(G-249) Validating in vitro models of breast to brain metastases across the blood-brain barrier

Introduction:: Metastatic breast cancer has no cure, and patients ultimately succumb to disease. Of metastatic breast cancer patients, up to 24% experience metastasis to brain [1]. To establish brain metastases, cancer cells must traverse the blood-brain barrier (BBB ), a layered structure primarily comprised of endothelial cells, pericytes, and astrocyte endfeet. Recent research indicates that crosstalk between native brain immune microglia and astrocytes may facilitate metastasis of cancer cells across the BBB [2]. Cellular crosstalk in metastasis can be studied using in vivo models, but they are expensive, time-consuming, and pose ethical challenges. Current in vitro models, such as 2D cultures or slice assays, either fail to recapitulate features of the brain tumor microenvironment (TME) or offer little engineering control of its features [3]. To address these limitations, we are developing brain-mimetic in vitro models to study the metastatic breast cancer niche. These models seek to elucidate the interactions between breast cancer cells, microglia, and astrocytes using cell spheroids and dispersed cells encapsulated in collagen-hyaluronic acid hydrogels. Collagen I comprises part of the vascular basement membrane that acts as a migration highway for cancer cells while hyaluronic acid is a primary component of the brain extracellular matrix (ECM) [4]. These models were previously validated and used to study glioblastoma brain cancers [5].


Materials and Methods:: Culture models consisted of breast cancer cell spheroids encapsulated with dispersed brain astrocytes and/or microglia or multicellular spheroids of cancer cells, astrocytes, and/or microglia. For breast cancer cell spheroids, metastatic breast cancer cells (MDA-MB-231) (2000 fluorescently labeled cells/well) were formed in 96 round-bottom well plates over 3 days. In some experiments, 24 µg/ml collagen I was also added after 1 day to promote spheroid formation. After 3 days, spheroids were encapsulated in collagen I (1 mg/ml) and hyaluronic acid (2 mg/ml) hydrogels using polyethylene glycol diacrylate crosslinker as previously described [5]. Fluorescently labeled human astrocytes (ScienCell) and/or microglia (HMC3) were also dispersed within the hydrogel solution at 500k cells/ml prior to gel addition. For multicellular spheroids, 2000 human astrocytes and/or microglia were seeded concurrently with breast cancer cells, prior to hydrogel encapsulation. Spheroid morphology, cancer cell migration, and cell-cell interactions were monitored using time-lapse fluorescent confocal microscopy over several days coupled with image analysis. These data were compared to histological analysis of breast-to-brain metastasis in mouse models. Models were further validated using gene expression and secretion assays that are being compared to patient databases.


Results, Conclusions, and Discussions:: MDA-MB-231 breast cancer cells are derived from a highly infiltrative metastatic lesion. Not surprisingly, breast cancer cells did not spontaneously form compact spheroids prior to gel addition (Fig. 1A). Interestingly, addition of collagen I or co-culture with astrocytes or microglia during spheroid formation led to significant spheroid compaction (Fig. 1 B,C). This may result from cell-cell signaling between the cancer cells and astrocytes/microglia or from the physical barrier provided by the native brain cells. In contrast, the addition of dispersed astrocytes or microglia to pre-formed breast cancer cell spheroids reduced compactness and increased cell spreading compared to breast cancer cell spheroids in the absence of native brain cells (Fig. 1D). To further assess the model fidelity, confocal reflectance and fluorescence microscopy were used to examine tumor morphology, cell distributions, and ECM components near the periphery of tumor metastases in mouse brain. In addition, secretion assays and gene expression arrays are being used to assess immune changes in native brain cells in the presence of metastatic breast cancer cells. These data are compared to previous results for brain cancer cells and against patient databases to further validate these models. Brain-mimetic 3D in vitro culture models offer a bridge between 2D cultures and in vivo mouse models in the study of metastatic breast cancer. Such models are ideal for elucidating cancer biology and are also suitable for high throughput drug screening and patient therapy selection.


Acknowledgements (Optional): :

References (Optional): : 1. Rostami, R., S. Mittal, P. Rostami, F. Tavassoli, and B. Jabbari, Brain metastasis in breast cancer: a comprehensive literature review. J Neurooncol, 2016. 127(3): p. 407-14.

2. Wasilewski, D., N. Priego, C. Fustero-Torre, and M. Valiente, Reactive Astrocytes in Brain Metastasis. Front Oncol, 2017. 7: p. 298.

3. Narkhede, A.A., A.A. Narkhede, L.A. Shevde, and S.S. Rao, Biomimetic strategies to recapitulate organ specific microenvironments for studying breast cancer metastasis Biomimetic strategies to study breast cancer metastasis. International journal of cancer, 2017. 141(6): p. 1091-1109.

4. Rao, S.S., J. DeJesus, A.R. Short, J.J. Otero, A. Sarkar, and J.O. Winter, Glioblastoma Behaviors in Three-Dimensional Collagen-Hyaluronan Composite Hydrogels. ACS Applied Materials & Interfaces, 2013. 5(19): p. 9276-9284.

5. Y. Cui, P. Lee, J.J. Reardon, A. Wang, S. Lynch, J.J. Otero, G. Sizemore, and J.O. Winter. Evaluating glioblastoma tumour sphere growth and migration in interaction with astrocytes using 3D collagen-hyaluronic acid hydrogels. Journal of Material Chemistry B. under revision