Open Access

Investigating the correlation between DNA methylation and immune‑associated genes of lung adenocarcinoma based on a competing endogenous RNA network

  • Authors:
    • Chun Chang
    • Wei Kong
    • Xiaoyang Mou
    • Shuaiqun Wang
  • View Affiliations

  • Published online on: August 19, 2020     https://doi.org/10.3892/mmr.2020.11445
  • Pages: 3173-3182
  • Copyright: © Chang et al. This is an open access article distributed under the terms of Creative Commons Attribution License.

Metrics: Total Views: 0 (Spandidos Publications: | PMC Statistics: )
Total PDF Downloads: 0 (Spandidos Publications: | PMC Statistics: )


Abstract

In recent years, there have been major breakthroughs in immunotherapies for the treatment of cancer. However, different patients have different responses to immunotherapy. Numerous studies have shown that the accumulation of epigenetic abnormalities, such as DNA methylation, serve an important role in the immune response of lung adenocarcinoma (LUAD). To investigate the effects of DNA methylation on tumor immunity with survival and prognosis, relevant studies can be performed based on the regulatory mechanisms of RNA molecules. For example, long non‑coding RNAs (lncRNAs), which regulate gene expression through epigenetic levels. By constructing an immune-associated competitive endogenous RNA (ceRNA) network, the present study identified the regulatory associations among 3 key immune‑associations mRNAs, 2 microRNAs (miRs) and 29 lncRNAs that were closely associated with the prognosis of patients with LUAD. The molecular biology analysis indicated that hypomethylation of the 1101320‑1104290 regions of chromosome 1 resulted in the low expression levels of LINC00337 and that LINC00337 may affect the expression levels of CHEK1 by competitively binding with human (has)‑miR‑373 and hsa‑miR‑195. Therefore, abnormal DNA methylation in lncRNA‑associated regions caused their abnormal expression levels, which further affected the interactions between RNA molecules. The interactions between these RNA molecules may have regulatory effects on tumor immunity and the prognosis of patients with LUAD.

Introduction

Studies have reported some key breakthroughs in tumor immunological therapies for lung cancer (1,2). For example, a blockade of immune checkpoint therapy regulated T cell activity to enhance the antibody-tumor immune responses. Significant advances in the molecular characterizations of lung cancer have led to the creation of effective immunotherapies that assist in the recognition of cancer as foreign by the host immune system, stimulate the immune system and relieve the inhibition that allows tumor growth and spread (3). However, immunotherapies have a number of limitations in clinical applications due to the existence of tumor heterogeneity and immunological suppression (4,5). In a previous study that focused on tumor immunological genomics, the assessment of tumor immune infiltrating cells served a key role in investigating the way in which the interactions between the tumor and the immune system affect patient outcome (6).

In an investigation into lung cancer and its immune regulations, a previous study demonstrated the significance of CpG islands in the lung immune responses of granulomatous lung diseases (7). Simultaneously, the study indicated that long non-coding RNAs (lncRNAs) serve an important role in directing the development of various immune cells and controlling dynamic transcriptional programs and were hallmarks of immune cell activation (7). In addition to the direct regulation of lncRNAs by DNA methylation via interactions with their promoter regions, previous studies have revealed several intricate regulatory associations between lncRNAs and DNA methylation (8,9). In the cis-methylated lncRNAs group, DNA methylation adjacent to a lncRNA locus, such as the promoter or the imprinting control region, directly modulate the expression levels of target lncRNAs as a cis-regulator (10). In trans-methylation due to lncRNAs, lncRNAs regulate the DNA methylation of a trans-genomic locus as an intermedium by recruiting DNA (cytosine-5-)-methyltransferase (11,12). In the trans-methylation-regulated lncRNAs group, an alteration in the DNA methylation state at a specific genomic locus regulates the transcription of its antisense-oriented lncRNAs (13,14). The lncRNA extracoding CEBPA transcribed from the CEBPA gene locus is reported to be critical for the regulation of DNA methylation at this site through interactions with DNA methyltransferase 1 (15).

Based on the interactions between DNA methylation and lncRNAs, it is easy to investigate the role of methylation on the tumor immune responses from the perspective of lncRNAs. Currently, the competitive endogenous (ce)RNA hypothesis provides an approach to study the regulatory mechanisms of RNA molecules (16). ceRNA is an RNA with a binding site that competitively binds to a micro (mi)RNA to inhibit its target gene regulation (16). Pseudogenes, lncRNAs, miRNAs and other types of RNAs can be regulated by ceRNA mechanisms (7). Therefore, the present study investigated the roles of DNA methylation in tumor immune cell infiltration abundance and its association with patient prognosis from the novel perspective of RNA molecules, DNA methylation and their regulatory mechanisms.

The present study investigated the cause behind the differences in tumor immune responses from a new perspective. The present study investigated how DNA methylation affected the interactions between tumors and immune cells and between RNA molecules. Starting with DNA methylation, the present study revealed DNA methylation abnormalities, differentially expressed RNA (DERNA) molecules and mutual regulatory associations between DNA methylation and RNA molecules and assessed potential associations between DNA methylation and tumor immune infiltration in LUAD.

