Methods for identifying inhibitors of amyloid protein aggregation

Abstract

Methods for identifying compounds that are inhibitors or are likely to be inhibitors of amyloid protein aggregation, as well as three-dimensional, non-crystallographic models (i.e. “pseudo-crystal structures”) of amyloid aggregation utilized in the methods, are described. Means for creating the three-dimensional, non-crystallographic models (i.e. “pseudo-crystal structures”) of amyloid aggregation are also described.

What is claimed is: 
     
         1 . A method of creating non-crystallographic models of amyloid protein aggregation comprising:
 applying molecular mechanics and/or dynamics to a three-dimensional model comprising a model of crystallized amyloid protein, a model of multiply anti-amyloid compound bound in the pocket of the crystallized amyloid protein, and a candidate model of an uncrystallized amyloid protein, such that a pocket on the uncrystallized amyloid protein is formed around the multiply anti-amyloid compound;   extracting the model of uncrystallized amyloid protein from said model; and   validating said model of uncrystallized amyloid protein with respect to experimental data so as to confirm its utility as a non-crystallographic model of amyloid protein aggregation.   
     
     
         2 . A non-crystallographic model of amyloid protein aggregation comprising:
 a three-dimensional model of a monomeric amyloid peptide;   and a three-dimensional model of an amyloid protein consisting of one or more amyloid peptides, said amyloid protein model to be positioned with respect to said amyloid peptide model such that it forms a pocket in conjunction with said amyloid peptide model, such that a candidate compound may be inserted into said pocket thereby modulating amyloid aggregation.   
     
     
         3 . The non-crystallographic model of amyloid protein aggregation of  claim 2 , wherein said amyloid peptide model and said amyloid protein model are both composed of beta-amyloid protein or an amyloid-forming fragment thereof. 
     
     
         4 . The non-crystallographic model of amyloid protein aggregation of  claim 2 , wherein said amyloid peptide model is substantially SEQ ID: 1. 
     
     
         5 . The non-crystallographic model of amyloid protein aggregation of  claim 4 , wherein said amyloid protein model is substantially SEQ ID: 2. 
     
     
         6 . The non-crystallographic model of amyloid protein aggregation of  claim 5 , wherein said amyloid peptide model is further positioned with respect to said amyloid protein model such that:
 the amino acid Val at position 8 of SEQ ID: 1 interacts with the amino acid Glu at position 6 of SEQ ID: 2;   the amino acid Gly at position 9 of SEQ ID: 1 interacts with the amino acid Glu at position 6 of SEQ ID: 2;   the amino acid Ser at position 10 of SEQ ID: 1 interacts with the amino acid Gly at position 9 of SEQ ID: 2;   the amino acid Ser at position 10 of SEQ ID: 1 interacts with the amino acid Ser at position 10 of SEQ ID:2;   and the amino acid Lys at position 12 of SEQ ID: 1 interacts with the amino acid Glu at position 6 of SEQ ID: 2.   
     
     
         7 . The non-crystallographic model of amyloid protein aggregation of  claim 5 , wherein said amyloid peptide model is further positioned with respect to said amyloid protein model substantially in the orientation shown in  FIG. 2 . 
     
     
         8 . A method of identifying compounds that modulate amyloid aggregation comprising the steps of:
 constructing the non-crystallographic model of amyloid protein aggregation according to  claim 7  in a computer modeling program;   selecting a list of candidate compounds;   constructing said candidate compounds in a computer modeling program;   performing an iterative docking and scoring of all candidate compounds, by means of docking each candidate compound into a pocket formed by said model and scoring each candidate compound to reflect its degree of complementarity with respect to said pocket; and   identifying compounds that modulate amyloid aggregation or better modulate amyloid aggregation by reference to a score cutoff that substantially distinguishes active compounds from inactive compounds, or more active compounds from less active compounds, respectively.   
     
     
         9 . A method of improving the potency of a compound known to modulate amyloid aggregation according to  claim 8 , wherein the list of candidate compounds includes both the compound known to be active and analogs of said compound, and the score cutoff is that score which belongs to the compound known to be active. 
     
     
         10 . The non-crystallographic model of amyloid protein aggregation of  claim 1 , wherein said amyloid peptide model and said amyloid protein model are both composed of alpha-synuclein protein or an amyloid-forming fragment thereof. 
     
     
         11 . The non-crystallographic model of amyloid protein aggregation of  claim 1 , wherein said amyloid peptide model and said amyloid protein model are both composed of tau protein or an amyloid-forming fragment thereof.