MicroRNA Involvement in Human Cancers.

Featured Article: Yanaihara N, Caplen N, Bowman E, Seike M, Kumamoto K, Yi M, et al. Unique microRNA molecular profiles in lung cancer diagnosis and prognosis. Cancer Cell 2006;9:189 –98. Recent evidence has revealed that microRNAs (miRNAs) function as tumor suppressors and oncogenes, and therefore miRNAs might prove useful in the diagnosis and treatment of cancer. In 2002, Croce’s group reported the first connection between miRNAs and cancer, showing that the MIR15A (microRNA 15a) and MIR16-1 (microRNA 16-1) genes were deleted and/or downregulated in the majority of B-cell lymphomas (1 ). The findings that miRNAs were frequently located at cancer-associated genomic regions, including fragile sites, prompted these investigators to propose that miRNA genes were a new class of genes involved in human tumorigenesis (2). The development of miRNA gene expression profiling allowed the identification of miRNA gene expression signatures unique to specific solid tumors (e.g., lung, breast, stomach, prostate, and colon) and their cellular origins (3). These reports indicated the extensive involvement of miRNAs in the pathogenesis of solid cancer and highlighted their potential as biomarkers for cancer. Additional studies have shown a strong correlation between impaired expression of miRNA genes and oncogenesis (4). For example, the RAS oncogenes are regulated by the let-7 family of miRNAs, and MIR155 and MIR gene clusters 17–92 are associated with the MYC [v-myc myelocytomatosis viral oncogene homolog (avian)] oncogene. These findings suggest that miRNAs play a substantial role in the pathogenesis of human cancers. In 2006, when our Cancer Cell report was published, the field was primarily focusing on known coding genes and proteins to define the molecular network of lung carcinogenesis. Both our group and another (5 ) investigated whether miRNAs might lead to a better understanding of the biology of lung cancer. The results published in our report showed that lung cancer had extensive alterations in miRNA gene expression. Furthermore, miRNA gene expression profiles correlated with survival of lung adenocarcinomas, including those classified as stage I, indicating that the miRNAs might help predict the prognosis for a patient. Ours was one of the first studies to link miRNAs to a patient’s prognosis. In our study, 104 pairs of primary lung cancers and corresponding noncancerous lung tissues were analyzed by miRNA microarray. First, we identified miRNA genes that were expressed differently in phenotypic and histologic classifications. When we compared miRNA gene expression in lung cancer tissues and noncancerous lung tissues, the 2 tissue groups showed statistically significant differences in expression for 43 miRNA genes. Interestingly, several of the miRNAs were located inside fragile sites and/or in the cancer-associated genomic regions. Our next question was, “Do the miRNA molecular profiles of lung cancer correlate with patient survival?” A univariate Cox proportional hazards regression model for lung cancer gene expression profiles indicated that 8 miRNAs were related to the survival of patients with adenocarcinoma. Furthermore, the survival analysis of only stage I cases showed that 3 miRNAs were associated with outcome. One was the miRNA encoded by the MIR21 (microRNA 21) gene, frequently highly expressed in major types of human cancer (3 ), has been found to be a robust prognostic classifier of stage I lung adenocarcinoma in 3 different cohorts (6 ). High MIR155 and low MIRLET7A2 [mir-100-let-7a-2 cluster host gene (non-protein coding)] expression has also been correlated with poor survival in both univariate and multivariate analyses for MIR155. The relationship between miRNA gene expression signature and outcome was confirmed by real-time reverse-transcription PCR analysis of precursor miRNAs and was cross-validated with an independent set of adenocarcinomas. To investigate the biological consequence of altered MIR155 expression, we collaborated with Dr. Stephens’s group in the Advanced Biomedical Computing Center at the National Cancer Institute, Frederick National Laboratory (SAIC-Frederick, Inc.). This group conducted a bioinformatics analysis grouping of the predicted targets of MIR155 by Gene Ontology terms. We found a 1 Laboratory of Human Carcinogenesis, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD; 2 Department of Obstetrics and Gynecology, Jikei University School of Medicine, Tokyo, Japan. * Address correspondence to this author at: Laboratory of Human Carcinogenesis, National Cancer Institute, National Institutes of Health, 37 Convent Dr., Bldg. 37, Rm. 3068, Bethesda, MD 20892-4255. Fax 301-496 0497; e-mail Curtis_ [email protected]. Received February 8, 2013; accepted February 19, 2013. Previously published online at DOI: 10.1373/clinchem.2012.198176 3 This article has been cited more than 1000 times since publication. 4 Human genes: MIR15A, microRNA 15a; MIR16-1, microRNA 16-1; MIR155, microRNA 155; MYC, v-myc myelocytomatosis viral oncogene homolog (avian); MIR21, microRNA 21; MIRLET7A2, mir-100-let-7a-2 cluster host gene (nonprotein coding). Clinical Chemistry 59:12 1811–1812 (2013) Citation Classic