Abstract
Colorectal cancer (CRC), including colon adenocarcinoma (COAD) and rectal adenocarcinoma (READ), is one of the most prevalent malignancies worldwide. N6-methyladenosine (m6A) is a ubiquitous RNA modification that plays a vital role in human tumors, but its expression patterns and prognostic value in CRC have not yet been determined. Here, we first used the Cancer Genome Atlas (TCGA), the Gene Expression Omnibus (GEO) and the Human Protein Atlas (HPA) databases and a tissue microarray (TMA) cohort to verify the expression of m6A-related genes at the mRNA and protein levels. We found that most m6A-related genes were substantially upregulated in tumor tissues compared with normal tissues, but METTL14, YTHDF3 and ALKBH5 were downregulated in CRC. There was no obvious difference in FTO. In addition, WTAP, METTL16, HNRNPC and YTHDC1 were abundantly expressed in COAD but not in READ. Moreover, immunofluorescence (IF) analyses of SW480 and HCT116 cells showed that most of the m6A-related proteins were expressed in the nucleus and cytoplasm. Survival analysis demonstrated that the expression levels of METTL3, METTL14, METTL16, FTO and ALKBH5 were associated with the clinical outcomes of CRC patients. Taken together, all the results revealed that m6A-related genes were dysregulated in CRC and might play a significant role in the progression of CRC.
Keywords: Colorectal cancer, m6A, TCGA, TMA, prognosis
Introduction
Colorectal cancer (CRC) ranks third in terms of incidence (10.2% of total cases) and is the second cause of cancer-related death (9.2% of all cases) worldwide [1]. CRC includes colon adenocarcinoma (COAD) and rectal adenocarcinoma (READ). Although considerable advancements in therapeutic strategies have been achieved, the survival rate of CRC remains far from satisfactory due to its late diagnosis, rapid development and easy metastasis [2,3]. Therefore, extensive and in-depth studies are needed for improvements in the diagnosis and treatment of CRC and for the prediction of its recurrence.
N6-methyladenosine (m6A) is the most abundant and evolutionarily conserved modification [4], occurring in nearly all types of RNAs and in most organisms, from bacteria to animals [5]. Many studies have established that the m6A modification is reversible and involves adenosine methyltransferases, demethylases, and RNA-binding proteins, which can add, remove, or recognize m6A-modified sites and alter important biological processes accordingly [6]. Adenosine methyltransferases, known as the “writer” complex, consist of METTL3/14/16, RBM15/15B, WTAP, and KIAA1429, which aid the deposition of m6As at the DRACH (D=A/G/U, R=A/G, H=A/C/U) consensus site on RNA polymerase II (pol II)-transcribed RNAs [7]. FTO and ALKBH5, which are considered m6A “erasers”, are selective demethylases capable of regulating gene expression and cell fate through oxidative removal of the methyl group in m6A-containing substrates. Furthermore, RNA-binding proteins, which are considered “readers” and incorporate the YTH and hnRNP domains, can selectively recognize mRNA m6A sites to mediate the degradation of mRNA [8].
Recent studies have demonstrated that m6A is associated with various human diseases and is particularly found in tumors. The linkages between m6A and human cancer types have been previously demonstrated in various cancers, including cervical cancer [9], prostate cancer [10], breast cancer [11], pancreatic cancer [12], and hepatocellular carcinoma [13]. For example, Tang found that Wilms’ tumor 1-associating protein promoted renal cell carcinoma proliferation by regulating CDK2 mRNA stability [14]. Zhao et al determined that the overexpression of YTHDF1 was associated with poor prognosis in patients with hepatocellular carcinoma [15]. However, the specific expression patterns and clinical value of m6A-related genes in CRC are largely unknown.
In the present study, we investigated two types of colorectal cancer, namely, COAD and READ. And analyses of The Cancer Genome Atlas (TCGA), the Gene Expression Omnibus (GEO), the Human Protein Atlas (HPA) databases and the tissue microarray (TMA) cohort revealed that m6A-related proteins were frequently dysregulated in COAD and READ patients at the mRNA and protein levels. Immunofluorescence (IF) analyses were performed to determine the localization of m6A-related genes in CRC cells. Furthermore, the correlation between m6A-related gene expression and many molecular/clinicopathological parameters was explored in CRC patients. The survival analysis and univariate and multivariate Cox regression analyses established that the expression of m6A-related genes had a critical influence on the overall survival (OS) and recurrence-free survival (RFS) of CRC patients (Figure 1). These results emphasized the significance of m6A-related genes in colorectal cancer.
Figure 1.
Study design and flow diagram. We focused on two types of colorectal cancer: COAD and READ. We first revealed the expression patterns of m6A-related genes at the mRNA and protein levels based on the TCGA, GEO and the Human Protein Atlas databases and the TMA cohort. An immunofluorescence analysis was performed to determine the localization of the expression of m6A-related genes in CRC cells. The correlation between m6A-related gene expression and clinicopathological features was analyzed using χ2 test. Furthermore, survival analysis and univariate and multivariate Cox regression analyses established that m6A-related gene expression exerted a critical influence on the overall survival (OS) and recurrence-free survival (RFS) of CRC patients.
