GD2, a disialoganglioside, is overexpressed on melanomas and neuroblastomas and has been defined being a relevant tumor antigen. The anti-GD2 antibody 14.18 is used for diagnosis of neuroblastoma, and its mouse/human chimeric form ch14.18 has already entered passive immunotherapeutic regimens in neuroblastoma phase II clinical trials. For developing an active immunotherapy, we defined peptide mimics of the 14.18 epitope on GD2. By biopanning a random peptide phage display library with the antibody, thirteen GD2 mimotope candidates could be identified. They were proven to be true epitope mimics by mimicry tests in the ELISA format. Using three-dimensional computer modeling (BALLView), we were able to demonstrate that two selected mimotopes fitted into the antigen binding pouch of a GD2 specific antibody in the same manner as original antigen GD2. We conclude that we succeeded in generating peptide mimics of the carbohydrate antigen GD2. In contrast to GD2 which is a poor immunogen, these peptides might be utilized to develop an active immunization against overexpressing GD2 tumors. Supported by BioLife Science GmbH, the Austrian National Bank grant #10965 and the H&B Moser fund.

[Proc Amer Assoc Cancer Res, Volume 46, 2005]