Maturation of the miR15a/miR16-1 Family Is Impaired In Chronic Lymphocytic Leukaemia

Abstract 53

A role as tumor suppressor has been proposed for the miR15a/miR-16-1 family of miRNA genes localized in a critical region in 13q14 that is deleted in more than 50% of CLL patients. These two miRNA genes regulate cell cycle progression and apoptosis in cell lines and in vivo, and knockout of the syntenic region in mice results in the development of a CLL-like phenotype.

We measured by qRT-PCR mir15a and miR-16 in CLL samples (n=38) and normal B-cells (n=14) and observed reduced levels in CLL independent of the 13q deletion status. We then assessed the levels of the primary precursors of the mature miRNAs (pri-miRNAs 15a, 16-1 and 16-2). No difference could be identified between patients with a deletion of 13q and patients with a retention of both copies of 13q14. However, CLL cells had increased levels of pri-miRNAs when compared to CD19-sorted normal B-cells. Moreover, in CLL cells levels of miR-16 and miR15a were inversely correlated to their pri-miRNA transcripts, while a miRNA used as a control (miR-155) had a positive correlation with its primary transcript as expected. In addition, also the levels of precursor miRNA molecules of the 13q14 miRNA genes were low in CLL patients. This strongly suggested a defect of miRNA maturation at the Drosha processing step that produces the pre-miRNA from the pri-miRNA molecule. Therefore, we grouped patients according to the observed levels of pri-miRNA and pre-miRNA: CLL patients with a ratio of these two processing intermediates (pri-/pre) above the average levels observed in non-malignant CD19-B-cells were included in the “high ratio” group (58% of our cohort CLL patients), while the rest was grouped in the “normal ratio” group (42%). Patients with high pri-/pre ratio had significantly lower mature miRNA levels compared to patients with normal ratio. In contrast, the respective ratio of the precursor molecules of miR-155 that were used as control did not differ in the two groups. These findings suggest that there is a processing defect that reduces maturation of the miR15a/miR-16 in a subset of CLL patients. In order to test the actual processing activity in CLL cells, we used a luciferase based in-vivo Drosha processing assay and could show a significant reduction of pri-16-1 processing in patients belonging to the “high pri-/pre ratio” group. These findings underline the role of miR15a/miR-16 in the pathomechanism of CLL and show a complementary route of inactivation that supplements genomic loss of the critical region in 13q14 and its transcriptional downregulation in CLL.