Role of downregulated ADARB1 in lung squamous cell carcinoma.

Non‑small cell lung cancer (NSCLC) is prevalent worldwide. Lung squamous cell carcinoma (LUSC) is one of the main subtypes of NSCLC yet, currently, few biomarkers are available for the diagnosis of LUSC. The present study aimed to investigate the expression and role of adenosine deaminase RNA specific B1 (ADARB1) in lung squamous cell carcinoma (LUSC). Integrative bioinformatics analysis was used to identify the effects of ADARB1 expression on the occurrence and prognosis of LUSC. The expression of ADARB1 was further examined by immunohistochemistry (IHC). Bioinformatics analysis suggested that ADARB1 was downregulated in LUSC, serving as a potential tumor suppressor, and these results were verified by IHC performed on a lung cancer tissue array. Clinical studies suggested that ADARB1 expression and methylation levels were significantly associated with patient characteristics in LUSC. Moreover, ADARB1 global methylation levels were upregulated in LUSC tissues compared with normal lung tissues. Higher methylation levels of cg24063645 were associated with shorter overall survival time of patients with LUSC. A negative correlation was identified between ADARB1 and epidermal growth factor receptor (EGFR) expression in LUSC. Using the Gene Expression Omnibus database, it was suggested that the expression of ADARB1 in LUSC was significantly different compared with that in lung adenocarcinoma. Furthermore, protein‑protein interactions were studied and a biological process annotation analysis was conducted. The present study suggested that ADARB1 was downregulated in LUSC; therefore, ADARB1 may serve as a specific biomarker and a potential therapeutic target for LUSC.

Introduction

Lung cancer was the leading cause of cancer-related death worldwide in 2015 and non-small cell lung cancer (NSCLC) accounts for ~85% of lung cancer cases (1). Lung squamous cell carcinoma (LUSC) is one of the main subtypes of NSCLC. A previous study indicated that the 5-year overall survival (OS) rate for patients with LUSC is 17.7% due to late diagnosis (2). In addition, there have been limited opinions for LUSC treatment compared to lung adenocarcinoma (LUAD) although progress has been made in the treatment of NSCLC (3–5). The reason is partly because there are more molecular abnormalities for LUSC in the discoidin domain receptor tyrosine kinase 2 and fibroblast growth factor receptor 1 genes (6). Early identification of the disease would reduce the mortality rate of patients with LUSC who could benefit most from effective therapies (7). However, few biomarkers are currently available for the effective diagnosis of LUSC, which would enable patients to receive timely treatment (8). Thus, identifying novel target molecules is essential for improving the diagnosis and treatment of LUSC.

Adenosine deaminase RNA specific B1 (ADARB1) is a member of the adenosine deaminases acting on RNA (ADAR) family, which exhibit enzymatic editing activity in a process known as adenosine-to-inosine (A-to-I) editing (9). Recently, research has indicated an association between ADARB1 and cancer. A previous study reported that ADARB1, an indispensable marker in brain development and function, could inhibit the proliferation and progression of glioblastoma (10). In terms of esophageal squamous cell carcinoma, Chen et al (11) reported that ADARB1 induces apoptosis and suppresses tumor growth by editing insulin-like growth factor binding protein 7 mRNA. Additionally, ADARB1 can regulate exon 12A, an exon from the heterogeneous nuclear ribonucleoprotein L like transcript, to promote cancer cell survival (12). However, the relationship between ADARB1 and LUSC has not been investigated.

Therefore, the present study aimed to investigate the role of ADARB1 in LUSC. The results suggested that ADARB1 was downregulated in LUSC tissues and cell lines. The expression and methylation status of ADARB1 were significantly associated with clinical characteristics and prognosis. Additionally, the GSE10245 dataset (13), acquired from the Gene Expression Omnibus (GEO) (14) database, suggested that ADARB1 expression levels in LUSC were lower than in LUAD. Finally, functional enrichment analysis of ADARB1-associated co-expression genes was conducted and a protein-protein interaction (PPI) network was established. The results of the Gene Ontology (GO) analysis suggested that the main biological processes (‘metabolic process’ and ‘biological regulation’), cellular components (‘nucleus’) and molecular functions (‘protein binding’) were associated with ADARB1. In addition, ADARB1 expression was significantly related to the Kyoto Encyclopedia of Genes and Genomes (KEGG) terms ‘ubiquitin mediated proteolysis’ and ‘Wnt signaling pathway’.

