Patterns of antigenic expression of human glioma cells.

Gliomas are known to express over a hundred antigens, and no doubt make many more unknown antigens. Major categories of glioma cell antigens include glial antigens, ECM antigens, muscle antigens, melanoma antigens, "tumor-specific" antigens, and cellular proliferation antigens. A strikingly low number of cultured gliomas express glial antigens. They commonly express not only ectodermal, but also mesenchymal ECM antigens. Tumor-specific antigens have been an elusive goal of neuro-oncologists, but there are bright new prospects in need of further study. These include direct screening of hybridoma supernatants on glioma tissue and targeting glycolipids, glycoproteins, and oncogene products. Cellular proliferation antigens will become increasingly important in predicting prognosis of gliomas. Proliferation antigens of cultured gliomas are under intense scrutiny at present. The extent and evolution of antigenic heterogeneity of neoplastic cells in gliomas raise basic biologic questions with profound clinical ramifications. Individual glioma cell lines may generate more than 30 subtypes of cells with minor to major differences in antigen expression. These include expression of antigens representing multiple different cell lineages. Mesenchymal drift is the tendency of gliomas to progressively lose glial and gain mesenchymal features. Models of in vivo mesenchymal drift occur in glioma cell culture where mechanisms are more easily investigated than in situ. Neither exogenous protein absorption nor fibroblast overgrowth explain the phenomenon. Cells with the mesenchymal marker, fibronectin, overgrow GFAP-positive cells during explanation of gliomas. Many of these fibronectin-positive cells express cytologic and growth characteristics of neoplasia. The source of these cells is unknown. A leading candidate for the source of these neoplastic fibronectin-positive cells is the proliferation of vascular and mesenchymal cell elements of glioma tissue commonly called "endothelial proliferations". However, these elements in tissue do not display the same abnormalities of neoplasia as the fibronectin-positive cells in culture. Understanding this "tissue/explant paradox" may solve the conundrum of mesenchymal drift. In the absence of a counterpart in tissue of these neoplastic fibronectin-positive cells so abundant in glioma cell cultures, mechanisms of mesenchymal drift other than overgrowth of neoplastic mesenchyme must be considered. The occurrence of "dual cells" which express antigenic markers of entirely different cellular lineages suggests the possibility that neoplastic glia generate mesenchymal drift by altered gene expression. Various studies which suggest the capacity of cultured gliomas to alter phenotypic expression of their genes are critically examined and their relevance to mesenchymal drift discussed.(ABSTRACT TRUNCATED AT 400 WORDS)