Lung cancer is a global public health challenge with its high mortality [
18]. LUAD, a fatal malignancy associated with poor prognosis and high mortality rates, accounts for more than 40% of the total incidence of lung cancer [
19,
20]. Although the diagnosis and treatment of LUAD have made great progress, the effective diagnosis and prognosis prediction of LUAD patients is still a major clinical challenge [
21]. Therefore, identifying key molecules and constructing a prediction model with high stability and effectiveness are conducive to the implementation of precise treatment and improve the prognosis of patients.
N6-methyladenosine (m6A) is the most abundant internal modification of eukaryotic mRNA. Almost every stage of mRNA metabolism is affected by m6A mRNA methylation [
22]. In addition, new evidence suggests that m6A RNA methylation plays a vital role in tumorigenesis and development [
23]. FTO and METTL3 have been reported as potential targets for the diagnosis and treatment of LUAD patients. It is reported that FTO facilitates LUAD cell progression by activating cell migration through m6A demethylation [
24]. High expression of METTL3 in LUAD is believed to promote the growth and invasion of cancer cells [
25]. However, there are few studies on the relationship between m6A related genes and LUAD. In this study, the expression of 21 m6A regulatory factors from LUAD and adjacent normal tissues were systematically analyzed. We observed the different expression levels of these genes between LUAD and normal tissues. Based on the expression of 21 m6A regulatory factors, TCGA patients were divided into two clusters utilizing consensus classification. The prognosis of patients in cluster 1 was better than that of patients in cluster 2. Given that the patients’ prognosis was probably related to tumor immune microenvironment, the differences of immune infiltration between two clusters were analyzed. We found that patients in cluster 1 had a higher immune score and immune cell infiltration. Previous studies indicated that the immune-inflamed phenotype shows the infiltration of massive amounts of immune cells in the tumor microenvironment, and correlated with better prognosis [
11,
26]. To further explore the difference between the two clusters, we identified 457 DEGs and analyzed their biological functions by GO enrichment analyses. Interestingly, the immune-related biological processes were mainly enriched for the up-expressed genes in cluster 1. Then, after selecting seven survival-associated DEGs by Lasso Cox regression, a reliable prognostic model was successfully established. The survival analysis demonstrated patients with a high-risk score have a worse prognosis. The ROC curve confirmed the predictive accuracy of this prognostic risk signature. In addition, higher immune scores and immune cell infiltration were founded in the low-risk group. In recent years, immune checkpoint inhibitors have attracted much attention due to their promising application in the immunotherapy of cancer [
27]. Therefore, expressions of immune checkpoint molecules like PD-L1, PD-1, PD-L2, and LAG3 in different groups were examined. And it was found that there existed a positive correlation between risk score and the expressions of immune checkpoint molecules. Therefore, these results demonstrated that the prognosis model might have a potential role in predicting the clinical response of immunotherapy.
Among the seven key molecules (CLEC3B, TENM3, IGF2BP1, E2F7, ANLN, ANKRD18B, and FBN2), only CLEC3B highly expressed in the low-risk group while other genes highly expressed in the high-risk group. CLEC3B, C-type lectin domain family 3 member B, is a member of the C-type lectin superfamily [
28]. It encodes tetranectin, a plasminogen kringle-4-binding protein in cells [
29]. It is reported that CLEC3B was down-regulated in several tumors and considered as a tumor suppressor in oral squamous cell carcinoma [
30]. A previous study demonstrated that the expression of CLEC3B is correlated with the level of immune infiltration in lung cancer, and it is promising to be the important marker for the early diagnosis of lung cancer [
31]. The protein encoded by TENM3 gene belongs to the teneurin family and is involved in tumorigenesis and drug resistance [
32]. It is reported that TENM3 was up-regulated in tumor tissues, and it may function as an oncogenic gene in esophageal cancer [
33]. However, the role of TENM3 in lung cancer is not yet clear. IGF2BP1 is an RNA-binding protein that participates in tumor progression, tumor cell proliferation and growth [
34,
35]. The let-7 family exerts its role in suppressing the migration and growth of tumor cells by inhibiting the expression of IGF2BP1 [
36]. Previous studies showed the up-regulation of IGF2BP1 in LUAD, which affects the progression of the disease [
37,
38]. Furthermore, high expression of IGF2BP1 was associated with poor OS in LUAD [
37]. E2F7 is a member of the E2F transcription factors family. It is reported that the mammalian E2F transcription factors play vital roles in the cell cycle, so they are closely related to cancer [
39]. In addition, E2F7 has been found up-regulated in various malignant tumors, such as acute myeloid leukaemia and cutaneous squamous cell carcinomas [
40,
41]. Knockdown of E2F7 can repress cell growth in endometrial carcinoma [
42]. However, how E2F7 participates in LUAD is still unknown. The ANLN gene encodes an actin-binding protein that contributes to cell growth and migration [
43]. The expression of ANLN is up-regulated in a variety of types of tumors, including lung cancer, and the development of cancers is related to the expression level of ANLN [
44]. Previous studies indicated the vital role of ANLN in cell proliferation, and lacking ANLN reduced cell migration and invasion [
45,
46]. ANLN has been reported to participate in the metastasis of LUAD by promoting the EMT of tumor cells [
47]. ANKRD18B is a member of ANKRD family that functions in the occurrence of cancer, evidence that over-expression of ANKRD18B suppressed the growth of lung cancer cells has been reported [
48]. FBN2 encoded the protein which belongs to the connective tissue microfibrils and participates in elastic fiber assembly. Nevertheless, there is no doubt to determine the impact of FBN2 in LUAD in the future.
Nevertheless, our study had a few limitations to be considered. Firstly, further experiments are necessary to verify our results because our study was only based on public databases. Secondly, larger sample size is necessary to confirm the predictive ability of our prognostic models. Thirdly, the biological roles of the seven key genes in LUAD require further experimental validation.