Recently, the enzymatic characterization has been investigated for FabZ enzymes from several different strains including Enterococcus faecalis (EfFabZ) [32, 33], Pseudomonas aeruginosa (PaFabZ) [34], Plasmodium falciparum (PfFabZ) [29, 35], and H. pylori (HpFabZ) [7]. The crystal structural analyses have been determined for PaFabZ and PfFabZ [6, 29, 34], while some inhibitors against PaFabZ and HpFabZ were also discovered [8, 29, 30, 36, 37]. In the current work, the crystal structure of HpFabZ/Emodin complex was determined, and two different binding Tariquidar models (models A and B) were put forwarded. In the models, the hydrophobic interactions between Emodin and

the selleck chemicals nearby residues of HpFabZ contributed to the major interaction forces. In model

A, the interaction between ring A of Emodin and residues Tyr100 and Pro112′ in sandwich manner is the main hydrophobic interaction force, resulting in better electron density map around ring A, while ring C at the other end of Emodin had only weak interactions with residues nearby. In model B, the whole molecule of Emodin dove deeply into the active tunnel forming intense hydrophobic interactions with the residues nearby, thus the electron density map around Emodin was continuous, completive and much better than the map in model A (Fig. 3). Additionally, this interaction has also made the average B factor Linifanib (ABT-869) of Emodin in model B better than in model A (The average B factor of Emodin was 45.03 in model A, while 39.24 in model B). In comparison with our recent mTOR inhibitor published crystal structure of HpFabZ in complex with compound

1 (PDB code 2GLP) [8], there are some differences concerning their binding features due to the longer molecule of compound 1 than Emodin. In model A, the pyridine ring of compound 1 was sandwiched between residues Tyr100 and Pro112′ linearly as ring A of Emodin, while the 2,4-dihydroxy-3,5-dibromo phenyl ring at the other end of compound 1 stretched into another pocket formed by Arg158, Glu159, Phe59′, Lys62′ through hydrophobic interactions, which can not be found in the binding model A of Emodin (Fig. 5A). In model B, compound 1 entered into the middle of the tunnel. Its pyridine ring accessed the end of the tunnel where the ring C of Emodin located in the model B, and stayed in the right place via hydrophobic interactions. However, the 2,4-dihydroxy-3,5-dibromo phenyl ring of compound 1 was too large to dive into the tunnel. Thus it had to adopt a crescent shaped conformation and stretched the 2,4-dihydroxy-3,5-dibromo phenyl ring out of the tunnel forming a sandwich conformation with residues Ile98 and Phe59′ via π-π interactions. Based on these additional interactions, compound 1 should have a better inhibition activity against HpFabZ than Emodin.