Volume 16, Issue 6, Pages 633-643 (June 2009) Substrate-Guided Design of a Potent and Selective Kallikrein-Related Peptidase Inhibitor for Kallikrein 4  Joakim E. Swedberg, Laura V. Nigon, Janet C. Reid, Simon J. de Veer, Carina M. Walpole, Carson R. Stephens, Terry P. Walsh, Thomas K. Takayama, John D. Hooper, Judith A. Clements, Ashley M. Buckle, Jonathan M. Harris  Chemistry & Biology  Volume 16, Issue 6, Pages 633-643 (June 2009) DOI: 10.1016/j.chembiol.2009.05.008 Copyright © 2009 Elsevier Ltd Terms and Conditions

Figure 1 Amidolytic Activity of KLK4 against Sparse Matrix Library pNA Substrates The y axis represents the rate of substrate cleavage in mOD/min(405), with the amino acid at position P2 indicated beside each y axis. The x axis labels the bars corresponding to substrates with each of the 25 combinations of the amino acids in the P4 and P3 positions. Arginine was kept constant in the P1 position while varying the amino acid in positions P2–P4 as indicated in the graph. Data represent mean and standard error of the mean (± SEM) of three experiments in triplicate. Chemistry & Biology 2009 16, 633-643DOI: (10.1016/j.chembiol.2009.05.008) Copyright © 2009 Elsevier Ltd Terms and Conditions

Figure 2 Inhibition of serine protease proteolytic activity by SFTI and SFTI-FCQR Kallikrein and trypsin-mediated proteolysis of fibrinogen assessed by SDS-PAGE. Proteolytic products are visualized by Coomassie blue staining following resolution on 12% polyacrylamide gels. (A) Effect of increasing concentrations of SFTI-FCQR on fibrinogen digestion by KLK4 and SFTIwt at a concentration of 2000 nM. (B) Trypsin digestion of fibrinogen in the presence and absence of 2000 nM SFTI-FCQR. (C–F) Digestion of fibrinogen by KLK2, 5, 12, and 14, respectively, in the presence of increasing concentrations of SFTI-FCQR. Gels are representative of three independent experiments. Chemistry & Biology 2009 16, 633-643DOI: (10.1016/j.chembiol.2009.05.008) Copyright © 2009 Elsevier Ltd Terms and Conditions

Figure 3 Stability of SFTI-FCQR in Contact with Prostate Cancer Cells In Vitro Residual SFTI-FCQR was assayed in cell culture media from prostate cancer cells treated with a single 1 μM dose of inhibitor at time 0. Endogenous inhibitors were removed by boiling followed by centrifugation to remove precipitated denatured protein. Stability was assessed against LNCaP (closed circles), 22RV1 (open circles), and PC3 cells (triangles). Data are mean ± SEM from three experiments in triplicate. Chemistry & Biology 2009 16, 633-643DOI: (10.1016/j.chembiol.2009.05.008) Copyright © 2009 Elsevier Ltd Terms and Conditions

Figure 4 Effect of SFTI-FCQR on Calcium Release in Cell-Based Assays Lung murine fibroblasts (LMF) stably expressing human PAR-2 were incubated after 30 s with either activated insect cell-derived KLK4 (iKLK4, 300 nM), trypsin (10 nM), or PAR-2-activating peptide (100 nM) with or without SFTI-FCQR (1 μM). The fluorescence at 510 nm was measured following alternating excitation at 340 and 380 nm (Em510[340/380]). The ratio of Em510(340/380) is proportional to intracellular Ca2+ ion concentration. The data are represented as mean of three experiments in triplicate. Treatment with (A) buffer or (B) SFTI-FCQR only had no effect whereas treatment with (C) trypsin, (D) iKLK4, (E) trypsin and SFTI-FCQR, or (F) PAR-2-activating peptide resulted in increased intracellular Ca2+ ion concentration. Calcium flux was restored in cells initially treated with iKLK4 and SFTI-FCQR at 30 s by addition of (G) trypsin or (H) PAR-2-activating peptide after 3.5 min. Arrows indicate injection points for solutions as indicated. Chemistry & Biology 2009 16, 633-643DOI: (10.1016/j.chembiol.2009.05.008) Copyright © 2009 Elsevier Ltd Terms and Conditions

Figure 5 Structural Characteristics of the Protease-Inhibitor Interfaces for the Trypsin-SFTI and the Modeled KLK4/SFTI-FCQR Complex Catalytic residues are labeled with an asterisk. SFTI residues that have been substituted in this study are labeled in green. (A) Superposition between trypsin (cyan)-SFTI (green) complex (1SFI) and KLK4 (1BD1, pink) in and around the active site, showing regions of structural differences, notably around the “99 loop” (L/T98; L99) and at residue 175. Side chains are shown and labeled for sequence differences or where significant structural differences are observed (KLK residue shown before trypsin residue in label). The S2 subsite is indicated. (B and C) Hydrogen bonds (magenta dashed lines) between trypsin (yellow) and SFTI (green) (B), and (C) modeled KLK4 (yellow) and SFTI-FCQR (green). Viewed from the inverse angle compared with all other figures in panel, i.e., view from protease toward inhibitor. (D and E) Molecular surfaces of (D) trypsin-SFTI and (E) modeled KLK4-SFTI-FCQR complexes colored according to electrostatic potential (red indicates negative, blue is positive). Subsites are labeled as S1, S2, etc. Structural differences between trypsin and KLK4, notably at L95 and L/T98, shown in A, cause an increase in the width and depth of the electronegative S2 pocket, allowing its occupation by the modeled glutamine (4Q) side chain of SFTI-FCQR. Chemistry & Biology 2009 16, 633-643DOI: (10.1016/j.chembiol.2009.05.008) Copyright © 2009 Elsevier Ltd Terms and Conditions