P01.01 * EXPRESSION OF FIBROBLAST ACTIVATION PROTEIN ALPHA IN HUMAN HIGH-GRADE GLIOMAS

Abstract

Fibroblast activation protein alpha (FAP) is a dual specificity serine protease that is strongly associated with the development and progression of human carcinomas and represents a potential stromal target for anticancer treatment. Due to its gelatinolytic activity, FAP may participate in matrix-degradation, thus contributing to cell invasion. In addition, by its dipeptidyl peptidase enzymatic activity FAP was reported to efficiently hydrolyze several bioactive peptides relevant for cancer progression. However, relatively little is known about the expression and function of FAP in human astrocytic tumors. In our study, we detected higher FAP mRNA and protein expression in high-grade gliomas compared to controls (pharmacoresistant epilepsy). Using immunohistochemistry, we identified two different FAP+ cell subpopulations in gliomas. One comprised of dispersed cells intimately encircling the dysplastic vessels; these cells were negative for the glial and endothelia markers, but expressed the mesenchymal markers SMA, TEM-1 and the fibroblast antigen detected by the TE-7 antibody. Expression of these markers was maintained in FAP+ cells isolated and propagated in pericyte medium in vitro, further supporting their probable mesenchymal origin. The other FAP+ subpopulation was represented by GFAP+ parenchymal cells, part of which also expressed the transcription factor SOX-2, a marker of multipotent neural and glioma stem cells. In a mouse model, FAP was variably expressed by orthotopically implanted human glioma cells and was frequently induced in the host cells surrounding the tumors. To determine, whether FAP enzymatic activity may promote the malignant behavior of glioma cells, enzymatically active or mutant, enzymatically inactive human FAP were transfected into FAP negative glioma cells. Surprisingly, transgenic expression of wild type FAP decreased the migration and matrigel as well as collagen type I invasion of glioma cells. Moreover, transgenic wild type but not mutant, enzymatically inactive FAP negatively influenced adhesion and spreading on collagen I, but not fibronectin, which was associated with decreased focal adhesion formation and FAK phosphorylation. Nevertheless, tumor size was not affected by transgenic FAP in an orthotopic xenotransplantation mouse model. To summarize, FAP expression is increased in human glioblastomas due to its upregulation in the transformed glioma cells and/or stromal elements including perivascularly localized mesenchymal cells and possibly stromal glial cells. Introduction of FAP into glioma cells was not sufficient to promote their malignant behavior and tumor growth. Nevertheless, considering the selective expression of FAP in the glioblastoma microenvironment, FAP may represent a possible therapeutic target. Supported by the grants IGA12237-5/2011, PRVOUK-P27/LF1/1, UNCE204013, SVV260032/2014.