Genomic Regions Supporting microRNA Expression and Their Biogenesis Are Globally Conserved in Diffuse Large B-Cell Lymphomas (DLBCL) and Display a Limited Number of Homozygous Deletions or Amplifications.

MicroRNAs are proposed to play a direct role in oncogenesis as they act as both oncogenes and/or tumour suppressor genes (TSG). MicroRNA deregulation can be explained in some instances by deletion, amplification, translocation or epigenetic mechanisms. However the cause of aberrant expression of specific microRNAs in cancer remains largely unknown. Chromosomal-comparative genomic hybridization (CGH) and higher resolution array-CGH experiments have shown that genomic gains and losses play a crucial role in the development of DLBCL. Whether these gene copy number aberrations may involve specifically microRNA coding regions and be therefore implicated in the dysfunctional expression of microRNAs is currently undetermined. To address this issue, we analyzed 65 de novo DLBCL by array-CGH, using a high resolution microarray that contains ∼ 43,000 probes, giving an average spatial resolution of 35 kB. For each of the 474 microRNAs identified in the human genome (microRNA.sanger, April 2007 registry), 286 probes located near microRNA coding genes or microRNA clusters were defined and used to detect DNA copy alterations involving microRNA coding genes. 21% of the microRNA associated probes displayed a deletion in more than 5% of cases, a rate lower than the rate observed for the entire genome (30%, p =.0008). By contrast, 58% of the microRNA coding regions displayed a DNA copy gain, similarly to the whole genome (61%). Notably, none of the genes coding for components of the biosynthetic pathway (including DGCR8, DICER1-2, Argonaute 1, EIF2C2-3-4, Exportin 5 and DROSHA) displayed recurrent deletions. These results indicate that genomic regions supporting microRNA expression and their biogenesis are globally conserved in DLBCL. However, homozygous microRNA deletions, occurring in more than 5% of cases were observed for some microRNA, namely mir-548a (6%), mir-588 (6%), mir-548b (6%), mir-587 (5%) (all located in the 6q21–23 region), and mir-31 (9p21, 5%), suggesting that these microRNA may act as TSG and/or are located in TSG containing regions. Mir-31 was constantly co-deleted with CDKN2A, a well-defined TSG located in the vicinity. The most frequent amplifications involved mir-92b (1q22, 9%), mir-210 (11p15, 6%), mir-196a-2 (12q13, 6%), mir-148b (12q13, 6%), mir-122a (18q21, 9%), mir-153-1 (2q35, 8%), mir-567 (3q13, 8%), mir-106b (7q22, 9%) and mir-126 (9q34, 8%). Similarly, these genes represent candidate oncogenes or are located in oncogene containing regions. For instance, mir122a was co-amplified with MALT1, located in a close vicinity in 6/6 cases. In addition, distinct patterns of microRNA coding gene copy abnormalities between GCB and ABC subtypes were identified. To conclude, we provide experimental genome wide documentation of DNA copy alterations involving microRNA coding genes in DLBCL and suggest that some still uncharacterized microRNAs may act as TSG or oncogenes. Concordance between microRNA gene copy changes and microRNA gene transcriptional expression is currently investigated.