Introduction
Colorectal cancer (CRC) ranks third in incidence and second in mortality. Although tremendous efforts have been made to facilitate screening strategies, the prevalence of CRC has been increasing, and 1.9 million new cases were estimated in 2020. Among them, a large population is diagnosed at an advanced stage; nevertheless, the efficacy of current therapies for late-stage CRC is limited [
1]. Thus, a deeper understanding of the molecular mechanisms of CRC might help to bring novel strategies to CRC prevention and therapy.
While DNA-related aberrations are widely studied in the literature, post-transcriptional alterations have been relatively less studied. Yet, these changes play a critical role in regulating the initiation and progression of CRC [
2,
3]. Among these post-transcriptional alterations, N6-methyladenosine (m
6A) is the most common internal modification in transcripts, comprising more than 50% of eukaryote methylations now identified as a new level of crucial epigenetic regulation for mRNA stability, splicing, and translation [
4], as well as the generation of small and long non-coding RNAs [
5]. Dynamic and reversible m
6A methylation consists of readers (m
6A-binding proteins), writers (methyltransferases), and erasers (demethylases) responsible for m
6A's functions, methylation, and demethylation, respectively [
6]. Importantly, m
6A mRNA regulation plays a vital role in tumorigenesis, tumor development, and metastasis, and the dysregulation of m
6A is closely correlated with the development and pathogenesis of CRC [
7‐
9].
Nowadays, emerging evidence has proven the critical role of tumor microenvironment (TME) in the progress and metastasis of tumor. In addition, single-cell transcriptomics further revealed the complex intercellular communication between diverse subtypes of TME cells and tumor cells [
10,
11]. The TME cells consists of multiple cell types in addition to tumor cells, such as cancer-associated fibroblasts (CAFs), tumor-associated macrophages (TAMs), T cells, and B cells. Recently, Huilong et al. discovered that ablation of Mettl3 in myeloid cells promotes tumor growth and metastasis in vivo [
12]. Dali et al. reported that loss of YTHDF1 in dendritic cells can lead to enhanced cross-presentation of tumor antigens and the cross-priming of CD8
+ T cells in vivo [
13]. However, less research was performed to investigate the cell–cell interaction between m
6A mRNA modification associated subtypes of TME cells with tumor cells.
Here, we investigated the influence of m
6A mRNA methylation on the main TME cells, including stromal cells, myeloid cells, T cells, and B cells, based on 65,362 single-cell sequencing data derived from 33 CRC tumor samples [
14]. By nonnegative matrix factorization (NMF) clusters of 23 m
6A RNA methylation regulators, as described previously [
15,
16], it was observed that different patterns of m
6A mRNA methylation in each CRC TME cell type subpopulations manifested extensive and diversity communication with tumor epithelial cells, and associated with different immune characteristics, metabolic pathways, transcription characteristics and prognosis. To the best of our knowledge, the present study reveals, for the first time, that m
6A mRNA may guide intercellular communication of TME cells with tumor cells to contribute to colorectal cancer progression based on our comprehensive single-cell analysis.
Discussion
To date, several studies have revealed the correlation between RNA m
6A modification and the pathogenesis of colorectal cancer [
36‐
40]. However, only a few have investigated the potential tumorigenic role of m
6A-modified single cells. In the present study, we, for the first time, have comprehensively explored m
6A modification regulators of main cell types in the TME of colorectal cancer and further identified the diversity cell–cell interaction between m
6A associated TME cell subtypes and tumor cells at the 10X Genomic single-cell sequence level. This unique and new perspective allowed us to understand how RNA m
6A modification of these diverse cellular components of TME affects the fate of individual CRC patients.
Cancer epithelial cells constitute the majority of tumor tissue and drive tumor development. Meanwhile, the heterogeneity of cancer epithelial cells indicates the responsiveness of patients to treatment and determines prognosis. Besides cancer epithelial cells, TME cells, such as multiple types of stromal cells, vascular endothelial cells and infiltrating immune cells, all support growth and promote immune evasion of solid tumors. In the study, we found that the TME cells, including stromal cells, macrophages, T cells and B cells all manifested the diverse m6A regulatory patterns and the extensive communication with tumor epithelial cells based on the single-cell analysis. Furthermore, Cellphone analysis showed that ligand-receptor pairs, including MIF − (CD74 + CXCR4), MIF − (CD74 + CD44), MDK–NCL, LGALS9 − CD45, CLEC2D − KLRB1, CLEC2C − KLRB1, CLEC2B − KLRB1, APP − CD74, CD99 − CD99, and ADGRE5 − CD55 mediated the communication between m6A associated subtypes of TME cells and tumor epithelial cells.
