Quantitative characterization of cellular membrane-receptor heterogeneity through statistical and computational...

Area of Science:

• Computational biology
• Cancer research
• Cellular heterogeneity

Background:

• Cell population heterogeneity influences cellular response and drug resistance.
• Existing techniques for analyzing heterogeneity and its link to population response are limited.
• High-throughput 'cellomic' approaches enable heterogeneity profiling across multiple scales.

Purpose of the Study:

• To develop and present comprehensive statistical and computational methods for analyzing cellomic data and integrating it into deterministic models.
• To investigate how considering cellular heterogeneity impacts computational models of anti-angiogenic drug effects.
• To identify specific tumor cell subpopulations and their VEGFR expression levels that affect anti-VEGF treatment efficacy.

Main Methods:

• Developed a novel method for optimizing statistical distribution fits to heterogeneous cellomic data, preserving data integrity and excluding outliers.
• Integrated cellomic data, focusing on vascular endothelial growth factor receptor (VEGFR) membrane localization in endothelial cells, into deterministic computational models.
• Compared different methods of representing heterogeneous data to assess their impact on model predictions.

Main Results:

• Methodology for representing heterogeneous data can alter model predictions by up to 3.9-fold.
• VEGF levels are more sensitive to VEGFR1 cell surface levels than VEGFR2; updating VEGFR1 resulted in a 64% change in free VEGF, versus 17% for VEGFR2.
• Tumor cell and tumor endothelial cell (tEC) subpopulations with high VEGFR expression (>35,000 VEGFR/cell) were found to negate anti-VEGF treatments.

Conclusions:

• Considering cellular heterogeneity provides significant modeling insights into anti-angiogenic drug effects.
• Lowering the VEGFR membrane insertion rate in high-VEGFR subpopulations can restore anti-angiogenic treatment efficacy, identifying new therapeutic targets.
• This novel characterization of heterogeneous distributions demonstrates how data representation impacts drug efficacy predictions.