Materials and methods

Data acquisition and reanalysis using bioinformatics methods

Several bioinformatics web resources were used to reanalyze data concerning ADARB1 (Table SI). The Genome Mining (GE-mini) database was employed to examine gene expression profiles over a number of tissue types, including tumors (15). The Cancer Cell Line Encyclopedia (CCLE) project is a public database containing detailed genetic and pharmacological characterization of numerous cell lines (16). The GE-mini and CCLE databases were used to identify the expression profiles of ADARB1 in LUSC tissues and cell lines. Oncomine (17), University of Alabama Cancer Database (UALCAN) (18) and Gene Expression Profiling Interactive Analysis (GEPIA) (19) were used to verify the results. The GEPIA database was also used to identify the Pearson correlation between ADARB1 and epidermal growth factor receptor (EGFR) in LUSC.

Wanderer is an interactive platform, which provides human cancer gene expression and DNA methylation data. This platform was used to evaluate the relationship between ADARB1 levels and patient clinical characteristics (20). DiseaseMeth (version 2.0) is a human disease methylation database, which was used to analyze ADARB1 global methylation levels in LUSC (21). The MethSurv tool was used to analyze the association between ADARB1 methylation and LUSC prognosis by using the mean value to dichotomize methylation profiles of patients (22). The GSE10245 dataset, downloaded from the GEO database, provided ADARB1 expression profiles in LUAD and LUSC samples.

A dataset LUSC (TCGA, Nature 2012) (23) from the cBioportal (24) web tool was used to obtain gene co-expression data for ADARB1 in LUSC. The Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database (25) and Cytoscape software (26) were then used to build a PPI network of the co-expression genes. Moreover, GO and KEGG (27) pathway analyses were conducted, using the Web-based Gene Set Analysis Toolkit (WebGestalt) (28) and the Database for Annotation, Visualization and Integrated Discovery (DAVID) (29) bioinformatics resources, respectively.

Immunohistochemistry (IHC) analysis of a lung cancer tissue array

The LUC481 lung cancer tissue array and information regarding the cancer subtypes of the array were obtained from Fanpu Biotech, Inc. IHC was performed using the Histomouse SP Broad Spectrum DAB kit (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. Following antigen retrieval which was performed by microwaving in 10 mM citrate buffer (pH 6.0), paraffin-embedded sections were immunostained using a streptavidin peroxidase procedure. After blocking the samples with 3% bovine serum albumin (100 ml, SW3015, Beijing Solarbio Science & Technology Co., Ltd.) for 15 min at room temperature, the primary antibody against ADARB1 (cat. no. sc-73409, Santa Cruz Biotechnology, Inc., 1:100 dilution) was added and samples were incubated at 4°C overnight. Following washing with PBS, the horseradish peroxidase (HRP)-conjugated polymer second antibody (PV-6000, OriGene Technologies, Inc.) was added and samples were incubated for 30 min at room temperature. The signal was examined using a 3,3′-diaminobenzidine solution. Subsequently, four fields of view of the stained sections were independently observed in a light microscope at magnifications of ×4 and ×20 by two pathologists. The staining intensity of each protein was divided into four grades (intensity scores): 0 (negative), 1 (weak brown), 2 (moderate brown) and 3 (strong brown). The percentage of positive cells was scored as 0 (≤10%), 1 (11–25%), 2 (26–50%), 3 (51–75%) and 4 (>75%) (30). The final staining score was calculated using the following formula: Intensity score × percentage score (1). A final score >1 was defined as high expression and final scores ≤1 were defined as low expression.

Statistical analysis

Statistical analyses were performed using SPSS software (version 12.0; SPSS, Inc.). The data are presented as the mean ± SD. A paired-sample t-test was used for normally distributed and continuous variables were statistically analyzed by one-way analysis of variance with LSD post hoc test. Multivariate analysis was accomplished utilizing the Cox regression model to identify independence of accepted clinical parameters and the ordinal variables were tested with Kruskall-Wallis test followed by Dunn's post-hoc test. The Kaplan-Meier analysis was statistically tested by the log-rank method. P<0.05 was considered to indicate a statistically significant difference.

