RNA m⁶A Modification in Cancers: Molecular Mechanisms and Potential Clinical Applications

Abstract

N⁶-Methyladenosine (m⁶A) RNA modification brings a new dawn for RNA modification researches in recent years. This posttranscriptional RNA modification is dynamic and reversible, and is regulated by methylases (“writers”), demethylases (“erasers”), and proteins that preferentially recognize m⁶A modifications (“readers”). The change of RNA m⁶A modification regulates RNA metabolism in eucaryon, including translation, splicing, exporting, decay, and processing. Thereby the dysregulation of m⁶A may lead to tumorigenesis and progression. Given the tumorigenic role of abnormal m⁶A expression, m⁶A regulators may function as potential clinical therapeutic targets for cancers. In this review, we emphasize on the underlying mechanisms of m⁶A modifications in tumorigenesis and further introduce the potential m⁶A regulators-associated therapeutic targets for tumor therapy.

Public Summary

• •N⁶-Methyladenosine (m⁶A) RNA modification brings a new dawn for RNA modification researches in recent years.
• •The dysregulation of m⁶A may lead to tumorigenesis and progression.
• •m⁶A regulators may function as potential clinical therapeutic targets for cancers.

Graphical Abstract

Main Text

From the 1950s, an increasing accumulated number of chemically modified nucleosides in RNA have been identified, which change the RNA structures, leading to different biological functions.^1,2 Many RNA modifications have been identified in eukaryotic transcripts (Figure 1). N⁶-Methyladenosine (m⁶A) RNA modification defines as a methylation of adenosine at the N⁶ position, brings a new dawn for RNA modification research since its first discovery in 1974.³ m⁶A has been identified as the most abundant mRNA modification, widely distributing in the majority of eukaryotic species including mammals,3, 4, 5, 6 plants,⁷ insects,⁸ yeast,⁹ and certain viruses.^10,11 The identifications and characterizations of “writer,” “eraser,” and “reader” proteins, and development of high-throughput sequencing provide new insight into RNA modification biology, especially m⁶A RNA modification.12, 13, 14

Figure 1.

RNA Modifications in Eucaryon

A schematic diagram of common RNA modifications in eukaryotic transcripts. m¹A, N¹-methyladenosine; hm⁵C, 5-hydroxymethylcytosine; m⁶A_m, N⁶,2′-O-dimethyladenosine

m⁶A was identified in about one-third of the mRNAs in mammals while an average number of 3 to 5 m⁶A modifications were found in each mRNA. Of note, a lot of m⁶A sites are evolutionally conserved between humans and mice. The deposition of m⁶A modification in the transcriptome is not random.¹⁵ A characteristic DRACH sequence (D = G, A, or U; R = G or A; H = A, C, or U) often enriched in the 3′ untranslated region (3′ UTR) and the coding sequence (CDS) when m⁶A modification occurs.^12,13 Besides, it has clearly known that m⁶A modifications exist on almost all types of coding and non-coding RNAs (ncRNAs) and dynamically regulate their relevant molecular processes, physiological and pathological functions.

In this review, we emphasize the underlying mechanisms of m⁶A modifications in tumorigenesis and further introduce the potential m⁶A regulator-associated therapeutic targets for tumor therapy.

Dynamic Regulation of m⁶A

The RNA m⁶A modifications are dynamically and reversibly regulated by two important catalytic proteins, methyltransferases and demethylases, which are also recognized as “writers” and “erasers,” respectively. In addition, a set of binding proteins (“readers”) function as decoding the m⁶A modifications and recruiting downstream functional complexes (Figure 2).

Figure 2.

Mechanism of RNA m⁶A Modifications

The m⁶A modification is catalyzed by the “writer” proteins including METTL3, METTL14, WTAP, METTL5, METTL16, ZCCHC4, VIRMA, RBM15/15B, HAKAI, and ZC3H13. The m⁶A methylation is removed by “eraser” proteins FTO or ALKBH5. “Reader” proteins recognize m⁶A modifications and determine targeted RNA destiny.

