Tumor microenvironment is now getting much attention as antitumor strategy. Tumor cells interact with many other surrounding cells such as immune, endothelial, and fibroblast-like stromal cells. Among them the stromal cells can modulate tumor growth and metastasis positively and negatively through direct adhesion and secreted factors. We have been studying such tumor microenviroment, especially focusing on stromal cells, to innovate a new antitumor strategy. We previously developed an in vitro coculture system of human prostate cancer cells and prostate stromal cells, and found a novel antitumor compound NBRI16716A from a fungus metabolite. Our preliminary experiments suggested that NBRI16716A modulates tumor-stromal cell interactions, we therefore have studied the action of NBRI16716A using new coculture systems of various types of cancer. We established green fluorescence protein-expressing cells of lung, gastric, colon, pancreatic, breast, and prostate cancer cell lines, and developed various coculture systems using each organ-derived stromal cells. Using these coculture systems, we have found that NBRI16716A inhibits the growth of A549 lung cancer cells in coculture more strongly than cancer cells alone as well as the coculture system of DU-145 prostate cancer cells. The conditioned medium prepared from the stromal cells pretreated with NBRI16716A also showed stronger growth inhibitory activity suggesting possible involvement of some secreted factors from the stromal cells. We are now trying to identify the secreted factors from the stromal cells. Collaborators: T Kato, Y Aritoku, MicroBiopharma Japan.