Materials and methods
CRC dataset acquisition and process
The TCGA-COAD and TCGA-READ datasets and all corresponding clinical data used in our study were downloaded from the TCGA data portal (http://gdc-portal.nci.nih.gov/). Seven sets of microarrays (GSE20916, GSE41258, GSE41328, GSE19249, GSE33113, GSE68204 and GSE87211) were extracted from the GEO database (http://www.ncbi.nlm.nih.gov/geo/). Their characteristics, including cohort ID, RNA-Seq platform, number of samples (normal and tumor samples), publication year and country, are summarized in Table S1. Mutation data were obtained from cBioPortal (https://www.cbioportal.org/). In addition, validation of the translation of m6A-related genes was performed using the Human Protein Atlas database (http://www.proteinatlas.org/).
Tissue samples
For TMA, tumor tissues including 22 COAD tissue specimens and 21 READ specimens with corresponding normal adjacent tissue specimens, were obtained from April 2016 to December 2016 at the First Affiliated Hospital of Zhengzhou University, Zhengzhou University, China. None of the patients was administered any chemotherapy, immunotherapy, or radiotherapy prior to surgery. Our study was approved by the Institutional Review Board of the First Affiliated Hospital of Zhengzhou University, and all the patients provided signed informed consent.
Immunohistochemistry (IHC)
An IHC analysis of m6A-related genes was performed using formalin-fixed, paraffin-embedded tissues according to the manufacturer’s instructions [16]. Briefly, the TMA sections were deparaffinized, and 0.3% hydrogen peroxide was applied to block endogenous peroxidase activity. After antigen retrieval, the sections were incubated overnight with primary antibody at 4°C and then with secondary antibody at room temperature. Subsequently, the SignalStain® DAB Substrate Kit (CST, USA) and Hematoxylin QS (Vector Laboratories) were used for the detection of immunoreactive cells. Two pathologists who were blind to the clinical parameters assessed the staining intensity of the reactions. The samples were scored based on the proportion of positive cells as follows: 0-none, 1 - <25%, 2 - 25-50%, 3 - 50-75%, and 4 - 75-100%. The staining intensity was evaluated as follows: 0 - none, 1 - weak, 2 - medium and 3 - strong. A total score was then calculated by multiplying the two sub-scores, and the samples with total scores of 0-6 and 7-12 were classified as low and high expression, respectively. The characteristics of the antibodies used in this study are summarized in Table S2.
Cell lines and culture
The human colorectal cancer cell lines (HCT116 and SW480) used in this study were obtained from the Cell Bank of the Chinese Academy of Science (Shanghai, China). The cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal bovine serum (FBS) (Gibco, Grand Island, NY, USA) and 100 U/mL penicillin/streptomycin (Corning, New York, NY, USA) in a humidified incubator with an atmosphere of 5% CO2 at 37°C. The cell lines were passed for less than 6 months in culture prior to the experiments.
Immunofluorescence
The cultured cells were inoculated into 24-well plates in DMEM with 10% FBS for 24 h, rinsed in phosphate-buffered saline (PBS), fixed with 4% paraformaldehyde for 15 min at room temperature, and permeabilized in 0.5% Triton X-100-PBS for 15 min. To block nonspecific binding sites, the cells were incubated with 1% bovine serum albumin (BSA) PBS. The cells were subsequently incubated with primary antibody (1:200 dilution) at 4°C overnight and then for 1 h with the appropriate secondary antibody (1:200 dilution). The nuclei were counterstained by mounting the cells in DAPI II (Abbott Molecular, Abbott Park, IL, USA). The immunofluorescent signals were then detected using a fluorescence microscope (Axio Observer A1). The characteristics of the antibodies used in our study are summarized in Table S2.
Statistical analysis
SPSS 23.0 software (IBM Corp., Armonk, NY, USA) and GraphPad Prism 7 (San Diego, CA, USA) were used for the statistical analyses. The TMA results were evaluated through the χ2 test or Fisher’s exact test. We analyzed the patients’ survival time through Kaplan-Meier and log-rank tests. The best cut-off value for each gene and its survival curves were obtained using R Studio. The details are described in the Supplementary Methods. In addition, univariate and multivariate Cox regression analyses were performed to screen for independent factors that critically influenced the OS and RFS. Student’s test was used for comparison between two groups. P (two-sided) values less than 0.05 were considered to indicate statistical significance.
Results
Expression pattern of m6A “writers” in colorectal cancer
To explore the expression of m6A-related proteins in human CRC, we first extracted and analyzed the expression of m6A-related genes from the TCGA database at the mRNA level. In COAD, all “writers” were substantially upregulated in tumor tissues compared with normal tissues, with the exception of METTL14, which was downregulated (Figure 2A). In contrast, the results showed that WTAP and METTL16 had no significant difference in READ tissues compared with normal tissues (Figure 2B). Furthermore, we used the GEO database to further validate the expression status of the m6A “writers”. Figure 2 showed a hierarchical clustering heatmap of m6A-related gene expression in COAD and READ, and the results presented a similar conclusion to that obtained from the analysis of the TCGA database.
