Role of N6-methyladenosine in tumor neovascularization

Abstract

Tumor neovascularization is essential for the growth, invasion, and metastasis of tumors. Recent studies have highlighted the significant role of N6-methyladenosine (m⁶A) modification in regulating these processes. This review explores the mechanisms by which m⁶A influences tumor neovascularization, focusing on its impact on angiogenesis and vasculogenic mimicry (VM). We discuss the roles of m⁶A writers, erasers, and readers in modulating the stability and translation of angiogenic factors like vascular endothelial growth factor (VEGF), and their involvement in key signaling pathways such as PI3K/AKT, MAPK, and Hippo. Additionally, we outline the role of m⁶A in vascular-immune crosstalk. Finally, we discuss the current development of m⁶A inhibitors and their potential applications, along with the contribution of m⁶A to anti-angiogenic therapy resistance. Highlighting the therapeutic potential of targeting m⁶A regulators, this review provides novel insights into anti-angiogenic strategies and underscores the need for further research to fully exploit m⁶A modulation in cancer treatment. By understanding the intricate role of m⁶A in tumor neovascularization, we can develop more effective therapeutic approaches to inhibit tumor growth and overcome treatment resistance. Targeting m⁶A offers a novel approach to interfere with the tumor’s ability to manipulate its microenvironment, enhancing the efficacy of existing treatments and providing new avenues for combating cancer progression.

Facts

• Neovascularization is a rate-limiting step in tumor progression.

• m⁶A modification participates in various aspects of cancer biology.

• Tumor neovascularization induces an immunosuppressive microenvironment.

• Immunosuppressive cells promote tumor neovascularization.

Open Questions

• What is the role of m⁶A modification in different modes of tumor neovascularization and associated pathways?

• What is the relationship between m⁶A modification and anti-angiogenic drug resistance?

• Can anti-angiogenic therapy be combined with immunotherapy by targeting m⁶A regulators?

• Can m⁶A targeting effectively improve the limited efficacy of current anti-angiogenic therapy?

Introduction

Tumor neovascularization ensures the acquisition of adequate oxygen and nutrients required for sustained tumor growth [1]. Notably, solid tumors tend to grow around blood vessels and cannot expand beyond 2 mm³ without vascularization [2, 3]. The induction of the “angiogenic switch”, which depends on a balance of angiogenic and anti-angiogenic factors, is a rate-limiting step in tumorigenesis, triggering exponential tumor growth [4, 5]. Neovascularization, considered as a hallmark of cancer, is indispensable for tumor proliferation, invasion, and metastasis [5]. Consequently, targeting tumor neovascularization has emerged as a crucial component of cancer therapy. Existing anti-angiogenic strategies primarily focus on the vascular endothelial growth factor (VEGF) or VEGF receptor (VEGFR) signaling pathway. Despite advancements, these approaches yield transitory benefits and often fail to achieve long-term clinical responses [6]. Increasing evidence indicates that tumor neovascularization is a complex process involving multiple components, underscoring the need to elucidate the underlying mechanisms to improve anti-angiogenic therapy efficacy [7–9].

Recently, researchers have proposed that non-mutational epigenetic reprogramming facilitates the acquisition of hallmark capabilities by tumors [10, 11]. Epigenetics refers to the study of heritable alterations that do not involve changes to the DNA sequence, including DNA and RNA methylation, nucleosome remodeling, and histone modifications [11, 12]. Over 170 RNA modifications have been identified, including N6-methyladenosine (m⁶A), 5-methylcytosine (m⁵C), N7-methylguanosine, and N1-methyladenosine [13]. The most prevalent RNA modification among these is m⁶A, first described in 1974, which occurs as an RNA methylation at the sixth nitrogen atom of adenosine [14]. m⁶A modification is a dynamic and reversible process that is installed by methyltransferases (“writers”), removed by demethylases (“erasers”), and recognized by RNA-binding proteins (“readers”) [15]. m⁶A participates in multiple aspects of RNA metabolism processes, including splicing, translation, stability, degradation, and nuclear export. It plays an essential role in reshaping the tumor microenvironment (TME), regulating cancer metabolism, and facilitating carcinogenesis [12, 15–17].

Recent evidence highlights the role of m⁶A in regulating tumor neovascularization. Our previous review linked m⁶A to immune reprogramming [16]. In this review, we aim to explore the regulatory role of m⁶A in tumor neovascularization, offering a comprehensive grasp of its significance in cancer therapy. This review introduces the diverse role of m⁶A in multiple modes of neovascularization and associated signaling pathways. Additionally, we concisely outline its contribution to vascular-immune crosstalk. Finally, we discuss the current development of m⁶A inhibitors and their potential clinical applications. This review clarifies the underlying mechanism of tumor neovascularization and provides novel insights into targeting m⁶A in anti-angiogenic therapy.

Tumor neovascularization

In 1971, Folkman proposed that solid tumor growth is always accompanied by the formation of new blood vessels, suggesting that inhibiting tumor vascularization could suppress tumor growth [2]. Since then, serveral modes of tumor neovascularization have been identified, including sprouting angiogenesis, vasculogenesis, intussusceptive angiogenesis, vasculogenic mimicry (VM), vessel co-option, and cancer stem cell (CSC)-derived vasculogenesis (Fig. 1) [7, 8]. The first three modes occur in both normal tissues and tumors, whereas the latter three are specific to tumor neovascularization. Among them, angiogenesis and VM are the most extensively studied.

Fig. 1. Modes of tumor neovascularization.

There are several modes of tumor neovascularization. a Angiogenesis: blood vessels form from preexisting vessels through sprouting; b vasculogenesis: endothelial progenitor cells derived from the bone marrow are recruited and differentiate into endothelial cells to form blood vessels; c intussusception: transcapillary tissue pillars insert into the lumen of existing vessels, undergo vascular splitting, and eventually fuse to remodel the vascular network; d vascular mimicry: tumor cells form vessel-like structures; e vessel co-option: tumor cells hijack the existing vasculature and migrate along the vessel surface or infiltrate non-malignant tissues between vessels; f CSC differentiate into ECs or PCs: cancer stem cell differentiate into endothelial cells or pericytes. CSCs cancer stem cells, ECs endothelial cells, EPCs endothelial progenitor cells, PCs pericytes.

