Amplification of 6p Associated with Familial CLL

Abstract 2432

Chronic lymphocytic leukemia (CLL) is one of the most familial of all cancers but the genetic basis of this heritability remains poorly characterized. Families with very strong inheritance of CLL have been described in the literature, and recently the occurrence of CLL in one such family was associated with a polymorphism in the DAPK gene. Here we report the genomic characterization of a family in which CLL appears to be inherited in a Mendelian autosomal dominant manner. Within this family, five of eleven siblings of the first generation were affected, and one of those affected siblings had five children, of whom three were also affected (the second generation). The children of the second generation are currently aged 20–30 and hence too young to know whether they will develop CLL. We performed high-density single-nucleotide polymorphism (SNP) array analysis and gene expression profiling on tumor and germline DNA from four of the offspring of the second generation, as well as six of their children. Analysis of the SNP array data revealed a significant germline amplification of 6p, spanning 0–720 Mb and encompassing a known copy number variant (CNV) region but significantly larger than the CNV region. This amplification was found in both affected individuals with samples available from the second generation, and was transmitted by each of them to one of their two children in the third generation. This amplification was absent from the two unaffected members of the second generation, their children, or any of the other 189 individuals with CLL who were analyzed in our high-density SNP array dataset. None of the unaffected individuals with or without the amplification had evidence of monoclonal B cell lymphocytosis (MBL) by highly sensitive flow cytometry. These unaffected individuals also lacked any PCR-detectable oligoclonal or monoclonal immunoglobulin heavy chain gene rearrangement suggestive of MBL. The region of amplification contains four protein-coding genes: EXOC2, DUSP22, HUS1B and IRF4. We sequenced the coding regions of these four genes and the 5` and 3` UTRs of IRF4 in all family members, but found no somatic mutations in this family. All four genes were also sequenced in 92 other familial CLLs, identifying no somatic mutations. We then analyzed our gene expression profiling data to assess whether any genes in this region were altered in the affected individuals with the amplification. This analysis revealed a significant 1.74X increase in IRF4 expression in the CLLs with the amplification compared to those without (q value < 0.001). By Western blotting, we confirmed that IRF4 protein was increased approximately two-fold in amplified compared to non-amplified samples. These data suggest that the amplification may target IRF4, which has been previously implicated in CLL by a genome wide association study that identified a tag SNP in its 3` UTR as a CLL risk allele. Further analysis of our SNP data demonstrated allele specific amplification in this region, and mass-spectrometric genotyping confirmed enrichment of the CLL risk allele in the individuals with amplification. We conclude that amplification of IRF4 carrying the risk allele for CLL appears likely to be the culprit predisposing to CLL in this family.