Transcriptional Expressions of YTHDF 1/2/3 as Prognosis Indicators for Survivals in Lung Adenocarcinoma Patients

Adeno-carcinoma


Introduction
Lung cancer (LC) is the leading cause of all cancer-induced deaths worldwide [1,2]. In all cases of lung cancer, non-small cell lung cancer (NSCLC) is responsible for a majority measure of approximately 84% [2]. NSCLC can be subdivided into 3 types: lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC), and large cell carcinoma (LCC) [3]. LUAD is the most common form of NSCLC, found in smokers and nonsmokers alike, as well as being found more commonly in women than in men and in younger generations under the age of 45 [4]. It compensates for around 50% of all NSCLCs while the next most common type, LUSC, is responsible for about 30% of cases [5]. LUSC begins its replication in the squamous cells that line the large airways of the lung and holds the strongest connection to smoking out of all the types of NSCLC [6]. The focus of our study will pertain to LUAD and LUSC exclusively. mRNA modifications play a critical role in diverse biological processes including cancer development and progression [7][8][9]. Several mRNA modifications have been identified, including N1-methyladenosine (m1A), N6-methyladenosine (m6A), 5-methylcytidine (m5C), and the 5′ cap modifications N 6,2-O-dimethyladenosine (m6Am) and 2′-O-methylation (2′-O-me) [10]. Among these, m6A is the bestcharacterized and most prevalent mRNA modification [10,11]. The process of m6A mRNA modification is found to be both dynamic and reversible through the work of methyltransferases (m6A writers), RNA binding proteins (m6A readers) and demethylases (m6A erasers) [11]. A 20 gene catalog of proteins primarily functioning as regulators of m6A methylation has been identified and curated, broken down into eleven readers, seven writers, and two erasers [8,11,12]. Consequently, m6A is recognized by the YT521-B homology (YTH) domain family of proteins which directs and utilizes different complexes for the purpose of regulating RNA signaling pathways such as RNA metabolism, RNA splicing, RNA folding, and protein translation [13,14]. YTH N6-Methyladenosine RNA Binding Proteins include three members (YTHDF1-3) in human tissues [14]. YTHDF1-3, containing the YTH domain, have been characterized as direct m6A readers and function together to mediate the degradation of m6A mRNAs [14,15]. Widespread genetic abnormalities in regulators, such as mutations and copy number variations, have been found across cancer types as emerging evidence continues to be compiled. This association has concluded a link to tumor proliferation, invasion, differentiation, tumorigenesis, and metastasis and functions as oncogenes or antioncogenes in malignant tumors. So far, knowledge regarding the reconstitutes of m6A, particularly the reader YTHDFs in lung cancer, is still lacking. In this study, we aimed to systematically characterize the molecular alterations and clinical relevance of YTHDF1, YTHDF2, and YTHDF3 between LUAD and LUSC patients.

Ethics Statement
This study was approved by the Ethics Committee of the Jilin province cancer hospital. We retrieved the original data from the online databases and analyzed them by using the online databases.

Types of Cancers
The mRNA expression of YTHDFs members in different types of cancer and related adjacent control tissues were analyzed by using the ONCOMINE database [16] (www.oncomine.org). Data of the mRNA expression in these tissues were analyzed by students' t-test, with p value < 0.01, fold change > 1.5 and gene rank as 10%.

Analysis of Gene Expression in Different Stages Via UALCAN
We used UALCAN database (http://ualcan.path.uab.edu/), a free online open-access platform with gene expression data sets from TCGA database [17], analyzed the relative mRNA transcriptional levels of YTHDFs genes of interest between normal and different stages of cancer tissues. Data was analyzed by students' t-test and p <0.01 was recognized as statistically significant.

Plotter
We used the Kaplan-Meier Plotter database (KM Plotter, http://kmplot.com/analysis/), an online database with different gene expression statuses and survival information of lung cancer patients and analyzed the prognostic values of YTHDFs members [18]. The mean expression of gene was used to split the mRNA level of each YTHDFs genes. The prognostic value of YTHDFs in lung cancer patients were analyzed and compared between the high expression cohort and low expression cohort.

cBioPortal, GePIA 2.0 and Networkanalyst Database
We analyzed mutations in the genomic profiles of YTHDFs family members by using the cBioportal database with the z-score threshold ±1.8 [19,20]. We also collected the 40 frequently altered neighbor genes of each YTHDFs member in LUAD patients by using similar gene detection functions on the Gepia2.0 database [21] (http://gepia2.cancer-pku.cn/#index). These frequently altered neighbor genes were pooled with YTHDFs members and applied for protein interaction networks by using the Networkanalyst platform [22].

Genomes (KEGG) Analysis
We applied 120 genes significantly associated with these mutations for GO and KEGG enrichment analysis via Metascape database with min overlap at 3, p-value cut off at 0.05, and min enrichment at 3 (https://metascape.org/gp/index.html#/main/ step1) [23]. These genes were administrated to GO enrichment analysis for biological processes (BP), cellular components (CC), molecular functions (MF) and KEGG pathway analysis individually.

