Cell surface antigens of transformed cells are the traditional targets for antibody-guided detection and therapy of solid neoplasms. Alternative targets may be found in the tumor stroma, which contains newly formed blood vessels, reactive fibroblasts, and extracellular matrix proteins. The F19 cell surface glycoprotein of reactive fibroblasts is a prototypic stromal antigen since it is found in the stroma of >90% of common epithelial cancers but is absent or expressed at low levels in normal tissues and benign epithelial tumors. In the present study, we define an additional tumor stromal antigen, FB5, that is selectively expressed in vascular endothelial cells of malignant tumors. Immunohistochemical analysis of 128 tumors identified FB5 in endothelial cells in 67% of the samples (including 41 of 61 sarcomas, 26 of 37 carcinomas, and 18 of 25 neuroectodermal tumors) whereas normal blood vessels and other adult tissues tested lacked FB5 expression. In vitro studies showed that FB5 is a Mr 165,000 cell surface glycoprotein, comprised of a Mr 95,000 core polypeptide and highly sialylated 0-linked oligosaccharides but few if any N-linked sugars, and that the FB5 gene is located on chromosome llql3-qter. The restricted tissue distribution of the FB5 protein, which we refer to as endosialin, suggests strategies for tumor imaging and therapy that are aimed primarily at the tumor vasculature.

No abstract

The human fibroblast activation protein (FAP) defined by monoclonal antibody (MAb) F19 is a cell surface antigen expressed in reactive stromal fibroblasts of breast, colorectal, lung and other epithelial cancers. In contrast to its stroma-specific localization in epithelial neoplasms, FAP is expressed in the malignant mesenchymal cells of bone and soft tissue sarcomas. FAP is transiently expressed in some fetal mesenchymal tissues but is absent or expressed at low levels in most adult tissues. FAP is induced in cultured fibroblasts and, in these cells, consists of a M(r) 95,000 subunit (FAP alpha) carrying the F19 epitope and a non-covalently bound M(r) 105,000 subunit (FAP beta) lacking the F19 epitope. Using MAb F19 and 5 newly derived MAbs, we identify 3 distinct epitopes on FAP alpha and tentatively assign one epitope to FAP beta. Analysis of detergent extracts of a FAP alpha high beta- sarcoma cell line by size exclusion-high performance liquid chromatography (HPLC) revealed that FAP alpha does not elute as a M(r) 95,000 species but as part of a high-molecular weight complex (M(r) > 400,000) that dissociates into M(r) 95,000 subunits in SDS gels. Immunoaffinity purification of FAP alpha followed by tryptic digestion, reversed-phase HPLC and microsequencing identified 3 unique FAP alpha peptides, with 2 showing sequence similarity (23/38 identical amino acids) to segments of CD26, a T-cell activation antigen. CD26 is a membrane-bound enzyme (dipeptidyl aminopeptidase IV), but immunopurified FAP alpha has little if any dipeptidase activity with typical CD26 substrates. Finally, studies with FAPlow leptomeningeal fibroblasts revealed that transforming growth factor-beta, 12-O-tetradecanoyl phorbol-13-acetate and retinoids can upregulate FAP expression, whereas serum and several other factors had no or little effect on FAP levels. FAP and CD26 may belong to a family of structurally related but functionally distinct activation proteins that are expressed on different cell types and show unique modes of regulation in normal and malignant cells.

PSMA is a widely used marker for prostate epithelial cells. Its up-regulation in association with cancer, particularly in advanced cancer, is ideal for application as a prognostic marker. A variety of promising clinical applications utilizing PSMA have been or are being developed. In the future, these promise to have an important impact on cancer diagnosis and patient treatment.

No abstract