Results

ADARB1 is downregulated in LUSC tissues and cell lines

By analyzing several databases, the transcriptional levels of ADARB1 in LUSC tissues and cell lines were compared with those observed in normal lung tissue and cells. GE-mini analysis revealed that ADARB1 mRNA expression was significantly decreased in LUSC tissues (Fig. 1A). The CCLE database suggested that ADARB1 mRNA expression was significantly downregulated in ~10 LUSC cell lines compared with immortalized lung epithelial cell lines (P=0.001; Fig. 1B). To further investigate these observations, the Oncomine, UALCAN and GEPIA databases were employed. Analysis with all three databases suggested that ADARB1 was expressed at significantly lower levels in LUSC tissues compared with in normal lung tissues (Fig. 1C-E). Furthermore, IHC analysis, based on a lung cancer array containing 16 normal samples, 12 LUSC samples and 10 LUAD samples, suggested that ADARB1 was significantly downregulated in LUSC samples compared with normal lung samples (P=1.58×10−8; Fig. 2A and B). In summary, the downregulation of ADARB1 expression levels in LUSC tissues and cell lines suggested that ADARB1 may be a promising anti-oncogene in LUSC.

Figure 1.

Downregulation of ADARB1 in LUSC tissues and cell lines, as indicated by several databases. The expression of ADARB1 was analyzed by the (A) Genome Mining, (B) Cancer Cell Line Encyclopedia, (C) Oncomine, (D) University of Alabama Cancer Database and (E) Gene Expression Profiling Interactive Analysis databases. ADARB1, adenosine deaminase RNA specific B1; LUSC, lung squamous cell carcinoma; TPM, transcripts per million.

Figure 2.

Downregulation of ADARB1 in the lung cancer tissue array. (A) Immunohistochemistry was used to examine the level of ADARB1 expression in a commercial lung cancer tissue array. (B) Protein levels of ADARB1 were significantly downregulated in the lung tissue of patients with LUSC or LUAD compared with normal lung tissue. ADARB1 expression was significantly decreased in LUSC tissues compared with LUAD tissue samples. ADARB1, adenosine deaminase RNA specific B1; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma.

ADARB1 could be used for the identification of LUSC and LUAD

The results of the IHC analysis further suggested that ADARB1 was downregulated in LUAD tissues compared with normal lung tissues, which was consistent with our previous study (P=0.024; Fig. 2A and B) (31). Furthermore, ADARB1 expression levels were significantly lower in LUSC tissues than in LUAD tissues (P=0.025; Fig. 2A and B). Subsequently, the GSE10245 dataset, obtained from GEO, suggested that ADARB1 was expressed at a lower level in LUSC compared with LUAD (P=0.031; Fig. 3A). Furthermore, this significant difference was verified by the CCLE database (P=0.008; Fig. 3B). Therefore, these results suggested that ADARB1 expression profiles may serve as indicators for the clinical diagnosis of LUSC and LUAD.

Figure 3.

ADARB1 expression is significantly different between LUSC and LUAD. (A) Dataset GSE10245, obtained from Gene Expression Omnibus, suggested that ADARB1 was expressed at a lower level in LUSC compared with LUAD. (B) Difference in ADARB1 expression in LUSC and LUAD cell lines, as determined by Cancer Cell Line Encyclopedia analysis. ADARB1, adenosine deaminase RNA specific B1; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma.

Relationship between ADARB1 expression and the clinical characteristics of patients with LUSC

The effects of ADARB1 expression on the clinical characteristics of patients with LUSC were examined. Clinical data were downloaded from the Wanderer database, and the clinical characteristic parameters are summarized in Table I. The expression of ADARB1 was significantly associated with the pathologic N stage (P=0.045), pathologic M stage (P=0.012) and tobacco smoking history indicator (P=0.008). Furthermore, multivariate analysis of the clinical data suggested that the pathologic N stage (P=0.020), pathologic M stage (P=0.003) and tobacco smoking history indicator (P=0.017) were independently associated with ADARB1 expression in patients with LUSC (Table II).

Table I.

Association between clinical characteristic parameters and the expression of adenosine deaminase RNA specific B1 in lung squamous cell carcinoma.

