Poster Board II-926
Multiple studies in closely related diseases such as Chronic Lymphocytic Leukemia (CLL) have revealed distinct miRNA profiles. This increasing appreciation for the role of miRNA expression in disease pathogenesis and homeostasis within cancer biology lead us to profile the miRNA expression of CD19+ bone marrow cells from 11 patients with Waldenstrom's Macroglobulinemia (MR) and 5 healthy donors. The median age for patients was 72 years (range 49-81), WM ISS Prognostic Score was 1 (range 0-4), bone marrow disease involvement was 40% (range 5-80%), and serum IgM was 3,330 (range 202-6,110 mg/dL). Five patients (45.5%) were previously treated and 4 (36.4%) had extramedullary disease.
CD19+ cells were selected using auto-MACs cell sorting (Miltayni Biotec) and total RNA was extracted with Trizol (Invitrogen). Micro-RNA profiling was conducted using TaqMan low density arrays (Applied Biosystems) allowing for stem-loop based qPCR detection of 670 miRNAs per sample. Results were validated using RT2 miRNA SYBR green based qPCR (SABiosciences). Relative quantification was calculated by ddCT using U6 endogenous controls and normalized to the first healthy donor sample. Results were analyzed using custom perl scripts running bootstrap calculated 95% CIs and approximate permutation testing from 10,000 resampling groups for both means and medians resulting in a robust and distribution independent characterization of each population. Additional Mann-Whitney-Wilcoxon, general linear modeling (GLM), ANOVA, and Spearman correlation testing was conducted using R (R Project for Statistical Computing).
We identified miR-29c (+3.2 fold), miR-339-5p (+2.0 fold), and miR-21 (+3.2 fold) as significantly up-regulated in WM patients (p<0.006 for all). Down-regulated miRNAs included miR-324-3p (-2.0 fold), miR 875-5p (-3.2 fold) and miR-638 (-7.2 fold) (p<0.006 for all). Analysis of these findings with clinical features revealed a positive correlation of miR-29c with serum IgM (rho=.65, p=0.03), using multivariate analysis that included age and previous treatment status (p<0.001). Both miR-875-5p and miR324-3p inversely correlated with bone marrow disease involvement (rho <-0.79, p<0.005 for both), and both of these were found to be predictive under multivariate analysis (p < 0.001 and < 0.01 respectively). Furthermore, mir-638 was found to correlate with the presence of extramedullary disease (p=0.026). As some miRNAs are known to destroy their target mRNAs, we compared our findings to our existing gene expression profiling data using the Sanger miRBase Target Database. miR-29c was predicted to target MAD2L1BP (AKA p31comet, CTM2), a tumor senescence inducing protein whose expression is down-regulated in WM by gene expression profiling (-2.1 fold, p<0.0001); in addition, miR-21 which is predicted to target IL-12A was also decreased by -2.0 fold by gene expression profiling (p=0.005).
The above findings demonstrate a distinct miRNA profile in WM, and implicate several miRNAs in the pathogenesis of WM, including genes involved in the regulation of tumor cell senescence and IL-12A. These findings provide a framework for the exploration of novel signaling pathways and therapeutic approaches in WM.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.