Angiogenesis

Angiogenesis is the traditional process by which new blood vessels form from preexisting vessels through sprouting. Initially, endothelial cells (ECs) loosen their junctions, increasing permeability and releasing plasma proteins. Subsequently, ECs sprout, with tip cells penetrating the basement membrane, and an imbalance between matrix metalloproteinase (MMP) and tissue inhibitor of metalloproteinase (TIMP) leads to extracellular matrix (ECM) degradation. Finally, ECs proliferate and migrate, accompanied by pericyte recruitment, resulting in the formation of new blood vessels [7, 18].

Vasculogenic mimicry (VM)

The classical theory of tumor angiogenesis proposes that blood vessels are generated through ECs sprouting. However, in 1999, a paradigm shift occurred with the introduction of the concept of VM in a study focused on melanoma [19]. Unlike traditional angiogenesis, VM is independent of ECs which forms vessel-like structures by tumor cells [8]. Subsequent studies revealed that VM occurs not only in melanoma but also in glioma, hepatocellular carcinoma (HCC), and prostate cancer [20–22]. Mechanistically, VM involves the epithelial-mesenchymal transition (EMT) process and differentiation of CSCs [23, 24]. During VM, epithelial cell markers like E-cadherin are downregulated, whereas mesenchymal cell markers like VE-cadherin, vimentin are upregulated [24, 25].

m⁶A components

m⁶A writers

m⁶A writers are methyltransferases responsible for installing m⁶A and modifications on RNA.The core components include methyltransferase-like 3 (METTL3), methyltransferase-like 14 (METTL14), and Wilms tumor 1-associated protein (WTAP). METTL3 and its homolog METTL14 form an asymmetric heterodimer in a 1:1 ratio [26]. METTL3 functions as the catalytic subunit, whereas METTL14 acts as an allosteric activator, enhancing METTL3’s catalytic activity by providing an RNA-binding scaffold [27]. With the assistance of WTAP, an adapter subunit, the METTL3- METTL14 complex is located in nuclear speckles [28]. Additionally, RNA-binding motif protein 15 (RBM15) and its paralog RBM15B interact with WTAP-METTL3, recruiting it to proximal m⁶A consensus sites [29]. Other writer components, including vir-like m⁶A methyltransferase associated (VIRMA), zinc finger CCCH-type containing 13 (ZC3H13), and cbl proto-oncogene like 1 (CBLL1), also interact with WTAP [15]. Among them, VIRMA mediates m⁶A in the 3′ untranslated region (3′UTR) and near the stop codon by acting as a scaffold to hold WTAP/CBLL1/ZC3H13 together [30]. Methyltransferase-like 16 (METTL16) is an independent RNA methyltransferase that installs m⁶A on molecules such as U6 small nuclear RNA (snRNA), methionine adenosyltransferase 2A (MAT2A) mRNA, and metastasis-associated lung adenocarcinoma transcript 1(MALAT1) [31].

In various tumor types, METTL3 typically functions as an oncogenic factor, including acute myeloid leukemia (AML), lung cancer, HCC, among others, while exhibiting an anti-oncogenic role in endometrial cancer [32]. Additionally, METTL3 is implicated in therapy resistance across cancers. For instance, it promotes chemoresistance in HCC, breast cancer, and colorectal cancer (CRC), and reduces the effectiveness of immunotherapy in CRC and melanoma [33]. However, despite METTL14 lacking catalytic activity and functions as an allosteric activator of METTL3, its impact on tumors may not consistently align with METTL3. METTL14 exerts an oncogenic effect in HCC and AML, but suppresses tumor progression in CRC and renal cell carcinoma (RCC) [32, 34]. Additionally, decreased METTL14 has been reported to promote radiotherapy resistance in esophageal squamous cell carcinoma (ESCC) and sorafenib resistance in HCC [33]. This discrepancy could arise from the distinct target sites methylated by each protein within the writer complex. m⁶A methylation is preferentially enriched at DRACH (D = A, G, or U; R = G or A; H = A, C or U) motifs, suggesting numerous potential methylation sites exist within transcripts. However, the writer complex exhibits site- and transcript-specific selectivity, which may be influenced by transcription factors or histone marks that recruit the writer complex to specific genomic loci [15, 35].

m⁶A erasers

The reversible nature of m⁶A modification is regulated by m⁶A erasers, which are limited in specific tissues and are context-dependent [35]. m⁶A demethylation is catalyzed by two enzymes, fat mass and obesity-associated (FTO) and AlkB homolog 5 (ALKBH5), both part of the ALKB enzyme family [36]. FTO is the first identified m⁶A eraser that is associated with obesity and energy homeostasis [37]. Initially, FTO was identified as a nucleus m⁶A demethylase, but later studies suggested its primary target might be N6,2′-O-dimethyladenosine (m⁶A_m) rather than m⁶A [37]. m⁶A typically occurs at internal sites within mRNA, whereas m⁶A_m is located near the 5′cap structure. Subsequent studies indicated that FTO regulates m⁶A_m modification of snRNA, thereby affecting the alternative splicing of mRNA [38]. Despite a lower response rate, FTO can demethylate m⁶A and exhibit an oncogenic role. In AML, FTO is abnormally localized in the cytoplasm and exerts oncogenic effects by altering m⁶A demethylation [39]. ALKBH5, another m⁶A demethylase, is highly expressed in the testis, lungs, and germ cells, but weakly in cardiac and cerebral tissues [40, 41]. Unlike FTO, ALKBH5 has no activity toward m⁶A_m and can be induced under hypoxia conditions in tumors, promoting self-renewal of glioblastoma and breast CSCs [42].

m⁶A readers

m⁶A readers are RNA-binding proteins that specifically recognize and bind to m⁶A-modified RNA molecules. The YTH domain-containing proteins, including YTH domain-containing 1–2 (YTHDC1-2) and YTH domain family 1–3 (YTHDF1–3), were the first identified m⁶A readers. YTHDC1 promotes RNA splicing and nuclear export, and YTHDC2 weakly binds m⁶A and promotes mRNA translation and degradation. The YTHDF proteins have three highly similar paralogues: YTHDF1 enhances mRNA translation, YTHDF2 promotes mRNA degradation, and YTHDF3 performs both functions [43]. Additionally, m⁶A can indirectly recruit RNA-binding proteins by remodeling RNA structure, a phenomenon known as the“m⁶A-switch” [15]. The insulin-like growth factor 2 mRNA binding protein (IGF2BP) family and heterogeneous nuclear ribonucleoprotein (HNRNP) family belong to this category. IGF2BP1-3 promote mRNA stability, HNRNPC/G facilitate RNA splicing, and HNRNPA2B1 promotes both RNA splicing and degradation [15].