Figure 2.
Heatmap showing the alterations in the mRNA expression of m6A-related genes in the TCGA and GEO datasets. The red color indicates upregulated expression; the green color indicates downregulated expression; the black color indicates no significant changes, and the white color indicates that the related gene is absent in the datasets. The data were statistically analyzed by Student’s t test (unpaired, two-tailed). A. mRNA expression patterns of m6A-related genes in COAD. B. mRNA expression patterns of m6A-related genes in READ.
Considering the difference between mRNA transcription and protein expression, the protein changes of the m6A “writers” in CRC were further addressed by analyzing the TMA cohort consisting of 22 pairs of COAD and 21 pairs of READ tissue samples. The protein expression of all the “writers” was detected by IHC analysis (Figure 3A and 3B). Subsequently, according to the staining intensity and the percentage of positive cells in the tissue sections, we categorized the patterns into high and low expression. In COAD, the protein expression of five (71.4%) “writers” (all except METTL14 and METTL16) was consistent with the gene expression levels. Specifically, METTL16 and METTL14 exhibited an obvious abundance and was weakly expressed at the mRNA level, respectively, but no significant difference was found at the protein level (Figure 3C). In READ, WTAP and METTL14 showed different expression patterns at the protein level, in contrast to the findings obtained at the mRNA level. WTAP was highly upregulated at the protein level but showed no significant difference at the mRNA level. Analogously with the findings in COAD, METTL14 was downregulated at the mRNA level but showed no significant changes at the protein level in READ (Figure 3D). Moreover, the IHC staining results and the patient data obtained from the Human Protein Atlas database also demonstrated the expression status of the m6A “writers”, as shown in Figure 3E and 3F. The protein expression levels of most m6A “writers” were in accordance with their transcriptional levels, but the database did not include any IHC information for METTL3.
Figure 3.
Protein expression patterns of m6A-related “writers” in CRC and normal tissues. A. Representative IHC staining of m6A-related “writers” in COAD in the TMA cohort. B. Representative IHC staining of m6A-related “writers” in READ in the TMA cohort. C. Comparison of the relative expression of m6A-related “writers” between COAD and normal tissues in the TMA cohort. D. Comparison of the relative expression of m6A-related “writers” between READ and normal tissues in the TMA cohort. E. Information on the IHC staining of several m6A-related “writers” in COAD in The Human Protein Atlas database. F. Information on the IHC staining of several m6A-related “writers” in READ in The Human Protein Atlas database. (*P<0.05, **P<0.01, ***P<0.001, N.S: no significance).
Expression status of m6A “readers” in colorectal cancer
We analyzed the mRNA expression of the “readers” using the TCGA database. As shown in Figure 2A, YTHDF3 and the other “readers” was downregulated and overexpressed in COAD, respectively. However, no obvious discrepancies in HNRNPC and YTHDC1 expression were found in READ tissues compared with normal tissues (Figure 2B). Similar results were observed from the GEO database (Figure 2A and 2B).
To further verify the expression patterns of m6A “readers”, we performed immunohistochemical analyses of COAD and READ TMAs (Figure 4A and 4B). In COAD, all the “readers” showed significant changes in protein expression, and the direction of these changes was consistent with that found for the gene expression changes, with the exception of YTHDF3 and YTHDC1. Among these exceptions, the protein expression of YTHDF3 was contrary to its mRNA expression, and YTHDC1 showed an obvious abundance at the mRNA level but no significant change at the protein level, respectively (Figure 4C). The expression patterns found in READ were similar to those found in COAD via the TMA cohort (Figure 4D). YTHDF3 was also overexpressed at the protein level but downregulated at the mRNA level. However, HNRNPC was upregulated at the protein level but showed no apparent difference in expression at the mRNA level. In addition, the results from the Human Protein Atlas database demonstrated that the protein expression of most “readers” was consistent with their transcriptional level (Figure 4E and 4F).
Figure 4.
Protein expression patterns of m6A-related “readers” in CRC and normal tissues. A. Representative IHC staining of m6A-related “readers” in COAD in the TMA cohort. B. Representative IHC staining of m6A-related “readers” in READ in the TMA cohort. C. Comparison of the relative expression of m6A-related “readers” between COAD and normal tissues in the TMA cohort. D. Comparison of the relative expression of m6A-related “readers” on between READ and normal tissues in the TMA cohort. E. Information on the IHC staining of m6A-related “readers” in COAD in The Human Protein Atlas database. F. Information on the IHC staining of m6A-related “readers” in READ in The Human Protein Atlas database. (*P<0.05, **P<0.01, ***P<0.001, N.S: no significance).
