Differential endostatin binding to bladder, prostate and kidney tumour vessels

OBJECTIVES To define the anti‐angiogenic mechanism and causes of the heterogeneous influence of endostatin, one of a group of matrix‐derived inhibitors of tumour angiogenesis of increasing significance in tumour treatment, on various tissue types. MATERIALS AND METHODS Variations in the binding behaviour of endostatin with vessels were assessed in different tumours (bladder, prostate and kidney) and compared with benign tissue vessels. Biotinylated endostatin was used and detected using extravidin CY3 and extravidin‐gold immunolabelling. RESULTS There were significant differences in the number of vessels showing endostatin binding among benign and malignant bladder, prostate and kidney tissues. While there was distinct endostatin binding on a mean ( sd ) of 94.2 (3.0)% bladder and 73.8 (19.5)% prostate tumour vessels, there was less binding, at 11.32 (3.9)%, on kidney tumour vessels. There was less binding to vessels of benign bladder, prostate and kidney tissue, at 2.0 (1.5), 1.7 (1.7) and 1.5 (1.7)%, respectively. At the ultrastructural level, different binding sites were detected both inside and outside the endothelial cells. CONCLUSION Endostatin binds more to all tumour tissues than to benign tissue, but the degree of binding in malignant kidney tissue was significantly less than that in malignant prostate and bladder tissues. These divergent vascular endostatin‐binding patterns could be responsible for a tumour type‐dependent anti‐angiogenic effect attributable to endostatin. Such selective behaviour would have therapeutic consequences for future anti‐angiogenic therapy, in which different kinds of tumours could be further classified into those responding to endostatin or not.