CRYSTAL COMPLEXES OF WILD TYPE AND MUTANT CI-2 INHIBITOR WITH SUBTILISIN CARLSBERG: THE STRUCTURAL BASIS FOR BINDING KINETICS AND A HIGH RESOLUTION STUDY. Wojciech R. Rypniewski¹, Claus von der Osten², Torben Halkier² & Keith S. Wilson^{1. 1}European Molecular Biology Laboratory, c/o DESY, Notkestrasse 85, D-22603 Hamburg, Germany; ²Novo Nordisk, Novo Alle, DK-2880 Bagsvaerd, Denmark.

Both wild type CI-2 and reactive loop mutant M59P form stable complexes with alcalase, a variant of subtilisin Carlsberg, but K_i values differ by 10⁵. The crystal structures of both complexes were solved using the synchrotron radiation on the EMBL beamlines at the DORIS synchrotron in Hamburg and examined to determine the basis of the differences in binding kinetics. As expected, the differences can be derived from the special structural properties of proline, but at first glance this was not clear and a simplistic explanation had to be abandoned. A detailed explanation followed from examining the high resolution structures.

The wild type structure has been solved at atomic resolution (1Å). The high quality and resolution of the data allowed a detailed analysis of this 32 kdal protein complex, with refined anisotropic temperature factors, direct visualisation of most hydrogen atoms, unrestrained refinement of some stereochemical parameters including peptide bond planarity and chiral volumes, and details of water structure and its interactions with the protein.