m⁶A Writer Proteins

The RNA m⁶A modification writer proteins are a kind of methyltransferase, consist of methyltransferase-like 3 (METTL3), methyltransferase-like 5 (METTL5), methyltransferase-like 16 (METTL16), zinc finger CCHC-type containing 4 (ZCCHC4), and additional partner proteins such as methyltransferase-like 14 (METTL14), Wilms tumor 1-associating protein (WTAP), Vir like m⁶A methyltransferase associated (VIRMA), zinc finger CCCH-type containing 13 (ZC3H13), and RNA-binding motif 15/15B (RBM15/15B).16, 17, 18, 19, 20, 21 METTL3 was found acting as the key methyltransferase for m⁶A methylation in 1997 and abnormal expression of METTL3 often affects the total level of m⁶A methylation.22, 23, 24 Meanwhile, METTL14 functions as a synergistical partner for METTL3, which has structural support for METTL3 when they combine into a core methyltransferase complex, and therefore facilitates RNA binding.^24,25 The main function of WTAP is binding to METTL3/METTL14 complex and is necessary for the recruitment of optimal substrate and the localization of METTL3/METTL14 complex.^17,26 RBM15/15B has been confirmed, helping the combination process between METTL3 and WTAP by specific binding to U-rich regions.²⁷ HAKAI, also known as Cbl Proto-Oncogene Like 1 (CBLL1), assists controlling nuclear m⁶A methylation. The function of ZC3H13 is similar to CBLL1, which helps nuclear localization.²⁰ VIRMA is very important for m⁶A modification, especially locating at the 3′-UTR and around stop codon sequences.¹⁹

Of all the m⁶A methyltransferases, METTL5, METTL16, and ZCCHC4 could function independent of other methyltransferase proteins. As for METTL5, it has been identified as a kind of independent m⁶A methyltransferase, catalyzing m⁶A on some certain structured RNAs, including 18S rRNA, 28S rRNA, and U6 small nuclear (snRNA).^21,28, 29, 30, 31 A special structure of heterodimeric complex with TRMT112 is formed by METTL5 and could enhance its own metabolic stability in cells.²⁸ Meanwhile, TRMT112 functions as a coactivator of METTL5, and the relationship between METTL5 and TRMT112 is similar to the relationship between METTL3 and METTL14.²⁸ Besides, the atomic resolution structure of METTL5-TRMT112 complex demonstrates the RNA-binding mode of the complex differs considerably from that of other m⁶A RNA methyltransferases.²⁸ Similar to METTL5, METTL16 is also demonstrated as an independent m⁶A methyltransferase, whose binding sites have no overlap with METTL3/METTL14 complex, regulating the stability and splicing of mRNAs.^31,32 Therefore, the overexpression of METTL16, with a mutation on catalytic domain, activates the splicing process.³³ Besides, METTL16 has been reported binding to U6 snRNA and numerous ncRNAs (ncRNAs), long ncRNAs (lncRNAs), and pre-mRNAs.^30,31 With respect to ZCCHC4, it has been identified as an RNA m6A methyltransferase, methylating human 28S rRNA. Besides, the rRNA m⁶A methylation mediated by ZCCHC4 makes a difference to the distribution of ribosome subunit, global translation, and cell proliferation, which may lead to tumorigenesis when aberrant ZCCHC4 expression occurs.²¹

m⁶A Eraser Proteins

The m⁶A eraser proteins, acting as demethylase, remove m⁶A modifications by raising ferrous iron as co-factor and α-ketoglutarate as co-substrate at the same time.³⁴ Unlike the numerous types of m⁶A writer proteins, only two m⁶A demethylase, including fat mass and obesity-associated (FTO) protein and AlkB homolog 5 (ALKBH5) protein, have been identified so far. In addition, m⁶A eraser proteins often work in the nucleus where the process of demethylation occurs.

FTO, the first recognized m⁶A eraser protein, belongs to AlkB family, which can mediate m⁶A modification by demethylating not only internal m⁶A but also N6, 2ʹ-O-dimethyladenosine (5ʹ cap m⁶A_m) mRNA.^35,36 Besides, FTO is nucleus localized to a large extent and regulates −10% of total m⁶A modification in almost all the cell lines and mRNA internal m⁶A is the major substrate of FTO.37, 38, 39 Moreover, FTO is highly expressed in the brain, a metabolically active organ. The processes of RNA transcription and translation regulate metabolism, and of course reversible m⁶A modification plays an important role in these processes. A recent study showed that NADP directly binds FTO, independently increases FTO activity, promoting the demethylation of RNA m⁶A modifications and adipogenesis.⁴⁰ The discovery of FTO proves the process of m⁶A methylation is reversed and controlled under physiological and pathological status.