Materials and methods

Acquisition and preprocessing of the data

LUAD transcriptome, miRNA, DNA methylation and associated clinical data were downloaded from The Cancer Genome Atlas (TCGA) website (TCGA-LUAD; portal.gdc.cancer.gov/). The transcriptome data contained 594 samples (59 normal samples and 535 tumor samples), the miRNA data contained 577 samples (46 normal samples and 531 tumor samples) and the DNA methylation data included 507 samples (30 normal samples and 477 tumor samples). These data were downloaded on April 1, 2018. For the original data, most of the data with RNA expression level=0 were deleted. For the transcriptome data, the RSEM-preprocessed per million transcript method was used to isolate and obtain lncRNA and mRNA expression data (17).

Extraction of the differential methylation regions and associated lncRNAs

For the preprocessed DNA methylation data, differential DNA methylation regions and associated lncRNAs were identified using the lncDM algorithm in a package in R 3.5.3 in March 11, 2019. LncDM is a novel computational method for identifying differentially methylated regions in specific diseases. LncDM uses the Illumina HumanMethylation450 BeadChip GENCODE of all annotation information based on the reannotation method and obtains the CpG values of the differential transcript (18). The gene functional regions were used to calculate the patterns of differential methylation. The screening criteria for the differentially methylated regions were adjusted to P<0.05 and β diffcut >0.3. The β diffcut value was used to reflect the differential situations between samples.

Extraction of the abnormally expressed RNA molecules

For those RNA molecules in original data that were subjected to deletion processing, DERNA were screened using the DEseq2 algorithm in R software. The DEseq2 method uses shrinkage to estimate fold change, thus improving the stability and interpretability of the DEseq-based estimation (19). The screening criteria for the DERNA were |log FC|>2 and P<0.05.

Selection of the immune cell-associated genes

To investigate the cause of tumor heterogeneity and its association with prognosis, the immune infiltration abundance of relevant LUAD tumor samples needed to be calculated. In this process, the TIMER 1.0 algorithm was used (20). First, the CHAT package was used to estimate the tumor purity of the samples (21), then the COMBAT package was used to combine all relevant immune cell gene expression data to eliminate the effect of batch processing of different data (22). Genes negatively correlated with tumor purity were selected and matched with some known immune cell (B cell, CD4+ and CD8+ T cell, neutrophil, macrophage and dendritic cell) tag genes to obtain 6 immune cell-associated genes.

Construction of the DNA methylation and immune-associated ceRNA network

To construct the ceRNA network, multiple databases were used to investigate the inter-relationships between DERNA molecules and DNA methylation. First, the differential methylation region (DMR)-DElncRNA relationship pairs were identified by employing the Lnc2Meth 1.0 (bio-bigdata.hrbmu.edu.cn/Lnc2Meth/) database (18). Next, the DElncRNA-DEmiRNA relationship pairs were identified by searching the miRcode 1.0 (mircode.org/) database (23). miRNAs were then annotated with 3p and 5p, which were used to distinguish between the 3′ and 5′ arms of miRNA precursors that target different sites, using the Starbase v2.0 (starbase.sysu.edu.cn/) database (24). Finally, the DEmiRNA-DEmRNA relationship pairs (2527) were identified using miRDB v3.0 (mirdb.org/), miRTarBase 1.0 (mirtarbase.mbc.nctu.edu.tw/) and TargetScan v7.2 (targetscan.org/vert_72/). The visualization of this ceRNA network was achieved using Cytoscape version 3.6.0 software (28). In order to study the overall survival for the associated RNA molecules in the network, the present study used the Renyi test (29), a weighted test method, to generate survival curves.

Clinical analysis of the immune cell genes associated with DNA methylation

In order to investigate the effects of DNA methylation-associated tumor immune infiltration on clinical data, a multivariate regression model with associated factors, such as infiltration abundance, tumor purity, stage, infection status, age and sex was constructed. In this process, a single-factor Cox analysis was performed on the differential DNA methylation-associated immune genes. Those associated mRNAs with P-value <0.0001 were screened as clinically relevant genes. A multivariate risk regression analysis was then performed based on these clinically relevant genes. The theoretical basis of this model was the semiparametric regression model proposed by the British statistician D.R. Cox (30). The dependent variables of this model were the overall survival rates and result of survival. This model had the advantage of analyzing the influences of multiple factors on overall survival rates and the truncation analysis of overall survival rates simultaneously. Finally, Kaplan-Meier curves were used to verify the overall survival rates within 5 years. Kaplan-Meier curves were generated using the log-rank test. These aforementioned processes were completed with the survival package 3.2–3 of R software (CRAN.R-project.org/package=survival).

Functional enrichment analysis

For the immune genes associated with DNA methylation, the biological functions of genes were analyzed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) functional enrichment (geneontology.org/). The KEGG (kegg.jp/) and GO functional enrichment analyses were performed by using the Database for Annotation, Visualization and Integrated Discovery (DAVID; v6.8) website (31).

Results

Extraction of DERNAs and differential DNA methylation regions

To understand the RNAs and methylation regions that have undergone significant changes in tumor and control tissues, differential data were first extracted. The extraction process of differential data involved extracting DERNAs, DNA methylation regions and the regulatory correlations between DNA methylation regions and lncRNAs. In the process of extraction, DERNA molecules in LUAD, such as DElncRNAs, DEmiRNAs and DEmRNAs were screened using the DEseq2 method in R software. The screening criteria for DERNA molecules were |log FC|>2 and P<0.05 (Fig. 1). A total of 1,187 DElncRNAs, 133 DEmiRNAs and 1,474 DEmRNAs were obtained. The extracted RNA molecules were increased or decreased expressed in the patients with LUAD. It can be inferred that the abnormal expression levels of RNA molecules were associated with LUAD.

