Integrative Analysis Reveals Multiple Alterations of p53 Signaling Pathway Components In Primary Diffuse Large B-Cell Lymphomas

Abstract 635

Diffuse large B-cell lymphoma (DLBCL) is a genetically heterogeneous disease with infrequent alterations of multiple apoptotic, developmental and signaling pathways. Inactivating somatic mutations of the p53 tumor suppressor are uncommon in DLBCL, prompting speculation regarding alternative mechanisms of modulating p53 activity in this disease. As part of a comprehensive analysis of genetic alterations in DLBCL, we recently integrated high-resolution copy number data and transcription profiles in a series of 81 newly diagnosed previously untreated DLBCLs. Copy number (CN) alterations were evaluated using Affymetrix SNP Array 6.0 data on over 1.8 million genomic loci (906,000 SNP probe sets and 946,000 copy number probes) and a data-processing pipeline incorporating the normalization, calibration, and segmentation of the CN profiles and their analysis by the GISTIC algorithm. The genes within the peak and region of each GISTIC-identified alteration were tested for association with the corresponding transcripts' expression and significance values were corrected for multiple hypothesis testing by the false discovery rate (FDR) procedure, with the correction accounting for both the multiple genes within each peak and the multiple peaks. The signature for a given copy number alteration was then defined as the set of within-peak and -region transcripts with FDR q-values < 0.25. The alteration signatures thus defined were then tested for pathway/gene set enrichment based on the hypergeometric distribution, using a compendium of gene sets from the MSigDB repository (Broad Institute). This analysis highlighted a genome-wide pattern whereby: i) individual pathways are targeted by multiple alterations at different loci (with different genes of the same pathway located within distinct CN alteration peaks or regions); and ii) members of multiple pathways are targeted by the same alteration (with genes co-located within a single CN alteration belonging to different pathways). The most highly significant enriched pathways (FDR < .05) included apoptotic, p53 signaling, and ARF pathways and identified multiple modulators and effectors of normal p53 activity. These include: deletion of the MDM2-inhibitor, ARF; amplification of the p53 E3 ligases, MDM2 and COP1; deletion of p53 itself; deletion of the ribosomal protein that enhances p53 translation following DNA damage, RPL26; and deletion of p53 effectors including BNIP3L and DFFB. Of note, the deletion of 17p13.1 encompasses both p53 and its positive modulator, RPL26. In all cases, the observed alterations in p53 pathway components would decrease functionally active p53 protein and p53 apoptotic effectors in primary DLBCLs. Taken together, these data identify alternative genetic mechanisms for reducing p53 activity in DLBCL and highlight the value of an integrated comprehensive analysis of genetic alterations in this disease.