Exploring the Flap Pocket of the Antimalarial Target Plasmepsin II: The “55 % Rule” Applied to Enzymes

With 300–660 million infections annually, malaria is a major health issue and threatens approximately 40% of the world’s population. In the search for new antimalarial targets, attention has turned to the degradation of human hemoglobin by the parasite Plasmodium falciparum. Several enzymes, including the three aspartic proteases plasmepsins (PMs) I, II, IV, and a histo-aspartic protease (HAP), are involved in the process, which is both vital and specific for the pathogen. As these proteases show overlapping substrate specificity, all of them have to be targeted by a potential antimalarial drug. Unfortunately, to date no crystallographic data are available for PM I or HAP in the protein data bank (PDB). The only two available Xray crystal structures of PM IV in complex with pepstatin A and an allophenylnorstatin-based inhibitor (PM IV of Plasmodium malariae) are showing a pepsinlike fold, which is also observed for the seventeen published X-ray crystal structures of PM II. However, three of these PM II structures (PDB codes 2BJU, 2IGX, 2IGY) feature a new cavity (flap pocket). This pocket is opened and shaped by the n-pentyl chain of the reported inhibitors. The existence of this pocket has been observed before in renin. It is mainly lined by hydrophobic amino acid residues, and only its occupancy resulted in potent and selective nonpeptidomimetic inhibitors for PMs and HAP. Related human aspartic proteases such as cathepsin D and E must not be affected, and indeed the activity of both was only insignificantly impaired. The exploration of the molecular recognition properties of the plasmepsin proteases, especially with respect to the flap pocket, is a major objective of our research, and in this work we describe the optimal filling of this hydrophobic cavity. Recently, a new class of nonpeptidomimetic inhibitors targeting the flap pocket was designed and synthesized in our laboratory (Figure 1). Inhibitor ( )-1 was the most active compound against PM II (IC50=130 (PM II) and IC50=50 nm ACHTUNGTRENNUNG(PM IV); IC50=concentration of inhibitor at which 50% of the maximum initial rate is observed), whereas compound ( )-2 was the most potent against PM IV (IC50=210 (PM II) and IC50=30 nm ACHTUNGTRENNUNG(PM IV)). Enantiomer ( )-2 binds much better to PM II (IC50=45 nm) and PM IV (IC50=10 nm) than (+ )-2 (IC50= 3260 and 33900 nm, respectively). The absolute configuration of the active enantiomer was tentatively assigned based on molecular modeling. We chose compound ( )-1 as a lead structure. With the residue R targeting the S1/S3 subsite left unchanged (Figure 1), the vector aiming for the flap pocket was varied systematically to explore the binding properties of this cavity, especially with respect to the volume that is available for binding. In this context, important concepts previously demonstrated in molecular recognition studies with synthetic receptors were transferred and applied to an enzyme environment. The 55% rule of Mecozzi and Rebek states that inclusion complexes are favored, when a guest occupies 55 9% of the available space within a host. This rule applies in particular to apolar binding processes. In the case of alkyl chains, it has been observed that residues apparently too long in their fully extended conformation for a certain cavity, adopt energetically less favorable conformations to fit into the available space. Unfavorable internal gauche strain in these contorted guests is compensated by burial of hydrophobic surfaces, chemical complementarity, and proper filling of space. Examples of nonstaggered alkyl residues bound by enzymes such as lipid-binding proteins are also known. It was further suggested that the principle of ideal volume occupancy is generally applicable, for example, to drug design. However, the 55% rule has to our knowledge not yet been applied to enzymes. In contrast to rigid synthetic container molecules, proteins are more flexible, and there are several cases known where the enzyme adapts itself to the—although contorted—ligand. It is therefore difficult to estimate the volume of an enzyme subpocket. The flap pocket of PM II however is relatively narrow and well defined, as it is observed in the three X-ray crystal structures of PM II Figure 1. General scaffold of our plasmepsin inhibitors, lead compound ( )-1, and inhibitor ( )-2. The (protonated) azanorbornane addresses the catalytic Asp dyad, the alkyl chain binds into the flap pocket, and the aromatic residue occupies the S1/S3 site.