m⁶A and tumor neovascularization

m⁶A and angiogenesis

Among m⁶A writers, METTL3 and METTL14 are primarily investigated for their roles in tumor angiogenesis (Table 1). Firstly, they facilitate the expression of angiogenic factors in various cancers. For example, METTL3 directly promotes the expression of hypoxia-inducible factor 1-alpha (HIF-1α), VEGFA and tyrosine kinase (TEK) in bladder cancer (BLCA) [44, 45]. Similarly, METTL14 induces the expression of basic leucine zipper ATF-like transcription factor 2 (BATF2), which indirectly upregulates VEGFA secretion and promotes angiogenesis in tongue squamous cell carcinoma (TSCC) [46]. Beyond VEGFA, METTL3 also regulates ECM components to modulate angiogenesis [47]. In HCC and prostate cancer, METTL3 stimulates angiogenesis by increasing the expression of MMP2 and MMP9 [48–50]. Moreover, in CRC, METTL3 enhances plasminogen activator, urokinase (PLAU) expression, which activates angiogenic factors stored in the ECM, thereby facilitating angiogenesis [51, 52]. Additionally, METTL3 regulates cell cycle-associated proteins. In gastric cancer (GC) and in head and neck squamous cell carcinoma (HNSCC), METTL3 upregulates centromere protein F (CENPF), ensuring an adequate blood supply for rapidly dividing tumor cells [53, 54]. METTL3 and METTL14 also affect metabolism and inflammation, indirectly regulating angiogenesis. In GC, METTL3 increases hepatoma-derived growth factor (HDGF) expression, promoting glycolysis, which subsequently contributes to angiogenesis and liver metastasis [55]. In RCC, METTL14 activates TNF receptor-associated factor 1 (TRAF1), and indirectly facilitates angiogenesis [56].

Table 1.

Tumor neovascularization regulated by m⁶A methylation in different types of cancer.

Cancer type	m6A regulator	Target molecules	Effect on tumor neovascularization	Types of tumor neovascularization	Reference
Breast cancer	YTHDF3	EGFR and VEGFA	Positive	Angiogenesis	PMID: 33125861
BLCA	METTL3	VEGFA and TEK	Positive	Angiogenesis	PMID: 33681207
	METTL3	BIRC5	Positive	Angiogenesis	PMID: 35749893
	ALKBH5	lncBLACAT3	Negative	Angiogenesis	PMID: 37612524
CRC	METTL3	PLAU	Positive	Angiogenesis	PMID: 35567945
	METTL3	VEGFA and EphA2	Positive	VM	PMID: 35595748
	WTAP	VEGFA	Positive	Angiogenesis	PMID: 37428639
	ALKBH5	circ3823	Negative	Angiogenesis	PMID: 34172072
	YTHDF3	circ3823	Negative	Angiogenesis	PMID: 34172072
	IGF2BP2	Cyclin D1	Positive	Angiogenesis	PMID: 36230970
	IGF2BP3	Cyclin D1 and VEGF	Positive	Angiogenesis	PMID: 32993738
GC	METTL3	CENPF	Positive	Angiogenesis	PMID: 37256823
	METTL3	HDGF	Positive	Angiogenesis	PMID: 31582403
	METTL3	ADAMTS9	Positive	Angiogenesis	PMID: 35574388
	IGF2BP3	HIF-1α	Positive	Angiogenesis	PMID: 34621671
Glioma (including GBM)	METTL3	HOTAM1	Positive	VM	PMID: 36086906
	METTL3	BUD13	Positive	VM	PMID: 36463205
	METTL3	MMP2, CDH1, CDH2,FN1	Negative	VM	PMID: 35261810
HCC	METTL3	YAP1	Positive	VM	PMID: 32920668
	METTL3	FOXO3	Negative	Angiogenesis	PMID: 32368828
	YTHDF2	IL11 and SERPINE2	Negative	Angiogenesis	PMID: 31735169
HNSCC	METTL3	CDC25B	Positive	Angiogenesis	PMID: 35287752
ICC	FTO	TEAD2	Negative	Angiogenesis	PMID: 31143705
Lung cancer	METTL3	VEGFA	Positive	Angiogenesis	PMID: 37103476
	IGF2BP2	FLT4	Positive	Angiogenesis	PMID: 37353784
	IGF2BP2	TK1	Positive	Angiogenesis	PMID: 33758932
	YTHDC2	lncZNRD1-AS1	Negative	Angiogenesis	PMID: 36581942
MM	ALKBH5	SAV1	Positive	Angiogenesis	PMID: 35414790
Pancreatic cancer	METTL3	lncLIFR-AS1	Positive	Angiogenesis	PMID: 34658294
RCC	METTL14	TRAF1	Positive	Angiogenesis	PMID: 35538475
	FTO	VHL	Positive	Angiogenesis	PMID: 32817424
	YTHDF2	circPOLR2A	Negative	Angiogenesis	PMID: 35840930
TSCC	METTL14	BATF2	Positive	Angiogenesis	PMID: 35949109

BLCA bladder cancer, CRC colorectal cancer, GBM glioblastoma, GC gastric carcinoma, HCC hepatocellular carcinoma, HNSCC head and neck squamous cell carcinoma, ICC intrahepatic cholangiocarcinoma, m⁶A N6-methyladenosine, MM multiple myeloma, RCC renal cell carcinoma, TSCC tongue squamous cell carcinoma, VM vasculogenic mimicry.

The regulation of tumor angiogenesis by m⁶A erasers varies across different cancer types. In multiple myeloma (MM), ALKBH5 promotes angiogenesis by elevating salvador family WW domain-containing protein 1 (SAV1) expression and activating the Hippo pathway [57]. Conversely, in BLCA, ALKBH5 suppresses angiogenesis and hematogenous metastasis by inhibiting lncBLACAT3 expression, which inactivates the NF-κB pathway [58]. In RCC, FTO promotes angiogenesis by inhibiting von Hippel-Lindau tumor suppressor (VHL) expression, whereas in intrahepatic cholangiocarcinoma (ICC), FTO suppresses angiogenesis by inducing TEA domain transcription factor 2 (TEAD2) expression [59, 60]. These findings demonstrate that the effects of m⁶A erasers on angiogenesis are specific to the cancer type.

