Traditional studies using cancer cell lines are often performed on a two-dimensional (2D) cell culture model with a low success rate of translating to Phase I or Phase II clinical studies. In comparison, with the advent of developments three-dimensional (3D) cell culture has been championed as the latest cellular model system that better mimics in vivo conditions and pathological conditions such as cancer. In comparison to biospecimens taken from in vivo tissue, the details of gene expression of 3D culture models are largely undefined, especially in mesothelioma – an aggressive cancer with very limited effective treatment options. In this study, we examined the veracity of the 3D mesothelioma cell culture model to study cell-to-cell interaction, gene expression and drug response from 3D cell culture, and compared them to 2D cell and tumor samples. We confirmed via SEM analysis that 3D cells grown using the spheroid methods expressed highly interconnected cell-to-cell junctions. The 3D spheroids were revealed to be an improved mini-tumor model as indicated by the TEM visualization of cell junctions and microvilli, features not seen in the 2D models. Growing 3D cell models using decellularized lung scaffold provided a platform for cell growth and infiltration for all cell types including primary cell lines. The most time-effective method was growing cells in spheroids using low-adhesive U-bottom plates. However, not every cell type grew into a 3D model using the the other methods of hanging drop or poly-HEMA. Cells grown in 3D showed more resistance to chemotherapeutic drugs, exhibiting reduced apoptosis. 3D cells stained with H&E showed cell-to-cell interactions and internal architecture that better represent that of in vivo patient tumors when compared to 2D cells. IHC staining revealed increased protein expression in 3D spheroids compared to 2D culture. Lastly, cells grown in 3D showed very different microRNA expression when compared to that of 2D counterparts. In conclusion, 3D cell models, regardless of which method is used. Showed a more realistic tumor microenvironment for architecture, gene expression and drug response, when compared to 2D cell models, and thus are superior preclinical cancer models.

No abstract