Abstract 1878: Profiling protein-protein interactions to predict sensitivity to drugs targeting BCL2 family members in lymphoma models

Abstract Introduction: Pharmacological inhibition of anti-apoptotic proteins, especially with BCL2 inhibitors (i), has proven its clinical efficacy for treating patients with hematological cancers. However, there are not yet widely available tools to predict the sensitivity of individual patients to the different drugs targeting BCL2, MCL1, or BCLXL. A new technology, Single-molecule Protein Interaction Detection (SPID), based on immunoprecipitation of anti-apoptotic proteins followed by the addition of fluorescently labeled probes and detection at single-molecule accuracy, has shown promising results in predicting the sensitivity of acute myeloid leukemia cells to the BCL2i venetoclax (Chun et al., 2024). Here, we explored SPID to perform extended profiling of protein-protein interactions of BCL2, BIM, BAX, BAK, BCLXL and MCL1 in models derived from mantle cell lymphoma (MCL), marginal zone lymphoma (MZL) including models of secondary resistance to BTK, PI3K and BCL2i. Methods: Cell viability was assessed by MTT assay after 3 days of exposure to increasing concentrations of drugs or DMSO. Sensitivity parameters included IC25, IC50, IC90, area under the curve, cell viability at 1µM of drug, or at the median of the concentration range used in the experiment. BCL2 family members’ levels as single proteins or in complexes were measured by SPID. Drug sensitivity parameters were integrated with protein data using Spearman Correlation and Multiple Linear Regression (MLR, Lasso). Results: We exposed MCL (n=10, JVM2, REC1, JEKO1, UPN1, SP49, SP53, MAVER1, GRANTA519, Z138, MINO), MZL (n=6, VL51, SSK41, Karpas1718, HC1, HAIRM, ESKOL) cell lines and MZL models of secondary resistance (MZL-RES) to idelalisib (n=2), copanlisib (n=2), ibrutinib (n=2), venetoclax (n=1) and copanlisib/venetoclax (n=2), to BCL2i (venetoclax, sonrotoclax), BCL2/BCXLi (navitoclax), and MCL1i (S63845). MCL models showed increased sensitivity to BCL2i, while MZL cells were more sensitive to MCL1i and BCLxLi. Elevated BCL2, BCL2-BAX, and, to a lesser extent, BCL2-BIM were associated with sensitivity to BCL2i, while high levels of MCL1-BIM, BCLxL and BCLxL-BIM were associated with BCL2i resistance. Expression of BCLxL, BCLxL-BAK and BCLxL-BIM predicted sensitivity to navitoclax, while resistance was observed with elevated MCL1-BIM. The response to the MCL1i did not correlate with MCL1 or MCL1-BIM levels, but an MLR model combining BCL2-BIM, BCLxL-BIM and MCL1 levels significantly enhanced the prediction of response. Conclusions: Anti-apoptotic inhibitors show variable heterogeneous activity in MCL and MZL models, including those resistant to PI3Ki and BTKi, which could be predicted quantifying the BCL2-family members levels and interactions using the SPID. Such an approach might have a clinical potential to predict the sensitivity of patients with mature B cell lymphomas to this class of agents. Citation Format: Alberto J. Arribas, Byungsan Choi, Elisa Civanelli, Eleonora Cannas, Camila Scalise, Saem Hong, Hongwon Lee, Francesco Bertoni. Profiling protein-protein interactions to predict sensitivity to drugs targeting BCL2 family members in lymphoma models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 1878.