Another eraser protein is ALKBH5, which has functional similarity with FTO in m⁶A demethylation. ALKBH5 is widely expressed in various tissues, especially in the testes.⁴¹ The differential expression of FTO and ALKBH5 among different tissues reveals that the two proteins each play their roles in different biological pathways. Specifically, m⁶A is the only known substrate of ALKBH5 to date.⁴² Even more encouraging is the crystal structures of ALKBH5 have been analyzed, providing new insights into the underlying mechanisms for the procedures of m⁶A recognition and demethylation.^41,43,44 The above findings greatly accelerate the development of targeted drugs for m⁶A demethylation inhibitors.

m⁶A Reader Proteins

The m⁶A-mediated posttranscriptional gene regulation has been further understood by identification and characterization of m⁶A readers. The m⁶A reader proteins control the destinies of the modified RNAs. Therefore irregulation of the proteins may cause misconception of the binding modified RNAs and following RNA metabolic disturbance. The m⁶A reader proteins consist of the YT521-B homology (YTH) domain family proteins (YTHDF1-3), YTH domain containing proteins (YTHDC1-2), insulin-like growth factor 2 mRNA-binding proteins (IGF2BPs), and heterogeneous nuclear ribonucleoproteins (HNRNPs, including hnRNPA2B1, hnRNPC, and HNRNPG).

The CCR4-NOT deadenylase complex mediates deadenylation of m⁶A modified RNAs.⁴⁵ Cytoplasmic YTHDF2, interacting with the SH domain of the CNOT1 subunit, recruits the CCR4-NOT complex and subsequently destabilizes m⁶A modified RNAs, leading to their deadenylation and decay.^45,46 On the contrary, the other two YTH domain family proteins, YTHDF1 and YTHDF3, are reported to facilitate translation by recruiting translation initiation factors.^47,48 In addition, YTHDF3 accelerates mRNA translation in conjunction with YTHDF1 and promotes the deadenylation and decay of m⁶A modified RNAs collaboratively with YTHDF2.^48,49

In the cell nucleus, YTHDC1 regulates mRNA splicing through recruiting and combining a certain splicing factor serine/arginine-rich splicing factor 3 (SRSF3).⁵⁰ Besides, YTHDC1 also helps mRNA exportation from nucleus to cytoplasm.⁵¹ Recently, YTHDC1 has been found to accelerate the decay of a subset of m6A modifications on chromosome-associated regulatory RNAs (carRNAs), especially the long interspersed element-1 (LINE1) elements, via the nuclear exosome targeting complex-mediated nuclear degradation.⁵² YTHDC2 enhances the translation efficacy of target RNAs while decreasing their abundance in the cytoplasm.⁵³

Currently, IGF2BPs are demonstrated to maintain the stable structures of target mRNA by m⁶A-dependent approaches under both normal and stressed circumstances.⁵⁴ HNRNPA2B1 recognizes the m⁶A signals of primary microRNA and interplays with DiGeorge syndrome critical region gene 8 (DGCR8), while selective recognition of m⁶A-dependent splicing in mRNA secondary structures was proven as the function of HNRNPC.55, 56, 57, 58, 59

m⁶A in Cancer

Previous studies have suggested the effects of m⁶A modification and its capacity to modulate and coordinate gene expression. The level of m⁶A may profoundly affect the characteristics of cancer. It is suggested that m⁶A may play an important role in carcinogenesis or inhibition in malignant tumor effect. Some proteins need to be modified by m⁶A to participate in the mechanism of carcinogenesis, but it is still not clear whether they take effect in the modification. The main causes of tumorigenesis in m⁶A-dependent manner (Figure 3) and the functions of the main m⁶A proteins in most of the tumor types (Figure 4) are shown.

Figure 3.

Precipitating Factors for Abnormal m⁶A Regulation in Cancer

Figure 4.

m⁶A Modifiers in Human Cancers

The sites of different cancers (A) and the roles of m⁶A genes in different cancers (B). HCC, hepatocellular carcinoma; PRC, prostate cancer; GC, gastric cancer; CVC, cervical cancer; CRC, colorectal cancer; PAC, pancreatic cancer; OVC, ovarian cancer; BRC, breast cancer; NSCLC, non-small cell lung cancer; AML, acute myeloid leukemia; RCC, renal cell carcinoma; GBM, glioblastoma; BLC, bladder cancer; ENC, endometrial cancer; MLM, melanoma; OST, osteosarcoma.

