Functional analysis of the human miRNome in non-small cell lung cancer unveils a novel miR-92b-3p/NOTCH3 axis that drives tumor progression

Abstract Lung cancer remains the leading cause of cancer-related mortality worldwide. MicroRNAs, endogenous non-coding RNA of ~22 nucleotides, play key regulatory roles in lung cancer–related mechanisms and represent promising diagnostic and prognostic biomarkers, as well as potential therapeutic agents. However, a comprehensive functional screening of human miRNAs in lung cancer was missing. To tackle this, we performed a lentiviral-based miRNA screening (~2500 microRNAs) of lung adenocarcinoma (LUAD) i.e. the most frequent lung cancer subtype. We discovered and annotated miRNAs functionally associated with proliferation, migration, and invasion of LUAD cells. We then integrated our functional miRNA screening with transcriptomic and clinical data from a large LUAD patient cohort (TCGA-LUAD), along with spatial transcriptomics (ST) of in-house LUAD samples. This approach enabled us to identify phenotype-specific miRNA signatures and develop a 15-miRNA risk score predictive of poor overall survival (HR = 2.48, p < 0.0001). Among these, miR-92b-3p, a less studied member of the oncogenic miR-17~92 cluster, emerged as a critical driver of LUAD progression. Transcriptomics analysis revealed that miR-92b-3p overexpression increase NOTCH3 signaling activation. The inhibition of NOTCH3 signaling, either genetically or pharmacologically, abrogated the pro-migratory and invasive effects of miR-92b-3p. In summary, our study provides a systematic functional annotation of the human miRNome in LUAD and identifies a novel miR-92b-3p/NOTCH3 axis that contributes to LUAD progression.