Expression of m6A “erasers” in colorectal cancer
The m6A “erasers” comprised FTO and ALKBH5, and an analysis of the TCGA database revealed that ALKBH5 showed obviously weaker mRNA expression in CRC than in normal tissue. No significant difference was found for the FTO gene in both COAD and READ. Similar results were found in the GEO database (Figure 2A and 2B). The IHC analysis revealed that ALKBH5 was prominently overexpressed at the protein level but downregulated at the mRNA level and that the protein expression of FTO was concordant with its mRNA expression in both COAD and READ (Figure 5A-D). We found similar expression patterns in the Human Protein Atlas with respect to the transcriptional expression of m6A “erasers” in CRC, with the exception of a deficiency of FTO in READ (Figure 5E and 5F).
Figure 5.
Protein expression patterns of m6A-related “erasers” in CRC and normal tissues. A. Representative IHC staining of m6A-related “erasers” in COAD in the TMA cohort. B. Representative IHC staining of m6A-related “erasers” in READ in the TMA cohort. C. Comparison of the relative expression of m6A-related “erasers” between COAD and normal tissues in the TMA cohort. D. Comparison of the relative expression of m6A-related “erasers” between READ and normal tissues in the TMA cohort. E. Information on the IHC staining of m6A-related “erasers” in COAD in The Human Protein Atlas database. F. Information on the IHC staining of several m6A-related “erasers” in READ in The Human Protein Atlas database. (*P<0.05, **P<0.01, ***P<0.001, N.S: no significance).
Localization of the expression of m6A-related genes in colorectal cells
Although the expression patterns of m6A-related genes in CRC at the mRNA and protein levels have been studied, information on the localization of m6A-related genes in CRC cells remains to be elucidated. Therefore, an IF analysis was performed to further identify the subcellular distribution of m6A-related proteins in CRC cells. As shown in Figure 6, most of the m6A-related genes were mainly expressed in the nucleus and cytoplasm. Specifically, some proteins (KIAA1429, RBM15, RBM15B, HNRNPA2B1, YTHDC1 and ALKBH5) showed strong nuclear staining as well as weak cytoplasmic staining. YTHDF1 and YTHDF2 were detected only in the cytoplasm, where as HNRNPC signals were found in the nucleus.
Figure 6.
Subcellular localization of m6A-related genes in SW480 and HCT116 cell lines. The cells were fixed and reacted with the corresponding antibodies. The secondary antibodies were anti-rabbit IgG-conjugated to fluorescein isothiocyanate and anti-mouse IgG-conjugated to rhodamine red. The nucleus was stained with DAPI (blue). The images were captured with a fluorescence microscope.
Relationship between m6A-related genes and clinicopathological features in CRC
We further analyzed the correlation between the expression of m6A-related genes and clinicopathological characteristics in COAD and READ to explore the clinical significance of m6A-related gene expression. As shown in Figure 7, in COAD, the KRAS mutation was associated with YTHDF1 expression (P=0.035) (Figure 7A), and the BRAF mutation could affect the expression of METTL3 (P=0.033), YTHDF1 (P<0.0001) and ALKBH5 (P=0.011) (Figure 7B). Moreover, the expression of KIAA1429 (P=0.036), RBM15B (P=0.003), YTHDF1 (P=0.022) and ALKBH5 (P=0.022) was correlated with age (Figure 7C). In addition, gender was related to WTAP (P=0.046), METTL16 (P=0.005), HNRNPC (P=0.018) and YTHDF1 (P=0.035) expression (Figure 7D). Race was found to be associated with WTAP (P=0.019) expression (Figure 7E), and the TNM stage was verified to have correlation with YTHDC1 (P=0.011) expression (Figure 7F). However, some differences were found in READ. The KRAS mutation was associated with YTHDF1 (P=0.005) expression (Figure 8A), and a significant relationship was found between BRAF mutation and the expression of RBM15 (P=0.025), METTL3 (P=0.03), METTL14 (P=0.007), YTHDF2 (P=0.043), YTHDF3 (P=0.018) and YTHDC1 (P=0.002) (Figure 8B). Moreover, age was found to be related to KIAA1429 (P=0.017), RBM15 (P=0.015) and METTL16 (P<0.0001) expression (Figure 8C), and no obvious association was found for gender, race and TNM stage with m6A-related gene expression (Figure S1).
Figure 7.
Relationship between m6A-related gene expression and molecular/clinicopathological features in COAD. A. YTHDF1 expression was associated with the KRAS mutation. B. The expression of METTL3, YTHDF1 and ALKBH5 was related to the BRAF mutation. C. A significant correlation was found between age and the expression of KIAA1429, RBM15B, YTHDF1 and ALKBH5. D. Gender was related to WTAP, METTL16, HNRNPC and YTHDF1 expression. E. WTAP expression was important to race. F. The TNM stage was correlated with YTHDC1 expression.
Figure 8.
Relationship between m6A-related gene expression and molecular/clinicopathological features in READ. A. The KRAS mutation was associated with YTHDF1 expression. B. An obvious linkage between the BRAF mutation and the expression of RBM15, METTL3, METTL14, YTHDF2, YTHDC1 and YTHDF3 was found. C. Age could affect the expression of KIAA1429, RBM15 and METTL16.
