Enhanced combinatorial analysis of tumor cell-ECM interactions using design-of-experiment optimized microarrays.

Abstract

The dysregulated and fibrotic tumor microenvironment of hepatocellular carcinoma delays diagnosis and presents many complex signals that drive disease progression. To better recapitulate this microenvironment, we have enhanced our established protein microarray platform by integrating design of experiments (DoEs) methodology with high-throughput cell microarray screening. This innovative approach systematically interrogates the intricate roles of matrix stiffness (spanning healthy and fibrotic conditions), extracellular matrix (ECM) composition, and protein concentration, while simultaneously examining their interdependent interactions. By leveraging DoE principles, we were able to explore 117 unique microenvironments on a single microscope slide, ultimately generating a comprehensive dataset of 234 different microenvironments without compromising statistical rigor. Our enhanced screening system enabled the identification of unique microenvironmental interactions critically significant in dictating cellular responses, including adhesion, survival, proliferation, epithelial-to-mesenchymal transition, and drug resistance markers. Utilizing advanced statistical techniques such as linear models and principal component analysis, we characterized phenotypic clusters defined by precise microenvironmental cues. This work presents a robust, high-throughput microarray screening system that comprehensively explores the contributions of nine physiologically relevant ECM proteins and matrix stiffness in modulating cellular behavior and disease progression through a methodologically sophisticated and statistically sound approach.