Elucidating conformational dynamics and ligand specificity of Schistosoma bovis 28 kDa glutathione transferases: a structural investigation.

Abstract

Schistosomiasis, caused by Schistosoma trematodes is the second most devastating neglected tropical disease. Schistosoma bovis largely affects ruminant hosts, significantly crippling the economic output of livestock farmers and food security. A promising anti-schistosomal drug target is the multifunctional detoxification enzyme Schistosoma bovis 28 kDa glutathione transferase (Sb28GST). While empirical structural techniques are insightful and used in drug design, they fall short in providing the dynamics of the protein of interest. Computer aided drug design (CADD) offers a time and cost-efficient workflow to screen and analyze drug candidates. This method has limitations, whose misinterpretation can hinder drug design and development. In this study we caution the careful selection of the crystal structure space groups used for CADD as isolated analyses influence the conformational states visualized and simulated as we search for Sb28GST drugs for Schistosomiasis intervention. Upon the successful expression and purification of Sb28GST, crystal conditions were manipulated to obtain the protein's crystal structure at 2.4 Å in an orthorhombic crystal system. A 500 ns molecular dynamic simulation consisting of eight stages was performed on the orthorhombic 8BHZ to compare to the monoclinic Sb28GST 8ALS. Two distinct trajectories were undertaken by the apo proteins, revealing significant differences in the dynamic behaviour of the protein systems. A library of flavonoid compounds was used to perform high-throughput virtual screening for the identification of potential Sb28GST ligands, whose outputs were filtered using standard precision and extra precision mode. The screening process showed a diverse selection of flavonoid compounds between the different space groups with strong favourable interactions implied from the low free binding. The only common ligand selected between the protein polymorphs was apigenin 7-O-(2G-Rhamnosyl) Gentiobioside however, quercetin-3-O-Beta-D-Glucose-7-O-Beta-D-Gentiobioside was the best performing ligand overall, with the lowest glide score of -15.6645 kcal/mol. The molecular dynamic simulations of the protein polymorphs with the selected ligands additionally highlighted different molecular interactions in which the proteins used to stabilize the proteins bond to the ligands. Concluding that the combined assessment of protein polymorphs provides more detailed insights into the conformation dynamics of the protein which would not be otherwise identified in a single space group analysis. To test the inhibitory potency of the selected ligands a CDNB-GSH assay showed that both apigenin and quercetin significantly reduced the activity of Sb28GST. The IC50 of the confirmed inhibitors showed that apigenin was a more potent inhibitor (IC50 = 0.130 mM) than quercetin (IC50 = 0.120 mM). The extrinsic fluorescence studies suggest that apigenin and quercetin both bind to the hydrophobic H-site and allosteric L-site at the dimer interface. Analysis of the thermal stability of Sb28GST showed that the presence of the ligands reduces the thermal stability of the protein in a very insignificant way. This study presents the first empirical validation of computationally selected lead flavonoid compounds as inhibitors of Sb28GST for schistosomiasis intervention. From this studies' findings, future research would entail validating the lead flavonoid inhibitors through kinetic and mechanistic studies. In addition, in vitro and in vivo testing of the flavonoid inhibitors would be expanded to other Schistosome GST isoforms to guide clinical candidate development.