Statistical Analysis
The mRNA expression of genes in lung cancer tissues was analyzed by Student's t-test. Kaplan-Meier survival plots were generated with survival curves compared using the log-rank test. P values less than 0.05 was considered as statistically different [24].

Members in NSCLC Patients
To explore the distinct prognostic and potential therapeutic value of different YTHDF members in patients with cancer, the mRNA expression was analyzed by the Oncomine database (www. oncomine.org). As shown in Figure 1

Subclasses
To further analyze the association of YTHDFs mRNA expression with various LUAD subclasses, we detected the survival effects of YTHDFs in different smoking status, different genders and different clinical stages. As shown in (Table 1), the high expression of YTHDF2 was correlated to better OS in LUAD patients both with and without smoking, and the high expression of YTHDF3 was correlated with better OS in LUAD patients with smoking. While the expression of YTHDF1 was not associated with the OS in LUAD patients with different smoking statuses. The high expression of YTHDF1/2/3 were associated with better OS both in female and male LUAD patients (Table 1). In addition to the LUAD patients from different clinical stages, the high expression of YTHDF1 or 2 were strongly correlated to better OS of LUAD patients in stage 1 and 2 but not in stage 3 ( Table 1). The high expression of YTHDF3 predicts better OS in LUAD patients from stage 1, 2 and 3. Thus, these results suggested the roles of YTHDFs as potential prognostic predictors in LUAD patients with different subclasses.

Predicted Functions and Pathways of the Mutations in YTHDFs and their 120 Frequently Altered Neighbor Genes in LUAD Patients
We further investigated the genetic alteration in YTHDFs in LUAD patients. As shown in Figure 5A, a high mutation rate (87%) of YTHDFs was observed in LUAD patients (cBioPortal). YTHDF1, YTHDF2, and YTHDF3 genes display high genetic alterations with mutation rates being 36%, 19%, and 32%, respectively. After  binding proteins, and demethylases [11,12]. The writer is composed of the enzymatic core components METTL3 and METTL14 and auxiliary proteins WTAP, VIRMA, FLACC, RBM15, and HAKAI. The writer is composed of YTHDF1/2/3, YTHDC1/2, and METTLs [11,12,25,26]. The binding of readers results in alterations of the translation efficiency and stability of m6A-containing RNAs [27].
FTO and ALKBH5 serve as eraser proteins that remove the methyl group [28].
Increasing studies on these regulators construct an increased association between m6A regulatory abnormalities or genetic alterations and various human cancers [7,8,29]. M6A demethylase FTO is reported as a prognostic factor in LUSC and facilitates cell proliferation and invasion despite inhibiting cell apoptosis by regulating MZF1 expressions [30]. METTL3 utilizes increased EGFR and TAZ expressions and promotes cell growth, survival, and invasion in order to facilitate a role as an oncogene in lung cancer [31]. mRNA circularization caused by METTL3-eIF3 promotes the translation and oncogenesis of LUAD [32]. However, the cohort numbers in patients from stage 3, and patients with chemotherapy were relatively small. Thus, the statistical analysis from these groups were not accurate. Further analysis by using larger cohorts would be necessary to address the prognostic role of YTHDFs in LUAD patients in these subclasses.
YTHDF1 is based in the cytoplasm, where the m6A binding protein facilitates the translation efficiency of m6A-modified mRNAs [36]. YTHDF2, after being targeted to a specific site via m6A recognition, recruits the CCR4-NOT deadenylase complex to destabilize and further decay target m6 A-modified transcripts [27]. YTHDF2 also serves m1A readers and destabilizes known m1A-containing RNAs [37]. YTHDF3 is another cytoplasmic m6A binder that promotes protein synthesis in synergy with YTHDF1 and affects methylated mRNA decay mediated through YTHDF2 [15]. This indicates that YTHDF3 plays critical roles in accelerating the metabolism of m6A-modified mRNAs in the cytoplasm while in conjunction with YTHDF1 and YTHDF2 proteins [15]. However, observances in m6A perturbations attributed to m6A regulators such as these implicate a close association between aberrant m6A modifications and human cancers. m6A reader YTHDF3 is reported to correlate with the activation of several oncogenic pathways including protein secretion, androgen response, and the TGF-β signaling pathway [35,38]. In this study, we proved that lower expressions of YTHDF1/2/3 predict worse survival conditions in LUAD patients, suggesting that YTHDFs suppressed the pathological progression in LUAD. YTHDF1/2 have been proved to inhibit the proliferation and migration of endosomal cancer cells via PHLPP2 and MTOR2 dependent pathways [7,39,40].
By using the protein network analysis, we identified that the expression of MTOR and other mRNA metabolic process genes were significantly related to YTHDFs mutations. This data suggests that the effect of YTHDFs on the suppression of LUAD may also occur through regulating MTOR dependent pathways. To address the detailed mechanisms of YTHDFs on LUAD progression, further molecular biological studies are still needed.
In summary, we systematically analyzed the expression profiles and prognostic values of the YTHDF1/2/3 in LUAD and LUSC patients. Our results revealed that YTHDF1, YTHDF2, and YTHDF3 might be useful markers for prognostic stratification for LUAD patients but not for LUSC patients. Our analysis also lays a foundation for the development of LUAD therapeutic strategies based on RNA methylation.