TRAF3 Loss-of-Function Reveals the Non-Canonical NF-Κb Pathway As a Therapeutic Target in Diffuse Large B-Cell Lymphoma

Introduction: Effective treatment of diffuse large B-cell lymphoma (DLBCL) remains hampered by extensive molecular, clinical and pathological heterogeneity. Oncogenic activation of the nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) pathway regulates gene expression programs contributing to tumor cell survival and differentiation. NF-κB signaling occurs via two branches, canonical and non-canonical, and is often enriched in somatic mutations of key pathway members in several lymphoid malignancies, including DLBCL. While canonical NF-κB activation has been extensively characterized as a key pathogenic driver of disease, the clinical significance and functional role of the non-canonical (NC) NF-κB pathway is less well understood.

Methods and Results:

By performing Affymetrix SNP6.0 copy number analysis on 347 de novo DLBCL samples from patients uniformly treated with rituximab, cyclophosphamide, doxorubicin, vincristine and prednisone (R-CHOP), we identified frequent, focal genomic losses of chr:14q32.31-32, which included the TRAF3 and RCOR1 loci (7%) in the minimally deleted region. FISH on select cases confirmed these alterations, including hemizygous and homozygous TRAF3 deletions. RNA-Seq performed on these DLBCL specimens revealed a significant reduction of TRAF3 mRNA in chr:14q32.31-32 deleted cases, compared to copy number neutral cases (p<0.0001).

TRAF3 is a known regulator of normal B-cell survival and differentiation. To investigate the specific role of TRAF3 in DLBCL, we generated isogenic TRAF3 knockout DLBCL cell lines (DOHH-2, OCI-Ly1, HBL1, OCI-Ly3). We interrogated the activation status of the canonical and NC NF-κB pathways by immunoblotting and luciferase reporter assays, and uncovered pronounced nuclear localization of NC NF-κB subunits RelB and p52, and increased luciferase activity in TRAF3 loss-of-function (LOF) cells, compared to wildtype (WT) cells. Conversely, nuclear localization of the canonical NF-κB transcription factor complex subunit RelA was not significantly affected by TRAF3 LOF. These results suggest TRAF3 LOF specifically upregulates the NC NF-κB pathway in DLBCL. Next, to elucidate the underlying transcriptional programs triggered by TRAF3 LOF, we performed RNA-Seq on our isogenic cell line models. Differential gene expression analysis revealed upregulation of NF-κB targets, such as IL10 and BIRC3, and NF-κB associated pathways (e.g., NIK/NF-κB signaling, TNF signaling pathway) in TRAF3-deficient cells. Furthermore, using a supervised analysis we found significant enrichment for NC NF-κB genes both in TRAF3 LOF cell models and in patient samples carrying TRAF3 deletions.

TRAF3 is a direct regulator of the steady-state levels of the central NC NF-κB kinase NIK. We observed differential accumulation of NIK at the protein level in TRAF3 LOF cells, prompting us to assess the therapeutic potential of NIK inhibitors. We performed dose-response analysis in vitro on DLBCL cells using two NIK inhibitors (isoquinoline-1,3(2H,4H)-dione and NIK SMI1), using viability assays as readouts. TRAF3 LOF cells were consistently more sensitive to NIK inhibition compared to WT cells in all of our isogenic models (e.g., isoquinoline-1,3(2H,4H)-dione IC₅₀ TRAF3 WT range 2.4-5.6µM vs IC₅₀ TRAF3 LOF range 1.6-1.8µM). Moreover, we observed reduced proliferation and an increase in apoptosis among TRAF3 LOF cells, following treatment with NIK inhibitors. To validate this finding through an orthogonal approach, we performed MAP3K14(NIK) shRNA knockdowns and observed a differential reduction in viability for TRAF3 LOF cells, relative to TRAF3 WT cells.

Conclusion: Our findings show that recurrent TRAF3 LOF events induce transcriptional reprogramming in DLBCL, and highlight NC NF-κB activation as a pathogenic dependency in a subset of tumors harboring these deletions. We provide proof-of-concept for targeting NC NF-κB deregulated lymphomas and uncover NIK as a potential therapeutically actionable target. Future studies will explore the utility of NC NF-κB inhibitors in vivo.

Scott:Incyte: Consultancy; Roche: Research Funding; AstraZeneca: Consultancy, Honoraria; Janssen: Consultancy, Research Funding; Abbvie: Consultancy; NanoString: Patents & Royalties. Steidl:Epizyme: Research Funding; Roche: Consultancy; Bristol-Myers Squibb: Consultancy; Bayer: Consultancy; AbbVie: Consultancy; Curis Inc: Consultancy; Seattle Genetics: Consultancy; Trillium Therapeutics: Research Funding.