Isolation of tumor cells with MBs and magnetic beads.

<p>Tumor cells were spiked into 1 ml of plasma-depleted mouse blood (see <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0058017#s2" target="_blank">Methods</a>). Mouse EpCAM-targeted MBs and beads were added at 1×10⁷ MB/ml. Following gentle mixing for 15 minutes, MB layer was separated from blood cells by centrifugation at 100 g for 2 minutes. Magnetic beads were separated with a neodymium magnet. <b><i>A</i></b>, Collection of the floating MB layer on top depends on the downstream analysis (see <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0058017#s2" target="_blank">Methods</a>); <b><i>B</i></b>, After the isolation of MBs and beads, the cells were placed on a hemocytometer grid (MBs were destroyed for 1 second with gentle sonication) and the GFP-positive cells were counted at low magnification (40×). MBs and beads showed similar numbers of GFP+ cells in the isolated (enriched) fraction and near absence of GFP+ cells in the blood cell (depleted) fractions. For size reference, red outline shows the major 5×5 square of the hemocytometer; <b><i>C</i></b>, Quantification of isolation efficiency of GFP-4T1 cells. <b><i>D</i></b>, Quantification of tumor cell depletion from blood cells with flow cytometry. One ml of blood cells was spiked with non-labeled 4T1 cells at 1×10⁶ cells/ml. Remaining blood cell fraction and the isolated MB fraction were analyzed for tumor cells after CD45 (leukocyte) and EpCAM staining (as described in <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0058017#s2" target="_blank">Methods</a>). Red blood cells were lysed and gated out (left image, FSC/SSC plot). MBs efficiently depleted 4T1 cells, and enriched them with 95.4% purity. A representative experiment out of two is shown.