Understanding Structure–Activity Relationships for Trypanosomal Cysteine Protease Inhibitors by Simulations and Free Energy Calculations

Abstract

The protozoan cysteine proteases cruzain in Trypanosoma cruzi and rhodesain in Trypanosoma brucei are therapeutic targets for Chagas disease and Human African Trypanosomiasis (HAT), respectively. A benzimidazole series was previously characterized as potent noncovalent competitive cruzain and rhodesain inhibitors with activity against trypanosomes. Common structure–activity relationships (SAR) trends and structural modifications leading to selectivity against each enzyme were described. However, some of these trends could not be understood based on the reported binding mode of lead compound 1. Therefore, we employed microsecond molecular dynamics simulations and free energy calculations to understand qualitative SAR trends and to quantitatively recapitulate them. Simulations revealed the most stable protein–ligand interactions and provided insights concerning enzyme selectivity. Calculated relative binding free energies of compound 1 analogs exhibited deviations of 1.1 and 2.2 kcal/mol from the experimental values for cruzain and rhodesain, respectively. These data encourage prospective thermodynamic integration (TI) studies to optimize this series and facilitate the prioritization of compounds for synthesis.

Elucidating the Dual Mode of Action of Dipeptidyl Enoates in the Inhibition of Rhodesain Cysteine Proteases

“…As commented in the Introduction, the inhibition of rhodesain depends on the peptidic framework of the inhibitor. [5,6,10,[28][29][30][31][32][33][34] In order to perform a more detailed analysis of this dependency, the QM/MM interaction energies (electrostatic plus Lennard-Jones) between residues of rhodesain and the inhibitor were computed as an average over 10000 structures from the AM1d/MM simulations (see Figures from S4, S5, S7 and S8 in the Supporting Information). The main favourable rhodesain-inhibitor interactions in the RS state of both inhibition mechanisms is shown in Figure 6.…”

“…[27] Several authors have demonstrated that the inhibition of rhodesain is dependent on the peptidic framework of the inhibitor. [5,6,10,[28][29][30][31][32][33][34][35] Diederich and co-workers, [28] in a structurebased study of nitrile-derived rhodesain inhibitors, showed that variations in the S1 pocket had weak effects on selectivity of this enzyme; moreover, rhodesain preferred extended hydrophobic for its S2 pocket, while the S3 pocket shows clear preference for aromatic substituents. Later, Schirmeister and coworkers confirm these results, [31] and concluded that the interaction between the S2 pocket of rhodesain and the P2 residue of the inhibitor is an important specificity determinant.…”

“…Several authors have demonstrated that the inhibition of rhodesain is dependent on the peptidic framework of the inhibitor [5,6,10,28–35] . Diederich and co‐workers, [28] in a structure‐based study of nitrile‐derived rhodesain inhibitors, showed that variations in the S1 pocket had weak effects on selectivity of this enzyme; moreover, rhodesain preferred extended hydrophobic for its S2 pocket, while the S3 pocket shows clear preference for aromatic substituents.…”

The gene repertoire of the main cysteine protease of Trypanosoma cruzi, cruzipain, reveals four sub-types with distinct active sites

“…On the other hand, there are enough substitutions in this region to suggest that inhibitors will have different a nities or even reach selectivity among the cruzipain sub-types. Throughout the literature, several studies have described cruzain inhibitors that are not effective against T. cruzi or discrepancies between the structure-activity relationships obtained for a series of compounds against cruzain and the parasite 40,42,43 . These differences are partially explained by the different chemical properties of the compounds, affecting features like membrane permeability, chemical stability, and metabolism.…”

“…That is the only difference in the S2 pocket of these parasitic enzymes, resulting in a more open pocket. While several classes of inhibitors binding both to cruzain and TbrCatL have been described 12 , studies with a series of benzimidazole inhibitors provide an example of different structure-activity relationships for the two enzymes [39][40][41] .…”