Redesigning Serine Proteases for Resistance to Inhibition

Using computational and molecular biology approaches, we are using the model serine protease, trypsin, to identify amino acid residues that are important in the selection and binding of macromolecular substrates and inhibitors. This work will help us engineer trypsin and similar proteases to be more resistant to inhibition so that they may be more effective as therapeutic agents. In several trypsin inhibitor co-crystal structures, the backbone carbonyl oxygen of phenylalanine 41 (F41) accepts a hydrogen bond from the P2’ backbone amide hydrogen of the inhibitor. F41 is conformational restricted by its interaction with the methylene groups of lysine 60, which presumably optimizes the backbone carbonyl oxygen of F41 for hydrogen bonding with the inhibitor. By making substitutions at position 41, we may be able to increase the conformational flexibility of its backbone carbonyl and reduce the affinity of trypsin for inhibitors. To test this hypothesis, we have made two trypsin variants, F41L and F41G. Both F41L and F41G trypsin variants were expressed in Pichia pastorisas zymogens,were purified using hydrophobic chromatography. The mature forms of both variants were obtained by auto-activation at pH 8.00. The mature variants were isolated using affinity chromatography. Lastly, activity assays were performed with and without inhibitors using 1nM enzyme and varying concentrations of Z-GPA-pNA at 25 °C. Variants were kinetically similar to wild-type trypsin in K_M (7.55 ± 1.82 µM), k_cat (4736 ± 200 min^-1), and k_cat/K_M (627±109 uM^-1 min^-1) and in the K_Ivalue for benzamidine, a small molecule inhibitor (6.1 ± 1.0 µM).  These data suggest that the substitutions do not affect the active site structure or the chemistry of the enzyme. We observed differences in the variants when we measured the effect of a macromolecular inhibitor (BPTI) on activity. The association rate for BPTI (k_on) with wild-type and F41L were similar (8110.5 ± 460.62 M^-1s^-1 versus 9262.8 ± 593.6 M^-1s^-1 respectively. However, the k_on value for F41G variant was significantly lower (4991.2 ± 187.67 M^-1s^-1). The results with BPTI suggest that glycine residue at position 41 decreased the affinity of trypsin for BPTI. Collectively, the data show that the substitutions do not affect the activity of the enzyme and support the hypothesis that increasing the flexibility at position 41 is a plausible approach to decreasing the sensitivity of trypsin (and trypsin-fold proteases) to inhibition by macromolecular inhibitors.