The processed DNA methylation data were used to screen for differential DNA methylation regions and the corresponding lncRNAs using the lncDM method in R software. Those methylation regions were removed when methylation differences were <0.05. The detailed information regarding the DMRs included the methylation regions, the associated lncRNAs and the methylation difference are presented in Table SI. During the process, 1,692 DMRs associated with the lncRNAs were extracted. It was hypothesized that the increased or decreased expression levels of lncRNAs were associated with abnormal methylation.

Construction of the immune-associated ceRNA network

In order to investigate the regulatory mechanisms of gene expression associated with immune cells from the perspective of DNA methylation, a DNA methylation and immune-associated ceRNA network was constructed based on the internal competitive RNA hypothesis (Fig. 2). By matching the Lnc2Meth 1.0 database, 1,692 DMR-DElncRNA relationship pairs were identified, of which there were 64 lncRNAs. In addition, 285 DElncRNA-DEmiRNA relationship pairs were identified by matching the Micode 1.0 database, of which there were 21 miRNAs and 64 lncRNAs in DElncRNA-DEmiRNA relationship pairs. In addition, 1,177 DEmiRNA-DEmRNA relationship pairs were obtained by searching the databases of interactions between miRNAs and mRNAs, of which there were 21 miRNAs and 92 mRNAs.

Then, the RNA molecules in the aforementioned relationship pairs were matched with the increased or decreased expressed RNA molecules yielding abnormally expressed relationship pairs. The abnormal expression levels of RNA molecules in relationship pairs are presented in Tables SII, SIII and SIV. The expression levels of CHEK1, SLC7A11 and CCNE1 were much lower in LUAD samples compared with normal samples (Table SII). The expression levels of hsa-miR-373 and hsa-miR-195 were much lower in LUAD samples compared with normal samples (Table SIII). The expression levels of LINC00377, LINC00473 and LINC00355 were much lower in LUAD samples compared with normal samples (Table SIV). The changes in DNA methylation regions and corresponding RNAs of the ceRNA network are presented in Table SV.

According to the abnormally expressed relationship pairs, the ceRNA network was constructed. In order to improve the current understanding of the ceRNA network, the visualization of this ceRNA network was achieved using Cytoscape version 3.6.0 software (27). This network included 3 mRNAs, 2 miRNAs and 29 lncRNAs. It can be inferred that low expression levels of the lncRNA LINC00337 were associated with hypomethylation in the corresponding DNA regions (Fig. 2). The miRNAs (hsa-miR-372, hsa-miR-373) can bind to mRNAs (SLC7A11, CHEK1 and CCNE1). The lncRNAs (LINC00337 and LINC00355) may bind to miRNAs (Fig. 2).

The associated RNA molecules in this network were critical for investigating immunity and DNA methylation. The present study used the Renyi test (29), a weighted test method, to generate survival curves for the associated RNA molecules in the network (Fig. 3). In the present study, the Renyi test weight is 1.

The criterion for expression levels of RNA molecules was derived from the criteria in the premier increased or decreased expression level extraction of RNA molecules. The screening criteria for DERNA molecules were |log FC|>2 and P<0.05, which indicated that the expression levels of the RNAs were evidently increased or decreased in LUAD samples. Survival curves of all the RNA in the network were generated and survival curves with P<0.05 were selected as this indicated that abnormal RNA expression levels were significantly associated with survival rates in the LUAD samples.

The patients with LUAD that exhibited low expression levels of mRNA CCNE1 (Fig. 3E), CHEK1 (Fig. 3F) and SLC7A11 (Fig. 3G) had improved survival rate. The patients with LUAD that exhibited high expression levels of TCEAL3-AS1 (Fig. 3C) had improved survival time. The patients with LUAD that exhibited low expression levels of lncRNA LINC00337 had improved survival rate (Fig. 3A). These results indicated that the increased or decreased expression levels of immune cells genes affected the survival rate of patients with LUAD and suggested that the interactions between DNA methylation and RNA molecules may affect the responses to immunotherapies. The associated mRNAs can be considered potential markers of immunotherapies in LUAD. It was also verified that the RNAs in the network influenced the survival rate of patients with LUAD.

In order to investigate the expressional relationships between the RNA molecules, correlational analyses of the potential mRNAs and lncRNAs in the network were performed. There may be several competitive relationships in these RNA molecules. The expression levels of CHEK1 were positively correlated with all the associated lncRNA molecules including LINC00473, LINC00337 and LINC00355. Among them, CHEK1 was strongly correlated with LINC00337 (Fig. 4A). There were also positive correlations between the expression levels of CCNE1 and the majority of its associated lncRNAs, including C20orf48, LINC00377 and TCL6 (Fig. 4B). These results verified the hypothesis of the present study based on the ceRNA hypothesis. It can be inferred that LINC00337 affects the expression levels of CHEK1 by competing with the binding of hsa-miR-373 and hsa-miR-195 from the ceRNA network and competitive relationships. Concerning the changes in DNA methylation regions and the corresponding lncRNAs, it can be deduced that the hypomethylation of the 1101320–1104290 regions of chromosome 1 resulted in the low expression levels of LINC00337, and that LINC00337 may have affected the expression levels of CHEK1 by competitively binding with hsa-miR-373 and hsa-miR-195, thereby affecting tumor immune cells and overall survival rates in patients with LUAD.

