Mutations in Multiple Genes Cause Deregulation of the NFkB Pathway in Diffuse Large B-Cell Lymphoma

Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous disease comprising multiple biologically and clinically distinct subgroups, including germinal center B cell-like (GCB) and activated B cell-like (ABC) DLBCL. Gene expression profile studies have shown that a key feature of its most aggressive subtype, ABC-DLBCL, is the constitutive activation of the NF-kB transcription complex. However, except for a small fraction of cases (Lenz et al., Science 2008), it remains unclear whether NF-kB activation in these tumors reflects an intrinsic program of the cell of origin or represents a primary pathogenetic event. To address this question, we first characterized 165 DLBCL samples (18 cell lines and 147 primary biopsies, including 26 ABC, 28 GCB, 10 unclassified and 83 not profiled) for the presence of active, nuclear NF-kB complexes by using immunohistochemical/immunofluorescence staining of NFKB1 p105/p50 (as a readout for the canonical pathway) and NFKB2 p100/p52 (as a readout for the non-canonical pathway). Nuclear localization of NF-kB, indicative of constitutive activity, was observed in 14/26 (54%) ABC-DLBCL and 8/28 (28%) GCB-DLBCL primary biopsies, as well as in 3/10 (30%) unclassified and 48/83 (58%) non-profiled cases, and correlated with significant enrichment in expression of NF-kB target genes, as assessed by gene set enrichment analysis (GSEA) of transcriptionally profiled cases. In addition, the more sensitive GSEA approach detected a gene expression signature of NF-kB activity in >90% ABC-DLBCLs and 53% GCB-DLBCLs, indicating that constitutive activation of this key signaling pathway is a common feature of ABC-DLBCL but can also be observed in a smaller fraction of GCB-DLBCL. To investigate whether NF-kB activity represents a primary pathogenetic event, we then screened for mutations the complete coding sequences of 31 genes encoding for NF-kB pathway components in a panel of 14 ABC-DLBCLs, which was expanded to 48 samples (12 ABC, 26 GCB and 10 not profiled) for validation of the mutated genes. The results showed that >50% of ABC-DLBCL (n=15/26) and a smaller fraction of GCB-DLBCL (n=8/26, 31%) carry somatic mutations in multiple genes, including negative (TNFAIP3/A20) and positive (CARD11, TRAF2, TRAF5, MAP3K7/TAK1 and TNFRSF11A/RANK) regulators of NF-kB. Of these, the A20 gene, which encodes for a ubiquitin-editing enzyme involved in termination of NF-kB responses, is the most commonly affected, with ~27% ABC-DLBCLs and 25% (8/32) immunohistochemically classified non-GC DLBCL displaying biallelic A20 inactivation by somatic mutations and/or deletions. Sequence changes include premature nonsense mutations, frameshift deletions/insertions and splice site mutations, leading to severely truncated proteins that lack functionally relevant domains and have thus lost their enzymatic activity. In virtually all mutated cases, FISH analysis revealed loss of the second allele, while 4 additional samples showed biallelic deletion of the gene. Thus, A20 is inactivated by a classic “two-hit” mechanism, suggesting a tumor suppressor role. Less frequently, missense mutations of CARD11 (10%) and TRAF2 (4%) produce molecules with significantly enhanced ability to activate NF-kB in transient transfection/reporter gene assays. Our results demonstrate that NF-kB activation in DLBCL is caused by genetic lesions affecting multiple genes, whose loss or activation may promote lymphomagenesis by leading to abnormally prolonged NF-kB responses. These findings provide the rationale and the assays for the identification of patients amenable to NF-kB targeted therapeutic intervention.