Biol. Chem., Vol. 386, pp. 339–349, April 2005 • Copyright  by Walter de Gruyter • Berlin • New York. DOI 10.1515/BC.2005.041 2005/173 Article in press - uncorrected proof Homology modeling and SAR analysis of Schistosoma japonicum cathepsin D (SjCD) with statin inhibitors identify a unique active site steric barrier with potential for the design of specific inhibitors Conor R. Caffrey 1 , Lenka Placha 2 , Cyril Barinka 2 , Martin Hradilek 2 , Jir ˇı´ Dosta ´l 2 , Mohammed Sajid 1 , James H. McKerrow 1 , Pavel Majer 3 , Jan Konvalinka 2 and Jir ˇı´ Vondra ´s ˇ ek 2,4, * 1 Sandler Center for Basic Research in Parasitic Diseases, University of California at San Francisco, Box 0511, San Francisco, CA 94143, USA 2 Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech Republic, Flemingovo n. 2, 166 10 Praha 6, Czech Republic 3 Guilford Pharmaceuticals Inc., 6611 Tributary Street, Baltimore, MD 21224, USA 4 Center for Biomolecules and Complex Molecular Systems, Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech Republic, Flemingovo n. 2, 166 10 Praha 6, Czech Republic * Corresponding author e-mail: jirka@uochb.cas.cz Abstract Proteases that digest the blood-meal of the parasitic fluke Schistosoma are potential targets for therapy of schistosomiasis, a disease of chronic morbidity in humans. We generated a three-dimensional model of the cathepsin D target protease of Schistosoma japonicum (SjCD) utilizing the crystal structure of human cathepsin D (huCD) in complex with pepstatin as template. A homology model was also generated for the related secreted aspartic protease 2 (SAP2) of the pathogenic yeast, Candida albicans. An initial panel of seven statin inhibitors, originally designed for huCD wMajer et al., Pro- tein Sci. 6 (1997), pp. 1458–1466x, was tested against the two pathogen proteases. One inhibitor showed poor reactivity with SjCD. Examination of the SjCD active-site cleft revealed that the poor inhibition was due to a unique steric barrier situated between the S2 and S4 subsites. An in silico screen of 20 potential statin scaffolds with the SjCD model and incorporating the steric barrier con- straint was performed. Four inhibitors (SJ1–SJ4) were eventually synthesized and tested with SjCD, bovine CD and SAP2. Of these, SJ2 and SJ3 proved moderately more specific for SjCD over bovine CD, with IC 50 values of 15 and 60 nM, respectively. The unique steric barrier identified here provides a structural focus for further development of more specific SjCD inhibitors. Keywords: active site; aspartyl protease; Candida; drug design; parasite. Introduction Schistosomiasis, caused by parasitic bloodflukes of the genus Schistosoma, is of serious public health concern in 74 tropical and sub-tropical countries. More than 200 million people are infected, with 20 million suffering from serious disease (Chitsulo et al., 2004). Adult worms reside within the blood vessels of the intestine and blad- der, and their eggs cause morbidity associated with a progressive, immunopathological reaction in various tis- sues including the liver and intestine (Pearce and Mac- Donald, 2002). Schistosomes feed on blood proteins, including hemo- globin, and express a range of gut-associated proteases to facilitate digestion. The proteases thought to be direct- ly involved in this process include orthologs of the mam- malian cysteine proteases, cathepsins B and L, and the aspartic protease, cathepsin D (see Caffrey et al., 2004 for a review). Because of their critical function in nutrition, therefore, such proteases are considered potential tar- gets for chemotherapeutic intervention (Caffrey et al., 2004) and, indeed, the cysteine proteases have been demonstrated as such (Wasilewski et al., 1996). The fully processed cathepsin D of the Asian blood- fluke Schistosoma japonicum (SjCD; accession number L41346, Becker et al., 1995, modified as U90750) shares 57.2% identity with human cathepsin D (huCD). Like this enzyme, SjCD displays an acid pH optimum (Caffrey et al., 1998; Brindley et al., 2001) and preferentially hydro- lyzes bonds between two hydrophobic residues (Brindley et al., 2001). However, the preferential cleavage sites in human hemoglobin for SjCD and huCD differ (Brindley et al., 2001), suggesting significant topographical differenc- es between the active-site clefts of both enzymes. Such differences might be exploited in the rational design of novel and selective drugs to treat schistosomiasis. In the absence of crystal structures of schistosome CDs, com- parative molecular modeling using the crystal structure of huCD wProtein Data Bank (PDB) ID: 1LYB; Baldwin et al., 1993x has been applied to explain active site differ- ences between parasite and human orthologs (Brink- worth et al., 2001; Silva et al., 2002). Thus far, however, no attempt has been made to screen, design or test small-molecule inhibitors of schistosome CDs either using structure-activity relationship (SAR) studies with a homology model or otherwise. For this report, we generated a homology model of SjCD in complex with pepstatin and compared the active-site amino acid composition and topography with both huCD and the related protease, secreted aspartic protease 2 (SAP2), from the pathogenic fungus Candida