4102

MAGEA3

CT1.3|HIP8|HYPD|MAGE3|MAGEA6;melanoma antigen family A, 3;MAGEA3;melanoma antigen family A, 3

MAGE-3 antigen|antigen MZ2-D|cancer/testis antigen 1.3|cancer/testis antigen family 1, member 3|melanoma-associated antigen 3

Xq28

protein-coding

BACKGROUND: Neuroblastoma is the most common solid extracranial tumor in childhood, still with poor survival rates for metastatic disease. Neuroblastoma cells are of neuroectodermal origin and express a number of cancer germline (CG)antigens. These CG antigens may represent a potential target for immunotherapy such as peptide-based vaccination strategies. OBJECTIVE: The purpose of this study was to analyze the presence of MAGE-A1, MAGE-A3/A6, and NY-ESO-1 on an mRNA and protein level and to determine the expression of MHC class I and MHC class II antigens within the same tumor specimens. METHODS: A total of 68 tumors were available for RT-PCR, and 19/68 tumors were available for immunohistochemical (IHC) analysis of MAGE-A1, MAGE-A3/A6, and NY-ESO-1. In parallel, the same tumors were stained with a panel of antibodies for MHC class I and MHC class II molecules. RESULTS: Screening of 68 tumor specimens by RT-PCR revealed expression of MAGE-A1 in 44%, MAGE-A3/A6 in 21%, and NY-ESO-1 in 28% of cases. Immunohistochemistry for CG antigens of selected tumors showed good agreement between protein and gene expression. However, staining revealed a heterogeneous expression of CG antigens. None of the selected tumors showed MHC class I or MHCclass II expression. CONCLUSIONS: mRNA expression of MAGE-A1, MAGE-A3/A6, and NY-ESO-1 is congruent with the protein expression as determined by immunohistochemistry. The heterogeneous CG-antigen expression and the lack of MHC class I and II molecules may have implications for T-cell-mediated immunotherapyin neuroblastoma.

cancer-germline genes (CGGs) code for immunogenic antigens that are present in various human tumors and can be targeted by immunotherapy. Their expression has been studied in a wide range of human tumors in adults. We measured the expression of 12 CGGs in pediatric brain tumors, to identify targets for therapeutic cancer vaccines. Real Time PCR was used to quantify the expression of genes MAGE-A1, MAGE-A2, MAGE-A3, MAGE-A4, MAGE-A6, MAGE-A10, MAGE-A12, MAGE-C2, NY-ESO-1 and GAGE-1,2,8 in 50 pediatric brain tumors of different histological subtypes. Protein expression was examined with immunohistochemistry. Fifty-five percent of the medulloblastomas (n = 11), 86% of the ependymomas (n = 7), 40% ofthe choroid plexus tumors (n = 5) and 67% of astrocytic tumors (n = 27) expressed one or more CGGs. Immunohistochemical analysis confirmed qPCR results. With exception of a minority of tumors, the overall level of CGG expression in pediatric brain tumors was low. We observed a high expression of at least one CGG in 32% of the samples. CGG-encoded antigens are therefore suitable targets in a very selected group of pediatric patients with a brain tumor. Interestingly, glioblastomas from adult patients expressed CGGs more often and at significantlyhigher levels compared to pediatric glioblastomas. This observation is in line with the notion that pediatric and adult glioblastomas develop along different genetic pathways.

MAGE-1 and MAGE-3 or -6 are genes encoding melanoma-rejection antigens recognized by cytotoxic T lymphocytes in an HLA-A1 restriction manner. MAGE-1 and MAGE-3 or-6 were expressed in 5/14 (36%) and 6/14 (43%) neuroblastoma (NB) cell lines, and in 20/41 (49%) and 24/41 (59%) clinical NB-related tumors, respectively. Additionally, they were also expressed in pediatric tumors of other types such as rhabdomyosarcoma and Wilms' tumor. MAGE-1 expression at a functional level in tumor cells was confirmed by the cytotoxicity assay using MAGE-1-specific tumor-infiltrating lymphocytes (TIL). In clinical NB-related tumors, MAGE-3 or -6 expression demonstrated an inverse correlation to clinical stage. Furthermore, although the sample number was small, the incidence of MAGE-1 and/or MAGE-3 or -6 expression was significantly correlated to the absence of metastasis and a more favorable clinical outcome (p < 0.05). These results may suggest that NB cells silent for the expression of MAGE genes escape from the host anti-tumor immune response and consequently retain a growth advantage. Finally, NB-related tumors could be reliable candidates for immunotherapy targeted towards MAGE gene products.