To identify miRNAs that preferentially repress the epidermal growth factor receptor (EGFR) pathway in glioblastoma (GBM).

GBM is the most common primary malignant adult brain tumor. Amplification of EGFR and overactivation of its pathway is present in 57% of primary GBMs, and is thought to drive tumor progression. However, targeted inhibition of EGFR has not shown clinical benefit. Additionally, the EGFR pathway is upregulated in GBM recurrence, even without a detectable EGFR amplification, suggesting an alternate mode of pathway regulation. As such, miRNA regulators of EGFR are of key clinical interest, both to explain the pathway overactivation in recurrence, and as therapeutic agents, though their therapeutic role may be limited by the pleiotropy of their inhibitory effects. Here, we computationally define a set of miRNAs that synergistically and selectively inhibit the EGFR pathway in GBM.

We analyzed three datasets with paired human mRNA-miRNA expression data (N = 119) of pre-treatment GBM. We identified the miRNAs with the highest covariance with EGFR mRNA expression and EGFR pathway expression and combined these across datasets. We refined this set of putative EGFR-regulating miRNAs to include only those with the strongest predicted inhibitory activity towards the EGFR pathway by integrating the known human protein-protein interaction network.

We identified a number of miRNAs predicted to regulate the EGFR pathway in GBM, with high exclusivity for this pathway. Top candidates, among which we will select for further in vitro singular and combinatorial validation, include: miR-204, miR-30a/e, miR-148a, miR-223, miR-221, miR-34a, miR-338, miR-242 and miR-195. 

A novel computational approach was used to identify miRNA selectively regulating the EGFR pathway, a key driver of GBM, presenting a path towards in vitro and in vivo validation of miRNA targeting as a potential therapeutic strategy in this disease.