DNA methylation-associated immune genes analysis

In the process of constructing the ceRNA network, the correlations between RNA molecules were obtained by searching a large number of databases. In addition, 181 mRNAs associated with tumor immunity and clinical outcomes were identified. In order to investigate how these associated mRNAs influence survival rates in patients with LUAD, a multivariate Cox regression model was performed using 8 mRNA expression values (SEMA7A, ENTPD1, ORMDL3, STK17B, B4GALT1, ALK, CCNB1, SMAD3) to reflect the associations between gene expression levels and the survival rates in patients with LUAD (Fig. 5A). A survival rates analysis between the high-risk group and the low-risk group was conducted (Fig. 5B). The samples were divided based on the median value of the risk values. The risk value was derived from the multivariate Cox regression analysis. The cut-off value for high-risk samples was a risk value >0.956. Based on the survival curves between high-risk and low-risk patients, it can be seen that low-risk patients have improved overall survival rates. In the present study, the area of the ROC curves was 0.716 (Fig. 5C). The area of the ROC curve indicated the accuracy of diagnostic methods. The criterion of trustworthy accuracy was when the area of the ROC curves >0.7. The prediction of survival outcomes had reliable accuracy in patients with LUAD. The results demonstrated that the expression levels of the DNA methylation-associated immune cells genes affected the prognosis of patients with LUAD.

Finally, the functional enrichment analysis of DNA methylation-associated immune cells genes was conducted using the DAVID website. Some pathways were found to be associated with immune processes, including the pathway terms ‘B cell receptor signaling pathway’ and the ‘T cell receptor signaling pathway’. Several biological processes involving immune processes were also obtained, including the GO terms ‘Positive regulation of cell migration’ and the ‘T cell receptor signaling pathway’ (Fig. 6). The results of the enrichment analysis also indicated that these associated genes served an important role in LUAD tumor immunity. From these results, it can be inferred that DNA methylation may have regulatory effects on the expression levels of the immune cells genes.

Discussion

Understanding tumor immune cells infiltration is crucial for investigating the differences in responses to cancer immunotherapies and for developing effective immunotherapies. DNA methylation may explain the variabilities in immune cells of lung cancer in response to antigens (7). However, it is not completely clear how DNA methylation affected tumor immune cell infiltration. The present study investigated how DNA methylation affected the interactions between tumors and immune cells from the perspective of DNA methylation affecting the interactions between RNA molecules. A ceRNA network with associated DNA methylation and immune cell genes was constructed to study the associations between DNA methylation and the expression levels of immune cell genes.

In the present study, differential DNA methylation regions and DERNA molecules in lung adenocarcinoma (LUAD) samples were extracted. lncDM software (lncdm.r-forge.r-project.org/) was used to extract differential DNA methylation regions and their corresponding lncRNAs (18). The DEseq2 method was used to extract DElncRNAs, DEmiRNAs and DEmRNAs in LUAD samples (19). Then, to study the effects of differential DNA methylation on tumor immune infiltrating cells, tumor immune cells infiltration abundance was calculated. When calculating the infiltration abundance of tumor immune cells, CIBERSORTx (cibersortx.stanford.edu/) outperformed other algorithms in eliminating noise, estimating unknown mixture contents and accuracy. However, the algorithm was prone to bias due to the statistical collinear effect generated by the regression analysis, whereas the TIMER 1.0 algorithm (cistrome.dfci.harvard.edu/TIMER/) blocked immunological marker genes and highly expressed genes to eliminate bias effects and collinearity between immune cells, ensuring the accuracy of reasoning (20). Thus, the TIMER 1.0 algorithm was applied in the present study. Databases of molecular associations between RNAs (Lnc2Meth 1.0, miRcode 1.0, Starbase v2.0, miRDB v3.0, miRTarBase 1.0 and TargetScan v7.2) were used to investigate the associations between RNA molecules and differential DNA methylation regions (13,2327). Based on the ceRNA hypothesis, a ceRNA regulatory network of immune genes was established to determine the regulatory associations between DNA methylation and RNA molecules in LUAD. By using mRNAs expressional data and clinical data associated with differential DNA methylation regions, a regression model was constructed to investigate the associations between methylation-associated tumor immune cells infiltration and overall survival rates (30).

Compared with previous studies (68) that focus on the tumor immune cells infiltration, the present study further analyzed the regulatory roles in it. It was hypothesized that the hypomethylation of the 1101320–1104290 regions of chromosome 1 may affect the expression levels of CHEK1, thereby affecting tumor immune cells and survival outcomes in patients with LUAD. The ceRNA network was further enriched from the perspective of DNA methylation. From the present study, it can be deduced that the hypomethylation of the 1101320–1104290 regions of chromosome 1 resulted in the low expression levels of LINC00337, and that LINC00337 may affect the expression levels of CHEK1 by competitively binding with hsa-miR-373 and hsa-miR-195.

The present study was performed based on theoretical research described in previous studies (79,16). There is evidence that lncRNAs play important roles in directing the development of various immune cells and controlling dynamic transcriptional programs, which are hallmarks of immune cell activation (7). Additionally, a ceRNA is an RNA with an miRNA binding site that competitively binds to an miRNA to inhibit its target gene regulation (16). Pseudogenes, lncRNAs, miRNAs and other types of RNA, can be regulated by ceRNA mechanisms (7). Therefore, the ceRNA hypothesis provided an approach to investigate the regulatory mechanisms of RNA molecules. In addition, it was possible to construct a ceRNA network using existing databases of the interactions between DNA methylation and RNA molecules. In addition, previous studies have shown that DNA methylation directly regulates lncRNAs through interactions with their promoter regions (8,9). For example, in the cis-methylated lncRNAs group, DNA methylation adjacent to an lncRNA locus, such as the promoter or the imprinting control region, directly modulates the expression levels of target lncRNAs as a cis-regulator (9).

