MicroRNA Profiling In Multiple Myeloma

Multiple Myeloma (MM) is a plasma cell malignancy, characterized by the accumulation of malignant plasma cells in the bone marrow (BM). Prognostic factors in MM include translocations and ISS stage; still, the clinical course is difficult to predict. MicroRNAs (miRNAs) are a class of small non-coding single stranded RNAs involved in posttranscriptional gene regulation, which may be of use in defining MM prognosis and outcome. MiRNAs regulate protein levels by binding to either partially or complete complementary sites in messenger RNAs (mRNAs), leading to translational repression or transcript degradation respectively. In this manner, miRNAs play a role in critical biological processes including cellular growth and differentiation. Specific disease related miRNAs in both acute myeloid leukemia and chronic lymphocytic leukemia have been found with specific miRNA signatures associated with different cytogenetic subtypes. Until now, information available for miRNA expression in MM is limited.

MiRNA expression profiling was performed in 45 newly diagnosed MM patients enrolled in the HOVON-65/GMMG-HD4 trial; a randomized, phase III trial performed to evaluate the efficacy of bortezomib prior to high-dose melphalan (HDM) for response, progression free survival (PFS) and overall survival (OS) in patients with newly diagnosed MM. As controls, four healthy BM samples were obtained from subjects undergoing BM harvest for allogeneic transplantation donorship. For all samples, BM derived CD138 selected plasma cells (PCs) with a minimum purity of > 80% were obtained. RNA was isolated using the miRVANA kit, with subsequent miRNA profiling by TaqMan Human MicroRNA Array v1.0. Unsupervised hierarchical clustering with the centered correlation metric with average linkage was performed using BRB-array tools 3.6.0. For survival analysis, miRNA expression was divided in quartiles: the top quartile vs the rest to identify cases with high expression and the bottom quartile vs the rest for cases with low expression. Log-rank tests for univariate association with PFS and OS were performed for each of the 365 miRNAs using the false discovery rate to correct for multiple testing. Chromosomal abnormalities t(4;14), t(11;14), t(14;16) and deletion 13q14 were determined by FISH analysis. MiRNA expression was compared to mRNA expression, available for 39 out of 45 MM patients, using a Spearman's rank correlation test. mRNA expression was determined by Affymetrix U133 Plus 2.0 arrays (Broyl et al., Blood 2010).

Clustering resulted in a dendrogram with 5 clear branches, consisting of 4 MM clusters and 1 normal BM cluster. The MM clusters are characterized by up- and downregulation of distinctive miRNAs: cluster A: upregulation of miRNA clusters miRNA-17∼92 and miRNA-106∼25 (n=23); cluster B: upregulation of miRNA-130a and miRNA-424 (n=8); cluster C: upregulation of miRNA-576 and miRNA-106b (n=9) and cluster D: upregulation of miRNA-372 and miRNA-200a (n=4). An additional cluster of one sample was not defined. MiRNAs predominantly expressed in normal BM were miRNA-28 and miRNA-30c. None of the miRNA clusters correlated with cytogenetic subgroups, i.e. deletion 13q14, t(4;14), t(11;14), and t(14;16) Still, a supervised approach showed significantly higher expression of miRNA-122a, miRNA-33, miRNA-489, miRNA-519e, and miRNA-555 in patients with t(11;14). Upregulation of let-7f, miRNA-194 and miRNA-296 expression were found to be associated with better OS with borderline significance (P = .06). Finally, a significant inverse correlation between miRNA-21 expression and gene expression of two of its validated targets, PDCD4 (P = 2× 10⁻⁴) and RECK (P = 8×10⁻⁴) was found. PDCD4 is a novel tumor suppressor, whose functions include inhibition of translation through eIF4A/eIF4G. RECK has been shown to be involved in angiogenesis, through inhibition of MMP2 and MMP9.

miRNA expression in MM is deregulated compared to normal PCs and MM patients can be classified according to their miRNA expression pattern in four clusters. Furthermore, a trend was found for high expression of miRNAs let-7f, miRNA-194 and miRNA-296 with increased OS. Integration of miRNA and mRNA data shows the putative interaction between miRNA-21 and two of its validated targets; tumor suppressor gene PDCD4 and RECK, suggesting a functional relationship between miRNA expression and development of MM.

Sonneveld:Johnson & Johnson: Membership on an entity's Board of Directors or advisory committees; Millennium Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees.