Genome-wide analysis of DNA copy number changes in multiple myeloma using high-density SNP arrays

Introduction. Multiple myeloma (MM) is characterized by deep genomicinstability that involves both ploidy and structural rearrangements.Nearly half of MM tumors are non-hyperdiploid and frequently showchromosome 13 deletion and the common chromosomal translocationsinvolving the immunoglobulin heavy chain (IGH) locus on chromosome14q32. The remaining tumors are hyperdiploid, showing low prevalenceof IGH translocations and chromosome 13 deletions. Despite therecent advances, the spectrum of genetic lesions leading to biological andclinical variability of MM has not been defined yet. Our study wasaimed at defining a genome-wide pattern of genetic lesions in a representativeand stratified panel of MM patients using novel high-throughputapproaches, to provide insights into the genomic heterogeneity associatedwith plasma cell neoplasms. Methods. Genome-wide profilingdata of 57 MM patients, 7 plasma cell leukemia (PCL) and 14 normalsamples were generated on the GeneChip® Human Mapping 50K XbaSNP arrays. The local DNA copy number variations were calculatedusing the DNAcopy Bioconductor package. The gene expression profilesof the tumor samples were generated on the GeneChip® HG-U133Aarrays. Hierarchical clustering algorithms were applied to identify thenatural grouping of gene expression and genome profiles. Results. Anunsupervised analysis on the genome-wide profiles of our dataset showed that chromosome 1q gains, chromosome 13 deletions andhyperdiploidy are the main genetic aberrations driving samples grouping.High frequencies of increased DNA copy number levels were foundin six extended regions, specifically chromosome 1q and those chromosomesinvolved in hyperdiploidy (7, 9, 11, 15 and 19). As regards theregions with decreased copy number, we found the involvement of thechromosomal regions 6q and 16q, together with the newly reported 18qand 4p that warrant further investigations. The depicted clusteringstrongly correlated with that generated on the gene expression profilesof the patients in the dataset. Furthermore, we identified different geneswith altered expression level corresponding to local copy number variations.Discussion. Our results showed that genomic structural abnormalitiesin multiple myeloma closely reflected in expression imbalances.Our data reinforce the importance of using novel high-throughputapproaches to provide insights into the characterization of novel potentialgenetic lesion in primary myeloma tumors.