Cancer-associated fibroblasts (CAFs), as one of the critical components of stromal cells, were classified as pan-myCAFs, pan-dCAFs, pan-iCAFs, pan-nCAFs, and pan-pCAFs, according to specific molecular characteristics [
35]. To date, few studies have revealed the potential role of RNA m
6A modification in CAFs. In our study, we found that m
6A-mediated fibroblasts manifested more extensive communications with tumor epithelial cells compared with non-m
6A-mediated fibroblasts. Furthermore, WTAP-fibroblasts had a strong relationship with inflammatory-CAFs and the elevated expression of proinflammatory factors, such as CXCL1, CXCL2, CXCL3, CCL2, IL-6 and IL-7. Pathway analysis also revealed the participation of CAFs in IL-17 signaling pathway, TNF signaling pathway and the innate immune associated pathways. CAFs may shape an immunosuppressive microenvironment through the secretion of CXCL1, IL6 and CCL2 [
41‐
44]. Therefore, we speculated that RNA m
6A modification CAFs may form the immunosuppressive interaction with tumor cells to promote the progress and metastasis of tumor.
Nowadays, increasing research has revealed the significant role of RNA m
6A methylation in the regulation and reprogramming of immune cells [
12,
13,
45‐
47]. In the study of Lihui Dong and colleagues [
45], C1q + TAMs, a subtype of TAMs, were found to be regulated by an RNA m
6A program and promote CD8 + T cell dysfunction by expressing Ebi3 transcript with decreased m
6A level. In addition, Lei Zhang et al. [
48] reported that SPP1 + TAMs had a pro-angiogenic signature and weakened tumor immunity. By NMF clusters, we found that m
6A-mediated subtypes of macrophages all manifested extensive communications with tumor cells. Correlation analysis showed that HNRNPA2B1 + mac-C3 was significantly related to SPP1 + and C1q + macrophages, and further prognosis analysis revealed the inverse correlation between the expression with survival probability. The metabolic process had a profound influence on TAMs and thus modulated cancer progression and immune responses, including glucose, glutamine and fatty acid metabolism [
49]. Our research found that m
6A-mediated macrophage, especially for HNRNPA2B1 + mac-C3 subtypes, manifested obvious activation of metabolism-related pathways, such as purine biosynthesis, gluconeogenesis, and cysteine and methionine metabolism et al. Moreover, except for macrophages, we found that m
6A-mediated T cells (CD8+, CD4+ and regulatory T cells), NK cells and B cells also showed extensive interaction with tumor cells. Furthermore, m
6A-mediated subtypes of four main T cells exhibited variable T cell active and inactive characteristics. These findings all indicate the significant role of RNA m
6A methylation in immune escape and the tumor-promoting effect of macrophages and T cells.
To identify cell-specific gene regulatory networks, we performed an analysis of TFs at the single-cell level. In general, each subtype of CAFs, macrophages, B cells and various types of T cells all manifested distinct TFs characteristics. For CAFs, WTAP-fibroblasts exhibited a unique TF gene signature, such as ETS1, CEBPB, IRF1, REL and NFKB1. Previous studies have revealed the relationship between m
6A modification and the expression of ETS1, CEBPB, IRF1 and REL, which suggests the role of m
6A in the regulation of CAFs [
50‐
53]. In addition, for macrophages, we observed higher activity of SPI1 and STA1 on HNRNPA2B1-mac. Similarly, the correlation between m
6A modification and both SPI1 and STA1 was reported in previous research [
52,
54‐
56]. Moreover, for B and T cells, we also found distinct TF characteristics of m
6A-mediated cell subtypes. To sum up, m
6A-mediated cell subtypes may modulate distinct TF regulatory networks to reshape and reprogram the TME. Finally, cell network analysis revealed that these m
6A-mediated TME cells were closely connected and communicated with tumor cells. Notably, either m
6A-mediated CAFs or immune cell subtypes had more communication with cancer epithelial cells, indicating that formation of an immunosuppressive tumor microenvironment might partially be determined by RNA m
6A methylation.
Considering the complex intrinsic patterns of RNA m6A methylation in TME cells, we comprehensive summarized the relationships of these sub clusters' scores with prognosis and immune response from the public bulk RNA-seq cohorts. Clearly, the patients with different domination of m6A regulators of the TME cells had huge prognosis differences of CRC and exceedingly distinguished the immune response for patients who underwent ICB therapy, especially for the CAFs and macrophages, which revealed that the critical role of TME m6A for CRC patients in further research.
As a preliminary study, the major limitations of our analysis were that the low depth of scRNA-seq and the inadequate samples, and our conclusion need to acquire verification in more patients. Compared with bulk RNA-seq, the scRNA-seq of some m6A regulators in CRC would typically be minor and had more zero observation, which might contribute to the bias of the clustering method in our study. Nonetheless, the scRNA-seq analysis still provides us a novel view to reveal the characteristics of m6A methylation regulators in various TME single cells to reduce the tumor heterogeneity in CRC, which is a key forward step for clinical practice.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.