AGO2-RIP-Seq reveals miR-34/miR-449 cluster targetome in sinonasal cancers

Abstract

Sinonasal tumours are heterogeneous malignancies, presenting different histological features and clinical behaviour. Many studies emphasize the role of specific miRNA in the development and progression of cancer, and their expression profiles could be used as prognostic biomarkers to predict the survival. Recently, using the next-generation sequencing (NGS)-based miRNome analysis the miR-34/miR-449 cluster was identified as miRNA superfamily involved in the pathogenesis of sinonasal cancers (SNCs). In the present study, we established an Argonaute-2 (AGO2): mRNA immunoprecipitation followed by high-throughput sequencing to analyse the regulatory role of miR-34/miR-449 in SNCs. Using this approach, we identified direct target genes (targetome), which were involved in regulation of RNA-DNA metabolic, transcript and epigenetic processes. In particular, the STK3, C9orf78 and STRN3 genes were the direct targets of both miR-34c and miR-449a, and their regulation are predictive of tumour progression. This study provides the first evidence that miR-34/miR-449 and their targets are deregulated in SNCs and could be proposed as valuable prognostic biomarkers.

Fig 1. Distribution of miR-34c and miR-449a in SNCs.

Expression of miR-34c and miR-449a in SNCs and non-malignant (NM) cells (A). Distribution of miR-34c and miR-449a in the SNC histotypes (B). Comparisons between and among groups were determined by Student’s t-test two-tailed (two groups) and by analysis of variance (ANOVA) followed by post-hoc Tukey analysis, respectively. Differences with p < 0.05 were considered statistically significant.

Fig 2. Analysis of miR-34c and miR-449a expression in NSSCC miR-34c/miR-449a mimic transfected cells and NSSCC control (empty vector transfection) cells.

Relative miR-34c and miR-449a expression was adjusted to miR-99b (A). Comparison among groups was determined by analysis of variance (ANOVA), followed by post hoc Tukey analysis. Differences with p < 0.05 were considered statistically significant. * miR-34c mimic and miR-449a mimic vs. CTRL. The RNA extracts of the input and AGO2-IgG-IP fractions were analysed by high-throughput sequencing (AGO2-RIP-Seq) and comparisons between AGO2-IP miRNAs and AGO2-IP scramble, IgG-IP miRNAs and IgG-IP scramble and input miRNAs and input scramble were performed (B). The data were from three independent AGO2-IP experiments. MA plot of the three comparisons performed for both miR-449a and miR-34c.

Fig 3. Distribution of STK3, C9orf78 and STRN3 gene expression in the SNC histotypes.

Expression of STK3 (A), C9orf78 (B) and STRN3 (C) in tumour with respect to their non-malignant counterparts (fold-change) in the SNC histotypes. Spearman’s correlation among STK3, C9orf78 and STRN3 and miR-34c and miR-449a in the SNACC histotype (D). Comparison among groups was determined by analysis of variance (ANOVA), followed by post hoc Tukey analysis. Differences with p < 0.05 were considered statistically significant. * SNACC and SNSCC vs. SNADC, SNUC and SNEC.

Fig 4. Kaplan-Meier survival curves for ITAC stratified for STK3, C9orf78 and STRN3 gene expression.

Low and high gene expression of STK3 (A), C9orf78 (B), and STRN3 (C) were associated with overall survival (OS). Comparisons between groups were made using the log-rank test and p < 0.05 were considered statistically significant.

Fig 5. Kaplan-Meier survival curves for ITAC stratified for STK3, C9orf78 and STRN3 gene expression.

Low and high gene expression of STK3 (A), C9orf78 (B), and STRN3 (C) were associated with disease-free survival (DFS). Comparisons between groups were made using the log-rank test and two-sided p < 0.05 were considered statistically significant.

Fig 6. MiR-34/miR-449-induced STK3, C9orf78, STRN3 expression in NSSCC cells.

NSSCC cells were transiently transfected with miR-34c/miR-449a mimic and evaluated for STK3, C9orf78 and STRN3 gene (A), and protein expression (B). The protein levels were evaluated by densitometry analysis of the bands relative to actin. The miR-34c/miR-449a-transfected NSSCC cells and their parental counterparts were investigated for cell migration/invasion (C), cell proliferation (D) and colony forming assay (E). Scale bar = 100 μm. Comparisons between and among groups were determined by two-tailed Student’s t-test and analysis of variance (ANOVA) followed by post-hoc Tukey analysis. Differences with p < 0.05 were considered statistically significant.

Fig 7. Silencing STK3, C9orf78, STRN3 in NSSCC cells.

NSSCC cells were silenced for STK3 (shSTK3), C9orf78 (shC9orf78) and STRN3 (shSTRN3) or for all genes (mix siRNA) and their expression evaluated (A). The silenced NSSCC cells and their parental counterparts were investigated for cell migration/invasion (B), cell proliferation (C) and colony forming assay (D). Scale bar = 100 μm. Comparisons among groups were determined by analysis of variance (ANOVA) followed by post-hoc Tukey analysis. Differences with p < 0.05 were considered statistically significant.