LAUSR

Intravital imaging can be performed in mice and has proven extremely useful in uncovering the role of immune cells in cancer metastasis, but requires specialized expertise and equipment. In the last decade, microfluidic models have been developed that replicate key aspects of the metastastic microenvironment, allowing for higher resolution imaging than in vivo studies, and for enhanced control over the biochemical and physical microenvironment of the assay at cellular length scales in 3D. This is particularly useful to studying metastasis, which is composed of several hidden steps such as intravasation or extravasation that can be rare and thus difficult to image in vivo; in contrast, this rate of occurrence can be more readily modulated in vitro. In this study, we use a microfluidic model to investigate the role of monocytes in cancer cell extravasation. In mouse studies, monocytes have recently been shown to have either pro- or anti-metastatic functions depending on their phenotype; however, the underlying mechanisms remain unclear, and in particular, whether the monocytes helped during extravasation or at later stages. We propose to investigate this in our microfluidic models, so as to find ways to therapeutically block immune-mediated extravasation as a means to reduce metastasis. A previously developed microfluidic assay was used to form a 3D microvascular network (MVN) that consists of 3 separate compartments, flanked by media channels connected to reservoirs. In the two lateral compartments, normal human lung fibroblasts were seeded at 5 million/ml, while in the central compartment GFP HUVECs were placed at 4 million/ml all in a 3D fibrin gel. After 4 days, the HUVECs self-assemble into a MVN, which forms lumens that open directly to the media channel. Therefore, cancer cells can be perfused within the MVN via the reservoirs, and their transendothelial migration is tracked by confocal imaging over time for up to 12 hours. Monocytes were perfused with cancer cells in the MVN to test whether the monocytes could affect cancer cell extravasation intra-luminally, as they would in the blood circulation. In addition, we sort the monocytes into classical or inflammatory monocytes and perfuse them separately with MDA-MB-231. Both monocytes and macrophages were derived from healthy donors blood. Monocytes were found to significantly decrease cancer cell extravasation. 11 hrs after cancer cells were perfused with monocytes inside the MVN, 19.8±6.1 % of MDA-MB-231 had extravasated vs. 53.5±11% when cancer cells were perfused alone (p=0.041, unpaired t-test). When monocytes were perfused alone, most of them had extravasated 4 days later (94±2%). Interestingly, 46.4±3.1 % stained positive for MRC-1, suggesting the monocytes had differentiated into M2-like macrophages following extravasation. In addition, MRC-1 positive macrophages were significantly closer to the vascular network than macrophages that did not express MRC-1 (37.1±7.1 vs. 65.9±9.7 µm, p =0.02, unpaired t-test). One day after perfusion, only 33% of classical monocytes had extravasated vs. 77% of inflammatory monocytes. Inflammatory monocytes migrated much faster (1.12±0.85 vs 0.33±0.20 µm.min, respectively, p =0.0005, unpaired t-test). Current work will reveal whether each subset has different effects on cancer cell extravasation. Citation Format: Alexandra Boussommier-Calleja, Roger Kamm. Role of monocytes in 3D microfluidic models of cancer cell extravasation. [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2016 Oct 20-23; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2017;5(3 Suppl):Abstract nr B22.