TL;DR
This paper develops a mathematical model to determine the relative importance ("weights") of two key anti-apoptotic proteins, Bcl-2 and Mcl-1, in controlling programmed cell death. By quantifying these weights, the researchers provide a predictive tool to identify optimal drug combination ratios for cancer therapy, potentially reducing the need for extensive experimental screening.
What Problem Does This Paper Address?
Cancer cells often evade programmed cell death (apoptosis) by overexpressing anti-apoptotic proteins, particularly those in the Bcl-2 family. Inhibiting these proteins is a promising cancer treatment strategy. The Bcl-2 family includes proteins like Bcl-2 and Mcl-1 that act as "brakes" on apoptosis. While it's known that targeting both Bcl-2 and Mcl-1 can be more effective than targeting just one, the precise contribution (or "weight") of each protein to apoptosis inhibition has remained unclear. This lack of understanding makes it difficult to determine the ideal ratio of drugs that inhibit Bcl-2 and Mcl-1 to achieve the best therapeutic outcome. This paper aims to quantify these weights and use them to predict optimal drug combination ratios.
Methods
The researchers employed a systems biology approach, building a mathematical model based on ordinary differential equations (ODEs) to simulate the mitochondrial apoptosis pathway. This model incorporates protein-protein interactions, protein production and degradation rates, and drug-target interactions.
Key methodological steps include:
- Mathematical Modeling: An ODE model was developed to represent the core components of the mitochondrial apoptosis pathway, focusing on the interactions between Bcl-2, Mcl-1, and effector proteins like Bax.
- Correlation and Sensitivity Analysis: These analyses were used to identify the most critical molecular determinants influencing Bcl-2 and Mcl-1 dependency and to understand their relative "weights" in controlling apoptosis.
- Parameterization: The model was parameterized using experimental data from literature and their own experiments, including protein expression levels, drug binding affinities, and drug effects on protein degradation.
- Experimental Validation: The model's predictions were validated against experimental data from various cancer cell lines treated with combinations of Bcl-2 and Mcl-1 inhibitors. This included apoptosis assays and combination index (CI) calculations.
- Equation Derivation: Based on the model's insights, two sets of equations were derived to predict the optimal drug combination ratio (λ) using protein levels of Bcl-2 and Mcl-1 as inputs.
Datasets
The study utilized several sources of data:
- Literature Data: Existing experimental data on protein levels, drug effects, and cell line dependencies were incorporated.
- Experimental Data: The authors performed their own experiments, including immunoblotting to assess Mcl-1 levels after drug treatment and Annexin V-FITC apoptosis assays to measure cell death.
- Cancer Cell Line Encyclopedia (CCLE) Database: This database was used to obtain protein and mRNA expression levels for various cell lines, which were then normalized to serve as model parameters.
Key Insights and Ideas
- Quantifying "Weights": The study successfully quantified the relative contributions ("weights") of Bcl-2 and Mcl-1 to apoptosis inhibition, moving beyond qualitative understanding.
- Dynamic Mcl-1 is Crucial: Unlike the more stable Bcl-2, Mcl-1's short half-life and dynamic regulation (production and degradation rates) are critical determinants of its role in apoptosis and drug response.
- Predictive Equations for Drug Ratios: The development of two predictive equations that use protein levels of Bcl-2 and Mcl-1 to forecast the optimal drug combination ratio is a significant advancement.
- Model-Guided Therapy: The model and derived equations offer a powerful tool to guide experimental design and potentially clinical treatment by predicting effective drug ratios, saving time and resources.
- Beyond Static Levels: The research highlights that simply looking at static protein levels isn't enough; understanding the dynamics (production/degradation) of Mcl-1 is essential for predicting drug response.
Key Results
- Model Accuracy: The mathematical model accurately predicted the dependency of 27 out of 30 cell lines on Bcl-2 and/or Mcl-1, demonstrating its robustness.
- Key Determinants Identified: Bcl-2 protein level, Mcl-1 protein level, and Mcl-1 production and degradation rates were identified as the major molecular determinants for Bcl-2 and Mcl-1 dependency.
- Experimental Validation: The model's predicted best drug ratios were experimentally validated in MCF-7 and MDA-MB-231 cell lines, showing good agreement with observed synergistic effects (low CI values). For example, a combination of ABT-737/A-1210477 at a 1:2 ratio showed a CI of 0.25 in MDA-MB-231 cells, indicating strong synergy.
- Predictive Equations Performance: The derived equations for predicting the best drug ratio achieved high correlation coefficients (R² of 0.85 and 0.71) when compared to model predictions, indicating their reliability.
Limitations and Future Work
- Model Simplification: The authors acknowledge that their model is a simplification of the complex apoptosis network. For instance, they did not explicitly model Bax retrotranslocation by Bcl-xL, which a recent report suggests can influence MOMP sensitivity. However, they argue this simplification might not significantly impact the prediction of drug concentrations due to the similarity in degradation rates between Bcl-xL and Mcl-1.
- Generalizability: While validated in several cell lines, further validation across a broader range of cancer types and patient-derived samples would strengthen the model's applicability.
- Future Directions: The authors suggest that their model can be extended to incorporate other Bcl-2 family members and to predict drug combinations involving novel therapeutic modalities like PROTACs by incorporating their specific degradation mechanisms. They also propose using protein array data from clinical samples to guide treatment.
Summary updated