Acute Myeloid Leukemia

Acute myeloid leukemia (AML) has obvious genetic variation due to uncontrolled proliferation and cell differentiation defects of myeloid leukocytes. So far, the treatment for AML is still unsatisfactory. Previous studies have demonstrated that MELLT3 and METTL14 promoted the translation of MYC, MYB, BCL2, SP1, and PTEN, increasing the level of phosphorylated AKT.^60,61 It has been found that FTO played an important role in the proliferation of AML cells.62, 63, 64, 65, 66 FTO can enhance leukemia oncogene-mediated cell transformation and leukemia, inhibit all trans retinoic acid–mediated differentiation of AML cells, and regulate the mRNA synthesis of target genes (such as ASB2 and RARA) by down-regulating the level of m⁶A.⁶⁶ Su et al.³⁹ found that R-2HG inhibited the proliferation of leukemia cells and resulted in cell-cycle arrest by inhibiting FTO activity, increasing m⁶A modification, reducing the stability of MYC/CEBPA transcription. Many studies have confirmed that YTHDF2 can promote tumor progression. Paris et al.⁶⁷ suggested that YTHDF2 was highly expressed in AML and played a key role in the initiation and proliferation of AML. They found that targeting YTHDF2 could prolong the half-life of the m⁶A modified transcripts, thus selectively destroying the initiation and proliferation of AML without affecting normal hematopoiesis. Besides, WTAP is upregulated in AML and the high expression of WTAP predicts poor outcomes in patients with AML.^68,69

Glioblastoma

Glioblastoma (GBM) is the deadliest primary brain tumor.⁷⁰ Research on the role of METTL3 in GBM has produced conflicting conclusions. The different phenotype associated with METTL3 can be illustrated by the different dependence and genetic heterogeneity of m6A modified RNA in different types of GBM cells. Cui et al.⁷¹ demonstrated that METTL3 and METTL14 inhibit the growth and tumorigenesis of glioblastoma stem-like cells (GSCs) by the down-regulation of ADAM19/EPHA3/KLF4 pathway. ALKBH5 was also highly expressed in GSCs.^72,73 Silencing ALKBH5 could inhibit the proliferation of GSCs. ALKBH5 can induce demethylation of FOXM, a transcription factor, and stimulate cell proliferation, resulting in increased FOXM1 expression.⁷³ Inhibiting the expression of ALKBH5 provides a new direction for the treatment of glioma. Visvanathan et al.⁷⁴ found that METTL3 was highly expressed in malignant GBM. METTL3 promoted tumor growth by targeting the 3′ UTR of SOX2 mRNA. Silencing METTL3 can inhibit tumor growth and enhance tumor radiosensitivity, which can be used as a potential molecular target for GBM therapy. Similar to AML, the high expression of WTAP also predicts poor prognosis in patients with GBM.^75,76

Lung Cancer

Lung cancer is the leading cause of cancer-related mortality worldwide.77, 78, 79, 80, 81, 82, 83 METTL3 is the oncogene of lung cancer through different mechanisms. Lin et al.⁸⁴ found that METTL3 can enhance RNA translation without the aid of methyltransferase and reader protein activity. METTL3 increases RNA translation by directly recruiting translation initiation factors. METTL3 increases the growth, survival and invasion of lung adenocarcinoma cells by increasing EGFR and TAZ.⁸⁴ Choe et al.⁸⁵ demonstrated that METTL3 could enhance translation, transform oncogenes, and form dense polyribosomes by interacting with EIF3H in primary lung cancer, which can be applied as a potential therapeutic target. METTL3 can also promote tumor progression by regulating some microRNAs (miRNAs). Du et al.⁸⁶ found that mir-33a can inhibit the proliferation of non-small cell lung cancer (NSCLC) by targeting METTL3 mRNA. Wei et al.⁸⁷ revealed that mir-600 can inhibit the migration and proliferation of lung cancer cells by down-regulating the expression of METTL3. Li et al.⁸⁸ found that FTO promotes the proliferation of NSCLC by increasing the expression of USP7. Liu et al.⁸⁹ found that overexpression of FTO can down-regulate the level of m⁶A in MZF1 mRNA transcripts, increase the stability of mRNA, and promote the expression of MZF1, leading to the proliferation and invasion of lung squamous cell carcinoma cells. In addition, m⁶A demethylase ALKBH5 was indicated to inhibit tumor growth and metastasis in patients with NSCLC by reducing the expression of YTHDFs-mediated YAP, whereas some studies suggested ALKBH5 promoted the progression of NSCLC.^90,91 As for YTHDF1, it is also controversial for its role in NSCLC. On one hand, YTHDF1 deficiency was proved to suppress NSCLC cell proliferation and xenograft neoplasia via affecting the translational efficacy of cyclin D1, CDK2, and CDK4. On the other hand, it was observed that the high expression of YTHDF1 associated with better prognosis and the depletion of YTHDF1 helped tumor cells resistant to cisplatin.⁹²

