Poster S13 - Reverse Engineering the Lung Metastatic Niche by Co-Culturing Patient Circulating Tumor Cells with Stromal Cells

Introduction: Interactions between circulating tumor cells (CTCs) and stromal cells at metastatic sites mediate survival and proliferation, contributing towards successful “homing” towards particular organs. In particular, estrogen receptor positive (ER+) breast cancers often metastasize to liver and lung, while others metastasize to bone, each representing very different niches with distinct stromal cell populations. Here, we use live cell imaging and RNA-seq to elucidate interactions between patient CTCs and stromal cells cultured as multicellular spheroids in low adhesion microwells. We investigate how these co-culture conditions influence proliferation and self-sorting of these cell types, as well as perturb these interactions with targeted inhibitors. We hypothesized that lung fibroblasts will promote CTC sorting and proliferation through secreted soluble and matrix factors.

Materials and

Methods: Patient-derived breast cancer CTC lines were kindly provided by S. Maheswaran (Massachusetts General Hospital), isolated using microfluidic devices from blood draws of metastatic breast cancer patients with informed consent and IRB approval. CTCs were cultured in Ultra Low Adhesion (ULA) plates and maintained in serum-free media supplemented with epidermal growth factor (EGF) and basic fibroblast growth factor (FGF) under hypoxic conditions (5% O2). Primary Lung Fibroblasts (hTERT Fibroblasts immortalized, ATCC) were cultured following ATCC protocols
ATCC cell culturing protocol. CTCs and lung fibroblasts were then seeded at varying ratios and seeding densities in Corning Eplasia plates for spheroid formation under low adhesion conditions. All spheroids were imaged in widefield and live cell confocal microscopy for varying time points. Morphological, growth rates, and fluorescence expression were analyzed utilizing the FIJI Software (for area and diameter) and a custom MATLAB code (fluorescence). At the completion of the experiment, spheroid samples were collected and prepared for flow cytometry and RNA-seq to profile gene expression in co-culture relative to monoculture. Spheroid growth and sorting in selected conditions of interest were perturbed with targeted inhibitors relative to vehicle control.

Results, Conclusions, and Discussions: CTC-only remained loosely aggregated over 1 week in culture, retaining rounded single cell morphologies, but gradually proliferating to fill the entire microwell (Figure 1A). In comparison, mixtures of CTCs with lung fibroblasts immediately self-sort into a core of fibroblasts and an outer shell of CTCs within 24 h (Figure 1B,C,D). Notably, spheroids with higher starting percentages of CTCs (e.g. 90%, 50% Brx50) also increase in size over time due to CTC proliferation (Figure 1B,C). In comparison, lung fibroblast only spheroids were somewhat elongated and tightly compacted at 24 h (Figure 1E). These spheroids with higher starting percentages of LF (10%, 0% Brx50) appear to decrease in overall size due to fibroblast compaction. Quantitative analysis reveals that CTCs proliferate slightly faster in co-culture, particularly when cultured with larger percentages of lung fibroblasts. In comparison, the diameter of the lung fibroblast core tends to decrease due to spheroid compaction due to cellular contractility. Based on these findings, it is possible that reciprocal stimulation of lung fibroblasts by CTCs also occurs. RNA-seq of lung fibroblasts only in 2D vs 3D culture suggest potential interactions based on secreted growth factors as well as extracellular matrix deposition, which were verified by immunostaining and perturbation using targeted inhibitors.

We show that co-culture of Brx50 with lung fibroblast results in self-sorting and alterations in proliferation rate, depending on the initial seeding density and ratios. These functional changes can be attributed to secreted soluble factors and matrix deposition, and further affect drug response of CTCs. We envision that this ex vivo culture platform can be utilized to predict patient response to drugs in the lung metastatic niche as a step towards personalized therapies.

Acknowledgements (Optional): We thank S. Maheswaran (Massachusetts General Hospital) for the gift of CTC lines (BRx-50) and funding from NIGMS under P20GM109035.