Survival analysis of m6A-related proteins in colorectal cancer
To evaluate the prognostic roles of m6A-related proteins in CRC progression, we first classified COAD and READ patients into two groups (high-expression group and low-expression group) according to the optimal cut-off value. The correlation of m6A-related gene expression with corresponding clinical follow-up information was determined through Kaplan-Meier analysis and a log-rank test. We first investigated whether the expression levels of m6A-related genes were correlated with the outcome of CRC patients. In COAD patients, low FTO expression predicted poor OS (Figure 9A), and patients with high METTL3 expression was associated with a shorter RFS compared with those with low METTL3 expression (Figure 9B). Moreover, as shown in Figure 9C-E, in addition to high ALKBH5 expression, low METTL14 and METTL16 expression in READ tissues were clearly associated with worse OS. However, no significant difference in METTL16 and FTO mRNA expression was found between tumor and normal tissues. Additionally, m6A-related genes did not predict RFS in READ (Figure S2). Further details were presented in Figures S3, S4, S5.
Figure 9.
Kaplan-Meier curves and univariate and multivariate Cox regression analyses of the TCGA database. A. Low FTO expression predicted poor OS in COAD patients. B. High METTL3 expression predicted a shorter RFS in COAD patients. C. Low METTL14 expression predicted poor OS in READ patients. D. Low METTL16 expression predicted poor OS in READ patients. E. READ patients with high ALKBH5 expression had a shorter OS compared with those with low ALKBH5 expression. F. Univariate Cox regression analysis of the RFS of COAD patients. G. Univariate Cox regression analysis of the OS of COAD patients. H. Univariate Cox regression analysis of the OS of READ patients. I. Univariate Cox regression analysis of the RFS of READ patients.
The univariate Cox regression analysis was performed to identify risk factors related to patient prognosis. Univariable analyzes of COAD revealed that the TNM (tumor, node, and metastasis) stage and high METTL3 expression were significant prognostic factors for RFS (Figure 9F) and that the TNM stage and age were prognostic factors for OS (Figure 9G). For READ patients, age, the TNM stage and the expression of METTL14, METTL16 and ALKBH5 were found to have a critical influence on OS (Figure 9H), and the TNM stage was the only prognostic factor for RFS (Figure 9I). The details are shown in Tables S3 and S4. Furthermore, the multivariate Cox regression analysis revealed that the TNM stage was an independent risk factor for OS (P<0.0001) and RFS (P<0.0001) in COAD (Tables 1 and 2). Age was also found to be a factor affecting RFS (P=0.001) (Table 2). In READ, the expression of METTL14 (P=0.004) and ALKBH5 (P<0.0001) and the TNM stage (P<0.025) were verified to be independent factors of OS (Table 3).
Table 1.
Multivariate Cox regression analysis of OS in COAD
P<0.01;
P<0.001.
Table 2.
Multivariate Cox regression analysis of RFS in COAD
| Parameters | HR | 95% CI | P-value |
|---|---|---|---|
| METTL3 | 1.356 | 0.975-1.886 | 0.07 |
| TNM stage | 2.676 | 1.905-3.758 | <0.0001*** |
P<0.001.
Table 3.
Multivariate Cox regression analysis of OS in READ
| Parameters | HR | 95% CI | P-value |
|---|---|---|---|
| METTL14 | 0.133 | 0.034-0.524 | 0.004** |
| ALKBH5 | 6.013 | 2.244-16.111 | <0.0001*** |
| METTL16 | 0.303 | 0.048-1.905 | 0.203 |
| Age | 0.353 | 0.081-1.530 | 0.164 |
| TNM stage | 3.298 | 1.165-9.336 | 0.025* |
P<0.05;
P<0.01;
P<0.001.
Discussion
M6A was initially reported by Ronald Desrosiers in 1974 [17], but the precise mechanism and regulatory function of the m6A modification remained largely unknown until recently [18]. Many studies have revealed that the m6A modification affects almost every aspect of RNA metabolism, including RNA expression, splicing, nuclear export, translation, decay and RNA-protein interactions (Figure 10) [19-21]. Studies conducted in recent years have demonstrated that m6A can regulate multiple spatial and temporalphysiological processes, including gametogenesis, sex determination, embryogenesis, cell fate determination, circadian rhythms, heat shock responses, DNA damage response, pluripotency, reprogramming and neuronal functions [5,22]. Furthermore, emerging evidence has revealed that m6A plays crucial roles in human diseases. For example, the m6A modification might lead to obesity [23], type 2 diabetes mellitus [24], and infertility [25], among other diseases. Although the acknowledgement of m6A methylation remains at controversial, advanced methods, such as high-throughput sequencing, have enabled researchers to explore the implication of m6A in human diseases, particularly cancer. An increasing number of studies have indicated that m6A plays an essential role in the initiation and progression of tumors. Additionally, aberrant m6A RNA methylation is closely associated with cancer, but the specific regulatory role of m6A in tumorigenesis and cancer progression needs to be fully elucidated. In this manuscript, we provided an overall summary of the roles of m6A in the regulation of CRC.
