DO GLIOMA-CELLS EXPRESS CYTOKERATIN

Abstract

There have been several recent reports of cytokeratin immunoreactivity in glial cells and tumors. We wanted to further test these tissues for cytokeratin immunoreactivity, and to determine whether antibody positivity corresponded to true cytokeratin expression. In the first set of experiments, a series of 10 formalin-fixed, frozen sections of glial tissue were employed; positive immunostaining with a cocktail of monoclonal anti-cytokeratin antibodies was seen only when a pepsin predigestion step was included in the immunostaining procedure. In the second set of experiments, 30 cases of malignant glioma fixed in both methacarn and formalin fixation were employed. Using a panel of three different anti-cytokeratin monoclonal antibodies (35 beta H11, 34 beta E12, CAM5.2), no positive immunostaining was observed in any of the methacarn-fixed tissues; positive immunostaining in the corresponding formalin-fixed tissue was frequently found, but only using the antibodies (35 beta H11, 34 beta E12) requiring enzyme predigestion. In the third set of experiments, immunoblots were performed on cytoskeletal extracts of human gliomas; no bands corresponding to known cytokeratins were observed in any cases. It is concluded that glial tissues and tumors do not, in fact, truly express cytokeratins, despite the fact that it is possible to obtain positive immunostaining of glial tumors and tissues using certain anti-cytokeratin antibodies under certain laboratory conditions.