Variable	n	Mean ± SD	P-value
Pathologic T			0.433
T1/T1a/T1b	100	8.148±0.924
T2/T2a/T2b	252	8.103±1.035
T3	48	8.367±0.967
T4	19	8.155±1.232
Pathologic N			0.045
N0	262	8.134±1.008
N1	111	8.236±0.955
N2	5	7.533±1.031
N3	6	9.050±0.584
Pathologic M			0.012
M0	357	8.113±0.996
M1/M1a	6	7.576±0.607
MX	50	8.511±1.081
Pathologic tumor stage			0.078
I/IA/IB	211	8.116±0.985
IIA/IIB	123	8.315±0.981
IIIA/IIIB	76	8.013±1.122
IV	6	7.576±0.607
Ethnicity			0.133
White	297	8.190±1.033
Black/African American	16	8.208±0.809
Asian	9	8.888±1.028
Tobacco smoking history indicator			0.008
Current reformed smoker for >15 years	71	8.404±0.991
Current reformed smoker for ≤15 years	211	8.018±1.007
Current reformed smoker, duration not specified	1	9.922±0.000
Lifelong non-smoker	17	8.533±1.042
Current smoker	112	8.137±0.977
Age			0.857
≤60 years	89	8.133±1.093
>60 years	322	8.155±0.994
Karnofsky score			0.236
0-70	47	7.939±1.233
80	14	7.910±0.884
90	22	8.173±0.948
100	10	8.687±0.957
Eastern Cooperative Oncology Group score			0.079
0	60	8.470±1.099
1	71	8.208±0.978
2	20	8.007±1.061
3	3	7.211±0.597

Values in bold are significant.

Table II.

Multivariate analysis of clinical data related to adenosine deaminase RNA specific B1 expression in lung squamous cell carcinoma.

Characteristic	Type III sum of squares	Degrees of freedom	Mean square	F-value	P-value
Pathologic N	11.32	4	2.83	2.96	0.020
Pathologic M	11.50	2	5.75	6.02	0.003
Tobacco smoking history indicator	9.88	3	3.30	3.45	0.017

ADARB1 methylation is related to the clinical characteristics of patients with LUSC

The ADARB1 global methylation levels in LUSC were evaluated using the DiseaseMeth database. ADARB1 displayed significantly higher levels of methylation in LUSC samples compared with in normal lung samples (P=1.62×10−12; Fig. 4A). The highest methylation value of cg24063645 in ADARB1 was identified using the Wanderer and MethSurv databases (P=3.60×10−19; Fig. 4B; Table SII). Subsequently, the association between cg24063645 methylation and the clinical characteristics of patients with LUSC was analyzed. The methylation of cg24063645 of ADARB1 was significantly associated with the Karnofsky score of patients with LUSC (P=0.025; Table III). Moreover, higher methylation levels of cg24063645 of ADARB1 were associated with shorter OS of patients with LUSC (P=0.044; Fig. 4C). Therefore, the results suggested that ADARB1 methylation levels may serve a role in the prognosis of patients with LUSC.

Figure 4.

Methylation values of ADARB1 in patients with LUSC. (A) Global methylation of ADARB1 in LUSC samples compared with normal samples, using DiseaseMeth. (B) Heat map () acquired from MethSurv described the methylation sites of ADARB1 in LUSC. Red represents upregulation and blue represents downregulation. (C) Relationship between methylation values of cg24063645 of ADARB1 and overall survival time in LUSC. ADARB1, adenosine deaminase RNA specific B1; HR, hazard risk; LR, log rank; LUSC, lung squamous cell carcinoma.

Table III.

Association between the clinical characteristics and methylation site value of cg24063645 of adenosine deaminase RNA specific B1 in patients with lung squamous cell carcinoma.

Variable	n	Mean ± SD	P-value
Sex			0.717
Male	215	0.983±0.042
Female	73	0.985±0.021
KRAS gene analysis indicator			0.760
Yes	6	0.988±0.002
No	197	0.983±0.038
EGFR mutation status			0.682
Yes	15	0.987±0.002
No	174	0.983±0.040
Treatment outcome: First course			0.718
Complete remission/response	55	0.983±0.022
Stable disease	2	0.989±0.001
Pathologic T			0.562
T1/T1a/T1b	76	0.986±0.009
T2/T2a/T2b	166	0.980±0.049
T3	37	0.987±0.004
T4	9	0.988±0.002
Pathologic N			0.117
N0	179	0.984±0.029
N1	79	0.985±0.014
N2	24	0.967±0.099
NX	6	0.984±0.004
Pathologic M			0.861
M0	233	0.984±0.035
M1/M1a	3	0.988±0.003
MX	50	0.986±0.004
Tumor stage			0.919
I/IA/IB	141	0.986±0.015
IIA/IIB	96	0.986±0.013
IIIA/IIIB	46	0.976±0.072
IV	3	0.988±0.003
Ethnicity			0.878
White	224	0.984±0.028
Black/African American	10	0.988±0.002
Asian	7	0.988±0.003
Karnofsky score			0.025
0-70	29	0.988±0.002
80	14	0.987±0.002
90	18	0.985±0.004
100	9	0.987±0.002
Age			0.341
≤60 years	59	0.987±0.003
>60 years	221	0.982±0.043

Values in bold are significant. EGFR, epidermal growth factor receptor.