Endometrial Cancer

Approximately 70% of patients with endometrial cancer show decreased m⁶A expression level because of either reduced expression of METTL3 or mutation (loss-of-function) in METTL14 and the tumorigenicity is associated with activation of the AKT pathway.⁹³ Down-regulation of METTL14 can reduce the level of m⁶A of RNA. Recent studies have shown that the decrease of m⁶A level is associated with endometrial cancer caused by METTL14 gene mutation.⁹³ They found that the deletion of METTL14 increased cancer cell proliferation, clone formation, and metastasis. They observed that the level of m⁶A in tumor tissue was lower than that in adjacent normal tissue. They believe that the METTL14 mutation reduces the level of m⁶A, which plays a key role in the progression of endometrial cancer. Liu et al.⁹³ also found that the down-regulation of METTL14 could lead to the decrease of the expression of PHLPP2, the negative regulator of AKT pathway, and increase the expression of mTORC2, the positive regulator of AKT pathway, which led to the proliferation of endometrial cancer cells. FTO can also participate in the occurrence and development of endometrial cancer through a variety of mechanisms. Zhu et al.⁹⁴ have shown that estrogen induces FTO nuclear aggregation through the mammalian target of Rapamycin signaling pathway, thus enhancing the proliferation and activity of endometrial cancer cells and promoting tumor progression. Another study showed that estrogen induces overexpression of FTO gene by activating PI3K/AKT and MAPK signaling pathways, which leads to the proliferation and invasion of endometrial cancer cells.⁹⁵

Cervical Cancer

The role of FTO in cervical cancer tumorigenesis has been uncovered, FTO functions as an oncogenic regulator for cervical cancer in terms of promoting tumor cell proliferation and migration by regulating E2F1 and Myc transcripts.⁹⁶ Long ncRNA GAS5-AS1 was found having a lower expression in cervical cancer when compared with that of adjacent normal tissues and played its role on YTHDF2-dependent pathway.⁹⁷ Besides, m⁶A “reader” IGF2BP family proteins were also reported serving as carcinogenic roles in CVC.^98,99

Ovarian Cancer

METTL3 promotes the maturation of miR-126-5p and therefore accelerates ovarian cancer (OVC) progression.¹⁰⁰ In addition, another study demonstrated that METTL3 facilitated OVC growth and invasion via activating epithelial to mesenchymal transition.¹⁰¹ The expression of ALKBH5 was found to be increased in epithelial OVC tissues when compared with the normal ovarian tissues and ALKBH5 was identified as a candidate oncogene in epithelial OVC.¹⁰² The “reader” protein IGF2BP1 enhanced SRC/MAPK-driven invasive growth of OVC cells and the high expression of IGF2BP1 was related to poor prognosis of patients with OVC.^103,104 Most recently, recruitment of YTHDF1 to m⁶A-modified TRIM29 was participated in accelerating TRIM29 translation in the OVC cells with cisplatin-resistance, which would be a potential therapeutic target.¹⁰⁵

Breast Cancer

There is now a compelling body of evidence demonstrating that epigenetic modifications including RNA m⁶A modification play a vital role in the tumorigenesis and progression of breast cancer.¹⁰⁶ Cai et al.¹⁰⁷ found that METTL3 interacts with hepatitis B X-interacting protein (HBXIP) in breast cancer cells, and HBXIP inhibits the expression of METTL3 by acting on 3′UTR of miRNA let-7g. At the same time, METTL3 promoted the expression of HBXIP through m⁶A modification. A positive feedback loop was formed between HBXIP/miRNA let-7g/METTL3, which promoted the proliferation of breast cancer cells. METTL14 was promoted by LINC00942, enhancing the initiation and progression of breast cancer.¹⁰⁸ Niu et al.¹⁰⁹ found that FTO is highly expressed in breast cancer tissues, and the higher the expression, the worse the prognosis of patients. FTO promotes the proliferation and metastasis of breast cancer cells by inhibiting BNIP3. In one study, Zhang et al.¹¹⁰ found that ALKBH5 mediated m⁶A demethylation of NANOG (an embryonic stem cell transcription factor) mRNA, which increased the expression of NANOG in breast cancer. Knockout of ALKBH5 can increase the demethylation of NANOG mRNA and decrease the expression level of NANOG, which can significantly inhibit lung metastasis of breast cancer. Subsequently, they found that HIFs promoted the invasion and metastasis of breast cancer cells by regulating ZNF217 to inhibit methylation and ALKBH5 induced demethylation.¹¹¹

