Identification of SRXN1 and KRT6A as Key Genes in Smoking-Related Non-Small-Cell Lung Cancer Through Bioinformatics and Functional Analyses

Abstract

Background: Lung cancer is the leading cause of cancer-related mortality worldwide. Although cigarette smoking is an established risk factor for lung cancer, few reliable smoking-related biomarkers for non-small-cell lung cancer (NSCLC) are available. An improved understanding of these biomarkers would further the development of new biomarker-targeted therapies and lead to improvements in overall patient survival.

Methods: We performed bioinformatic analysis to screened potential target genes, then quantitative PCR, western, siRNA, CCK-8, flow cytometry, tumorigenicity assays in nude mice were performed to validated the function.

Results: In this study, we identified 83 smoking-related genes (SRGs) based on an integration analysis of two Gene Expression Omnibus (GEO) datasets, and 27 hub SRGs with potential carcinogenic effects by analyzing a dataset of smokers with NSCLC in The Cancer Genome Atlas (TCGA) database. A survival analysis revealed three genes with potential prognostic value, namely SRXN1, KRT6A and JAKMIP3. A univariate Cox analysis revealed significant associations of elevated SRXN1 and KRT6A expression with prognosis. A receiver operating characteristic (ROC) curve analysis indicated the high diagnostic value of SRXN1 and KRT6A for smoking and cancer. Quantitative PCR and western blotting validated the increased expression of SRXN1 and KRT6A mRNA and protein, respectively, in lung cancer cell lines and NSCLC tissues. In patients with NSCLC, SRXN1 and KRT6A expression was associated with the tumor-node-metastasis (TNM) stage, presence of metastasis, history of smoking and daily smoking consumption. Furthermore, inhibition of SRXN1 or KRT6A suppressed viability and enhanced apoptosis in the A549 human lung carcinoma cell line. Tumorigenicity assays in nude mice confirmed that the siRNA-mediated downregulation of SRXN1 and KRT6A expression inhibited tumor growth in vivo.

Conclusions: In summary, SRXN1 and KRT6A act as oncogenes in NSCLC and might be potential biomarkers of smoking exposure and the early diagnosis and prognosis of NSCLC in smokers, which is vital for lung cancer therapy.

Keywords: KRT6A; SRXN1; lung cancer; non-small-cell lung cancer; smoking-related gene.

Figure 1

Flowchart of the study. GSE18385 and GSE76324, Gene Expression Omnibus cohorts; TCGA, The Cancer Genome Atlas.

Figure 2

Identification of hub smoking-related genes (SRGs). (A, B) Volcano plots of differentially expressed genes in the GSE18385 and GSE76324 cohorts. Log2 (FC) vs. -log10 (adj.P.Val) for differentially expressed genes. Red and blue dots represent upregulated and downregulated genes, respectively (log2|FC| > 1, adj.P.Val < 0.05). (C) Venn diagram of SRGs in GSE18385 and GSE76324. (D) Venn diagram of upregulated differentially expressed genes in GSE18385 and GSE76324. (E) Venn diagram of downregulated differentially expressed genes in GSE18385 and GSE76324. (F) Venn diagrams plotted to showcase overlaps between the 83 SRGs and TCGA dataset. N represents 83 SRGs from healthy smokers, while I, II, III and IV represent clinical stages I, II, III and IV among smokers with lung cancer, respectively. (G) Significantly enriched Kyoto Encyclopedia of Genes and Genomes pathways corresponding to the 27 hub SRGs.

Figure 3

Associations of two top hub smoking-related genes (SRGs) with overall survival in lung cancer patients via Kaplan–Meier and Cox regression analyses. (A) Kaplan–Meier curves of 27 hub SRGs; (B) receiver operating characteristic (ROC) curves of SRXN1 in the GSE18385 and GSE76324 datasets. (C) ROC curves of KRT6A in the GSE18385 and GSE76324 datasets. (D) ROC curves of SRXN1 in The Cancer Genome Atlas (TCGA). (E) ROC curves of KRT6A in TCGA.

Figure 4

Strong expression of SRXN1 and KRT6A in non-small-cell lung cancer (NSCLC) tissues and lung cancer cell lines. (A) SRXN1 expression was detected in NSCLC and normal tissues by RT-qPCR. (B) KRT6A expression was detected in NSCLC and normal tissues by RT-qPCR. (C, D) Higher expression of SRXN1 and KRT6A in patients with clinical stage III/IV NSCLC vs. those with clinical stage I/II NSCLC. (E, F) Higher expression of SRXN1 and KRT6A in patients with cM+ NSCLC vs. patients with cM0 NSCLC. (G, H) Higher expression of SRXN1 and KRT6A in patients with cN+ NSCLC vs. patients with cN0 NSCLC. (I, J) Higher expression of SRXN1 and KRT6A in smokers with NSCLC vs. never-smokers with NSCLC. (K) Higher expression of SRXN1 in LUAD patients vs. LUSC patients. (L, M) Scatter plots of the correlations between SRXN1 and KRT6A expression and daily smoking consumption. (N, O) RT-qPCR and western blot analyses of the SRXN1 and KRT6A mRNA and protein levels in 95D, A549 and Beas-2B-NNK cells vs. with Beas-2B cells. *P < 0.05, **P < 0.01.

Figure 5

SRXN1 or KRT6A inhibition suppressed cell viability and promoted cell apoptosis. (A) RT-qPCR and western blot analyses were performed to test the effect of SRXN1 interference. (B) RT-qPCR and western blot analyses were performed to test the effect of KRT6A interference. (C) Cell viability in response to SRXN1 depletion was monitored using a CCK-8 assay. (D) Cell viability in response to KRT6A depletion was monitored using a CCK-8 assay. (E) Representative flow cytometry plots of cell apoptosis in response to SRXN1 depletion and a graph of apoptosis rates per group. (F) Representative flow cytometry plots of cell apoptosis in response to KRT6A depletion and a graph of apoptosis rates per group. *P < 0.05, **P < 0.01. NC, negative control; ns, no significance.

Figure 6

SRXN1 or KRT6A knockdown suppressed tumor growth in vivo. (A) Tumor growth curve in response to SRXN1 depletion. (B) Tumor growth curve in response to KRT6A depletion. (C) Representative images of xenograft tumors from nude mice and comparison of tumor weights in response to SRXN1 depletion. (D) Representative images of xenograft tumors from nude mice and comparison of tumor weights in response to KRT6A depletion. (E, F) Representative images of hematoxylin–eosin-stained tumor samples from SRXN1 and KRT6A knockdown samples, respectively, relative to controls and negative controls (NC). (G, H) Representative images of Ki67-immunostained tumor samples from SRXN1 and KRT6A knockdown samples, respectively, relative to controls and NCs. *P < 0.05.