DERNA molecules and DEDNA methylation regions associated with the lncRNAs were extracted in the present study to understand the RNAs and methylation regions that have undergone significant changes in tumor and control tissues. In order to investigate the regulatory mechanisms of genes expression levels associated with immune cells from the perspective of DNA methylation, a DNA methylation and immune-associated ceRNA network was constructed based on the internal competitive RNA hypothesis. Next, the associations between the expression levels of RNA molecules in the network and the survival rates of patients were investigated. Then, correlational analyses of potential lncRNA-mRNA pairs in LUAD was conducted to further investigate the competitive relationships. Subsequently, how these associated genes affected survival rates was investigated in the present study, and analyses of the DNA methylation-associated immune cells genes were performed. Finally, enrichment analyses of DNA methylation-associated immune cells genes were performed.

According to the analysis of the results, it can be inferred that hypomethylation of the 1101320–1104290 regions of chromosome 1 resulted in the low expression levels of LINC00337, and that LINC00337 may affect the expression levels of CHEK1 by competitively binding with hsa-miR-373 and hsa-miR-195, thereby affecting tumor immune cells and survival outcomes in patients with LUAD.

To verify the hypothesis, several studies (3242) on the RNA molecules in the ceRNA network were searched. Previous studies (3239) have also indicated that these RNAs showed important correlations in cancer, tumor immunity and DNA methylation. For example, CCNE1 is significantly associated with TRIM58 methylation and the treatment of lung squamous cell carcinoma and may be used as a potential prognostic biomarker (32). Previous studies (3335) have also shown that CCNE1 is the most common amplified gene in lung cancer, and that its gene amplification was a therapeutic target for survival and lung cancer. Therefore, CCNE1-targeted therapy may be beneficial for CCNE1 amplification in patients with lung cancer (33). HOXC13, which is directly targeted by miR-141, is highly expressed in LUAD and promotes the proliferation of LUAD by modulating the expression of CCNE1 (34). Cyclin E1, encoded by the CCNE1 gene, promotes G1/S phase transition and chromosome instability (35). Cyclin E1 is downregulated by both miR-497 and miR-34a, which synergistically retard the growth of human lung cancer cells (35).

In addition, smoking can induce the high expression levels of SLC7A11 in oral cancer cells, which suggested that the expression levels of SLC7A11 may encourage lung tumor progression (3638). In addition, a previous study (36) demonstrated that SLC7A11 regulates metabolic requirements during lung cancer progression and is a potential therapeutic target in non-small cell lung cancer. The transporter gene SLC7A11 exhibits significant changes in enhancer DNA methylation and gene expression levels during pregnancy (37). In the present study, marker SLC7A11 expression levels were determined in lung cancer cell lines using microarray data and Kaplan-Meier survival analyses were performed for each marker using patient clinical data (38). High expression levels of the marker SLC7A11 are significantly associated with worse survival rates (38).

In a previous study (39), among the five associated pathways in LUAD, the p53 signaling pathway was the most significant, with CHEK1 being identified as a hub gene. Previous studies (3943) have also shown that miRNAs (hsa-miR-195 and hsa-miR-372) in the network play important roles in tumor immunity and DNA methylation. For example, hsa-miR-195 suppresses tumor growth and is associated with improved survival outcomes in several types of malignancy, including non-small cell lung cancer. The expression of hsa-miR-195 is lower in tumor tissues and is associated with poor survival outcomes (39). The overexpression of miR-195 suppresses tumor cell growth, migration and invasion. CHEK1 is a direct target of miR-195, which decreases the expression levels of CHEK1 in lung cancer cells. Therefore, the high expression levels of CHEK1 in lung tumors are associated with poor overall survival rates (39).

In addition, hsa-miR-195 suppresses non-small lung cell cancer and is a predictive factor for lung cancer prognosis (39). A previous study (40) demonstrated that the expression levels of hsa-miR-195 and O-6-methylguanine-DNA methyltransferase methylation status are associated with the clinical outcomes of the primary malignancy (40). In addition, miR-372-3p enhances the proliferation and invasion of lung squamous cell carcinoma cells by inhibiting FGF9 (41). A previous study (42) conducted a comparative proteomic analysis of non-small-cell lung carcinoma (NSCLC) CL1-0 cells expressing miRNA-372 and/or a vector only using two-dimensional gel electrophoresis, two-dimensional difference gel electrophoresis and liquid chromatography tandem mass spectrometry. Proteins identified as up- or downregulated were further classified according to their biological functions (42). A number of the proteins identified in the study (42) may be potential diagnostic biomarkers of NSCLC, particularly phosphorylated eIF4A-I. In the corresponding study (43), to clarify the molecular mechanisms underlying the tumor suppressive roles of propofol in the human lung cancer cell line A549, the study detected the expression levels of miR-372 and analyzed the mediating effects of miR-372 on the proliferation and metastasis of A549 cells (43). A previous study demonstrated that the expression levels of LINC00473 can serve as a robust biomarker for the tumor LKB1 functional status that can be integrated into clinical trials for patient selection and treatment evaluation, and implicated LINC00473 as a therapeutic target for LKB1-inactivated NSCLC (44).