Colorectal Cancer

Epigenetic alterations exist in various aspects of colorectal tumorigenesis.¹¹² In colorectal cancer, METTL3 acts as a functional oncogene in an m⁶A-IGF2BP2/3-dependent manner.¹¹² Oncogene c-Myc can promote the expression of YTHDF1, induce the proliferation and metastasis of cancer cells, and increase their resistance to chemotherapy drugs.¹¹³ Knockdown of c-Myc can inhibit the expression of YTHDF1 in colorectal cancer through HIF-1α. Besides, oncogene c-Myc can promote the expression of YTHDF1, inducing the proliferation and metastasis of cancer cells, and increasing their resistance to chemotherapy drugs.¹¹³ YTHDF3 negatively regulated lncRNA GAS5 through GAS5-YAP-YTHDF3 axis both in vivo and in vitro.¹¹⁴ The expression of YTHDC2 is positively correlated with the stage and metastasis of colon cancer.¹¹⁵ Knockdown the expression of YTHDC2 can inhibit the metastasis of tumor cells in vivo and in vitro through HIF-1α. Except all the preceding m⁶A-associated oncogenes, METTL14 is demonstrated as a tumor suppresser, decreasing the proliferation and tumor metastasis of colorectal cancer via different molecular mechanisms.116, 117, 118, 119

Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is the most common primary neoplasm of the liver.¹²⁰ Chen et al.¹²¹ found that METTL3 can promote the progression of liver cancer cells through YTHDF2-dependent transcriptional regulation of SOCS2 silencing. Further study showed that METTL14 interacted with DGCR8 and positively regulated the expression of miRNA 126.¹²² Overexpression of METTL14 can inhibit the metastasis of liver cancer cells in mice. According to these results, they speculated that METTL14 may regulate the expression of miRNA 126 through the modification of m⁶A, and then regulate its downstream target to inhibit the metastasis of HCC. Therefore, METTL14 may be an important adverse prognostic factor for HCC.^122,123 As for IGF2BP family, IGF2BP1,124, 125, 126 IGF2BP2,¹²⁷ and IGF2BP3^128,129 were all identified as oncogenes to promote the carcinogenesis of HCC. YTHDF2 not only acts as a tumor activating protein, but also acts as a tumor suppressor protein. Some studies have shown that, hypoxia can induce the decrease of the expression of YTHDF2 in hepatocellular carcinoma.^130,131 They also found that overexpression of YTHDF2 inhibited the proliferation of HCC cells and activated MEK and ERK. YTHDF2 can directly act on the 3′UTR m⁶A modification site of EGFR mRNA, resulting in the degradation of EGFR mRNA. In addition, the phosphorylation of ERK induced by hypoxia was also blocked by YTHDF2, suggesting that hypoxia can down-regulate the phosphorylation of ERK induced by YTHDF2. YTHDF2 inhibits ERK/MAPK signal transduction by reducing the stability of EGFR mRNA in HCC, thus inhibiting the proliferation of hepatoma cells.¹³¹ WTAP was examined to determine whether it was related to clinicopathological factors of patients with HCC, and the results showed that the expression level of WTAP was increased more in HCC than in para-carcinoma tissues and associated with worse prognosis.^132,133 Besides, VIRMA was also defined as an oncogene in HCC.^134,135

Pancreas Cancer

Pancreas cancer is recognized as a kind of high-grade malignant neoplasm. He et al.⁵⁴ revealed that ALKBH5 may be a potential therapeutic target for pancreatic cancer by down-regulating methylation of lncRNA KCNK15-AS1 in pancreatic cancer cells and inhibiting cell motility. Taketo et al.¹³⁶ found that pancreatic cancer cells with low METTL3 expression are more sensitive to chemotherapy drugs such as gemcitabine, 5-fluorouracil, and cisplatin and radiotherapy, it provides a potential target for the treatment of pancreatic cancer. Many studies have proved that IGF2BP2 was overexpressed in pancreas cancer and promoted cancer proliferation.^137,138 A case-control study demonstrated that variants in the FTO gene was associated with pancreatic cancer risk.¹³⁹ YTHDF2 promoted proliferation while inhibiting migration and invasion in pancreatic cancer cells.¹⁴⁰

