The design of a three-dimensional spheroidal model of tumor angiogenesis and its implementation in anti-angiogenesis studies

Angiogenesis is critical for cancer development and progression. Anti-angiogenic agents have been shown to have clinical activity in several tumor types, including melanoma. Unfortunately these therapies are effective only in a minority of patients and are toxic and expensive. Development of a predictive technology that could identify those patients most likely to respond would decrease the risk to benefit ratio. We constructed a three-dimensional model of tumor angiogenesis and used the model to demonstrate (1) that a subset of cancer cells expressing thrombospondin-1 (TSP1) could sensitize vascular endothelial cells (VECs) to anti-angiogenic therapy, (2) that VECs sensitive to anti-angiogenic therapy also have inherent properties, including up-regulation of TSP1, and (3) the importance of using a model of angiogenesis which incorporates a tumor compartment.;TSP1 is a natural inhibitor of angiogenesis. We hypothesized that VEC diapedesis would be blunted and that VECs would be more sensitive to anti-angiogenic agents when VECs were co-cultured with tumor spheroids producing high levels of TSP1 versus low levels of TSP1. We found high TSP1 to be related to decreased VEC survival and increased sensitivity to anti-angiogenic agents in our in vitro model.;Resveratrol, a natural polyphenol, has been shown to have both a cardio-protective and anti-angiogenic effect. We showed that while resveratrol had a protective/proliferative effect on endothelial cells grown by themselves in monolayer, it promoted endothelial cell apoptosis when endothelial cells were grown with cancer cells in our co-culture model. These phenomena are reminiscent of what is seen in vivo and demonstrate the importance of using a model of tumor angiogenesis that incorporates a tumor compartment.