Stromelysin results

In the experimental studies [2], acetohydroxamic acid (compound Sl) was bound in the active site to prevent proteolysis as well as provide the lead for the primary site. Computationally, acetohydroxamic acid was docked to stromelysin in the protonated form to simplify the protein-ligand model. It is conceivable that there could be protonation state changes upon binding of the ligand to the protein, as well as charge transfer between the acetohydroxamic acid and the metal center of the protein. Figure 5a shows the lowest energy configuration of the acetohydroxamic acid, compound S1. This position is very similar to the binding orientation shown in Hadjuk et al [2]. Despite using such a simple model, we obtain reasonable results for predicting the structure and the binding affinity (see Table 4) of the acetohydroxamic acid.

The compounds S2-S26 shown in Figure 3 were docked in the absence and in the presence of the acetohydroxamic acid, compound S1. Interestingly, these compounds bound in the same hydrophobic pocket regardless of the presence or absence of a ligand in the primary binding site. This is in contrast to the FKBP docking studies where all ligands preferred to bind in the primary site where the trimethoxyphenyl pipecolinic acid bound. Table 4 shows the predicted binding affinities for these compounds in the presence of the acet-

ohydroxamic acid as well as the free energy of binding for acetohydroxamic acid. The predicted free energies reproduce the experimental values [2] well, except for compound S4 which is predicted to bind several kilocalories more tightly than the free energies estimated from NMR experiments. Examining the compounds in the SAR series described in the Hadjuk paper [2], it is not so clear why some compounds in the series bound tightly and others did not in the NMR experiment. Given the hydrophobic nature of the molecules in general and the fact that some molecules contain a polar or ionizable group, it is possible that there could be some solubility difficulties due to solute aggregation. If there is solute aggregation, the estimated free ligand concentration will be lower than predicted and the dissociation constant will be an upper bound. It is also plausible that the scoring function incorrectly ranks this compound, which demonstrates a limitation in the scoring function that needs to be addressed.

Figure 5b shows the lowest energy structure of compound S8 bound to stromelysin in the presence of acetohydroxamic acid. The distance from the methyl carbon of the acetohydroxamic acid to the hydroxyl oxygen of compound S8 was 3.51 Â. Compounds S49-S52 were created [2] from compound S1 and compound S8 with various length hydrocarbon linkers. Compound S53 was created [2] from compound S1 and compound S26. The predicted free energies are found in Table 4 and agree well with experiment. Figure 5c shows the lowest energy configuration of compound S50. The docking of these composite compounds was done with a fully flexible ligand, unlike the larger composite compounds of the FKBP study. Although the acetohydrox-amic acid moiety remains bound to the zinc, it moves slightly closer to the pocket containing the hydrophobic moiety, possibly in an attempt to bury the hydrophobic linking group.

0 0

Post a comment