MALT1 Degradation with a Proteolysis-Targeting Chimera for the Treatment of Activated B-Cell Type Diffuse Large B-Cell Lymphoma

Introduction

ABC-DLBCL is characterized by chronic activation of NFκB, which is dependent on the CARD11-BCL10-MALT1 (CBM) complex. MALT1 activates NFκB transcription factors via distinct protease and scaffold functions. Clinical allosteric MALT1 protease inhibitors are in development, however these agents partially preserve NFκB activation via MALT1 scaffold function. In addition, MALT1 protease is critical for regulatory T cell (Treg) function, and its inactivation is associated with increased inflammatory cytokine expression by T-effector cells. However, overt inflammation is not seen with MALT1 knockout. It is therefore possible that MALT1 protease inhibition with preserved scaffold function will lead to autoimmune toxicities. As an alternative strategy, we developed MALT1-directed proteolysis targeting chimera (PROTAC) compounds which consist of a MALT1-binding domain linked to an IMiD-based cereblon (CRBN) binding domain, which induce ubiquitination and proteasomal degradation of MALT1. Here we present preliminary studies of a lead compound, PS-II-115, to assess feasibility of this approach for ABC-DLBCL.

Results

To assess MALT1 protein degradation, OCI-Ly3 cells were treated with DMSO or PS-II-115 10 µM for 24 hours and immunoblots of cell lysates were performed. PS-II-115 induced degradation of MALT1 (degradation ±SEM 73.94 ±9.8% vs DMSO, N = 3). No significant degradation of CRBN neosubstrates IKZF1, IKZF3, or GSPT1 was seen. To assess effects downstream of MALT1, we treated OCI-Ly3 cells with PS-II-115 for 23.5 hours followed by stimulation with PMA/IO for 30 minutes, and immunoblots were performed. We found increased IκB, an inhibitor of NFκB which is degraded following MALT1 scaffold mediated IKK activation, in PS-II-115 treated cells (IkB 206% compared to DMSO), suggesting that MALT1 degradation inhibits MALT1 scaffold-mediated NFκB activation pathways.

To measure MALT1 protease inhibition, we used the GloSensor split luciferase method, in which a chimeric protein is cleaved by MALT1 resulting in functional luciferase. Raji cells expressing this construct were treated with DMSO or PS-II-115 for 23.5 hours followed by stimulation with PMA/IO for 30 minutes, and GloSensor assay was performed. PS-II-115 induced dose-dependent inhibition of MALT1 protease activity compared to DMSO (IC ₅₀ 1.203 µM, 95% CI 0.374-4.230, N=4).

To assess growth inhibition, ABC-DLBCL and GCB-DLBCL cell lines were treated with DMSO or PS-II-115 for 96 hours, and CellTiter Glo assay was performed. GCB-DLBCLs are not dependent on chronic NFκB activation and were expected to be resistant to PS-II-115. PS-II-115 induced growth inhibition of ABC-DLBCL cell lines (OCI-Ly3 IC ₅₀ 2.54 µM, OCI-Ly10 IC ₅₀ 9.65 µM, TMD8 IC ₅₀ 1.87 µM, N=4) more potently than GCB-DLBCL cell lines (OCI-Ly1 IC ₅₀ 34.6 µM, OCI-Ly7 IC ₅₀ >50 µM, N=4), suggesting the effect of PS-II-115 is based on MALT1 rather than off-target effects.

We then assessed the impact of PS-II-115 on T cell phenotypes. Human T cells were isolated from healthy donors' peripheral blood. Cells were stimulated with CD3/CD28 beads and treated with DMSO, PS-II-115, or an allosteric MALT1 inhibitor for 48 or 96 hours and flow cytometry was performed. We found a significant decrease in CD4+ FOXP3+ CD25+ CD127- Tregs among cells treated with PS-II-115 1 µM or allosteric inhibitor compared to DMSO at 48 hours (PS-II-115 3.89%, allo inh 3.85%, DMSO 9.72% Tregs among CD4+ T cells, P<0.005). We also found decreased pro-inflammatory CD4+ CD45RA+ CCR7- terminal effector T cells expressing RA (TEMRA) among cells treated with PS-II-115 20 µM for 96 hours compared to DMSO (35.67% vs 56.97% TEMRA among CD4+ T cells, p<0.05). The allosteric inhibitor did not affect TEMRA cells.

Conclusion

We describe a MALT1-directed PROTAC which induces (1) degradation of MALT1 with preservation of other IMiD-associated CRBN neosubstrates, (2) blockade of MALT1 scaffold mediated NFκB activation pathways, (3) MALT1 protease inhibition, (4) selective growth inhibition of ABC-DLBCL cell lines, (5) decrease in Tregs, and (6) decrease in CD4+ TEMRA cells. These results provide proof of principle that MALT1 degradation using a PROTAC compound is a feasible strategy against ABC-DLBCL, and may have distinct effects on T-effector phenotypes compared to MALT1 protease inhibition. Further studies are needed to elucidate immunologic effects of MALT1 degradation and to validate findings in vivo.