Molecular basis for the specific binding of different α-amylase inhibitors from Phaseolus vulgaris seeds to the active site of α-amylase

In search of a possible mechanism of inhibition which might be responsible for the different specificities of the three isoforms of the bean (Phaseolus vulgaris) α-amylase inhibitor α-AI1, α-AI2 and α-AIL (EC 3.2.1.1), the two isoforms α-AI2 and α-AIL were modelled from the atomic co-ordinates of α-AI1 in the α-AI1/PPA complex and docking experiments were performed with pig pancreatic α-amylase (PPA) and the modelled amylase from Zabrotes subfasciatus (ZSA). The modelled α-AI2 penetrates without any steric hindrance in the substrate cleft of both enzymes but the possible hydrogen bonds between PPA and α-AI2 seem too few to maintain the stability of the complex. α-AIL, which differs from α-AI1 and α-AI2 by the absence of post-translational proteolytic cleavage and the occurrence of two additional loops of fifteen and six residues, creates steric clashes with PPA and ZSA that prevent its penetration into the substrate cleft of the enzyme. Docking experiments explain at the molecular level the specificity of α-amylase inhibitor isoforms towards enzymes of different origins. In addition, they explain why, according to its unprocessed and more bulky character, α-AIL was previously shown to be inactive on all α-amylases assayed. In fact, this last isoform is now considered as an evolutionary intermediate between phytohaemagglutinins, arcelins and α-amylase inhibitors.