Abstract
Multiple myeloma (MM) is characterized by a profound genomic instability and a highly variable clinical course. To comprehensively analyze the effects of copy number alterations on gene activity and prognosis, we applied novel bioinformatics and computational approaches to integrate high-resolution aCGH data (Agilent 244K array) and gene expression profiling data (Affymetrix U133Plus2.0 array) from CD138-selected plasma cells from 92 newly diagnosed MM treated with high-dose therapy and stem cell transplantation. We identified what we have termed the “atom region” (AR). ARs are found throughout the genome and define recurrent breakpoint boundaries of sub-chromosomal loss or gain. Genes coding for miRNAs were significantly enriched in regions of the genome defined by ARs. Although many AR-related breakpoints lie in intergenic regions, and thus capture whole genes, a substantial number were identified within genes, making these genes strong candidate disease genes. We found that gain of 1q and loss of 1p represent the most significant copy number alterations linked to early disease-related death. Importantly, 1q gains and 1p loss were significantly linked to a recently described 70-gene expression model of high-risk disease predominated by increased expression of genes mapping to 1q and reduced expression of genes mapping to 1p (1). Gain of 1q and loss of 1p were also highly correlated with a gene-expression-based proliferation index widely used across many tumor types. We provide evidence that gain of 8q24 was also highly correlated with the 70-gene high-risk score, but not the proliferation index. 8q24 gains and survival were linked to elevated expression of EIF2C2/AGO2, a master regulator of microRNA expression and maturation and an overexpressed gene previously identified in our 70-gene high-risk model (1). Interestingly, although the expression of MYC, which also maps to 8q24, was related to copy number changes, its expression was not related to survival. Given the importance of AGO2 in B-cell development (2), and miRNAs in cancer, detailed functional studies of the potential role of AGO2 in MM are underway.
Disclosures: No relevant conflicts of interest to declare.
References
Author notes
Corresponding author