes) in synovial tissue (97). B lymphocytes are conducted to antigen presentation and immune complex formation, then internalised by dendritic cells and macrophages through an FCγR-mediated process (98). This conclusion has been verified in a murine model of autoimmune arthritis, in which both T cells (CD4+) and antigen-presenting cells jointly promote the occurrence and development of the disease. Still, the premise is that B-cell or IgM-deficient mice will not develop arthritis even under the condition of immunity (92).

Previous studies have illustrated that activated B cells release cytokines to interact with T cells and neutrophils (99). Stimulated T cells react to other chemokines and guide them to search for their B-cell partners. The migration route of B cells with sufficient antigen binding from the naive B cells was redirected and moved to the boundary of B cells and T cells. Similarly, B cells receive the help of T cells, which assist in forming autoantibodies in an HLA-DR dependent manner through class switching and somatic hypermutation (100). B cells and neutrophils together bring about a vicious circle. B-cell-derived IL-8 recruits neutrophils into the synovium for activation. In turn, primed neutrophils activate B cells in the synovium by continuously releasing cytokines, such as BAFF and IL-21 (101).

To date, there is no literature about the impact of RNA modification on immune cells (B cells) in synovium, especially pseudouridine. However, only a few studies have examined the impact of RNA modification on B cells in the cancer micro-environment. Kang and colleagues elucidated stage-dependent reliance on METTL3-mediated m6A of B cell development (95). They demonstrated that in the pro-B stage, METTL3 deficiency would affect the development and function of B Cells and the fibrogenic activity of B cells in liver fibrosis (102).

New research has indicated that METTL14-mediated m6A modification plays a crucial part in the response of gastric cancer B cells (101). Further demonstrated that the Mettl14/m6A/Ythdf2 axis inhibits negative regulators’ expression through mRNA decay, which is mandatory for the selective favour of GC B cells. Furthermore, inhibition of gene expression by M6A modification guides suitable determination of B cell fate at the beginning of the adaptive immune response. For example, METTL3 mediated m6A modification is required for maintaining GC stability through the function of different m6A readers (103, 104). The m6A binder IGF2BP3 is necessary to stabilise the expression of MYC mRNA and the target gene, while the m6A reader YTHDF2 acts indirectly to regulate the expression of oxidative phosphorylation. Grove et al. found YTHDF2 inhibits the genetic program of plasma cells and promotes the formation of GC at the initial stage of B cell immunological response (104, 105). It is worth noting that according to the analysis of CIBERSORT and TIMER databases, METTL16 (m⁶A “writer”) expression exhibits a significantly positive association with the invasion of B cells and CD8+T cells (106).

Chuanxi Yang and coworkers systematically analysed the prognostic and immunological roles of the ac4C regulator NAT10 in Pan-cancer (69). Contrasting with normal tissues, the expression of NAT10 significantly affects the prognosis of patients with pan-cancer. There is a positive association between NAT10 expression and infiltration of tumor immune cells, consisting of B cells, fibroblasts, CD8⁺ T cells, CD4⁺ T cells, macrophages, and neutrophils in LIHC. Therefore, it also represents a valuable research direction for tumor research.