Figure 10.
Mechanism of m6A modification in CRC. M6A RNA methylation is a dynamic and reversible process that affects almost every aspect of RNA metabolism and involves adenosine methyltransferases, demethylases, and RNA-binding proteins.
Analyses of the TCGA and GEO databases revealed that most of the m6A-related genes were dysregulated in CRC. The results revealed that higher expression of KIAA1429, RBM15B, RBM15, HNRNPA2B1, YTHDF1, YTHDF2 and METTL3 and weaker expression of ALKBH5, YTHDF3 and METTL14 in COAD and READ. In addition, we performed an IHC analysis to further substantiate the m6A-related gene expression patterns at the protein level based on the Human Protein Atlas and TMA cohort. The results were almost consistent and revealed that m6A-related genes were dysregulated in CRC tissues, which indicated that most of these genes might play an oncogenic role in CRC. Similar results were previously reported. Specifically, Chen et al found that METTL3 was upregulated in liver cancer and promoted liver cancer progression through the YTHDF2-dependent post-transcriptional silencing of SOCS2 [26]. Joao Lobo reported that RBM15B was highly expressed in urological tumors, such as prostate cancer, testicular germ cell tumors and papillary renal cell carcinoma [27]. In addition, METTL14 exhibited low expression in hepatocellular carcinoma [28], glioblastoma [29] and breast adenocarcinoma [30]. And ALKBH5 reportedly downregulated the motility of pancreatic cancer by demethylating the long non-coding RNA (lncRNA) KCNK15-AS1 [31]. Nishizawa Y revealed that high YTHDF1 expression was associated with poor prognosis and that its overexpression was driven by c-MYC in CRC [32]. We also demonstrated a significant relationship between the expression of many m6A-related genes and molecular/clinicopathological features in COAD, such as the KRAS and BRAF mutations, age, gender, race and TNM stage. However, we did not find a correlation between m6A-related gene expression and gender, race and TNM stage in READ. In addition, Kowk et al demonstrated that genetic alterations in m6A regulators could predict poorer survival and were associated with TP53 mutations in acute myeloid leukemia [33]. This study first verified the clinicopathological features related to the regulation of m6A-related genes and provided novel insights for further study, and the results demonstrate that m6A-relatedgenes might play a vital role in CRC.
Abundant studies have reported that the dysregulation of m6A-related genes is related to poor prognosis. For example, Liu et al verified that the m6A demethylase FTO facilitated tumor progression in lung squamous cell carcinoma by regulating the expression of MZF1 [34]. Chen et al revealed that bladder cancer patients with positive WTAP expression had a higher postoperative recurrence risk than those with negative WTAP expression [35]. In addition, ALKBH5 was reportedly a novel prognostic biomarker that predicted the prognosis of pancreatic cancer [36]. Ma et al found that hepatocellular carcinoma patients with reduced METTL14 expression experienced more frequent recurrence and poorer survival [28]. Consistent with these findings, using TCGA data, we found that high METTL3 expression and decreased regulation of METTL14, METTL16, FTO and ALKBH5 were positively correlated with poor prognosis in CRC patients. Additionally, univariate and multivariate analyses showed that age, the TNM stage and the expression of METTL14 and ALKBH5 were independent prognostic factors in CRC.
Conclusions
M6A-related genes were dysregulated in CRC, and their expression was associated with CRC progression and poor prognosis. The results of this study showed the value of m6A-related genes as clinical biomarkers in CRC and emphasized their potential as prognostic biomarkers in CRC patients.
Acknowledgements
This study was supported by funds from the National Natural Science Foundation of China (81702757, 81702346, 81600506, 81702927); The Joint research fund of the First Affiliated Hospital of Zhengzhou University and Dalian Institute of Chemical Physics Chinese Academy of Sciences (RRS). National S&T Major Project (SQ2018ZX100301). National Engineering Laboratory for Internet Medical System and Application open fund project (NELIMSA2018P03). The funding body had no role in the design of the study, in the collection, analysis, and interpretation of the data, or in the manuscript writing.
All the patients provided written informed consent, and the project was performed in accordance with the Helsinki Declaration of 1975. The clinical information of the patients utilized for this study is maintained in the databases of the First Affiliated Hospital of Zhengzhou University.
Disclosure of conflict of interest
None.
