Harnessing repressive LEF1/β-catenin complexes to overcome drug resistance in chronic lymphocytic leukemia Free

Background and Significance. LEF1 is overexpressed and often used as diagnostic marker for CLL. LEF1 is a transcriptional target and canonical partner of β-catenin, which acts as a transactivator of WNT/β-catenin signaling. Consistent with LEF1-expression, previous reports showed activation of β-catenin in CLL although genetic lesions stabilizing β-catenin were not found. We recently discovered that CLL express 80-200-fold lower levels of β-catenin protein than solid tumors and depend on efficient β-catenin degradation (Cosgun et al., Biorxiv 2023). Here, we identified pharmacological accumulation of β-catenin as a promising strategy for the treatment of refractory and Richter Transformation (RT) CLL.

Results. To study the role of β-catenin in CLL, we introduced B-cell specific deletion of β-catenin in Eμ-Tcl1 CLL model (Mb1-Cre x Ctnnb1^fl/fl). β-catenin was dispensable for CLL development and progression and CD5+ CD19+ CLL burden and survival were not changed by B-cell-specific β-catenin deletion. In a converse experiment, tamoxifen induced stabilization of β-catenin (Mb1-Cre-ER^T2 x Ctnnb1^ex3/+) subverted leukemogenesis: while 7 out of 11 wildtype littermates developed CLL (median survival=426 days), none of 11 mice with B-cell-specific β-catenin stabilization succumbed to CLL (P=0.02). Interestingly, β-catenin stabilization in B-cells resulted in selective depletion of CD5+ CD19+ CLL cells, which express high levels of nuclear LEF1 (intracellular FACS) and spared CD5- CD19+ normal B-cells that lack LEF1 expression. For this reason, we hypothesized that LEF1 may be mediating β-catenin induced toxicity in CLL cells. Indeed, sensitivity to GSK3B-inhibition on β-catenin accumulation strictly correlated with LEF1 expression levels (P=0.0004).

In patient derived CLL cells (n=10), GSK3B-inhibition with LY2090314 mediated stabilization of β-catenin induced cell death at low nanomolar concentrations (IC50=3.5 nM). Three GSK3B inhibitors LY2090314, AZD1080 and CHIR99021 were assessed in Phase 1/2 clinical trials and achieved favorable safety and PK/PD profiles. Chemogenomic CRISPR screens for LY2090314, AZD1080 and CHIR99021 confirmed β-catenin accumulation and LEF1 expression as central mechanism of action with guide RNAs targeting CTNNB1 and LEF1 as top-ranking rescue hits. Loss of LEF1 in leukemia cells subverted nuclear translocation of β-catenin, reverted transcriptional programs induced by β-catenin activation and completely rescued β-catenin induced toxicity. Transcriptomic analysis identified that although β-catenin and LEF1 interaction led to an increase in expression of β-catenin target genes, MYC was repressed. This was surprising as in other cell types, β-catenin interacts with TCF/LEF family factors to activate MYC. CUT&RUN experiments confirmed loss of H3K27Ac signal at MYC enhancer regions upon β-catenin stabilization, which was reverted by LEF1 deletion restoring expression of MYC. Globally, β-catenin activation led to an increase of H3K27Ac signal at 10,383 regions while in 2,877 regions H3K27Ac was decreased. Further analysis of the decreased regions identified that these regions were co-bound by LEF1 and NuRD repressor components. LEF1 and β-catenin interactions led to recruitment of Nucleosome Remodeling and Deacetylase complex (NuRD), which resulted in loss of H3K27Ac signal and gene repression. Co-IP experiments using anti-β-catenin antibody confirmed interaction with LEF1, Ikaros factors and NuRD complexes in B-cell leukemia while LEF1 deletion resulted in loss of these interactions, which was particularly prominent at MYC superenhancer regions. Consistent with this model, we identified that deletion of LEF1 restored loss of H3K27Ac signal, Myc repression and rescued β-catenin mediated toxicity.

Conclusions. In summary, we discovered that β-catenin is dispensable for CLL development. Unlike solid tumors and myeloid leukemias, inducible stabilization of β-catenin leads to selective elimination of CLL cells expressing high levels of LEF1. Mechanistically, LEF1 mediates nuclear translocation of β-catenin, enabling assembly of repressive complexes with NuRD factors which induce Myc repression and cell death. Three GSK3B inhibitors LY2090314, AZD1080 and CHIR99021 achieved favorable safety and PK/PD profiles in previous clinical trials for solid tumors and neurological disorders. Our results provide a rationale for repurposing GSK3B-inhibitors for the treatment of refractory and RT CLL.