Correlations between tumor angiogenesis and m⁶A readers have been observed in both the IGF2BP and YTH domain-containing proteins, which exhibit opposing functions. The IGF2BP family exerts a pro-angiogenic effect by enhancing the stability of downstream genes. For instance, in lung cancer, IGF2BP2 increases the stability of fms-related tyrosine kinase 4 (FLT4; also known as VEGFR3) or thymidine kinase 1 (TK1) mRNA, leading to tumor angiogenesis and aggressiveness [61, 62]. Similarly, in CRC, IGF2BP2 and IGF2BP3 stabilize cyclin D1 and VEGF mRNA, thereby promoting angiogenesis [63, 64]. In contrast, the YTH domain-containing proteins have been demonstrated to suppress tumor angiogenesis. In lung cancer, YTHDC2 enhances the translation efficiency of lncZNRD1-AS1, further suppressing angiogenesis and tumorigenesis through the miR-942/Tensin 1 axis [65]. Additionally, YTHDF2 inhibits angiogenesis in clear cell RCC (cRCC) and HCC by facilitating the degradation of target genes. In cRCC, YTHDF2 inhibits angiogenesis by increasing circPOLR2A degradation [66]. In HCC, it increases the degradation of IL11 and serpin family E member 2 (SERPINE2) and thus contributing to vascular normalization [67].

m⁶A and vasculogenic mimicry

The well-studied m⁶A regulator METTL3, is associated with vasculogenic mimicry (VM) in CRC, glioma, and HCC [68–70]. In CRC, METTL3 promotes VM indirectly by targeting Eph receptor A2 (EphA2) and VEGFA, enhancing their stability through IGF2BP2 and IGF2BP3, respectively [69]. Elevated METTL3 levels in glioma contribute to VM through targeting HOXA transcript antisense RNA myeloid-specific 1 (HOTAIRM1) [70]. Similarly, in HCC, inhibition of METTL3 impairs VM-related tumor vasculature formation, indicating a positive correlation between m⁶A levels and VM [68]. These findings suggest that METTL3 regulates target genes at the translational level, ultimately inducing VM.

However, the effect of METTL3 on VM in glioblastomas (GBM) appears to be different. One study found that METTL3 enhances the stability of BUD13 homolog (BUD13), promoting the translation of cyclin-dependent kinase 12 (CDK12) and muscleblind-like splicing regulator 1 (MBNL1). This cascade results in the upregulation of MMP2 and laminin subunit gamma 2 (LAMC2), promoting VM [71]. Conversely, another study indicates that reduced METTL3 facilitates VM and is correlated with higher histopathological grade and lower overall survival [72]. The contradictory results may be attributed to differences in cell line selection and sample size. Further research is needed to clarify the specific role of m⁶A on VM in different tumors.

m⁶A and stemness-associated factors

m⁶A and Oct4

Octamer-binding transcription factor 4 (Oct4), a member of the POU transcription factor family, is essential for stemness maintenance and differentiation of CSCs [73]. Recent studies have also linked it to angiogenesis [74]. Our previous research demonstrated that Oct4 regulates the differentiation of liver CSCs into tumor ECs [75]. YTHDF2 interacts with the 5′UTR of Oct4 mRNA to increase its expression, thus maintaining stemness and promoting lung metastasis in HCC [76]. ALKBH5 is positively correlated with Oct4 in MM and non-small cell lung cancer. Suppression of ALKBH5 reduces Oct4 expression and inhibits CSC characteristics, suggesting that ALKBH5 may induce neovascularization through CSC-derived vasculogenesis manner in these tumors [57, 77].

m⁶A and Sox2

SRY-box transcription factor 2 (Sox2) is a transcription factor that is essential for the self-renewal and pluripotency of stem cells. It has been demonstrated that Sox2 is capable of promoting tumor neovascularization. In ESCC, Sox2 promotes angiogenesis by inducing suprabasin expression [78]. Furthermore, it contributes to VM in CRC [79]. On the other hand, tumor neovascularization increases Sox2 expression, which in turn helps to maintain the CSC phenotype. In skin tumors, CSCs are found in proximity to ECs and reside within a perivascular microenvironment [80]. Besides, in retinoblastoma, VEGF has been found to stimulate Sox2 expression and enhance tumor invasiveness [81]. These findings demonstrate a reciprocal relationship between Sox2 and tumor neovascularization, where both factors reinforce the aggressive behavior of the tumor.

Sox2 is a downstream target of METTL3, with IGF2BP2 recognizing methylated Sox2 transcripts and preventing their degradation [82, 83]. METTL3 sustains Sox2 expression through an m⁶A-mediated mechanism, thereby maintaining stemness and metastasis in CRC [82]. In GBM, METTL3 binds to the 3′UTR of Sox2 mRNA, thus maintaining stemness and radioresistance [83]. Additionally, ALKBH5 has been reported to facilitate Sox2 expression in lung cancer, MM and endometrial cancer [57, 84, 85]. For instance, in lung cancer, ALKBH5 counteracts YTHDF2-mediated degradation of Sox2 and thereby promoting tumor aggressiveness [84].

m⁶A and tumor neovascularization-associated pathways

m⁶A and VEGF

VEGF, an extensively studied angiogenic factor, is produced by various cell types. The VEGF family consists of six members, with VEGFA being the most critical for tumor neovascularization. VEGFR are categorized into three subtypes, with VEGFR1 and VEGFR2 being the most prevalent in vascular ECs. VEGFR1 is responsible for hematopoiesis, and VEGFR2 is involved in vasculogenesis and angiogenesis. VEGFR3, mainly expressed in lymphatic ECs, is associated with lymphangiogenesis [86]. When VEGF binds to VEGFR, it triggers TEK phosphorylation, activating intracellular signaling pathways that regulate the proliferation, migration, survival, and penetration of vascular ECs, ultimately leading to neovascularization [18]. Moreover, VEGF infulences tumor neovascularization thorugh various downstream signaling pathways, including PI3K/AKT, MAPK, PLC, and SRC [87].