The present study was primarily focused on the biological information analyses section, and so all the results were from the biological information analyses without biological verifications; for example, the direct validation experiments of the associations between DNA methylation and the expression levels of immune genes, the direct validation experiments of the associations between DNA methylation and tumor immune cells and survival outcomes in patients with LUAD. However, in order to verify the hypothesis of the present study, relevant studies were searched and some analyses were added in this discussion section.

The present study investigated the way in which DNA methylation affects the reactions between tumors and immune cells by constructing a ceRNA network. To further determine how the DNA methylation regions influenced the expression levels of immune genes by regulating the ceRNA network, the associations between the expression levels of RNAs, the overall survival rates of patients with LUAD and the competitive relationships between RNAs were investigated. Finally, a multivariate Cox regression model was constructed using 8 mRNAs expression values to investigate how these associated genes affected survival in patients with LUAD. The results revealed that hypomethylation of the 1101320–1104290 regions of chromosome 1 resulted in the low expression levels of LINC00337 and that LINC00337 may affect the expression levels of CHEK1 by competitively binding with hsa-miR-373 and hsa-miR-195, thereby affecting tumor immune cells and survival outcomes in patients with LUAD. The RNA molecules in the ceRNA network served important roles in DNA methylation and immunotherapies and can be used as biomarkers of immunotherapies in patients with LUAD. The results of the present study may provide a new perspective for the exploration of cancer-immune interactions.

Supplementary Material

Supporting Data
Supporting Data
Supporting Data
Supporting Data
Supporting Data

Acknowledgements

The authors would like to thank Dr Deng Jin and Mr Wei Boyang (both Shanghai Maritime University, Shanghai, China) for their technical assistance and writing assistance during the research process.

Funding

The present study was supported by the National Natural Science Foundation of China (grant no. 61803257) and Natural Science Foundation of Shanghai (grant no. 18ZR1417200).

Availability of data and materials

LUAD transcriptome, miRNA, DNA methylation data and associated clinical data can be downloaded from The Cancer Genome Atlas (TCGA) website (TCGA-LUAD; portal.gdc.cancer.gov/).

Authors' contributions

CC, WK, XM and SW conceived and designed the present study. CC and WK conducted analyses. All authors read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References

1 

Sharma P, Wagner K, Wolchok JD and Allison JP: Novel cancer immunotherapy agents with survival benefit: Recent successes and next steps. Nat Rev Cancer. 11:3173–812. 2011. View Article : Google Scholar

2 

Pardoll DM: The blockade of immune checkpoints in cancer immunotherapy. Nat Rev Cancer. 12:252–264. 2012. View Article : Google Scholar : PubMed/NCBI

3 

Carter BW, Halpenny DF, Ginsberg MS, Papadimitrakopoulou VA and de Groot PM: Immunotherapy in Non-small cell lung cancer treatment: Current status and the role of imaging. J Thorac Imaging. 32:300–312. 2017. View Article : Google Scholar : PubMed/NCBI

4 

Di Giacomo AM, Calabrò L, Danielli R, Fonsatti E, Bertocci E, Pesce I, Fazio C, Cutaia O, Giannarelli D, Miracco C, et al: Long-term survival and immunological parameters in metastatic melanoma; patients who responded to ipilimumab 10 mg/kg within an expanded access; programme. Cancer Immunol Immunother. 62:1021–1028. 2013. View Article : Google Scholar : PubMed/NCBI

5 

Prieto PA, Yang JC, Sherry RM, Hughes MS, Kammula US, White DE, Levy CL, Rosenberg SA and Phan GQ: CTLA-4 blockade with ipilimumab: Long-term follow-up of 177 patients with metastatic melanoma. Clin Cancer Res. 18:2039–2047. 2012. View Article : Google Scholar : PubMed/NCBI

6 

Liu XS and Mardis ER: Applications of immunogenomics to cancer. Cell. 168:600–612. 2017. View Article : Google Scholar : PubMed/NCBI

7 

Yang IV, Konigsberg I, Macphail K, Li L, Davidson EJ, Mroz PM, Hamzeh N, Gillespie M, Silveira LJ, Fingerlin TE and Maier LA: DNA methylation changes in lung immune cells are associated with granulomatous lung disease. Am J Respir Cell Mol Biol. 60:96–105. 2019. View Article : Google Scholar : PubMed/NCBI

8 

Gudrun B, Rowley MJ, Kuciński J, Zhu Y, Amies I and Wierzbicki AT: RNA-directed DNA methylation requires stepwise binding of silencing factors to long non-coding RNA. Plant J Cell Mol Biol. 79:181–191. 2014. View Article : Google Scholar

9 

Hui Z, Shangwei N, Xiang L, Yuyun L, Wei W and Xia L: A novel reannotation strategy for dissecting DNA methylation patterns of human long intergenic non-coding RNAs in cancers. Nucleic Acids Res. 42:82582014. View Article : Google Scholar : PubMed/NCBI

10 

Bohne F, Langer D, Martiné U, Eider CS, Cencic R, Begemann M, Elbracht M, Bülow L, Eggermann T, Zechner U, et al: Kaiso mediates human ICR1 methylation maintenance and H19 transcriptional fine regulation. Clin Epigenetics. 8:472016. View Article : Google Scholar : PubMed/NCBI

