The mechanism of uncompetitive inhibition was determined at the intersection of the Cornish–Bowden plot occurring in the second quadrant and the intersection of the Dixon plot occurring at y = −∞(Ki>>>Ki′). Fisetin (3,3′,4′,7-tetrahydroxyflavone) was four times more potent than

quercetin (3,3′,4′,5–7-pentahydroxyflavone), which indicates that the hydroxyl at position 5 may not be necessary to inhibit arginase. Moreover, quercetin, which has a hydroxyl at position 3, is twice as potent as luteolin (3′,4′,5–7-tetrahydroxyflavone). Direct comparison of fisetin with luteolin indicates that the hydroxyl at position 3 of fisetin provides CCI-779 in vivo an inhibition ten times greater than that when the hydroxyl is at position 5 in luteolin. Surprisingly, tetrahydroxyflavone fisetin was expected to have the optimal number and a better distribution of hydroxyls, but we found that 7,8-dihydroxyflavone (IC50 = 12 μM) presented an IC50 close to the IC50 of luteolin (tetrahydroxyflavone). A key feature of the inhibitors is the presence of the catechol group because, in its absence, the compounds apigenin, vitexin and isovitexin displayed no significant inhibition. A comparative analysis of quercetin and structurally related compunds (kaempferol and galangin) showed that the catechol group is more important for inhibition than are the phenol TSA HDAC molecular weight and

benzyl groups. The 7,8-dihydroxyflavone, in which the aromatic ring has no hydroxyls, contrasts with the other inhibitors. The hydroxyl at position 8 exhibited ten times the inhibition of galangin (benzyl group) and five times the inhibition of kaempferol (phenol

group). The optimum docking protocol was constructed with flexibility in the enzyme binding pocket (‘induced fit’). Each compound was docked with softened potentials (steric, hydrogen bonding, and electrostatic forces) and, at this point, the enzyme residues were kept rigid at their default conformation. Then, all residue sidechains that were close enough to the compound to interact with it were Leukocyte receptor tyrosine kinase energy-minimized. The final step was the energy minimization of the compounds. The docking scores of the interactions between the arginase from L. amazonensis and the target compounds, as well as the hydrogen bonding, steric, and van der Waals energy contributions and the number of possible atom–atom free torsions, are shown in Table 2. Fig. 2A shows the solvent-accessible surface of arginase from L. amazonensis, including the docked compounds in the binding pocket. A close up of the docking interaction with the enzyme is shown in Fig. 2B and C. Fig. 3 shows a 2D-representation of the flavonoid-enzyme interactions ( Schomburg, Ehrlich, Stierand, & Rarey, 2010). The intermolecular hydrogen bonds are shown as black dashed lines in both Fig. 2B and Fig. 3.