The Target Genes and Prognostic Significance of Mir-150 in Chronic Lymphocytic Leukemia

Abstract 3859

Chronic lymphocytic leukemia (CLL) is the first disease in which miRNAs (hsa-miR-15a-16–1) were directly linked to cancer pathogenesis (Calin et al. PNAS, 2002). We and others have also shown that expression of certain miRNAs associates with disease activity in patients with CLL (Calin et al. NEJM, 2005; Mraz et al. Blood, 2012; Mraz et al. Leukemia, 2009). Moreover, patients with more aggressive disease have CLL cells that generally express unmutated IGHV and/or ZAP-70 and have a miRNA expression profile that differs from that of CLL cells from patients with indolent disease (Calin et al. NEJM, 2005). However, we still have very limited understanding of how miRNAs affect CLL cell-biology and expression of genes that play a critical role in either promoting or arresting the disease. We used pooled samples from 10 CLL patients to screen (TaqMan miRNA Cards-ABI, 750 miRNAs) for abundantly expressed miRNAs that could hypothetically influence CLL B cell biology. We identified miR-150 as the most abundant miRNA in CLL cells and also as being strongly expressed when compared to CD19+ blood lymphocytes of normal adults (N=5, P=0.008). This miRNA already has been reported to influence the differentiation and gene expression of normal B cells (Xiao et al. Cell, 2007) suggesting its possible relevance for CLL B cell biology. We examined additional CLL cell samples (N=168) and confirmed high miR-150 levels and also noted heterogeneity in its expression between CLL cells of patients with aggressive versus indolent disease. In our cohort, CLL cells of patients that expressed ZAP-70 (20% cut-off, N=74) or had unmutated IGHV (N=72) expressed significantly lower median-levels of miR-150 (fold change −1.7 and −2.0 respectively, p<0.005). Moreover, the lower levels of miR-150 also directly associated with higher response to stimulation of B-cell receptor (BCR) on CLL cells with anti-IgM (P<0.05, N=36, quantified by flow cytometric measurement of calcium mobilization). To understand the gene network regulated by miR-150 in CLL we performed array-based transcriptome analyses (HG-U133 Plus 2.0, Affymetrix) of 110 patient samples, which identified differential expression of 215 genes between CLL cells expressing low versus high levels of miR-150 (SAM analysis of upper and lower terciles). Thirty-eight of these 215 genes (17%) are predicted targets of miR-150 (determined by TargetScan, www.targetscan.org). Two well annotated genes (GAB1 and FOXP1) have evolutionary conserved binding sides for miR-150 in their 3‘UTRs, suggesting the possible importance of miR-150 in their regulation. GAB1 is an adaptor molecule and plays a key role in variety of cell signaling pathways (PLCγ, Ras/Erk, PI3K/Akt, CrkL). Interestingly, GAB1 modulates PI3K/Akt-pathway through binding domain identical to Bruton’s tyrosine kinase (Rameh et al. JBC, 1997) and is a key molecule involved in regulating BCR-signaling (Ingham et al. JBC, 1998, 2001), a process that factors prominently in the pathogenesis and progression of CLL. FOXP1 is an essential participant in the transcriptional regulatory network of B lymphopoiesis and has been identified as playing a role in disease progression of other B-cell lymphomas (Hu et al. Nat Immunol, 2006). The immunoblot analysis of GAB1 and FOXP1 in CLL cells confirmed their higher protein levels in cases with low miR-150 expression (P<0.005, fold change >10.0). Importantly, cells with higher expression of GAB1 or FOXP1 were more responsive to BCR stimulation in vitro (P<0.01, N=36) and higher expression of each associates with shorter overall survival (OS) (13.9 vs. 22.7 years, 13.9 vs. 21.1 years; N=168; P<0.05). Most notably, a reverse trend was observed for miR-150, where higher levels (>median) were associated with significantly longer OS (not-reached vs. 13.9 years, N=168, P=0.006). Additionally, the expression level of miR-150 was an independent predictor of OS and time to first treatment (TTFT) in multivariate analyses, which included IGHV status, ZAP-70, CD38, Rai stage, gender, and age (OS HR: 3.4 [CI 1.4–8.6], P=0.009; TTFT HR: 2.3 [CI 1.3–4.2], P=0.004). We conclude that there is an inverse association between high-risk disease and expression of miR-150, which may reflect its capacity to regulate the expression of genes encoding proteins that may contribute to BCR-signaling and/or survival of CLL B cells.