Reactivation of multiple fetal miRNAs in lung adenocarcinoma

MicroRNAs (miRNAs) play vital roles in the regulation of normal developmental pathways. However, cancer cells can co-opt these miRNAs, and the pathways that they regulate, to drive pro-tumourigenic phenotypes. Characterization of the miRNA transcriptomes of fetal organs is essential for identifying these oncofetal miRNAs, but it has been limited by fetal sample availability. As oncofetal miRNAs are absent from healthy adult lungs, they represent ideal targets for developing diagnostic and therapeutic strategies. We conducted small RNA sequencing of a rare collection of 25 human fetal lung (FL) samples, and compared them to two independent cohorts (n = 140, n = 427), each comprised of adult non-neoplastic lung (ANL) and lung adenocarcinoma (LUAD) samples. We identified 13 oncofetal miRNAs that were expressed in FL and LUAD but not in ANL. These oncofetal miRNAs are potential biomarkers for LUAD detection (AUC = 0.963). Five of these miRNAs are derived from the imprinted C14MC miRNA cluster at the 14q32 locus, which has been associated with cancer development and abnormal fetal and placental development. Additionally, we observed the pulmonary expression of 44 previously unannotated miRNAs. The sequencing of these fetal lung samples also provides a baseline resource against which aberrant samples can be compared. Overall design: We used a sequencing-based approach to profile the miRNA transcriptomes of a group of 165 samples, consisting of 25 fetal lung (FL) samples, 77 adult non-malignant lung (ANL) samples, and 63 lung adenocarcinoma (LUAD) samples, in order to identify lung oncofetal miRNAs. The miRNA expression of 73/77 ANL and 59/63 LUAD samples in this study was previously analyzed using different methods, and that resulting data was deposited in GEO (GSE62182). Sample titles for these samples directly correspond to those used in GSE62182. Samples from GSE62182 that yielded less than 5 million sequencing reads were excluded from this study.