Lung cancer is the leading cause of cancer mortality with poor prognosis due to late stage diagnosis, drug resistance and high risk of relapse. There is a high need for tissue engineered 3D models that can recapitulate tumor heterogeneity and complexity to understand the cellular mechanisms leading to lung tumorigenesis, metastasis and drug responses. Future of precision oncology is envisioned with the ability to perform therapeutic tests on pre-clinical tumor models that can faithfully recapitulate the native tumor microenvironment combined with the use of patient-derived cells that would inform the treatment decision making. Patient-derived xenografts (PDX) and tumor-derived organoids (TDO) have emerged as methods to provide reliable pre-clinical models that account for genomic diversity and cellular heterogeneity especially for cancers with lack of established cell lines. However, PDX models are costly with low tumor formation rate that limit medium to high-throughput screening approaches. On the other hand, TDO allowed in vitro culturing of both solid tumor and liquid biopsy-derived cells in Matrigel and revealed preservation of genomic diversity. Nevertheless, Matrigel culturing hinders a systematic study of the role of tumor extracellular matrix (ECM) components and mechanics on tumor cell growth, phenotype, metastatic potential and drug responses.
Lung tumors are marked by an increase in tissue stiffness as well as changes in the biochemical composition (i.e. increase in cell-instructive ECM ligands). We developed a biomaterial-based human in vitro human non-small cell lung adenocarcinoma model to study the effect of aberrant tumor matrix characteristics in a controlled manner on the phenotype and malignancy of pulmonary epithelium. We built models for healthy and tumorous lung matrix from hydrogels of decellularized native lung extracellular matrix (ECM) with differing ligand content and tissue stiffness. We then encapsulated non-small cell lung adenocarcinoma cells (A549) in both healthy-like (low stiffness, low ligand content) and tumor-mimetic lung matrices (high stiffness, high ligand content) and monitored cell growth and phenotype over 4 weeks. We performed analyses on gene (qRT-PCR, RNAseq) and protein expression (Western blot, immunofluorescence) to investigate the signaling mechanisms involved in the tumor matrix-mediated effects on cell growth and phenotype. We performed loss of function (small molecule, shRNA) and overexpression studies to validate proposed mechanisms triggered in the cells in the respective engineered microenvironments.
Lung tumor cells in tumor-mimetic matrices demonstrated significantly higher cell growth with increased number of larger and invasive-looking (decreased circularity) colonies. Gene and protein expression analyses revealed an upregulation of the epithelial-mesenchymal transition program with significantly higher expression of known markers including N-Cadherin, Zeb-1 and Twist-1 as well as lung adenocarcinoma markers such as TTF-1 when compared to heathy-like microenvironments. Mechanistic studies revealed ECM-ligand mediated induction of aberrant growth and tumorigenic phenotype that synergizes with the increased mechanical stiffness in the microenvironment.
Understanding the key characteristics of lung tumor microenvironment and recapitulating the compositional and mechanical differences in tissue engineered models hold a great importance towards achieving cellular responses seen in patients to steer therapeutic approaches for better clinical outcome.