11 

Vennin C, Spruyt N, Robin YM, Chassat T, Le Bourhis X and Adriaenssens E: The long non-coding RNA 91H increases aggressive phenotype of breast cancer cells and up-regulates H19/IGF2 expression through epigenetic modifications. Cancer Lett. 385:198–206. 2016. View Article : Google Scholar : PubMed/NCBI

12 

Yu F, Chen B, Dong P and Zheng J: HOTAIR epigenetically modulates PTEN expression via MicroRNA-29b: A novel mechanism in regulation of liver fibrosis. Mol Ther. 25:205–217. 2017. View Article : Google Scholar : PubMed/NCBI

13 

Gregory MC and Loeb DM: Hypoxia-sensitive epigenetic regulation of an antisense-oriented lncRNA controls WT1 expression in myeloid leukemia cells. PLoS One. 10:e01198372015. View Article : Google Scholar : PubMed/NCBI

14 

Tianyi G, Bangshun H, Yuqin P, Yeqiong X, Rui L, Qiwen D, Huilin S and Shukui W: Long non-coding RNA 91H contributes to the occurrence and progression of esophageal squamous cell carcinoma by inhibiting IGF2 expression. Chinese J Clin Lab Sci. 54:359–367. 2015.

15 

Ruscio AD, Ebralidze AK, Benoukraf T, Amabile G, Goff LA, Terragni J, Figueroa ME, Pontes LLDF, Alberich-Jorda M, Zhang P, et al: DNMT1-interacting RNAs block gene-specific DNA methylation. Nature. 503:371–376. 2013. View Article : Google Scholar : PubMed/NCBI

16 

Salmena L, Poliseno L, Tay Y, Kats L and Pandolfi PP: ceRNA hypothesis: The Rosetta stone of a hidden RNA language? Cell. 146:353–358. 2011. View Article : Google Scholar : PubMed/NCBI

17 

Li B and Dewey CN: RSEM: Accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics. 12:323. 2011. View Article : Google Scholar : PubMed/NCBI

18 

Zhi H, Li X, Wang P, Gao Y, Gao B, Zhou D, Zhang Y, Guo M, Yue M, Shen W, et al: Lnc2Meth: A manually curated database of regulatory relationships between long non-coding RNAs and DNA methylation associated with human disease. Nucleic Acids Res. 46:D133–D138. 2017. View Article : Google Scholar

19 

Love MI, Wolfgang H and Simon A: Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15:5502014. View Article : Google Scholar : PubMed/NCBI

20 

Li B, Severson E, Pignon JC, Zhao H, Li T, Novak J, Peng J, Hui S, Aster JC, Rodig S, et al: Comprehensive analyses of tumor immunity: Implications for cancer immunotherapy. Genome Biol. 17:1742016. View Article : Google Scholar : PubMed/NCBI

21 

Li B and Li JZ: A general framework for analyzing tumor subclonality using SNP array and DNA sequencing data. Genome Biol. 15:4732014. View Article : Google Scholar : PubMed/NCBI

22 

Chao C, Kay G, Judith B, Dandan Z, Elliot G, Li J and Chunyu L: Removing batch effects in analysis of expression microarray data: An evaluation of six batch adjustment methods. PLoS One. 6:e172382011. View Article : Google Scholar : PubMed/NCBI

23 

Ashwini J, Marks DS and Erik L: miRcode: A map of putative microRNA target sites in the long non-coding transcriptome. Bioinformatics. 28:2062–2063. 2012. View Article : Google Scholar : PubMed/NCBI

24 

Li JH, Liu S, Zhou H, Qu LH and Yang JH: starBase v2.0: Decoding miRNA-ceRNA, miRNA-ncRNA and protein-RNA interaction networks from large-scale CLIP-Seq data. Nucleic Acids Res. 42((Database Issue)): D92–D97. 2014. View Article : Google Scholar : PubMed/NCBI

25 

Agarwal V, Bell GW, Nam JW and Bartel DP: Predicting effective microRNA target sites in mammalian mRNAs. Elife. 4:e050052015. View Article : Google Scholar

26 

Chou CH, Shrestha S, Yang CD, Chang NW, Lin YL, Liao KW, Huang WC, Sun TH, Tu SJ, Lee WH, et al: miRTarBase update 2018: A resource for experimentally validated microRNA-target interactions. Nucleic Acids Res. 46:D296–D302. 2018. View Article : Google Scholar : PubMed/NCBI

27 

Wong N and Wang X: miRDB: An online resource for microRNA target prediction and functional annotations. Nucleic Acids Res. 43((Database Issue)): D146–D152. 2015. View Article : Google Scholar : PubMed/NCBI

28 

Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B and Ideker T: Cytoscape: A software environment for integrated models of biomolecular interaction networks. Genome Res. 13:2498–2504. 2003. View Article : Google Scholar : PubMed/NCBI

29 

Gill RD: Censoring and stochastic integrals. Statistica Neerlandica. 34:124. 1980. View Article : Google Scholar

30 

Fox J: Cox proportional-hazards regression for survival data. See Also. 371–372. 2012.

