Oncogenic Role of ZFAS1 lncRNA in Head and Neck Squamous Cell Carcinomas

Abstract

Background: Head and neck squamous cell carcinoma (HNSCC) is a heterogeneous disease with high mortality. The identification of specific HNSCC biomarkers will increase treatment efficacy and limit the toxicity of current therapeutic strategies. Long non-coding RNAs (lncRNAs) are promising biomarkers. Accordingly, here we investigate the biological role of ZFAS1 and its potential as a biomarker in HNSCC.

Methods: The expression level of ZFAS1 in HNSCC cell lines was analyzed using qRT-PCR. Based on the HNSCC TCGA data, the ZFAS1 expression profile, clinicopathological features, and expression of correlated genes were analyzed in patient tissue samples. The selected genes were classified according to their biological function using the PANTHER tool. The interaction between lncRNA:miRNA and miRNA:mRNA was tested using available online tools. All statistical analyses were accomplished using GraphPad Prism 5.

Results: The expression of ZFAS1 was up-regulated in the metastatic FaDu cell line relative to the less aggressive SCC-25 and SCC-040 and dysplastic DOK cell lines. The TCGA data indicated an up-regulation of ZFAS1 in HNSCCs compared to normal tissue samples. The ZFAS1 levels typically differed depending on the cancer stage and T-stage. Patients with a lower expression of ZFAS1 presented a slightly longer disease-free survival and overall survival. The analysis of genes associated with ZFAS1, as well its targets, indicate that they are linked with crucial cellular processes. In the group of patients with low expression of ZFAS1, we detected the up-regulation of suppressors and down-regulation of genes associated with epithelial-to-mesenchymal transition (EMT) process, metastases, and cancer-initiating cells. Moreover, the negative correlation between ZFAS1 and its host gene, ZNFX1, was observed. The analysis of interactions indicated that ZFAS1 has a binding sequence for miR-150-5p. The expression of ZFAS1 and miR-150-5p is negatively correlated in HNSCC patients. miR-150-5p can regulate the 3'UTR of EIF4E mRNA. In the group of patients with high expression of ZFAS1 and low expression of miR-150-5p, we detected an up-regulation of EIF4E.

Conclusions: In HNSCC, ZFAS1 displays oncogenic properties, regulates important processes associated with EMT, cancer-initiating cells, and metastases, and might affect patients' clinical outcomes. ZFAS1 likely regulates the cell phenotype through miR-150-5p and its downstream targets. Following further validation, ZFAS1 might prove a new and valuable biomarker.

Keywords: HNSCC; ZFAS1; ZNFX1 antisense RNA 1; biomarker; head and neck cancers; lncRNA; non-coding RNA.

Figure 1

(A) Microscopic pictures of dysplastic oral keratinocyte (DOK), SCC-25, SCC-040, and FaDu cell lines, magnification 20×; (B) the capacity of spheres forming and (C) expression level of ZFAS1 lncRNA presented as mean with SEM; one-way ANOVA; ** p < 0.01, *** p < 0.001.

Figure 2

The expression level of ZFAS1 in head and neck squamous cell carcinoma (HNSCC) patients. (A) Expression in normal (n = 44) and cancer (n = 520) tissues; (B) Expression depending on HNSCC localization (n = 520); Graphs from UALCAN database, modified; Un-paired T-test; the graphs show mean of value presented as transcripts per million; and box and whiskers with 5–95 percentile, one-way ANOVA obtained using Dunn’s multiple comparisons tests; ns—no significant, **** p < 0.0001.

Figure 3

Disease-free survival (DFS) and overall survival (OS) in HNSCC patients with low (n = 130) and high (n = 130) expression levels of ZFAS1; a—Log-rank (Mantel-Cox) test, b—Gehan-Breslow-Wilcoxon test; p < 0.05 considered as significant.

Figure 4

The expression level of ZNFX1 in HNSCC patients. (A) Correlation between ZNFX1 and ZFAS1 in HNSCC patients; Graph from StarBase v3.0 database, modified; (B) Expression in normal (n = 44) and cancer (n = 520) tissues; **** p < 0.0001; (C) Expression of ZNFX1 in cancer samples (n = 520); (D) Expression depending on HNSCC localization (n = 520); Graphs from UALCAN database, modified; Un-paired T-test; the graphs show mean of value presented as transcripts per million; and box and whiskers with 5–95 percentile, one-way ANOVA obtained using Dunn’s multiple comparisons tests; ns—no significant, *** p < 0.001; (E) DFS and OS in HNSCC patients with low (n = 130) and high (n = 130) expression levels of ZNFX1; a—Log-rank (Mantel-Cox) test, b—Gehan-Breslow-Wilcoxon test; p < 0.05 considered as significant.

Figure 5

ZFAS1 regulation of miR-150-5p and its targets. (A) Possible interaction between lncRNA ZFAS1 and miR-150-5p sequences and co-expression of ZFAS1 and miR-150-5p in HNSCC patients; from StarBase v3.0 database. (B) Predicted miR-150-5p targets and position of regulation in their mRNA sequences. (C) The division to the groups of HNSCC patients: (i) with high level of miR-150-5p (n = 30) and low ZFAS1, and opposite (ii) with a low level of miR-150-5p and high ZFAS1 (n = 30); from StarBase v3.0 database. (D) The expression level of the predicted miR-150-5p targets in groups of patients (n = 60) with different expression levels of ZFAS1 and miR-150-5p; expression level presented as mean with SEM; un-paired T-test; ns – no significant, *** p < 0.001, **** p < 0.0001.

Figure 6

The proposed mechanism of the oncogenic role of lncRNA ZFAS1 in HNSCC. ZFAS1 acts as a molecular sponge and down-regulates abundance of miR-150-5p. The low level of suppressor miR-150-5p causes up-regulation of oncogenic targets such as EIF4E, which in turn up-regulates expression of genes connected with EMT, metastasis and poor patient outcome.