ADARB1 expression is negatively correlated with EGFR expression in LUSC tissues

EGFR is an oncogenic driver that contributes to the activation and development of lung cancer (7). Previous studies have indicated a poor correlation between EGFR expression and the overall response rate in lung cancer, particularly in LUSC (7,32,33). The GEPIA database was employed to evaluate the relationship between ADARB1 and EGFR. A negative correlation between ADARB1 and EGFR transcript levels in tissues of patients with LUSC was identified (P=7.80×10−6; R=−0.16; Fig. 5), which suggested that ADARB1 acted as an anti-cancer gene and might be linked to EGFR.

Figure 5.

ADARB1 expression is negatively correlated with EGFR expression in LUSC. The Gene Expression Profiling Interactive Analysis database identified an association between the transcript levels of ADARB1 and EGFR in tissues derived from patients with LUSC. ADARB1, adenosine deaminase RNA specific B1; EGFR, epidermal growth factor receptor; LUSC, lung squamous cell carcinoma; TPM, transcripts per million.

Network analysis of genes co-expressed with ADARB1

To further investigate the biological function of ADARB1, a functional enrichment analysis of its associated co-expressed genes was conducted. Firstly, 14,102 genes co-expressed with ADARB1 in LUSC were identified using the cBioPortal database, and a volcano plot was generated to display the association between groups with altered and unaltered ADARB1 expression (Fig. 6A). Based on the criteria of |log ratio|>0.3 and P<0.05, 158 ADARB1-associated, co-differentially expressed genes (co-DEGs) were identified (Table SIII). A protein-protein interaction network was generated using the STRING database and Cytoscape software (Fig. 6B). Moreover, WebGestalt was used to analyze GO terms (Fig. 6C) and indicated that the main biological processes of the co-DEGs were primarily enriched in ‘metabolic process’ and ‘biological regulation’. Moreover, Gan et al (34) reported that ADARB1 participated in the regulation of pancreatic islet and β-cell function. Furthermore, deficient or hyperactive ADARB1 expression is associated with a number of human diseases, including immunological disorders, amyotrophic lateral sclerosis, epilepsy and cancer (35). Therefore, it was speculated that ADARB1 might serve a role in LUSC via metabolic processes and biological regulation. In addition, for cellular components and molecular functions, the co-DEGs were primarily enriched in ‘nucleus’ and ‘protein binding’, respectively. Finally, KEGG pathway enrichment was analyzed using the DAVID database, and ‘ubiquitin mediated proteolysis’ and ‘Wnt signaling pathway’ were identified as significantly enriched pathways related to ADARB1 expression (Table SIV).

Figure 6.

Functional enrichment analysis of ADARB1-associated co-DEGs in LUSC samples. (A) Volcano plot displayed the genes co-expressed with ADARB1 in LUSC (red represents upregulation, green represents downregulation and black is no change). (B) A protein-protein interaction network of ADARB1-associated co-DEGs was generated using the Search Tool for the Retrieval of Interacting Genes/Proteins database and Cytoscape software (square nodes represent hub genes, round nodes represent co-expression genes; the color represents the degree score; degree score <0.5 represent low values (colored yellow), degree score ≥0.5 represent high values (other colors). (C) Gene Ontology analysis of ADARB1-associated co-DEGs. ADARB1, adenosine deaminase RNA specific B1; co-DEGs, co-differentially expressed genes; LUSC, lung squamous cell carcinoma.

Discussion

The present study was conducted to evaluate the use of ADARB1 as a potential biomarker for patients with LUSC. To the best of our knowledge, the present study was the first to use bioinformatics methods to suggest that ADARB1 was downregulated in LUSC tissues and cell lines. Moreover, the expression of ADARB1 in LUSC was significantly different to that in LUAD. In addition, the global methylation levels of ADARB1 were upregulated in LUSC and a higher methylation value of cg24063645 was related to a shorter OS in patients with LUSC.

