Conformetrix Technology

 

Conformetrix 3D data out-performs the market alternatives

 

Ligand-receptor interactions are dependent upon the 3D structures of the constituent interacting molecules. Designing drug molecules based on the structure of a receptor's natural ligand represents an excellent start point. Conformetrix have solved the 3D and 4D structures of angiotensinl (a decapeptide) and lisinopril (illustrated opposite), the natural inhibitor of angiotensin converting enzyme (ACE) and a previous best-selling ACE-inhibitor (ACEi), respectively.

 

The root mean squared deviation (RMSD) of the Conformetrix lisinopril structure (blue) from the available protein-bound crystal structures (grey, RCSB accession codes 1O86, 2C6N and 1J36) is 0.67 Angstroms (Figure 1, opposite). This is significantly better than the RMSD of the small molecule X-ray crystallographic structure (Figure 2) and the previous best NMR structure (Figure 3), which are shown in red and have RMSD values of 2.63 Angstroms and 1.76 Angstroms, respectively.

 

Figures 4 and 5 and 6 illustrate the lowest energy lisinopril geometries obtained from CORINA, multiple theoretical molecular dynamics simulated annealing runs and a stochastic search algorithm, respectively. Again, these are shown relative to protein-bound crystal structures (grey) and perform significantly worse than the Conformetrix structure in terms of RMSD from the protein-bound crystal structure. This data illustrates that Conformetrix provides the most relevant starting point for rational drug design processes, such as docking studies.

 

When docked to the ACE crystal structure active site the Conformetrix coordinates (C4X) bound as well as the positive control co-complex ligand (RCSB accession code 1O86 and denoted XTAL:B in the graph below) and two other crystal structures of protein-bound lisinopril (26CN and 1J36, denoted XTAL:A and XTAL:C, respectively).

 

 

 

 

The Conformetrix 3D structure out-performed previous experimentally-determined NMR and small molecule crystal structures (denoted NMR and CCD, respectively) in the docking exercise. It also bound more tightly to the active site than did 3D structures generated theoretically by CORINA, molecular dynamics simulated annealing (SIM) and a stochastic search algorithm (STOCH), all of which failed to reproduce the experimentally observed binding pose.

 

 

A pharmacophore distance map (DMAP, above), based on features of the lisinopril electrostatic potential (the relationship between the location of local minima and maxima associated with pharmacophoric features) showed the Conformetrix structure (C4X, blue) to reproduce best the properties of the protein-bound structure (1O86, grey) when compared with the small molecule crystal structure (CCD, yellow) and previous NMR structure (NMR, green).