Abstract 2157 Strengthening Oral Defenses through Modification of Cystatin-SA: A Molecular exploration of Protease Interactions and Hydroxyapatite Affinity

Cystatin-SA is a naturally occurring protein present in saliva, it functions as a thiol protease inhibitor (TPI), and plays an important role in creating a salivary film (pellicle) on enamel surfaces capable of surface exclusions of transient pathogen microbes, protecting the oral cavity against microbial and chemical attacks. Breach of this film by thiol proteases results in adhesion and colonization of pathogenic bacteria resulting in formation of caries, and periodontal inflammation. Dual roles of cystatin involve interactions with a) the hydroxyapatite (HA) of the enamel surface and b) with thiol proteases from oral microbiome pathogens. To understand structural aspects of cystatin, we have undertaken computational and wet lab experiments designed to explore the following questions: 1) What aspects of cystatin enhance binding to HA, 2) what residues in cystatin provide affinity for target thiol proteases; do they determine TPI specificity? 3) Does cystatin immobilization on HA impact its TPI activity. Our computational work suggests negatively charged carboxyl groups (8-10Å apart) contribute to HA binding affinity. Several such groups are located near the inhibitory loops of cystatin and hence HA binding might impact TPI activity. To explore this, we have assayed TPI activity of cystatin using the model thiol protease, papain, in the presence and absence of HA. Control experiments suggest that papain binding to HA may impact protease activity. To explore cystatin interactions with thiol proteases, we have used HawkDock MM/GBSA analysis of the tarocystatin-papain complex (3IMA), comparing it to a model we created of human cystatin-SA in complex with papain. This analysis showed the tarocystatin-papain complex is more stable than the human cystatin-SA-Papain complex and that there were distinct differences in the residues in papain involved in complex formation, suggesting different forms of cystatin have different protease specificity. In particular, G32, R74 and E75 in human cystatin-SA contributed negatively to the complex with papain. Computational mutation of these three residues to delete G32, create R43A and E44Q triple mutant followed by energy minimization of the papain complex and subsequent HawkDock analysis showed the affinity for papain increased from -67.33 kcals/mole to -92.71 kcals/mole, comparable to that seen for the tarocystatin-Papain complex (93.54 kcals/Mole). Specific site directed mutations of cystatin have been constructed to confirm the computational predictions regarding cystatin-protease specificity/affinity. Analysis of the structures of human cystatin SA and papain indicate a number of negatively charged groups with approximately 10Å spacings likely to provide an HA binding site. In papain, several of these are located near the cystatin binding pocket suggesting that HA bound papain may have different TPI properties. This has been explored experimentally using standard cystatin inhibition curves conducted in stopped time assays of papain in the presence or absence of a fixed amount of HA. Since a long term goal of this project is to optimize cystatin, the removal of the carboxyl groups from the protease inhibitory region while adding appropriately spaced carboxyl groups would be necessary to enhance HA binding. Our current studies provide a solid basis for such future endeavors.