Filopodia are actin-filled projections that can have diameters of 0.2?m, extend up to three cell diameters in length [76] and expand into pseudopods [77]. and growth [54C56]. Second, interactions between cancer and endothelial cells play a major role in metastasis [57]. Metastatic cells shed from tumors migrate to and intravasate blood vessels, a process facilitated by macrophages [58]. Intravasated cancer cells are then swept by blood flow to specific anatomical locations, where they extravasate through the vessel walls, entering body locations to form secondary tumors [5,59]. While there BX-912 has been intense investigation of the regulatory, cell surface and signal molecules involved in adhesion during both the vascularization of tumors [60C64] and cancer cell intravasation of blood and lymph vessels [65C70], the dynamic behavior of interacting cancer and endothelial cells in a 3D matrix has received much less attention. This deficit is in part due to the underutilization of computer-assisted 3D reconstruction systems of live cells over time. We have, therefore, begun to develop transparent 3D preparations, and LSCM and BX-912 computer-assisted reconstruction techniques, that facilitate investigations of cellular behaviors basic to tumorigenesis and metastasis. We first applied these methods to analyze cancer cells in a transparent 3D Matrigel environment in the absence of other cell types, and identified specialized behaviors and unique cell types which mediated aggregate coalescence preceding spherule formation [26,28]. We also demonstrated using this model that a minority of cancer cells can recruit nontumorigenic cells into the tumor-like aggregates, a possible explanation for the cellular heterogeneity of tumors [27]. The model was then used to test the activity of 266 mAbs primarily against cell surface molecules, for their ability to block aggregation in a 3D Matrigel matrix [31]. Of the 266 tested mAbs, only those against the two components of integrin 31 and against CD44 exhibited blocking activity [31]. Recently, the model was employed to assess interactions between breast cancer cells and fibroblasts [29]. The results revealed both reciprocal signaling and direct physical interactions [29]. Here, we have IL4 used this general model to assess the behavioral interactions between breast cancer cells and multicellular reticulated networks of human umbilical vein endothelial cells (HUVECs). In contrast to our previous studies, we added laser scanning confocal microscopy, which allowed us to import optical sections of differentially colored MB-231 cells and HUVECs through multiple channels. These images could then be used to generate LSCM 3D images of live cells over time, or by bitmap algorithms and edge detection, 3D reconstructions of the HUVEC network and MB-231 cells over time. The latter J3D-DIAS reconstructions could be used to quantitate behavior. MB-231 cell behavior in the 3D model Mammary tumor-derived MB-231 cells dispersed in a 3D Matrigel environment in the absence of HUVECs were motile, but nondirectional BX-912 and did not coalesce into large aggregates during the initial 72?hours of analysis [26]. When in the vicinity of a HUVEC network, however, MB-231 cells moved in a directed fashion toward, and attached to, the reticulated multicellular HUVEC network within 8?hours of incubation. Directional movement toward the network proceeded within a distance of approximately 60?m from the network, a distance of roughly three or more cell diameters. The Matrigel region in this zone became relatively devoid of MB-231 cells as they translocated toward and attached to the static HUVEC network. These behavioral characteristics differed markedly from those of MB-231 cells that are dispersed above a fibroblast monolayer [29]. In contrast to the stable network of HUVECs, the fibroblasts moved up from.
