PRIMING IN THE BRAIN, AN IMMUNOLOGICALLY PRIVILEGED ORGAN, ELICITS ANTI-TUMOR IMMUNITY Hassan M. FATHALLAH-SHAYKH *, Wei GAO, Michael CHO and Maria Alejandra HERRERA Department of Neurology, University of Texas Southwestern Medical Center at Dallas, Dallas, TX, USA A crucial question in the study of tumor neuro-immunology concerns the capacity of the central nervous system to initiate and execute an immune response. In a 100%fatal rat malignant glioma model, genetically modified tumors secret- ing INF- intracerebrally generate an immune response resulting in a substantial increase in survival time, tumor rejection and specific systemic immunity. Tumors modified to secrete IL-2 alone do not change the biologic behavior of transfected gliomas. IN F- induces elevated expression of major-histocompatibility-complex-class-I and -class-II mol- ecules in microglia throughout the brain and invokes en- hanced tumor infiltration by CD4, CD8 and NK cells. These findings demonstrate successful immunization against a cen- tral-nervous-system tumor by direct priming in the brain with a live growth-competent tumor vaccine. Int. J. Cancer 75:266–276, 1998.  1998 Wiley-Liss, Inc. Malignant brain tumors progress rapidly and are almost invari- ably fatal. Over the past decade, despite aggressive combined- modality treatments, little improvement has been made in the prognosis and survival of patients suffering from malignant glial neoplasms. The location of tumors in the brain, an organ believed to be ‘‘immunologically privileged,’’ is thought to help them evade immune surveillance. The classical idea that the mammalian brain is an immunologi- cally privileged site was initially based on observations that: (i) allografts of carcinomas are more successful in the brain than in the sub-cutaneous space (Barker and Billingham, 1971); (ii) a wide variety of tissues and organs from embryonic donors grow successfully in the brain (Barker and Billingham, 1971); (iii) the brain lacks defined lymphatic drainage (Yoffey and Courtice, 1970); (iv) the expression of major-histocompatibility-com- plex(MHC)-class I and -II molecules in the brain is low (Daar et al., 1984); and (v) only activated T lymphocytes cross the blood-brain barrier (Hickey et al., 1991). These facts are consistent with the concept that the privileged status of the brain is related to an interruption in the afferent pathway of the immune response (Barker and Billingham, 1971; Yoffey and Courtice, 1970). In the study of systemic tumor immunology, a 3-stage process is required for the generation of effector T cells after peripheral vaccination: (i) tumor antigen uptake and processing at the site of injection by professional antigen-presenting cells (APC); (ii) migration of APC into regional draining lymph nodes, where T-cell priming occurs; (iii) circulation of activated T cells that either perform or initiate effector mechanisms leading to tumor-cell destruction (Maass et al., 1995). Although, when transplanted to a naive brain, skin homografts survive, when transplanted to specifi- cally immunized animals they break down. This disparity of growth suggests that the immune privileges of the brain are related to the absence of typical lymphatic drainage (Medawar, 1948). However, the privileges of the brain are far from complete, since, after irradiation of established experimental gliomas, treated ani- mals reject 37% of second intracerebral implants and 43% of s.c. implants (Scheinberg et al., 1965). Typical lymphatic vessels do not exist in the brain, and the major outflow pathway for extracellular fluid is passage through the arachnoid villi into blood of the dural sinus (Weller et al., 1992; Yamada et al., 1991). Nevertheless, following injection of radiola- belled protein into the CNS, a significant fraction (14–47%) of this tracer leaving the CNS appears in the cervical lymph nodes (Weller et al., 1992; Yamada et al., 1991). Moreover, the injection of sheep erythrocytes into the lateral ventricles of rabbits (Jankovic et al., 1961) and the micro-infusion of human BSA into rat brains lead to the production of specific antibodies in serum. The latter response was reduced about 10-fold by cervical lymph obstruction (Harling- Berg et al., 1989). Considerable gaps in our understanding of tumor neuro- immunology relate to the identity of the brain-resident cell able to present antigens to T cells, and the site of antigen presentation. Recently, microglia have emerged as potential candidates for the role of APC of the CNS (Cash and Rot, 1994). When resting microglia are activated by treatment with interferon gamma (INF-), they become fully competent to present antigens to lymphocytes in an appropriately-restricted manner and acquire the capacity to activate an immune response in vitro (Hickey and Kimura, 1988). However, the site where microglia could present antigens to T cells in vivo has not been identified. It is not known whether antigen presentation takes place at a peripheral lymphatic organ, in an intravascular compartment or within the CNS (Hart and Fabry, 1995). In the study of peripheral immunology, both interleukin 2 (IL-2) and INF- favor the development of Th1 cells. However, the intracerebral co-transplantation into syngeneic animals of fibro- blasts or of endothelial cells transduced with viral vectors to secrete IL-2 and/or INF- with murine glioma cells does not elicit tumor rejection (Lichtor et al., 1995). Whether ‘‘high’’amounts of INF- and IL-2, secreted in the brain by the tumor itself, could induce tumor rejection and protective immunity remains unknown. We hypothesized that, given appropriate stimulus, the brain is capable of mounting an anti-tumor immune response. In this study, experiments were designed to investigate the biologic effects of priming directly into the brain by implanting genetically modified malignant glial cells secreting ‘‘high’’quanti- ties of INF- and IL-2. Responses were measured by: (i) survival, (ii) rejection and (iii) protective immunity against challenge by the parental line. This investigation led to successful immunization against a CNS tumor by direct priming in the brain with a live growth-competent tumor vaccine. MATERIALS AND METHODS Plasmid construction Human IL-2 cDNA, a generous gift from Dr. B. Cullen, (Duke University, Durham, NC), was removed from the plasmid BC12/ CMV/IL-2 (Cullen et al., 1988). Murine INF- cDNA was obtained from the plasmid pms 10 (ATCC, Rockville, MD). Sub-cloning was carried out by standard recombinant DNA tech- niques. The backbone of pGINF/IL2 was derived from pREP9 (InVitrogen, San Diego, CA) after digestion with SalI. The Contract grant sponsor: American Cancer Society; Contract grant number: 1RG-142L. *Correspondence to: Department of Neurology, University of Texas Southwestern Medical Center, Dallas, TX 75235-9036, USA. e-mail: fathalla@utsw.swmed.edu Received 9 July 1997; Revised 9 September 1997 Int. J. Cancer: 75, 266–276 (1998)  1998 Wiley-Liss, Inc. Publication of the International Union Against Cancer Publication de l’Union Internationale Contre le Cancer