Supporting Information
References
- 1.Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68:394–424. doi: 10.3322/caac.21492. [DOI] [PubMed] [Google Scholar]
- 2.Compton CC. Colorectal carcinoma: diagnostic, prognostic, and molecular features. Mod Pathol. 2003;16:376–388. doi: 10.1097/01.MP.0000062859.46942.93. [DOI] [PubMed] [Google Scholar]
- 3.Sun Y, Yang B, Lin M, Yu H, Chen H, Zhang Z. Identification of serum miR-30a-5p as a diagnostic and prognostic biomarker in colorectal cancer. Cancer Biomark. 2019;24:299–305. doi: 10.3233/CBM-182129. [DOI] [PubMed] [Google Scholar]
- 4.Wang X, Lu Z, Gomez A, Hon GC, Yue Y, Han D, Fu Y, Parisien M, Dai Q, Jia G, Ren B, Pan T, He C. N6-methyladenosine-dependent regulation of messenger RNA stability. Nature. 2014;505:117–120. doi: 10.1038/nature12730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Sun T, Wu R, Ming L. The role of m6A RNA methylation in cancer. Biomed Pharmacother. 2019;112:108613. doi: 10.1016/j.biopha.2019.108613. [DOI] [PubMed] [Google Scholar]
- 6.Pinello N, Sun S, Wong JJ. Aberrant expression of enzymes regulating m(6)A mRNA methylation: implication in cancer. Cancer Biol Med. 2018;15:323–334. doi: 10.20892/j.issn.2095-3941.2018.0365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Tuncel G, Kalkan R. Importance of m N(6)-methyladenosine (m(6)A) RNA modification in cancer. Med Oncol. 2019;36:36. doi: 10.1007/s12032-019-1260-6. [DOI] [PubMed] [Google Scholar]
- 8.Berlivet S, Scutenaire J, Deragon JM, Bousquet-Antonelli C. Readers of the m(6)A epitranscriptomic code. Biochim Biophys Acta Gene Regul Mech. 2019;1862:329–342. doi: 10.1016/j.bbagrm.2018.12.008. [DOI] [PubMed] [Google Scholar]
- 9.Zhou S, Bai ZL, Xia D, Zhao ZJ, Zhao R, Wang YY. FTO regulates the chemo-radiotherapy resistance of cervical squamous cell carcinoma (CSCC) by targeting beta-catenin through mRNA demethylation. Mol Carcinog. 2018;57:590–597. doi: 10.1002/mc.22782. [DOI] [PubMed] [Google Scholar]
- 10.Liu ZX, Li LM, Sun HL, Liu SM. Link between m6A modification and cancers. Front Bioeng Biotechnol. 2018;6:89. doi: 10.3389/fbioe.2018.00089. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Cai X, Wang X, Cao C, Gao Y, Zhang S, Yang Z, Liu Y, Zhang X, Zhang W, Ye L. HBXIP-elevated methyltransferase METTL3 promotes the progression of breast cancer via inhibiting tumor suppressor let-7g. Cancer Lett. 2018;415:11–19. doi: 10.1016/j.canlet.2017.11.018. [DOI] [PubMed] [Google Scholar]
- 12.Taketo K, Konno M, Asai A, Koseki J, Toratani M, Satoh T, Doki Y, Mori M, Ishii H, Ogawa K. The epitranscriptome m6A writer METTL3 promotes chemo- and radioresistance in pancreatic cancer cells. Int J Oncol. 2018;52:621–629. doi: 10.3892/ijo.2017.4219. [DOI] [PubMed] [Google Scholar]
- 13.Cheng X, Li M, Rao X, Zhang W, Li X, Wang L, Huang G. KIAA1429 regulates the migration and invasion of hepatocellular carcinoma by altering m6A modification of ID2 mRNA. Onco Targets Ther. 2019;12:3421–3428. doi: 10.2147/OTT.S180954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Tang J, Wang F, Cheng G, Si S, Sun X, Han J, Yu H, Zhang W, Lv Q, Wei JF, Yang H. Wilms’ tumor 1-associating protein promotes renal cell carcinoma proliferation by regulating CDK2 mRNA stability. J Exp Clin Cancer Res. 2018;37:40. doi: 10.1186/s13046-018-0706-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zhao X, Chen Y, Mao Q, Jiang X, Jiang W, Chen J, Xu W, Zhong L, Sun X. Overexpression of YTHDF1 is associated with poor prognosis in patients with hepatocellular carcinoma. Cancer Biomark. 2018;21:859–868. doi: 10.3233/CBM-170791. [DOI] [PubMed] [Google Scholar]
- 16.Zhu W, Li H, Yu Y, Chen J, Chen X, Ren F, Ren Z, Cui G. Enolase-1 serves as a biomarker of diagnosis and prognosis in hepatocellular carcinoma patients. Cancer Manag Res. 2018;10:5735–5745. doi: 10.2147/CMAR.S182183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Desrosiers R, Friderici K, Rottman F. Identification of methylated nucleosides in messenger RNA from Novikoff hepatoma cells. Proc Natl Acad Sci U S A. 1974;71:3971–3975. doi: 10.1073/pnas.71.10.3971. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Roundtree IA, Evans ME, Pan T, He C. Dynamic RNA modifications in gene expression regulation. Cell. 2017;169:1187–1200. doi: 10.1016/j.cell.2017.05.045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Gilbert WV, Bell TA, Schaening C. Messenger RNA modifications: form, distribution, and function. Science. 2016;352:1408–1412. doi: 10.1126/science.aad8711. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Roignant JY, Soller M. m(6)A in mRNA: an ancient mechanism for fine-tuning gene expression. Trends Genet. 2017;33:380–390. doi: 10.1016/j.tig.2017.04.003. [DOI] [PubMed] [Google Scholar]
