The most promising approach in Alzheimer disease immunotherapy is represented by amyloid beta derivatives with low intrinsic neurotoxicity and minimal overall T cell responses. To avoid toxicity and autoimmune response, we have designed a new class of Abeta derivatives through segmentation of the original Abeta[1-42] peptide and application of the glycine substitution modification technology. Abeta[1-16], Abeta[13-28] and Abeta[25-42] fragments were selected in order to retain the major immunogenic sites of the Abeta[1-42] peptide. All peptides showed comparable immunogenicity, and raised antibodies were all able to cross-recognize both Abeta[1-42] and Abeta[1-40] synthetic amyloid forms. Polyclonal antibodies produced against the simplified variants were able to recognize the parent peptide, but not the opposite simplified forms, in strict agreement with the model of independent surfaces of recognition. All Abeta simplified derivatives showed reduced fibrillogenic properties, thus underlining that the introduction of glycine residues in alternating positions allows to obtain modified peptides maintaining the main immunogenic properties of the parent peptides, but with reduced ability to adopt a beta-sheet conformation and therefore a much lower risk of toxicity in humans. In addition, in vitro studies on peripheral blood mononuclear cells (PBMCs) from healthy donors showed that only the Abeta[13-28]+G peptide failed to induce IFN-gamma production, thus suggesting that this molecule could represent a good candidate for potentially safer vaccine therapy to reduce amyloid burden in Alzheimer's disease instead of using toxic Abeta[1-42].