Identification of NAE1-dependent β-catenin neddylation as selective vulnerability in B-cell malignancies

Abstract

Abstract Background and Significance. In epithelial cells, β-catenin promotes canonical WNT signaling and expression of MYC. Mutations that stabilize β-catenin protein expression function as oncogenic drivers in solid tumors and myeloid leukemia. In contrast, B-cell tumors lack activating β-catenin lesions. Despite similar mRNA levels, β-catenin protein levels in B-cell malignancies are 80- to 200-fold lower. We recently discovered that in B-cell tumors, β-catenin protein is constitutively phosphorylated by GSK3B and poised for proteasomal degradation (Cosgun et al., Biorxiv 2023). Because B-cells have evolved and critically depend on highly efficient mechanisms of β-catenin degradation, we performed genome-wide CRISPR-KO screens in β-catenin fluorescent reporter knock-in B-cell lines to identify components of this previously unrecognized B-cell-specific high-efficiency degradation complex. Results. Using HDRT, we engineered fluorescent reporter knockin alleles in B-ALL (NALM6) and mantle cell lymphoma (JEKO) cell lines for detection of β-catenin protein (mScarlet) in single living B-ALL and MCL cells. β-catenin-mScarlet signal was not visible at baseline levels. However, treatment with GSK3B inhibitor induced stabilization of β-catenin within hours, resulting in a 20-fold increase in β-catenin-mScarlet signal. We introduced iCas9 and CRISPR-KO Brunello library and sorted β-catenin-mScarlet+ cells to identify enriched guides by next generation sequencing. As expected, guides targeting classical components of the β-catenin destruction complex, APC, GSK3B, AXIN1 were strongly enriched in mScarlet+ cells. Interestingly, guides targeting the neddylation pathway, including NEDD8 activating enzyme (E1) subunits (NAE1, UBA3), NEDD8 E3 ligase (FBXW11) and NEDD8 itself were most prominently enriched in β-catenin-mScarlet+ cells, which identifies neddylation as an essential mechanism controlling β-catenin levels in B-cells. Engineering β-catenin reporter cells with single CRISPR-guides targeting NAE1, UBA3, FBXW11, and NEDD8, we validated each of these components by Western blot and flow cytometry, resulting in accumulation of β-catenin and acute cell death. Based on Nae1fl/fl mice, we generated a genetic model for inducible, Cre-mediated deletion of Nae1 in BCR-ABL1- and NRASG12D transformed murine B-ALL cells, which phenocopied CRISPR-mediated deletion in human B-ALL and MCL cells, resulting in acute cell death, loss of colony formation and leukemia initiation ability. Hematopoietic reconstitution of congenic CD45.1 mice with CD45.2 Nae1fl/fl hematopoietic progenitor cells resulted in multilineage reconstitution from CD45.2 Nae1fl/fl progenitors. Underscoring B-cell-selective activity of NAE1, acute ablation of Nae1 had no significant effect on myeloid, erythroid, NK and T-cell development but caused a near-complete block during early B-cell development at the pro- to pre-B cell transition. Translational Implications. The NAE1-inhibitors pevonedistat and TAS4464 have been studied in clinical trials hematological malignancies, including myeloid leukemia, multiple myeloma and lymphoma. However, β-catenin protein degradation was not previously considered a mechanistic target. Interestingly, publicly available compound screening data (CTD2; DepMap) revealed previously unrecognized B-cell-selectivity of both pevonedistat (P=2e-10) and TAS4465 (P=0.03). Treatment of B-cell malignancies with pevonedistat and TAS4464led to rapid accumulation of β-catenin and acute cell death at low nanomolar concentrations. Toxicity induced by NAE1-inhibitors was largely rescued by β-catenin deletion, confirming mechanism of action. Since β-catenin is phosphorylated by GSK3B, followed by neddylation and proteasomal degradation, we examined potential synergism between GSK3B- and NAE1-inhibitors. Importantly, both pevonedistat and TAS4464 strongly synergized with the GSK3B inhibitor LY2090314 in patient-derived B-ALL, mantle cell lymphoma and CLL samples. Conclusion. Based on CRISPR-screens, we discovered a novel B-cell-selective high-efficiency complex for β-catenin protein degradation, which depends on NAE1-dependent neddylation. Given the surprising B-cell-selectivity of this mechanism, our findings support a rationale to repurpose clinically tested NAE1-inhibitors for the treatment of refractory B-cell malignancies as single agent or in combination with GSK3B inhibitors.