Gastric Carcinoma

With improved diagnostic strategy, patients who diagnosed with early-stage gastric carcinoma (GC) is increasing. Previous studies have shown that epigenetics may play an important role in the genesis and growth of GC. Li et al.¹⁴¹ demonstrated the high expression of FTO and ALKBH1 mRNA indicates poor prognosis of GC through mining TCGA database. METTL3 promotes GC angiogenesis and glycolysis by increasing the stability of HDGF mRNA and activating the AKT signaling pathway.¹⁴² ALKBH5 promotes invasion and metastasis of GC by decreasing methylation of the lncRNA NEAT1.¹⁴³ Some studies demonstrated that IGF2BP3 functioned as an oncogene to promote tumor progression in GC.^144,145 Knockdown of METTL14 (m⁶A suppression) promotes GC development through activating the Wnt/PI3K-AKT signaling pathway, whereas increasing m⁶A levels reversed these phenotypical and molecular changes.¹⁴⁶

Prostate Cancer

Prostate cancer (PRC) is one of the most common malignancies worldwide.¹⁴⁷ Some studies have shown that down-regulation of YTHDF2 can inhibit the proliferation and invasion of PRC cells.^148,149 There was a negative correlation between YTHDF2 and miR-493-3p. Knockout of YTHDF2 could increase the level of m⁶A and inhibit the proliferation and invasion of GC cell lines. Besides, METTL3 and VIRMA were also reported promoting the development and progression of PRC in different manners.150, 151, 152, 153

Renal Cell Carcinoma

METTL3 and METTL14 can not only promote the proliferation and metastasis of tumor cells, but also inhibit the progress of tumor.^154,155 Some studies have shown that low expression of METTL3 is associated with larger tumors and higher histological grade in mice.¹⁵⁶ The survival time of renal cell carcinoma patients with METTL3 overexpression was significantly prolonged.¹⁵⁶ WTAP also plays a role in tumor growth by binding with mRNA and enhancing the stability of mRNA. WTAP can combine with cyclin dependent protein kinase (CDK) 2 transcripts, enhancing its stability, delaying cell apoptosis and promoting the proliferation of renal cell carcinoma.¹⁵⁷ On the contrary, ALKBH5 and FTO was identified as a tumor suppresser for renal cell carcinoma.158, 159, 160

Potential Applications of Cancer Treatment Based on m⁶A

Many researchers have demonstrated that m⁶A has emerged as a widespread regulatory mechanism that controls gene abnormal expression in diverse pathological pathways, leading to tumorigenesis.^161,162 Therefore, m⁶A regulators may function as potential clinical therapeutic targets for cancers. Since the first m⁶A demethylase FTO was identified, FTO has become the most striking target for developing targeted drugs against tumors.¹⁶³ Several FTO-targeted inhibitors have already been developed, including MO-I-500, meclofenamic acid (MA), FB23, R-2HG, rhein, and so on. MO-I-500, a selective inhibitor, inhibits the enzyme activity of FTO and was reported to suppress the proliferation of breast cancer cell lines.^109,164,165 MA also proved to be a selective inhibitor for treating GBM through inhibiting FTO over ALKBH5.^71,166 Recently, FB23 and FB23-2 were designed as small-molecule FTO inhibitors, which have achieved remarkable inhibitory effect in AML models.⁶⁴ As for nonselective inhibitors for FTO, rhein was the first to be uncovered and competitively binds to specific site of FTO.¹⁶⁷ R-2HG, highly expressed by mutant isocitrate dehydrogenase 1/2 (IDH1/2) enzymes, was demonstrated to play important antitumor effect in glioma and leukemia cells.³⁹

Multidisciplinary therapy, including neoadjuvant therapy, surgery, adjuvant chemoradiotherapy, targeted therapies, and immunotherapies, has been widely adopted in cancer treatment. However, drug resistance maintains a dominating hindrance to curative treatment, leading to treatment failure and tumor progression.^168,169 Recently, the dysregulation of m⁶A regulators has been found related to the advent of treatment resistance.170, 171, 172 As for immunotherapy, it is an emerging way of dealing with cancer. YTHDF1 was proved to regulate antitumor immunity and had synergetic effect on immunotherapy by improving the therapeutic effect of PD-L1 inhibitor.¹⁷³ In addition, Yang et al.¹⁷⁴ found that m⁶A mRNA demethylation by FTO promoted melanoma growth and suppressed response to anti-PD-1 blockade, which indicated the combination of FTO inhibitors and anti-PD-1 blockade may help to promote sensibilization of immunotherapy in melanoma. Furthermore, the detailed information on drug resistance caused by m⁶A dysregulation is listed in Table 1. These studies indicated that m⁶A RNA modification signatures may provide important information for predicting drug resistance and help clinicians to adjust the treatment plan in time.

