Quantitative structure-activity relationship of human neutrophil collagenase (MMP-8) inhibitors using comparative molecular field analysis and X-ray structure analysis.

A set of 90 novel 2-(arylsulfonyl)-1,2,3, 4-tetrahydroisoquinoline-3-carboxylates and -hydroxamates as inhibitors of the matrix metalloproteinase human neutrophil collagenase (MMP-8) was designed, synthesized, and investigated by 3D-QSAR techniques (CoMFA, CoMSIA) and X-ray structure analysis. Docking studies of a reference compound are based on crystal structures of MMP-8 complexed with peptidic inhibitors to propose a model of its bioactive conformation. This model was validated by a 1. 7 A X-ray structure of the catalytic domain of MMP-8. The 3D-QSAR models based on a superposition rule derived from these docking studies were validated using conventional and cross-validated r2 values using the leave-one-out method, repeated analyses using two randomly chosen cross-validation groups plus randomization of biological activities. This led to consistent and highly predictive 3D-QSAR models with good correlation coefficients for both CoMFA and CoMSIA, which were found to correspond to experimentally determined MMP-8 catalytic site topology in terms of steric, electrostatic, and hydrophobic complementarity. Subsets selected as smaller training sets using 2D fingerprints and maximum dissimilarity methods resulted in 3D-QSAR models with remarkable correlation coefficients and a high predictive power. This allowed to compensate the weaker zinc binding properties of carboxylates by introducing optimal fitting P1' residues. The final QSAR information agrees with all experimental data for the binding topology and thus provides clear guidelines and accurate activity predictions for novel MMP-8 inhibitors.

Molecular design and characterization of an alpha-thrombin inhibitor containing a novel P1 moiety.

An inhibitor of alpha-thrombin was designed on the basis of the X-ray crystal structures of thrombin and trypsin. The design strategy employed the geometric and electrostatic differences between the specificity pockets of the two enzymes. These differences arise due to the replacement of Ser 190 in trypsin by Ala 190 in thrombin. The new inhibitor contained a tryptophan side chain instead of the arginine side chain that is present in the prototypical thrombin inhibitors. This inhibitor had a Ki value of 0.25 microM, displayed more than 400-fold specificity for thrombin over trypsin, and doubled the rat plasma APTT at a concentration of 44.9 microM. The X-ray crystal structure of the inhibitor/alpha-thrombin complex was determined. This represents the first reported three-dimensional structure of a thrombin/ inhibitor complex where the specificity pocket of the enzyme is occupied by a chemical moiety other than a guanidino or an amidino group. As was predicted by the molecular model, the tryptophan side chain docks into the specificity pocket of the enzyme. This finding is in contrast with the indole binding region of thrombin reported earlier [Berliner, L. J., & Shen, Y. Y. L. (1977) Biochemistry 16, 4622-4626]. The lower binding affinity of the new inhibitor for trypsin, compared to that for thrombin, appears to be due to (i) the extra energy required to deform the smaller specificity pocket of trypsin to accommodate the bulky indole group and (ii) the favorable electrostatic interactions of the indole group with the more hydrophobic specificity pocket of thrombin. The neutral indole group may be of pharmacological significance because the severe hypotension and respiratory distress observed following the administration of some thrombin inhibitors have been linked to the positively charged guanidino or amidino functionalities.

Synthesis and comparison of tripeptidylfluoroalkane thrombin inhibitors.

Fluorinated ketone thrombin inhibitors based on the peptide sequence methyl-(D)-Phe-Pro-Arg-CF2R were synthesized: MDL 73,446 (1, R = F); MDL 73,775 (2, R = CF3); and MDL 75,579 (3, R = CH2CH2CH3). These were shown to be highly effective, slow binding inhibitors of thrombin. Anticoagulant activity was dose-dependent with 3 > 2 > 1 at doubling thrombin time and APTT, respectively. Anticoagulant activity corresponded with efficacy in a platelet-dependent (FeCl3-induced) rat carotid artery thrombosis model. Arterial occlusion was dose-dependently prolonged with 3 > 2 > 1 at doubling the occlusion time.

Synthesis of peptidyl fluoromethyl ketones and peptidyl alpha-keto esters as inhibitors of porcine pancreatic elastase, human neutrophil elastase, and rat and human neutrophil cathepsin G.

Comparison of MeO-Suc-Val-Pro-Phe-CO2Me (29) and MeO-Suc-Ala-Ala-Pro-Phe- CO2Me (25) with their corresponding trifluoromethyl ketones 9a and 9b, respectively, in rat and human neutrophil cathepsin G assays showed the alpha-keto esters to be more potent inhibitors. Likewise, Ac-Pro-Ala-Pro-Ala-CO2Me (21) was more potent than its corresponding trifluoromethyl ketone (9c) in both porcine pancreatic elastase and human neutrophil elastase assays. Within a set of Ala-Ala-Pro-Val-CF3 elastase inhibitors, the carbobenzyloxy (Cbz) N-protecting group conferred greater potency as a P5 site recognition unit for elastase than did dansyl, methoxysuccinyl, or tert-butyloxycarbonyl. Initial inhibition of elastase was greater when trifluoromethyl ketone 9f was added from a stock solution of dimethyl sulfoxide than when it had been buffer-equilibrated prior to assay, which suggests that the nonhydrated ketone is the more effective form of the inhibitor. The most potent elastase inhibitor we report is Na-(Ad-SO2)-N epsilon-(MeO-Suc)Lys-Pro-Val-CF3 (16) which has a Ki of 0.58 nM.