B16 mouse melanoma sublines in culture spontaneously shed intact plasma membrane vesicles. These vesicles can be fused with the plasma membrane of cells from homologous and heterologous B16 sublines by using polyethylene gycol and phytohemagglutinin-P. Fusion of vesicles from a highly metastatic subline (FlO) that localizes exclusively in the lung with cells from a poorly metastatic subline (Fl) The arrest of circulating tumor cells in the microcirculation is an essential step in the formation of hematogenous metastases (1-3). Little is presently known about the mechanisms involved in this process or the factors that influence blood-borne tumor cell arrest in different regions of the microcirculation. Clinical and experimental observations indicate that in many metastatic tumors the distribution and arrest patterns of circulating tumor cells are nonrandom (1-3), and malignant cells may localize preferentially in particular organs (4-11). A striking illustration of this phenomenon is provided by experiments showing that tumor cell lines that ordinarily metastasize to the lung also stop in pieces of lung, but not other organs, implanted at ectopic sites (9, 12; I. R. Hart, personal communication). The importance of tumor cell properties in determining cell arrest patterns in the microcirculation is also suggested by recent work in which a series of B16 mouse melanoma cell variant lines that metastasize preferentially to lung (13), liver (14), ovary (15), and brain (16) have been isolated from the same parent B16 line.In this paper we show that differences in the ability of B16 melanoma sublines to become arrested in the lung microcirculation are determined in part by the surface properties of these cells. By transferring plasma membrane components from a highly metastatic B16 subline (FlO) that localizes exclusively in the lung to another B16 subline (Fl), which ordinarily forms few metastases in the lung or other organs, we have been able to significantly increase the ability of F1 cells to become arrested in the lung and form pulmonary metastases. (FlLr, F1OLr) selected from these two sublines have been described in detail (13,18). All cells were grown in Eagle's minimal essential medium plus 10% fetal calf serum as described and were found to be uncontaminated by mycoplasma species by several assays (19)(20)(21)(22). Plasma membrane vesicles were harvested from the culture medium of actively growing roller bottle cultures of the various B16 sublines and purified as described (17). In certain experiments, vesicle surface proteins were labeled with 1251 or fluorescein isothiocyanate by using methods described previously (23, 24). Protein content was measured by the method of Lowry et al. (25).Vesicle-Cell Interactions. Purified vesicle preparations containing 30040 g of protein were incubated in suspension with 1 X 107 melanoma cells in minimal essential medium containing phytohemagglutinin-P (PHA; Difco) at 50 ,g/ml and 40% (vol/vol) polyethylene glycol, molecular weight 6000 (PEG) (PHA/PEG) for 5 mi...