In lung cancer, METTL3 binds to the A859 site within the internal ribosome entry site of VEGFA 5’UTR, recruiting YTHDC2/eIF4GI complex to promote VEGFA translation and increase its expression [88]. This promoting effect of METTL3 on VEGFA is also observed in CRC, pancreatic cancer and BLCA [45, 69, 89]. However, in sorafenib-resistant HCC, METTL3 exerts an opposite effect. Depletion of METTL3 increases the expression of VEGFA and other angiogenic factors [90]. This may result from the alterations in the recognition site of the m⁶A-modified target gene following drug resistance, which affecting their binding capacity. In RCC, FTO is mutually exclusive with VHL. Elevated FTO inhibits VHL expression and increases VEGFA secretion [59]. IGF2BP3 has been reported to promote angiogenesis by interacting with VEGFA in CRC and GC [63, 69, 91]. Notably, it is IGF2BP3, rather than other IGF2BP members, that specifically binds to VEGFA [69]. Further research is required to elucidate the selectivity of m⁶A readers.

m⁶A and EGFR

Epidermal growth factor receptor (EGFR) is a membrane receptor on the surface of epidermal cells that belongs to the tyrosine kinase receptor family. Upon activation by EGF, it initiates tyrosine kinase activity and activates downstream signaling such as PI3K/AKT, MAPK, and JAK/STAT pathways, which regulate various biological processes [92].

In breast cancer with brain metastases, YTHDF3 enhances the translation of EGFR and VEGFA mRNA by binding to eukaryotic translation initiation factor 3 subunit A (eIF3a). Suppressing YTHDF3 reduces blood vessel density and impairs brain endothelial tube formation, thereby decreasing brain metastasis and prolonging survival [93]. Under hypoxic conditions, YTHDF2 is downregulated in HCC. However, when overexpressed, it binds to the 3’UTR of the EGFR mRNA and accelerates its degradation. Consequently, this process suppresses the MAPK/ERK pathway, thereby inhibiting cell proliferation and tumor growth [94].

m⁶A and PI3K/AKT

The PI3K/AKT signaling pathway is crutial in cancer development and progression. PI3K consists of a regulatory subunit (p85) and a catalytic subunit (p110). When activated, PI3K converts PIP2 to PIP3, subsequently activating PDK1 and AKT, while PTEN can counteract these effects. Upon activation, AKT stimulates mTOR, which in turn phosphorylates downstream substrates and regulates diverse biological processes [95]. The PI3K/AKT pathway is widely involved in tumor neovascularization. For instance, the inactivation of the p110 subunit impedes functional vessel formation, thereby restraining tumor growth. Activated AKT in tumor ECs results in an increase in nitric oxide levels, forstering vascular permebility. Conversely, the deletion of PTEN delays pericyte maturation, resulting in defective vascular remodeling [96–98].

In CRC, METTL3 promotes VM by activating the PI3K/AKT and ERK1/2 pathways [69]. Similarly, in GC, METTL3 promotes tumor angiogenesis through the ADAM metallopeptidase with thrombospondin type 1 motif 9 (ADAMTS9)-mediated PI3K/AKT pathway [99]. In pancreatic cancer, METTL3 increases the stability of lncLIFR-AS1 and indirectly promotes VEGFA expression. This, in turn, activates the AKT/mTOR pathway and further promotes tumor progression [89]. Using a bioinformatics database, Chen et al. found that METTL3 also regulates the PI3K/AKT pathway in BLCA, silencing METTL3 exerts an inhibitory effect on angiogenesis [45]. Furthermore, METTL14 is upregulated in sunitinib-resistant RCC compared to sensitive ones. It increases TRAF1 mRNA stability in an IGF2BP2-dependent manner, activating the AKT/mTOR/HIF-1α pathway and facilitating angiogenesis [56]. Similarly, in lung adenocarcinoma, IGF2BP2 upregulation in metastatic subpopulations is associated with a poor prognosis. It enhances FLT4 mRNA stability and activates the PI3K/AKT pathway, thereby promoting angiogenesis [61].

m⁶A and MAPK

MAPK, a serine-threonine protein kinase, regulates biological processes such as cell proliferation, differentiation, and migration. The MAPK family comprises ERK, p38, JNK, and BMK1, which represent four distinct MAPK pathways [100]. Among these, the Ras/Raf/MEK/ERK pathway has been extensively investigated and is closely linked to cancer [101].

Phosphatidylethanolamine binding protein 1 (PEBP1) inhibits the Raf/MEK/ERK pathway by binding to Raf and disrupting the Raf/MEK complex. In cRCC, YTHDF2 increases PEBP1 expression, leading to ERK pathway inactivation. Specifically, circPOLR2A facilitates the interaction between PEBP1 and ubiquitin protein ligase E3C (UBE3C), thereby promoting PEBP1 degradation. YTHDF2 negatively regulates circPOLR2A, resulting in ERK pathway inactivation and ultimately suppressing angiogenesis, metastasis, and tumor growth [66]. In GC, METTL3 upregulates CENPF expression through HNRNPA2B1. Elevated CENPF binds to focal adhesion kinase (FAK) and promotes its nuclear export, thereby activating the MAPK pathway and thus promoting angiogenesis and liver metastasis [53]. Similarly, METTL3, WTAP, and YTHDC1 are involved in MAPK pathway activation, enhancing tumor neovascularization in CRC [51, 69, 102].

m⁶A and Hippo

The Hippo pathway, comprising FZD2, LATS1/2, SAV1, and YAP/TAZ, is essential for regulating CSC maintenance, angiogenesis, and drug resistance. FZD2 has been reported to induce VM and maintain stemness in HCC, while YAP is associated with angiogenesis and VM in various cancers [103, 104]. Recent studies indicate that the Hippo pathway is involved in m⁶A-mediated tumor neovascularization, especially in VM.

