| 摘要 |
The present invention relates to computational methods for identifying the bioactive conformations of peptide domains, in particular the geometries of complexes of neurotrophins and neurotrophin receptors, and the geometries of neurotrophin receptors and ligands. The invention includes a method for identifying and theoretically modelling a receptor binding site for neurotrophins, such as NGF, BDNF, NT-3 and NT4/5, of the common neurotrophin receptor p75NTR. The principal residues of the p75NTR binding site are Asp47p, Lys56p, Asp75p, Asp76p, Asp88p and Glu88p of the second and third cysteine-rich domains. These residues interact with residues of variable loop regions I and V and other neighboring residues of each of the neurotrophins. The invention provides a method of designing a ligand for binding with common neurotrophin receptor p75NTR including computationally evolving a ligand having effective moieties located relative to each other in the ligand so that the moieties bind to at least two of p75NTR binding loop 2A including region Cys39p to Cys58p, p75NTR binding loop 2B including region Cys58p to Cys78p, and p75NTR binding loop 3A including region Cys79p to Cys94p. The invention further provides a method of identifying such a ligand encoded in a data base containing molecules coded for spatial occupancy, relative atomic position, bond type and/or charge. The designed or identified ligand may be an agonist or antagonist of p75NTR. |
