Analysis of the Polycomb-related lncRNAs HOTAIR and ANRIL in bladder cancer

Abstract

Background: Long non-coding RNAs (lncRNAs) have been claimed as key molecular players in gene expression regulation, being involved in diverse epigenetic processes. They are aberrantly expressed in various tumors, but their exact role in bladder cancer is still obscure. We have recently found a major role of the Polycomb repression complex in recurrence of non-muscle-invasive bladder cancer. Here, we report the xpression of Polycomb-related lncRNAs:antisense noncoding RNA in the INK4 locus (ANRIL) and HOX antisense intergenic RNA (HOTAIR) in these tumors.

Findings: We studied a dataset of non-invasive bladder cancer samples by quantitative reverse transcription PCR (RT-qPCR) and analyzed also invasive bladder cancer samples using TCGA data. Our results showed that, while ANRIL seemed not to have a determining role, an increased HOTAIR expression appeared in recurrent and high-graded tumors associated with poor prognosis. In addition, through genome-wide transcriptome analyses, we observed that HOTAIR-EZH2-complex-regulated genes can efficiently discriminate between non-tumoral, recurrent, and non-recurrent bladder cancer samples. We also observed a significant correlation between EZH2 and HOTAIR expression levels. Using overexpression, knockdown, and pharmacological approaches in bladder cancer cell lines, we also observed that EZH2 regulates HOTAIR expression.

Conclusions: Our findings indicate that HOTAIR expression has prognostic value for bladder cancer progression, recurrence, and survival and suggest that HOTAIR plays active roles in modulating the cancer epigenome, becoming an interesting candidate as a target for cancer diagnosis and therapy. The observed HOTAIR regulation by EZH2 and the possibility of modulating EZH2 activity with specific inhibitors open new possible paths to be explored in bladder cancer therapy.

Keywords: ANRIL; Bladder cancer; Epigenetics; HOTAIR; LncRNA; Recurrence.

Fig. 1

Expression of lncRNAs HOTAIR and ANRIL in NMIBC. a qPCR analyses showing the correlation between EZH2 and HOTAIR expression in 64 NMIBC samples. b,c qPCR analyses showing the expression of HOTAIR in normal and NMIBC samples according the tumor grade (b) and stage (c). d qPCR analyses showing the correlation between EZH2 and ANRIL expression in 64 NMIBC samples. e,f qPCR analyses showing the expression of ANRIL in normal and NMIBC samples according the tumor grade (e) and stage (f ). Correlations were calculated using Pearson correlation coefficient. Comparisons between gene expression levels were done using the Kruskal-Wallis test followed by Dunn’s multiple comparison. p values are provided as follows: *p < 0.05; **p < 0.01; ***p < 0.005. TBP was used as normalizer gene [18]

Fig. 2

HOTAIR mediates recurrence and progression in NMIBC. a qPCR analyses showing the expression of HOTAIR in NMIBC samples according the tumor recurrence. b Kaplan-Meier analysis showing that patients with higher HOTAIR expression (according the median) showed an earlier recurrence (p value was obtained by the log-rank test). c Heat map showing the distribution of genes (rows) and samples following unsupervised clustering (Pearson correlation and average linkage method) of 28 tumors and 10 normal samples [16] according the expression of genes previously identified by the binding of HOTAIR-EZH2 complexes [7, 24]. A red (overexpressed) to blue (downregulated) scheme following the above scale limits (in log₂ scale) is shown. Note that recurrent and non-recurrent tumors, besides of non-tumor bladder tissue, could be efficiently discriminated. d qPCR analyses showing the expression of HOTAIR in recurrent NMIBC samples according the tumor progression. e Kaplan-Meier analysis showing that patients with higher HOTAIR expression (according the median) showed an earlier progression upon recurrence (p value was obtained by the log-rank test)

Fig. 3

HOTAIR expression in MIBC from the TCGA portal. a Expression of HOTAIR in normal and MIBC tumor samples. Comparison was performed by Limma test. b Expression of HOTAIR in MIBC samples according tumor stage. Comparison was performed by Limma test. c Kaplan-Meier analysis of MIBC patient overall survival according HOTAIR expression (median discrimination) (p value was obtained by the log-rank test). RNA-seq data were downloaded from the TCGA portal (https://tcga-data.nci.nih.gov/tcga/)

Fig. 4

Functional evidence of HOTAIR regulation by EZH2. a Expression of EZH2 protein as assessed by immunoblot in the quoted non-invasive bladder cancer cell lines in parallel with HOTAIR (bar graph, assessed by RT-qPCR). b Expression of EZH2 protein (immunoblot) and HOTAIR (bar graph) in 5637 MIBC cell lines upon knockdown mediated by two different shRNA constructs. c Expression of EZH2 protein (immunoblot) and HOTAIR (bar graph) in RT112 NMIBC cell line upon transfection with CMV-EZH2-coding plasmid. d Expression of the quoted proteins and HOTAIR (bar graph) in MGH-U4 NMIBC cell line upon treatment (24 h) with NVP-BEZ35 (50 nM), rapamycin (50 nM), tyrphostin (100 μM), SB31542 (10 nM), DZNeP (10 μM), and PD98059 (10 μM). Note that HOTAIR expression is only significantly reduced upon treatment with the EZH2-specific inhibitor DZNeP. e Expression of EZH2 protein (immunoblot) and HOTAIR (bar graph) in MGH-U4 NMIBC cell line upon treatment for different time periods with DZNeP (10 μM). GAPDH and ACTIN were used for loading control in immunoblots, and TBP was used as a normalizer gene for RT-qPCR