| |
[摘要]：We present a new multidisciplinary strategy integrating computational biology with high-throughput microarray analysis aimed to translate molecular understanding of protein-antibody recognition into the design of efficient and selective protein-based analytical and diagnostic tools. The structures of two proteins with different folds and secondary structure contents, namely, the beta-barrel FABP and the alpha-helical S100B, were used as the basis for the prediction and design of potential antibody-binding epitopes using the recently developed MLCE computational method. Starting from the idea that the structure, dynamics, and stability of a protein-antigen play a key role in the interaction with antibodies, MLCE integrates the analysis of the dynamical and energetic properties of proteins to identify nonoptimized, low-intensity energetic interaction-networks on the surface of the isolated antigens, which correspond to substructures that can aptly be recognized by a binding partner. The identified epitopes were next synthesized as free peptides and used to elicit specific antibodies in rabbits. Importantly, the resulting antibodies were proven to specifically and selectively recognize the original, full-length proteins in microarray-based tests. Competition experiments further demonstrated the specificity of the molecular recognition between the target immobilized proteins and the generated antibodies. Our integrated computational and microarray-based results demonstrate the possibility to rationally discover and design synthetic epitopes able to elicit antibodies specific for full-length proteins starting only from three-dimensional structural information on the target. We discuss implications for diagnosis and vaccine development purposes. |
|