A-to-I RNA editing is a post-transcriptional modification mediated by the ADAR family, which is comprised of three members, ADAR1, ADARB1 and ADAR3 (36,37). The expression levels of ADARB1 determine the RNA editing activity at a number of A-to-I sites, which are involved in several diseases, including amyotrophic lateral sclerosis, autism and Fragile-X Syndrome (38). Recently, several studies have indicated that ADARB1 participates in tumorigenesis as a cancer suppressor gene, which is consistent with the results of the present study. By analyzing the influence of A-to-I RNA modification on cancer progression, Fritzell et al (39) reported that ADARB1 was expressed at low levels in glioblastoma, indicating that ADARB1 downregulation could induce malignant phenotypes. Additionally, using next-generation sequencing transcriptomics, Chan et al (40) reported that ADARB1 contributed to inhibitory mechanisms in gastric cancer via its catalytic deaminase. However, Altadill et al (41) demonstrated that ADARB1 was upregulated in endometrial cancer, and the increased expression was positively associated with disease aggression, suggesting that ADARB1 functions as an oncogene in endometrial cancer. The inconsistency between the observations of these aforementioned studies and the present study may be due to the diverse pathological states of different diseases; therefore, further investigation is required.

ADARB1 is a therapeutic target in a number of human diseases. Uchida et al (42) reported a significant increase in ADAR2-dependent A-to-I editing at the glutamate ionotropic receptor AMPA type subunit R/G, 5-hydroxytryptamine 2c receptor D and COPI coat complex-α I/V sites, which suggested that ADARB1 was a regulator of neuropathic pain after peripheral nerve injury and could be used as a target for the treatment of neuropathic pain. In addition, Galeano et al (43) reported that ADARB1 could inhibit glioblastoma growth by the regulation of the cell division cycle 14B/S-phase kinase associated protein 2/p21/p27 axis; therefore, ADARB1 or its substrates represented suitable targets for the treatment of glioblastoma. In the present study, ADARB1 was significantly downregulated in LUSC and was associated with clinical characteristics of patients with LUSC. Furthermore, the hypermethylation of ADARB1 was related to the poor prognosis of patients with LUSC. Therefore, it was speculated that ADARB1 may represent a potential target for the treatment of LUSC; however, this requires further investigation.

LUSC and LUAD are the two main subtypes of NSCLC, and the accurate differentiation of LUSC and LUAD is important for therapy stratification (44). Although p40 is regarded as the gold standard for the differentiation of LUSC and LUAD by the World Health Organization and the European Society for Medical Oncology (45–47), the identification of further biomarkers is required for distinguishing LUSC from LUAD. At present, numerous immune markers have been reported for the differentiation of LUSC and LUAD, including NK2 homeobox 1, napsin A, keratin 5 and p63. Kriegsmann et al and van Zyl et al (48,49) reported that all four aforementioned markers were differentially expressed in LUSC and LUAD, and could therefore be used as biomarkers for the two cancer subtypes. By immunohistochemistry and western blotting, Wang et al (50) reported that the expression of the transcriptional co-factor tafazzin was associated with subtypes of LUSC and LUAD. Additionally, spermatogenesis associated serine rich 2, which is highly expressed in both LUSC and LUAD, displayed significantly higher expression in LUSC than LUAD, and thus, could be used as a marker to discriminate between the two subtypes (51,52). Similarly, in the present study, ADARB1 expression was significantly higher in LUAD tissues and cell lines compared with in LUSC tissues. Therefore, the present study suggested that ADARB1 might be used to distinguish between LUSC and LUAD.

EGFR has been demonstrated to be a driving factor in various types of cancer. A previous study indicated that, following phosphorylation on a regulatory tyrosine, EGFR can be activated in a ligand-dependent or -independent manner to induce tumor progression (53). Additionally, EGFR has been reported to be upregulated in a number of different types of cancer and has been associated with cancer metastasis, drug resistance, poor prognosis and lower survival rates (54). Moreover, the combination of aberrantly expressed EGFR and transforming growth factor (TGF)-α secretion by NSCLC cells can form a TGF-α-EGFR autocrine ring, which promotes tumor development (55,56). EGFR can also activate the downstream PI3K/Akt/mTOR signaling pathway, leading to cell growth and proliferation during NSCLC (57). The present study identified a negative correlation between ADARB1 and EGFR expression in LUSC. Thus, it was hypothesized that ADARB1 could disturb the EGFR carcinogenic effect via metabolic processes and biological regulation in LUSC. This observation suggested that ADARB1 might act as an anti-oncogene by influencing EGFR expression; therefore, ADARB1 may be a useful molecular target for LUSC therapeutics.

In summary, the expression and role of ADARB1 in LUSC was investigated using a novel bioinformatics method. The results suggested that ADARB1 may be an anti-oncogene, and that it could be a promising biomarker in the tumorigenesis of LUSC and development of novel therapeutics for the disease. Therefore, understanding the function and mechanisms of ADARB1 in LUSC would provide a useful basis for the diagnosis and treatment of LUSC.