Evaluating Protein Interaction Potential of BCL2 Protein for the Direct Assessment of Inhibitor-Target Engagement in Hematological Cancer

Introduction: Understanding the pharmacodynamics (PD) of drugs is critical for designing effective clinical trials. Compounds targeting anti-apoptotic proteins, such as venetoclax, exert their therapeutic effects by inhibiting interactions with pro-apoptotic proteins. Consequently, changes in the levels of protein complexes between anti- and pro-apoptotic proteins are used as surrogate PD biomarkers. However, the levels of these protein complexes can vary significantly across individuals, and their formation is highly dependent on specific cancer contexts. These variations complicate the assessment of exact drug-target binding status in clinical specimens. Single-molecule Protein Interaction Detection (SPID) technology enables a comprehensive characterization of anti-apoptotic proteins such as BCL2. Utilizing single-molecule fluorescence imaging technology, SPID offers more than 100-fold enhanced sensitivity compared to traditional immunoassays, allowing for the detection of anti- and pro-apoptotic protein complexes more accurately. Here, we propose a strategy how to characterize BCL2 to dissect its engagement with applied BH3-mimetics.

Methods: To profile the molecular information of BCL2, HL60 and KMS12-BM cell line was selected. The protein interactions of BCL2 were modulated by treating the cell line with BCL2-specific compounds—venetoclax, sonrotoclax, and S65487—during cell culture for 4 hours. The compound-treated cells were harvested and subjected to SPID analysis. For the direct competition assay, we applied our novel probe binding assay (PBA) to BCL2. Specifically, BCL2 proteins collected from the prepared samples were immobilized on the surface of our SPID chip. We then employed a BH3 domain of BIM conjugated with Cy3 (BIM_BH3-Cy3), mimicking the endogenous interactor BIM. Simultaneously, we introduced the BCL2-specific compound at various doses. This mixture led to a competitive interaction between BIM_BH3-Cy3 and the introduced compound. The BIM_BH3-Cy3 count was assessed using SPID analysis to validate the direct competition at the same binding site on BCL2.

Results: The SPID analysis for protein complexes confirmed that each compound displaced the target anti-apoptotic protein complexes, showing less than 20% of baseline levels. The probe used in our PBA can interact with BCL2 in its unoccupied form endogenously. By utilizing the PBA with BIM_BH3-Cy3, we demonstrated that BCL2 in venetoclax-treated samples can interact with the applied BIM_BH3-Cy3. However, sonrotoclax treatment led to the generation of an interaction-less form of BCL2. To explore whether the BCL2 compounds compete with BIM_BH3-Cy3 at the same binding site on BCL2, we applied a mixture of the probe and the compound for the PBA. The PBA, along with dose titration, revealed that all three compounds successfully inhibited the interaction between BIM_BH3-Cy3 and BCL2. Remarkably, low PBA counts remained only in the sonrotoclax-treated condition when the compounds were applied prior to the probe application. This result indicated that venetoclax is more reversible than sonrotoclax after engaging with the BCL2 protein.

Conclusion: We present a precise assessment of unoccupied amounts of BCL2 protein using SPID analysis. Since current BCL2-specific compounds are designed to mimic a BH3 domain, our probe competes directly with these compounds at the same binding site on BCL2. Notably, sonrotoclax exhibited prolonged engagement with BCL2 even after cell lysis. Our analysis can be expanded to other anti-apoptotic proteins easily. Our approach can provide kinetic characteristics of BH3 mimetics as well as pharmacodynamic information valuable for clinical trial optimization.