31 

Huang DW, Sherman BT and Lempicki RA: Bioinformatics enrichment tools: Paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 37:1–13. 2009. View Article : Google Scholar : PubMed/NCBI

32 

Zhang W, Cui Q, Qu W, Ding X, Jiang D and Liu H: TRIM58/cg26157385 methylation is associated with eight prognostic genes in lung squamous cell carcinoma. Oncol Rep. 40:206–216. 2018.PubMed/NCBI

33 

Nakayama K, Rahman MT, Rahman M, Nakamura K, Ishikawa M, Katagiri H, Sato E, Ishibashi T, Iida K, Ishikawa N and Kyo S: CCNE1amplification is associated with aggressive potential in endometrioid endometrial carcinomas. Int J Oncol. 48:506–516. 2016. View Article : Google Scholar : PubMed/NCBI

34 

Yao Y, Luo J, Sun Q, Xu T, Sun S, Chen M, Lin X, Qian Q, Zhang Y, Cao L, et al: HOXC13 promotes proliferation of lung adenocarcinoma via modulation of CCND1 and CCNE1. Am J Cancer Res. 7:1820–1834. 2017.PubMed/NCBI

35 

Han Z, Zhang Y, Yang Q, Liu B, Wu J, Zhang Y, Yang C and Jiang Y: miR-497 and miR-34a retard lung cancer growth by co-inhibiting cyclin E1 (CCNE1). Oncotarget. 6:13149–13163. 2015. View Article : Google Scholar : PubMed/NCBI

36 

Ji X, Qian J, Rahman SMJ, Siska PJ, Zou Y, Harris BK, Hoeksema MD, Trenary IA, Chen H, Eisenberg R, et al: xCT (SLC7A11)-mediated metabolic reprogramming promotes non-small cell lung cancer progression. Oncogene. 37:5007–5019. 2018. View Article : Google Scholar : PubMed/NCBI

37 

Simner C, Novakovic B, Lillycrop KA, Bell CG, Harvey NC, Cooper C, Saffery R, Lewis RM and Cleal JK: DNA methylation of amino acid transporter genes in the human placenta. Placenta. 60:64–73. 2017. View Article : Google Scholar : PubMed/NCBI

38 

Cohen AS, Khalil FK, Welsh EA, Schabath MB, Enkemann SA, Davis A, Zhou JM, Boulware DC, Kim J, Haura EB and Morse DL: Cell-surface marker discovery for lung cancer. Oncotarget. 8:113373–113402. 2017. View Article : Google Scholar : PubMed/NCBI

39 

Ben L, Jinli Q, Fangxiu X, Yan G, Yu W, Herbert Y and Biyun Q: MiR-195 suppresses non-small cell lung cancer by targeting CHEK1. Oncotarget. 6:9445–9456. 2015. View Article : Google Scholar : PubMed/NCBI

40 

Lakomy R, Sana J, Hankeova S, Fadrus P, Kren L, Lzicarova E, Svoboda M, Dolezelova H, Smrcka M, Vyzula R, et al: MiR-195, miR-196b, miR-181c, miR-21 expression levels and O-6-methylguanine-DNA methyltransferase methylation status are associated with clinical outcome in glioblastoma patients. Cancer Sci. 102:2186–2190. 2011. View Article : Google Scholar : PubMed/NCBI

41 

Wang Y, Hu Y, Wu G, Yang Y, Tang Y, Zhang W, Wang K, Liu Y, Wang X and Li T: Long noncoding RNA PCAT-14 induces proliferation and invasion by hepatocellular carcinoma cells by inducing methylation of miR-372. Oncotarget. 8:34429–34441. 2017. View Article : Google Scholar : PubMed/NCBI

42 

Lai JH, She TF, Juang YM, Tsay YG, Huang AH, Yu SL, Chen JJW and Lai CC: Comparative proteomic profiling of human lung adenocarcinoma cells (CL 1-0) expressing miR-372. Electrophoresis. 33:675–688. 2012. View Article : Google Scholar : PubMed/NCBI

43 

Sun H and Gao D: Propofol suppresses growth, migration and invasion of A549 cells by down-regulation of miR-372. BMC Cancer. 18:12522018. View Article : Google Scholar : PubMed/NCBI

44 

Chen Z, Li JL, Lin S, Cao C, Gimbrone NT, Yang R, Fu DA, Carper MB, Haura EB, Schabath MB, et al: cAMP/CREB-regulated LINC00473 marks LKB1-inactivated lung cancer and mediates tumor growth. J Clin Invest. 126:2267–2279. 2016. View Article : Google Scholar : PubMed/NCBI

Related Articles

Journal Cover

October-2020
Volume 22 Issue 4

Print ISSN: 1791-2997
Online ISSN:1791-3004

Sign up for eToc alerts

Recommend to Library

Copy and paste a formatted citation
x
Spandidos Publications style
Chang C, Kong W, Mou X and Wang S: Investigating the correlation between DNA methylation and immune‑associated genes of lung adenocarcinoma based on a competing endogenous RNA network. Mol Med Rep 22: 3173-3182, 2020
APA
Chang, C., Kong, W., Mou, X., & Wang, S. (2020). Investigating the correlation between DNA methylation and immune‑associated genes of lung adenocarcinoma based on a competing endogenous RNA network. Molecular Medicine Reports, 22, 3173-3182. https://doi.org/10.3892/mmr.2020.11445
MLA
Chang, C., Kong, W., Mou, X., Wang, S."Investigating the correlation between DNA methylation and immune‑associated genes of lung adenocarcinoma based on a competing endogenous RNA network". Molecular Medicine Reports 22.4 (2020): 3173-3182.
Chicago
Chang, C., Kong, W., Mou, X., Wang, S."Investigating the correlation between DNA methylation and immune‑associated genes of lung adenocarcinoma based on a competing endogenous RNA network". Molecular Medicine Reports 22, no. 4 (2020): 3173-3182. https://doi.org/10.3892/mmr.2020.11445