Abstract Background: The tumor microenvironment (TME) is a 3D, dynamic interaction between tumor and stromal cells, extracellular matrix, and soluble factors which can promote tumor progression and resistance to therapy. Significant efforts are being made to develop long-term in vitro TME models to study cancer biology, improve drug development, and guide clinical decisions. We have developed a cost-effective micro-bioreactor culture device which enables a wide variety of dynamic cellular and TME interactions that can be monitored with in situ, label free, and non-lytic microscopic analysis. Using the 3DKUBE™, we developed a 3D breast cancer model composed of MCF-7 cells and fibroblasts in segregated co-culture and identified a highly active PI3K inhibitor that had limited activity in 2D monoculture. Based on a more complex morphologic and functional 3D mammary gland model that incorporates multiple human primary cells including epithelial cells (HuMEC), mammary fibroblasts, and adipocytes, our long-term 3D perfused tri-culture breast cancer model is designed to evaluate drug response continuously using non-lytic analysis. We theorize that different breast cancer epithelial cells will be supported in a standard 3D stromal platform which supports mammary gland function. Herein, we describe the 3DKUBE™ modifications and early optimization of a defined stromal TME which supports HuMECs in 3D perfusion. Materials & Methods: Primary human cells were used to optimize seeding conditions for silk fibroin scaffolds containing Matrigel™/collagen hydrogels, including, media composition, cellular ratios, flow rates, and analytical methods for the 3DKUBE™. Destructive analytical methods included morphology (H&E), phenotype (histology), viability & survival (CyQuant, TUNEL, Ki-67), and ER, PR, and casein expression (RT-PCR). Non-lytic assays included metabolism (resazurin reduction), cytotoxicity (LDH release), soluble biomarker expression (xMAP™), and two-photon emission fluorescence via multiphoton microscopy (MPM). Results: The 3DKUBE™ imaging window (1.25 mm thick with a refractive index of 1.590, transmittance (~90%) at wavelengths >290 nm) was replaced with a No. 1.5 glass, high resolution window allowing non-destructive, in situ microscopic imaging. MPM images obtained provided cellular detail (granularity, nuclear-cytoplasmic ratio, size). Optimization of the media conditions and cell numbers revealed that a tri-culture ratio of 1:1:1 was stable in the silk-Matrigel-collagen scaffold during prolonged static and perfusion culture in mixed media. Differentiation of adipocytes had to be performed separately from the tri-culture due to the cytotoxicity of the differentiation media to the epithelial cells. This was achieved via a serial seeding and differentiation method. Our initial drug studies revealed that cytotoxicity could be monitored over the course of a week by multiple methods, including LDH secretion, Presto blue, and CyQuant. Though non-specific drug binding (CPT-11) was seen in the complex 3D matrix, it did not materially impact total drug concentrations. Conclusions: Our current optimization results indicate our ability to grow a bioengineered 3D human breast tri-culture model under perfusion using the 3DKUBE™ while monitoring tissue structure, function, and drug response through both lytic and non-lytic methods. Future Development: The optimized perfused stroma will be used to develop ex vivo breast microtumors of clinically relevant subtypes (ER+, Her+, triple negative) validated against targeted agents. If validated, these 3D microtumors may be applied as predictive phenotypic screens to improve preclinical and clinical breast cancer drug development. Citation Format: Teresa M. DesRochers, Stephen Shuford, Terri F. Bruce, Matthew R. Gevaert, Chaitra Cheluvaraju, Irene Georgakoudi, David L. Kaplan, Davie E. Orr, Howland E. Crosswell. Development of a perfusion-based 3D human tri-culture breast microtumor. [abstract]. In: Abstracts: AACR Special Conference on Cellular Heterogeneity in the Tumor Microenvironment; 2014 Feb 26-Mar 1; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(1 Suppl):Abstract nr A40. doi:10.1158/1538-7445.CHTME14-A40
Read full abstract