- 21.Meyer KD, Jaffrey SR. The dynamic epitranscriptome: N6-methyladenosine and gene expression control. Nat Rev Mol Cell Biol. 2014;15:313–326. doi: 10.1038/nrm3785. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Pan Y, Ma P, Liu Y, Li W, Shu Y. Multiple functions of m(6)A RNA methylation in cancer. J Hematol Oncol. 2018;11:48. doi: 10.1186/s13045-018-0590-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zhou Y, Hambly BD, McLachlan CS. FTO associations with obesity and telomere length. J Biomed Sci. 2017;24:65. doi: 10.1186/s12929-017-0372-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Bohnsack MT, Sloan KE. The mitochondrial epitranscriptome: the roles of RNA modifications in mitochondrial translation and human disease. Cell Mol Life Sci. 2018;75:241–260. doi: 10.1007/s00018-017-2598-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ding C, Zou Q, Ding J, Ling M, Wang W, Li H, Huang B. Increased N6-methyladenosine causes infertility is associated with FTO expression. J Cell Physiol. 2018;233:7055–7066. doi: 10.1002/jcp.26507. [DOI] [PubMed] [Google Scholar]
- 26.Chen M, Wei L, Law CT, Tsang FH, Shen J, Cheng CL, Tsang LH, Ho DW, Chiu DK, Lee JM, Wong CC, Ng IO, Wong CM. RNA N6-methyladenosine methyltransferase-like 3 promotes liver cancer progression through YTHDF2-dependent posttranscriptional silencing of SOCS2. Hepatology. 2018;67:2254–2270. doi: 10.1002/hep.29683. [DOI] [PubMed] [Google Scholar]
- 27.Lobo J, Barros-Silva D, Henrique R, Jerónimo C. The emerging role of epitranscriptomics in cancer: focus on urological tumors. Genes (Basel) 2018;9 doi: 10.3390/genes9110552. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ma JZ, Yang F, Zhou CC, Liu F, Yuan JH, Wang F, Wang TT, Xu QG, Zhou WP, Sun SH. METTL14 suppresses the metastatic potential of hepatocellular carcinoma by modulating N(6)-methyladenosine-dependent primary MicroRNA processing. Hepatology. 2017;65:529–543. doi: 10.1002/hep.28885. [DOI] [PubMed] [Google Scholar]
- 29.Cui Q, Shi H, Ye P, Li L, Qu Q, Sun G, Sun G, Lu Z, Huang Y, Yang CG, Riggs AD, He C, Shi Y. m(6)A RNA methylation regulates the self-renewal and tumorigenesis of glioblastoma stem cells. Cell Rep. 2017;18:2622–2634. doi: 10.1016/j.celrep.2017.02.059. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Wu L, Wu D, Ning J, Liu W, Zhang D. Changes of N6-methyladenosine modulators promote breast cancer progression. BMC Cancer. 2019;19:326. doi: 10.1186/s12885-019-5538-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.He Y, Hu H, Wang Y, Yuan H, Lu Z, Wu P, Liu D, Tian L, Yin J, Jiang K, Miao Y. ALKBH5 inhibits pancreatic cancer motility by decreasing long non-coding RNA KCNK15-AS1 methylation. Cell Physiol Biochem. 2018;48:838–846. doi: 10.1159/000491915. [DOI] [PubMed] [Google Scholar]
- 32.Nishizawa Y, Konno M, Asai A, Koseki J, Kawamoto K, Miyoshi N, Takahashi H, Nishida N, Haraguchi N, Sakai D, Kudo T, Hata T, Matsuda C, Mizushima T, Satoh T, Doki Y, Mori M, Ishii H. Oncogene c-Myc promotes epitranscriptome m(6)A reader YTHDF1 expression in colorectal cancer. Oncotarget. 2018;9:7476–7486. doi: 10.18632/oncotarget.23554. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Kwok CT, Marshall AD, Rasko JE, Wong JJ. Genetic alterations of m(6)A regulators predict poorer survival in acute myeloid leukemia. J Hematol Oncol. 2017;10:39. doi: 10.1186/s13045-017-0410-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Liu J, Ren D, Du Z, Wang H, Zhang H, Jin Y. m(6)A demethylase FTO facilitates tumor progression in lung squamous cell carcinoma by regulating MZF1 expression. Biochem Biophys Res Commun. 2018;502:456–464. doi: 10.1016/j.bbrc.2018.05.175. [DOI] [PubMed] [Google Scholar]
- 35.Chen L, Wang X. Relationship between the genetic expression of WTAP and bladder cancer and patient prognosis. Oncol Lett. 2018;16:6966–6970. doi: 10.3892/ol.2018.9554. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Cho SH, Ha M, Cho YH, Ryu JH, Yang K, Lee KH, Han ME, Oh SO, Kim YH. ALKBH5 gene is a novel biomarker that predicts the prognosis of pancreatic cancer: a retrospective multicohort study. Ann Hepatobiliary Pancreat Surg. 2018;22:305–309. doi: 10.14701/ahbps.2018.22.4.305. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.