Table 1.

m⁶A Alteration in Drug-Resistant Cancer Cells

Cancer Type	Therapeutic Agents	m6A Proteins Involved	Change of m6A Proteins	Reference
AML	R-2HG	FTO	Down-regulate	Su et al.39
AML	TKI	FTO	Up-regulate	Yan et al.175
GBM	γ-irradiation	METTL3	Up-regulate	Visvanathan et al.74
HCC	Sorafenib	METTL3	Down-regulate	Lin et al.176
CRC	Doxorubicin	METTL3	Up-regulate	Uddin et al.172
NSCLC	Cisplatin	METTL3	Up-regulate	Jin et al.170
NSCLC	Cisplatin	YTHDF1	Down-regulate	Shi et al.92
NSCLC	Afatinib	m6A	Up-regulate	Meng et al.177
NSCLC	Crizotinib	METTL3, WTAP	Up-regulate	Ding et al.178
PAC	cisplatin, 5-Fu and gemcitabine	METTL3		Taketo et al.136
PRC	Cisplatin	VIRMA	–	Su et al.179
CSCC	cisplatin, irradiation	FTO	Up-regulate	Zhou et al.180
OVC	PARP inhibitor	FTO, ALKBH5	Down-regulate	Fukumoto et al.181
MLM	anti-PD-1 blockade	FTO, YTHDF2	Up-regulate	Yang et al.174

AML, acute myeloid leukemia; GBM, glioblastoma; HCC, hepatocellular carcinoma; CRC, colorectal cancer; NSCLC, non-small cell lung cancer; PAC, pancreatic cancer; PRC, prostate cancer; CSCC, cervical squamous cell carcinoma; OVC, ovarian cancer; MLM, melanoma.

Conclusions and Future Perspectives

Different from DNA, RNA has much more intricate posttranscriptional processing: RNA splicing, RNA editing, and RNA chemical modifications. RNA chemical modifications, most of which have no influence on nucleoside sequence, are structurally diverse and functionally multiple, indicating their functional importance.

In recent years, epigenetics, especially m⁶A RNA modification, has been further understood and explored with the rapid advances in detection methods and high-throughput sequencing techniques. It has been widely illustrated that the dysregulation of m⁶A RNA modification is related to various types of cancers, as well as the drug resistance to antitumor therapy.¹⁸² However, as the evidence (m⁶A plays important role in tumors) began to accumulate, the underlying mechanisms of m⁶A in cancer still have not yet been fully realized. To be specific, in many kinds of cancers, such as NSCLC, breast cancer, acute myelocytic leukemia, and some gynecological tumors, overexpression of either writer or eraser proteins can play tumorigenesis roles, while controversies of the roles of m⁶A regulators still exist in other cancer types.¹⁸³ Similar to DNA methylation signatures (hmc, 5hmc, and so on),184, 185, 186, 187, 188, 189 m⁶A RNA modifications of certain genetic loci could function as biomarkers associated with prognosis, molecular subtyping, and precise diagnosis.^68,190 Many researchers focus on the identification of m⁶A-targeted mRNAs in diverse diseases, especially in cancers. Moreover, how m⁶A RNA modification affects the functions of ncRNAs still remains unclear and ncRNAs have been confirmed having a close link with a wide variety of tumors.^15,191 carRNAs recently were found to have m⁶A modifications on them, which were deposited by METTL3, including repeat RNAs, promoter-associated RNAs, and enhancer RNAs.⁵² Thus, the direct crosstalk between the m⁶A carRNA modifications and chromatin state opens a new door on how m⁶A carRNA modifications regulate transcription. The Nobel Prize in Chemistry in 2020 was announced “for the development of a method for genome editing” (discovering the CRISPR/Cas9 genetic scissors). Therefore, whether a technology could developed for epitranscriptome editing is also worthy of expectation.

Considering that the dysregulation of m⁶A regulators is related to treatment resistance and cancer immunity, it is attractive to develop m⁶A-based targeted drugs for oncotherapy. Besides, m⁶A-based targeted small-molecule drugs could function as a kind of sensitizer, improving the therapeutic effects of chemotherapy, radiotherapy, and even immunotherapy in the future. In general, RNA m⁶A modification in cancers is an emerging field of cancer epigenetic research, providing not only new insights into the potential molecular mechanisms of tumorigenesis and cancer progression, but novel strategies for drug exploitation and clinical cancer therapies. At present, our understanding of RNA m⁶A modifications for cancers is still in its infancy and further studies are desperately required for the rosy scenario.