SNP Array Analysis Reveals Copy Number Alterations and Uniparental Disomy in Mantle Cell Lymphomas at High Resolution.

Abstract

Mantle Cell Lymphoma (MCL) is an aggressive B-Cell Non Hodgkin Lymphoma which is genetically characterized by the translocation t(11;14). This translocation leads to juxtaposition of the Cyclin D1 gene and the IgH locus, resulting in constitutive overexpression of Cyclin D1 and consecutive cell cycle dysregulation. Apart from this typical structural genetic alteration, several studies using conventional or array-based comparative genomic hybridization (CGH) reported a high number of secondary numerical genetic alterations contributing to MCL lymphomagenesis and influencing the clinical behavior. Increasingly, there is evidence that loss of heterozygosity (LOH) without copy number changes (e.g. caused by mitotic recombination between the chromosomal homologues, also referred to as acquired (partial) uniparental disomy (a(p)UPD), is an important alternative mechanism for tumor suppressor gene inactivation. However, this phenomenon is undetectable by CGH techniques. Single Nucleotide Polymorphism (SNP) based arrays allow - in addition to high resolution copy number (CN) analyses and SNP genotyping - in the same experiment the analysis of loss of heterozygosity (LOH) events and hereby enable the detection of copy neutral LOH. We analyzed the 3 t(11;14)-positive MCL cell lines Granta 519, HBL-2 and JVM-2 and 5 primary tumor specimens from untreated MCL patients with both the Affymetrix GeneChip®Human Mapping 100K and 500K array sets. In the 3 cell lines, we found an excellent agreement between the copy number changes obtained by SNP array analysis and previously published array CGH results. Extending published results (Nielaender et al., Leukemia 2006), we found regions of pUPD in all 3 MCL cell lines, which often affected regions reported as commonly deleted in MCL. Intriguingly, HBL-2 that is characterized by relatively few chromosomal losses, carries an increased number of large regions showing copy neutral LOH. Furthermore, we compared the results obtained by the 100K and 500K mapping array sets from 5 primary MCL tumor specimens with previously published conventional CGH data. All cases showed genetic alterations in both conventional CGH and SNP array analysis. The total number of copy number alterations detected by conventional CGH was 35, including 23 losses, 10 gains and 2 amplifications. The total number of CN alterations detected by the mapping 100K and 500K array sets was 81 (50 losses, 26 gains and 5 amplifications) and 82 (50 losses, 27 gains and 5 amplifications), respectively. We found an excellent agreement in the large CN alterations detected by conventional CGH and both SNP array platforms. Furthermore, we identified >40 mostly small CN alterations that have not been detected by conventional CGH (median size <5MB for losses and <3Mb for gains). The CN alterations detected by the 100k and the 500K array sets were highly identical. Importantly, we discovered regions of partial UPD in 4 of the 5 MCL cases (size range from around 2Mb up to a single region >40Mb). In conclusion, the results demonstrate the capability of SNP array analysis for identifying CN alterations and partial UPD at high resolution in MCL cell lines as well as in primary tumor samples.