PROJECT SUMMARY Metastasis is the primary cause of cancer mortality, yet few breast cancer drugs effectively inhibit metastasis. Breast cancer cells use collective migration to remodel and align surrounding extracellular matrix (ECM) fibrils, which facilitates invasion. Aligned tumor stroma topography can induce cluster budding and dissemination of breast cancer cells. The goal of this project is to identify chemotherapeutic drugs using engineered biomimetic tumor invasion models and to evaluate therapeutic feasibility of inhibiting the target genes involved in breast cancer dissemination. To achieve this goal, we developed a quasi-3D nanotopographically patterned substrate and are incorporating it into a nanopatterned impedance electrode array (nanoIEA) to quantify collective cell migration and proliferation in real-time at high-throughput. We are validating a 3D aligned collagen fiber hydrogel model with control over fiber alignment that recapitulates the fiber dimension and orientation of in vivo breast tumor stroma. These models markedly promote breast cancer cluster dissemination and increase its resistance to chemotherapy. In our preliminary study, we have identified differentially expressed genes via RNA-seq between ‘disseminated tumor cell clusters’ and ‘non-disseminated tumor cells’ using the quasi-3D model. We will pursue three aims that leverage our expertise in cancer molecular biology/genomics (Ahn), tissue engineering (Kim), machine learning (ML)-based image analysis (Lee), cancer organoids/metastasis (Ewald), and pharmacology/drug development (Liu). Human breast cancer patient-derived xenograft (PDX) cell clusters/organoids will be investigated in this project. In Aim 1, we will evaluate effects of the following drugs on collective cell migration and on growth using the nanoIEA: [a] the 23 oncology drugs (out of 147 drugs we tested) which most significantly inhibited the viability of breast cancer cells in the quasi-3D model, [b] the 73 non- oncology drugs which inhibited the viability of 22 breast cancer cell lines by at least 4-fold in conventional 2D culture, and [c] the 95 inhibitors of target genes (CYP1A1, CYP1A2, CYP1B1). In Aim 2, we will characterize phenotypic responses of breast cancer cells/organoids to the identified drug candidates from Aim 1 using live cell microscopy and ML analyses. Phenotypic changes (e.g., motility, morphology) will be quantified to contrast subpopulations (non-invasive vs. invasive) and drug-treated cells vs. untreated. In Aim 3, we will evaluate therapeutic feasibility of regulating the target genes to inhibit cancer invasion. We will determine expressions of target genes at protein levels in PDX organoids, then correlate these with organoid invasiveness in the 3D model. We will then determine how inhibition of the target genes influences chemosensitivity of PDX organoids and suppresses their invasiveness. This project will increase our understanding of the mechanisms of topography- induced breast cancer dissemination and establish our tumor ECM-mimetics, nanoIEA, and ML imaging analysis as a preclinical cancer invasion model/assay to characterize heterogenous cell populations with different metastatic phenotypes and to identify chemotherapeutic agents that directly inhibit breast cancer invasion.

PROJECT NARRATIVE Metastasis is the main cause of cancer mortality, yet the development of treatments directly inhibiting metastasis is still an unmet clinical need. This project aims to identify chemotherapeutic drugs using engineered biomimetic tumor invasion models and to evaluate therapeutic feasibility of inhibiting the target genes involved in breast cancer dissemination to suppress cancer invasion. This study has implications for the preclinical screening of anti-cancer drugs, for characterizing subtypes of breast cancers, and will increase our understanding of the molecular mechanisms of breast cancer invasiveness.