YAP exhibits widespread association with VM and can be mediated by m⁶A methylation. In HCC, METTL3 promotes YAP mRNA splicing and enhances its translation efficiency, thereby facilitating VM [68]. In pancreatic cancer, inhibition of YTHDF2 increases YAP expression and promotes EMT [105]. Using verteporfin, a YAP inhibitor, neovascularization is suppressed, as evidenced by reduced levels of angiopoietin-2 (Ang2), MMP2 and VE-cadherin [106]. Given that EMT serves as the mechanism underlying VM, it is speculated that YTHDF2 may inhibit VM in pancreatic cancer by suppressing the Hippo pathway. Similarly, in CRC, IGF2BP2 binds to YAP mRNA to promote its translation, and verteporfin reduces the number of cancer-associated fibroblasts, inhibiting angiogenesis and tumor progression [107–109]. Apart form YAP, SAV1 has been implicated in promoting stem cell phenotype and neovascularization in MM. Inhibition of ALKBH5 decreases SAV1 mRNA stability, suppresses the Hippo pathway and angiogenesis [57]. Considering the intimate connection among CSC, EMT and VM, it is worthwhile to investigate the role of m⁶A in regulating tumor neovascularization, especially in VM through the Hippo pathway.

m⁶A and Wnt/β-catenin

The Wnt signaling pathway is important for regulating CSC maintenance and tumor metastasis, with three distinct pathways indentified: Wnt/β-catenin, Wnt/planar cell polarity, and Wnt/calcium. In the classical Wnt/β-catenin pathway, Wnt binds to the Frizzled receptor and activates the TCF/LEF transcription factor. Subsequently, TCF/LEF binds to β-catenin and promotes the transcription of various downstream genes [110].

In CRC, upregulated circ3823 functions as a competing endogenous RNA, disrupting the suppressive effect of miR-30c-5p on TCF7, consequently activating the Wnt/β-catenin pathway. Suppression of YTHDF3 or ALKBH5 increases circ3823 expression, thus promoting angiogenesis, metastasis, and tumor growth [111] (Fig. 2).

Fig. 2. m⁶A regulators in various tumor neovascularization-associated signaling pathways.

m⁶A “writers” (yellow circles), “erasers” (red circles), and “readers” (blue circles) selectively target specific signaling components, leading to the activation/inactivation of multiple intracellular signaling pathways (including Wnt/β-catenin, VEGFR, EGFR, Hippo, PI3K/AKT and MAPK pathways) associated with tumor neovascularization.

m⁶A in tumor vascular-immune crosstalk

Tumor angiogenesis creates an immunosuppressive microenvironment, aiding tumors in evading immune surveillance. Additionally, immunosuppressive cells can stimulate blood vessel formation, establishing abberrant communication between vascular and immune cells [112]. The synergistic effect of combining anti-angiogenic therapy with immunotherapy has demonstrated significant efficacy in treating various cancers, including RCC, HCC, GC, and endometrial cancer [113–116]. Recent studies highlight the role of m⁶A in regulating both tumor neovascularization and the immune microenvironment [117].

m⁶A-mediated tumor vasculature effects on immune cells

Immune cells extravasation into the TME requires adherence to ECs. However, ECs create an immune barrier by expressing programmed cell death-1 ligand 1 (PD-L1) and FAS ligand (FASL), as well as inhibiting adhesion factors such as intercellular adhesion molecule 1 (ICAM1), vascular cell adhesion molecule 1 (VCAM1), and P-selectin (Fig. 3) [118, 119]. Additionally, angiogenic factors like VEGF impede dendritic cells (DCs) maturation, impairing antigen presentation, suppressing tumor-specific cytotoxic T lymphocytes (CTLs) activation, or promoting the accumulation of immunosuppressive cells such as myeloid-derived suppressor cells (MDSCs) and regulatory T cells (Tregs) [120]. METTL3 and IGF2BP3 promote VEGF expression in various cancers, potentially contributing to the immunosuppressive TME [44, 63, 69, 88, 91].

Fig. 3. m⁶A regulators in vascular-immune crosstalk.

Tumor neovascularization contributes to the establishment of an immunosuppressive tumor microenvironment, while immunosuppressive cells facilitate tumor angiogenesis through the secretion of pro-angiogenic factors. Immune effector cells like CD8 + CTL, M1 TAM, mDC contribute to the establishment of an anti-angiogenic TME, whereas immunosuppressive cells, such as M2-like TAM, MDSC, and iDC, promote angiogenesis. Meanwhile, VEGF reduces the expression of endothelial adhesion molecules (ICAM1, P-selectin, E-selectin) expression, inhibits DC maturation and CTL activation, increases the abundance of MDSC, consequently impedes the infiltration of immune cells. Furthermore, reduced pericyte coverage hinders blood vessel integrity and immune infiltration. m⁶A regulators play a role in modulating immune-vascular crosstalk by influencing various components such as immune cells, vascular-associated cells, angiogenic factors, and endothelial adhesion molecules. CTL cytotoxic T lymphocyte, FASL FAS ligand, iDC immature DC, M1 TAM M1-like tumor-associated macrophage, M2 TAM, M2-like tumor-associated macrophage, mDC mature DC, MDSC myeloid-derived suppressor cell, PD-L1 programmed cell death-1 ligand 1, ICAM1 intercellular adhesion molecule 1, VEGF vascular endothelial growth factor.

Tumor blood vessels exhibit chaotic, leaky, and highly permeable characteristics, leading to increased interstitial fluid pressure that hampers immune cell infiltration. Inhomogeneous blood flow reduces perfusion and oxygenation, adversely impacting the delivery of anticancer drugs [121]. Pericytes are crucial for maintaining blood vessel structural integrity and have dual effects on tumor neovascularization. Generally, pericytes produce angiogenic factors that stimulate neovascularization [122, 123]. However, under certain conditions, increased pericyte coverage can normalize tumor vasculature [5]. In HCC, low levels of YTHDF2 due to hypoxia result in decreased pericyte coverage. Elevated YTHDF2 facilitates the degradation of IL11 and SERPINE2, reducing vascular density and permeability, thereby favorably impacting vascular normalization [67].

m⁶A-mediated immune cell effects on tumor neovascularization

Immune cells regulate tumor angiogensis through the secretion of cytokines and chemokines, while m⁶A modification indirectly modulates this process by recruiting and activating immune cells. Tumor-associated macrophages (TAMs) exhibit distinct M1 or M2 phenotype, with M1-like TAMs promoting inflammation and inhibiting angiogenesis, and M2-like TAMs displaying an immunosuppressive phenotype that fosters angiogenesis [124]. ALKBH5 promotes M2-like TAMs recruitment in glioma [125]. In pancreatic cancer, lncPACERR promotes M2-like TAMs polarization via IGF2BP2, thereby driving tumor progression [126]. The impact of DCs on tumor angiogenesis depends on their maturation status. Mature DCs (mDCs) inhibit tumor angiogenesis, while immature DCs (iDCs) exhibit a pro-angiogenic effect [112, 120]. In GC, YTHDF1 deletion increases mDCs recruitment, implying a pro-angiogenic role of YTHDF1 [127]. In addition, other immunosuppressive cells such as MDSCs, Tregs, and Tie2-expressing macrophages also contribute to tumor angiogenesis [9, 112]. In CRC, METTL3 recruits MDSCs and promotes tumor growth [128]. Consistently, in cervical cancer, METTL3 positively correlates with CD33⁺ MDSCs and predicts unfavorable prognosis [129]. Conversely, in ICC, ALKBH5 decreases MDSC-like cell accumulation, enhances PD-L1 expression, thus facilitating immune infiltration [130].

