STUDIES OF PAPAIN IN COMPLEX WITH EPOXYSUCCINYL-PEPTIDE INHIBITOR

M.Kozak¹, E.Jankowska², J.Ciarkowski², Z.Grzonka², M.Jaskólski^1,3

¹Deptartment of Crystallography, A.Mickiewicz University, Poznan, Poland,
²Deptartment of Chemistry. University of Gdansk, Poland, ³Center for Biocrystallographic Research, Institute of Bioorganic Chemistry, Polish Academy of Sciences, Poznan, Poland
mariusz@krystal.amu.edu.pl

Keywords: cysteine proteases, papain, cystatins, covalent inhibitors, oxirane, cryobiocrystallography, semiempirical calculations, molecular dynamics

Cysteine (or thiol) proteases are the major component of lysozomal proteolytic systems and in physiological conditions their activity is controlled by a complicated system of inhibitors. Disruption of this fine-tuned balance can lead to many serious conditions, including chronic inflammation, periodontal disease, pulmonary emphysema, muscular dystrophy, osteoporosis and tumor. The catalytic center of cysteine proteases consists of a triad of residues, C-H-N, the thiol group of which is the reactive nucleophile. In addition to protein inhibitors (cystatins) of thiol proteases, there are also small-molecule inhibitors which can be of natural (E64) or synthetic (e.g.E64c) origin. A member of the cystatin family, cystatin C, uses two loop segments and, importantly, the N-terminal segment to block the enzyme's active site cleft forming a tight but reversible complex. In contrast, E64-based inhibitors contain an oxirane ring which inactivates the enzyme by forming an irreversible covalent link with the SH nucleophile.

We have prepared a synthetic inhibitor which contains an oxirane moiety connected to a peptidic fragment (ZRLV) mimicking the N-terminal binding sequence of human cystatin C. The inhibitor was used to inactivate the thiol protease papain and the complex was subsequently crystallized in the historically first polymorphic form of the enzyme. The 1.9-A room-temperature structure determined for a freshly grown crystal revealed the covalent attachment of the inhibitor to Cys-25, its general placement in the direction of the unprimed sites of the enzyme and a degree of disorder that is proportional to the distance from the point of attachment. However, the S2 pocket, earlier postulated to be responsible for tight binding and specificity, is not penetrated by the inhibitor. We have also analyzed the conformational behavior of the inhibitor through molecular dynamics simulations of its complex with papain. The results indicate that only in close vicinity of the covalent attachment to Cys-25 is the conformation of the inhibitor well defined and allows for conserved association with the enzyme. Farther away, the conformations available to the inhibitor diverge widely and in particular the enzyme's S2 pocket does not seem to be capable of ordering the inhibitor.

The crystal structure was later redetermined for an aged (one year old) crystal using low-temperature data extending to 1.6 A resolution. Unexpectedly, the electron density maps indicate that the original inhibitor is no longer present in the active site which is now covalently modified by a short, most likely -CH₂CH₂OH, moiety. Considering the harsh conditions of the crystallization medium (pH 9.2, aminoethanol, 60% ethanol/methanol), a hypothetical secondary solid-state reaction has been proposed that leads to the removal of the original inhibitor from the active site. A possible mechanism of such reaction has been confirmed through semiempirical calculations using the PM3 method.

Research sponsored in part by KBN (grant 279/PO4/97/13).