Abstract
In bulk meningioma (MNG) tumors, a biomarker based on the expression of 34 transcripts (34HR-MNG) has recently been described to be able to predict their outcome, including recurrence. To better study the molecular mechanisms regulating the expression of the 34HR-MNG transcripts and predict their functional involvement in MNG recurrence, we built a competitive endogenous RNA (ceRNA) network through an in silico approach. MiRNAs targeting 34HR-MNG transcripts and corresponding sponging circRNAs were retrieved through MiRTarbase and ENCORI databases, respectively. The expression of candidate circRNA host genes belonging to the 34HR-MNG transcripts was correlated with specific molecular and clinical features of 89 and 20 WHO grade 1 and 2 MNGs, respectively, by querying the RNA-seq dataset GSE189672. The expression of candidate circRNAs and their host gene was validated through qRT-PCR. Among the 34HR-MNG transcripts, the Pim-1 proto-oncogene, serine/threonine kinase (PIM1) was significantly upregulated in (i) WHO grade 2 vs. grade 1 and (ii) recurrent vs. not recurrent WHO grade 2 MNGs. PIM1 expression positively and negatively correlated with that of Ki-67 and NF2, respectively, in recurrent WHO grade 2 MNGs. CircRNAs 0076215 and 0076216, both generated from the PIM1 host gene, were predicted to sponge miRNAs 16-5p and 195-5p, two tumor suppressors in MNG, in turn targeting PIM1. The expression of circRNAs 0076215 and 0076216, validated for the first time in a set of 19 physiological human tissues, positively correlated with that of their host gene (Rho value = 0.579 and 0.681, p-value = 0.026 and 0.013, respectively). Our data suggest that PIM1 is an oncogene involved in the recurrence of WHO grade 2 MNG and that the upstream ceRNA network, comprising circRNAs 0076215 and 0076216 and miRNAs 16-5p and 195-5p, is responsible for its upregulation through a feed-forward loop.