Considering adaptive immunity, CD8⁺ CTLs release IFN-γ to normalize blood vessels, while helper T (Th) cells, including Th1, Th2, and Th17 subsets, actively participate in angiogenesis by releasing chemokines and activating M2-like TAMs [9, 112]. In CRC and melanoma, deletion of METTL3 and METTL14 increases CD8⁺ CTL levels and enhances response to anti-PD1 therapy [131]. Similarly, silencing of YTHDF1 in GC increases CD8⁺ CTL and mDCs proportions, thereby restoring sensitivity to immunotherapy (Table 2) [127].

Table 2.

Fundamental functions of m⁶A regulators and their effect on tumor neovascularization.

Type	Regulator	Role	Regulation manner on tumor neovascularization	Reference
m6A writers	METTL3	Catalyze m6A modification	Promote angiogenesis in BLCA	PMID:35749893; PMID: 33681207
			Promote angiogenesis in CRC	PMID: 35567945
			Promote angiogenesis in GC	PMID:37256823; PMID: 31582403; PMID: 35574388
			Promote angiogenesis in HNSCC	PMID: 35287752
			Promote angiogenesis in lung cancer	PMID: 37103476
			Promote angiogenesis in pancreatic cancer	PMID: 34658294
			Inhibit angiogenesis in HCC	PMID: 32368828
			Promote VM in CRC	PMID: 35595748
			Promote VM in GBM	PMID: 36463205
			Promote VM in glioma	PMID: 36086906
			Promote VM in HCC	PMID: 32920668
			Inhibit VM in GBM	PMID: 35261810
			Recruit MDSCs in CRC	PMID: 35700773
			Recruit MDSCs in cervical cancer	PMID: 33059689
			Reduce CD8+CTL levels in CRC	PMID: 32964498
			Facilitate sorafenib resistance in HCC	PMID: 33222692; PMID: 32368828
			Facilitate apatinib resistance in HCC	PMID: 35503144
			Facilitate lenvatinib resistance in HCC	PMID: 36764493; PMID: 36898427; PMID: 36932115
	METTL14	Enhance the catalyze activity of METTL3	Promote angiogenesis in TSCC	PMID: 35949109
			Promote angiogenesis in RCC	PMID: 35538475
			Reduce CD8+CTL levels in CRC	PMID: 32964498
			Facilitate sunitinib resistance in RCC	PMID: 35538475
	WTAP	Help the localization of METTL3-METTL14 heterodimer	Promote angiogenesis in CRC	PMID: 37428639
			Facilitate lenvatinib resistance in HCC	PMID: 36932115
m6A erasers	FTO	Remove m6A modification	Promote angiogenesis in RCC	PMID: 32817424
			Inhibit angiogenesis in ICC	PMID: 31143705
	ALKBH5	Remove m6A modification	Promote angiogenesis in MM	PMID: 35414790
			Inhibit angiogenesis in CRC	PMID: 34172072
			Inhibit angiogenesis in BLCA	PMID: 37612524
			Recruit M2-like TAMs in glioma	PMID: 35444654
			Reduce MDSCs accumulation in ICC	PMID: 34301762
m6A readers	YTHDC1	Promote the splicing and nuclear export of RNA	Promote angiogenesis in CRC	PMID: 37428639
			facilitate sunitinib resistance in RCC	PMID: 35974388
	YTHDC2	Promote mRNA degradation and translation efficiency	Promote angiogenesis in lung cancer	PMID: 37103476; PMID: 36581942
	YTHDF1	Promote mRNA translation	Suppress mDCs recruitment and reduce CD8+CTLs proportion in GC	PMID: 35193930
	YTHDF2	Promote mRNA degradation	promote vascular normalization in HCC	PMID: 31735169
			inhibit angiogenesis in RCC	PMID: 35840930
			inhibit VM in RCC	PMID: 37037853
			Suppress pazopanib resistance in RCC	PMID: 37037853
	YTHDF3	Promote the translation and degradation of mRNA	Promote angiogenesis in breast cancer with brain metastases	PMID: 33125861
			Inhibit angiogenesis in CRC	PMID: 34172072
	IGF2BP2	stabilize mRNA and promote its translation	Promote angiogenesis in CRC	PMID: 36230970
			Promote angiogenesis in lung cancer	PMID: 37353784; PMID: 33758932
			Promote angiogenesis in RCC	PMID: 35538475
			Promote VM in CRC	PMID: 35595748
			Enhance M2-like TAMs polarization in pancreatic cancer	PMID: 35526050
			Facilitate sunitinib resistance in RCC	PMID: 35538475
	IGF2BP3	Stabilize mRNA and promote its translation	Promote angiogenesis in CRC	PMID: 32993738
			Promote angiogenesis in GC	PMID: 34621671
			Promote VM in CRC	PMID: 35595748
	HNRNPA2B1	Promote RNA splicing and degradation	Promote angiogenesis in GC	PMID: 37256823

BLCA bladder cancer, CRC colorectal cancer, CTL cytotoxic T lymphocyte, GBM glioblastoma, GC gastric carcinoma, HCC hepatocellular carcinoma, HNSCC head and neck squamous cell carcinoma, ICC intrahepatic cholangiocarcinoma, m⁶A N6-methyladenosine, mDC mature dendritic cell, MDSC myeloid-derived suppressor cell, MM multiple myeloma, RCC renal cell carcinoma, TAM tumor-associated macrophage, TSCC tongue squamous cell carcinoma, VM vasculogenic mimicry.

m⁶A methylation in cancer therapy

Clinical treatment prospects of new strategies

The advancement of targeted therapy strategies focused on m⁶A regulators is rapidly progressing, driven by their crucial regulatory functions and precise recognition abilities. Dysregulation of enzymes and binding proteins involved in RNA methylation has been linked to various human cancers, suggesting a promising novel approach for clinical intervention [34].

m⁶A inhibitors

Recent studies indicate that inhibitors targeting RNA methylation regulatory factors have potential in cancer therapy. For instance, SAH and the broad-spectrum 2-OG oxygenase inhibitor IOX1 inhibit cancer development by targeting METTL3-METTL14 and ALKBH5, respectively [132]. Additionally, inhibitors like FB23-2, FG-2216/IOX3, Rhein, Entacapone, and meclofenamic acid inhibit FTO activity, preventing the self-renewal and tumorigenic properties of AML and GBM [132–134]. In mouse xenograft models, FTO inhibition resulted in reduced tumor growth and prolonged survival.

