Expression of the MiR-17-92 cluster in DLBCL and B-CLL and its relationship with disease outcome and established prognostic factors

• Main Author: Culpin, Rachel Emily
• Published: University of Newcastle Upon Tyne 2012
• Subjects: 616.9946
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.582656

Diffuse large B-cell lymphoma is an aggressive malignancy of large transformed B- lymphocytes. The disease is biologically and clinically highly heterogeneous and this nature has impacted on patients response to treatment. Gene profiling studies have identified two biologically diverse sub-groups, that exhibit different prognoses, namely; the germinal centre (GC)- and activated B-cell (ABC)- like DLBCL. B-cell chronic lymphocytic leukemia is also a heterogeneous disease. Patients may follow an indolent clinical course and survive for >20 years, but others have a rapidly progressive disease requiring aggressive treatment, with some dying a few years after diagnosis. It is possible to identify some of the high-risk patients with either DLBCL or B-CLL on the basis of clinical parameters. However, these do not always reflect the molecular heterogeneity of the diseases and have not impacted on the routine treatment or clinical management. Therefore, there is a need to identify molecular factors that represent these diseases more accurately and provide better prognostic information. Recent studies implicate microRNAs in the regulation of gene expression. The microRNAs are endogenous, non-coding RNAs that play key regulatory roles in a diverse range of pathways, including development, cell proliferation, differentiation and apoptosis. These 18-24 nucleotide single-stranded RNAs are involved in post- transcriptional gene regulation. They bind to complimentary sites in the 3' untranslated region (UTR) of messenger RNAs to induce gene silencing. In 2002, Calin et al directly associated deregulated expression of miR -15 and miR -16 in the pathogenesis of B-CLL. Since this time, a number of microRNAs have been implicated in haematological neoplasms, including members of the oncogenic miR-17-92 cluster, encoded by the CJ3orj25 gene located on chromosome 13q31. In the present study; i) microRNA microarray profiling was employed to identify microRNA signatures that are accurately assigned to the DLBCL prognostic sub-groups and ii) qRT-PCR technology was used to examine the role of expression of the mature microRNAs of the miR-17-92 cluster in the prognostication of DLBCL and B-CLL. Using a number of well characterised DLBCL cell lines, a 9 microRNA signature was identified that discriminated between GC-like and ABC-like prognostic sub-groups. This included 4 microRNAs of the miR-17-92 cluster and 3 newly identified microRNAs, not previously associated with DLBCL and predicted to target genes that are deregulated in lymphoma. A 4- and 18- microRNA signature was identified that differentiated between DLBCL and FL cell lines and between lymphoma and normal B-cells, respectively. These signatures have yet to be assessed in clinical material. Due to the emerging role of the miR-17-92 cluster in B-cell malignancy, using robust qRT-PCR technology the expression levels in B-cell lymphoma cell lines, DLBCL and B-CLL clinical samples were investigated. It was observed that, despite being transcribed from the same parental cluster, the expression levels of mature microRNAs varied, suggesting the existence of additional mechanisms of regulation in the processing of these molecules from their transcription to the mature form. In DLBCL clinical samples, high expression of miR-17-5p or miR-92 were identified as significant high risk factors when analysed as continuous variables (OS: miR-17-5p p = 0.025 and miR-92 p = 0.014, PFS: miR-92 p = 0.017). Similarly, as dichotomous variables, high expression of miRs -17-5p (p = 0.001), -18 (p = 0.051),- 19a (p = 0.011), -20 (p = 0.010) or -92 (p = 0.050) predicted for shorter OS and in addition, high expression of miRs -17-5p (p = 0.006) or -20 (p = 0.038) for shorter PFS. Expression of microRNAs of the cluster further risk stratified immunophenotypic- or Ann Arbor disease stage-based prognostic sub-groups. In multivariate analysis, miR-17- 5p (p = 0.012) and IPI (p = 0.009) were identified as independent prognostic factors. In B-CLL, when samples were analysed as dichotomous variables, high expression ofmiR-18 (p = 0.013) and miR-19a (p = 0.009) predicted for shorter TFS in the entire cohort and miR-19a (p = 0.008) retained prognostic significance in Binet stage A group of patients. Conversely, high expression of miR-17-5p (p = 0.044) or miR-92 (p = 0.052) predicted for longer TFS and remained predictive in Binet stage A (miR-17-5p: p = 0.022, miR-92: p = 0.014). In multivariate analysis CD38 (p < 0.001) and Binet disease stage (p = 0.002) were identified as independent prognostic factors but in the Binet stage A group, only CD38 expression (p = 0.011) retained prognostic significance. In conclusion: i) microRNA array-based signatures provide a useful mechanism to classify lymphoma sub-groups. Investigation into the role of these microRNAs may provide further insights into disease pathophysiology, ii) varying expression levels of mature miR-17-92 microRNAs in B-cell lymphoma and B-CLL suggest additional regulatory mechanisms in the processing of this cluster and iii) mature miR-17-92 microRNAs may act as either high, or low risk factors, depending upon the cellular context and provide useful prognostic information for patients with DLBCL or B-CLL. Given the recent identification of the involvement of miR-17-92 oncomiR-1 in the activation of anti-apoptotic and pro-proliferative pathways, the role of this cluster in DLBCL and B-CLL pathophysiology warrants further investigation.