Clinical trials and preclinical research

Ongoing clinical trials of m⁶A emphasize the potential of targeting m⁶A modifications in cancer treatment. STC-15, as an oral small molecule inhibitor targeting METTL3, is a first-in-class inhibitor of RNA modification. In November 2022, it advanced into Phase I clinical trials, representing the first RNA methyltransferase inhibitor to enter clinical development. STC-15 has demonstrated efficacy in suppressing AML or suppress tumor growth through anticancer immune responses, promising a novel avenue for cancer therapy [135].

Another promising candidate is STM2457, a highly selective METTL3 inhibitor with minimal effects on other methyltransferases, indicating its potential as a targeted cancer therapy. Preclinical research indicates that STM2457 specifically inhibits key stem cell populations in AML without significant toxicity to normal haematopoiesis, promoting cell differentiation, inducing apoptosis, and inhibiting tumor growth [136]. These findings support targeting m⁶A modification as a promising strategy for anticancer therapy.

Potential of post-transcriptional modifications as biomarkers

Post-transcriptional modifications (PTMs) have been demonstrated to be associated with various human diseases. RNA modifications can be potential biomarkers for monitoring cancer progression through regulating mRNA stability, translation efficiency, and other RNA metabolic processes. Distinct modification patterns in different cancer types influence tumor cell behaviors such as proliferation, apoptosis, invasiveness, and metabolic activity, which makes them important indicators in cancer research. For instance, the m⁵C-based signature is an independent prognostic factor associated with immunotherapy efficacy and drug susceptibility in RCC [137]. The interaction network of m⁶A/m⁵C/m¹A regulated genes is reported to assess the prognosis of HCC [138]. These imply the potential of RNA modification in clinical applications.

Clinically, detecting m⁶A often requires high-throughput sequencing (such as MeRIP-seq) and specific immunoprecipitation techniques [15]. These methods help clinicians analyze m⁶A levels in samples to assess tumor status and treatment response. As biomarkers, m⁶A modifications have the advantage of dynamically reflecting biological changes within tumor cells, allowing real-time monitoring of tumor progression and treatment response through non-invasive samples (such as blood) [139]. For instance, in gastrointestinal cancer, m⁶A levels were elevated compared to adjacent tissues and the serum of healthy individuals, and decreased post-surgery [140]. However, the dynamic and variable nature of m⁶A and other RNA modifications can complicate result interpretation. Moreover, the need for specialized techniques and equipment limits the widespread clinical application of m⁶A as a biomarker. Overall, research on RNA modifications as potential cancer biomarkers is still in its early stages, requiring further study to overcome current technical barriers and enhance their clinical accuracy and feasibility.

Conclusions and perspectives

This review emphasizes the significant role of m⁶A modifications in tumor neovascularization. We aim to explore how m⁶A influences various modes of neovascularization and its interactions with multiple signaling pathways and components of the TME. By inhibiting key m⁶A writers or erasers, it may be possible to suppress pro-angiogenic factors, thereby reducing tumor vascularization and growth. This approach could potentially enhance the efficacy of existing treatments, such as VEGF inhibitors.

Although studies using m⁶A inhibitors to directly target tumor neovascularization are limited, growing evidence suggests that m⁶A is widely involved in anti-angiogenic resistance. Specifically, in HCC and RCC, both of which are characterized by high vascular density. Increased m⁶A levels contribute to resistance against sorafenib, apatinib, and lenvatinib in HCC. Under normoxic conditions, METTL3 increases resistance to sorafenib and lenvatinib through the Wnt/β-catenin pathway, while WTAP facilitates lenvatinib resistance under hypoxic conditions. Knockdown of METTL14 restores sensitivity to sorafenib by upregulating hepatocyte nuclear factor 3 gamma (HNF3γ). In RCC, downregulation of METTL14 enhances sensitivity to sunitinib by reducing TRAF1 expression. YTHDC1 increases sensitivity to sunitinib by inhibiting histone deacetylase 2 (HDAC2). Conversely, in pazopanib-resistant cRCC, YTHDF2 fails to recognize lncIGFL2AS1 for degradation, promoting drug resistance. These findings suggest targeting m⁶A might provide a novel approach to anti-angiogenic therapy. In conclusion, the regulatory role of m⁶A in tumor neovascularization is critical and warrants further investigation to develop potential treatment strategies. Additional research is needed to fully understand its clinical potential, especially its interactions with various pathways and immune cells. Despite the limited studies targeting tumor neovascularization with m⁶A inhibitors, growing evidence suggests that this approach may offer promising therapeutic potential.

Author contributions

CX and HLL conceptualized the paper, HLL gave the outline of the paper. LZ searched and sorted out relevant literatures, LZ and QSL co-wrote the manuscript. YY, XXC, and SJZ each created Figs. 1–3, respectively, and QSG enhanced their presentation. TLZ prepared Table 1, XL and JG revised Table 1 in the revised manuscript. QF and TTD jointly prepared and revised Table 2. YJ polished the language of the article, XW and MXZ revised the manuscript. HDT supervised the revised manuscript and LZ wrote the final version of the manuscript.

Funding

This study was financially supported by the National Natural Science Foundation of China (Grant numbers: 82173378, U22A20374, 81960476), National High Level Hospital Clinical Research Funding (No. 2023-NHLHCRF-DJMS-06), and Elite Medical Professionals Project of China-Japan Friendship Hospital (No. ZRJY2021-TD02).

Competing interests

The authors declare no competing interests.