No abstract

The crystal structures of the molecular complexes between two serine proteinases and two of their protein inhibitors have been determined: subtilisin Carlsberg with the recombinant form of eglin-c from the leech Hirudo medicinalis and subtilisin Novo with chymotrypsin inhibitor 2 from barley seeds. The structures have been fully refined by restrained-parameter least-squares methods to crystallographic R factors (sigma[[Fo[ - [Fc[[/sigma[Fo[) of 0.136 at 1.8-A resolution and 0.154 at 2.1-A resolution, respectively. The 274 equivalent alpha-carbon atoms of the enzymes superpose with an rms deviation of 0.53 A. Sequence changes between the enzymes result in localized structural adjustments. Functional groups in the active sites superpose with an rms deviation of 0.19 A for 161 equivalent atoms; this close similarity in the conformation of active-site residues provides no obvious reason for known differences in catalytic activity between Carlsberg and Novo. Conformational changes in the active-site region indicate a small induced fit of enzyme and inhibitor. Some conformational differences are observed between equivalent active-site residues of subtilisin Carlsberg and alpha-chymotrypsin. Despite differences in tertiary architecture, most enzyme-substrate (inhibitor) interactions are maintained. Subtilisin Carlsberg has a rare cis-peptide bond preceding Thr211 (Gly211 in Novo). Both enzymes contain tightly bound Ca2+ ions. Site 1 is heptacoordinate with the oxygen atoms at the vertices of a pentagonal bipyramid. Site 2 in Carlsberg is probably occupied by a K+ ion in Novo. Conserved water molecules appear to play important structural roles in the enzyme interior, in the inhibitor beta-sheet, and at the enzyme-inhibitor interface. The 62 equivalent alpha-carbon atoms of the inhibitors superpose with an rms deviation of 1.68 A. Sequence changes result in somewhat different packing of the alpha-helix, beta-sheet, and reactive-site loop relative to each other. Hydrogen bonds and electrostatic interactions supporting the conformation of the reactive-site loop are conserved. The 24 main-chain plus C beta atoms of P4 to P1' overlap with an rms deviation of 0.19 A. Features contributing to the inhibitory nature of eglin-c and CI-2 are discussed.

Chymotrypsin inhibitor 2 (CI-2), a serine proteinase inhibitor from barley seeds, has been crystallized and its three-dimensional structure determined at 2.0-A resolution by the molecular replacement method. The structure has been refined by restrained-parameter least-squares methods to a crystallographic R factor (= sigma parallel Fo magnitude of-Fo parallel/sigma magnitude of Fo) o of 0.198. CI-2 is a member of the potato inhibitor 1 family. It lacks the characteristic stabilizing disulfide bonds of most other members of serine proteinase inhibitor families. The body of CI-2 shows few conformational changes between the free inhibitor and the previously reported structure of CI-2 in complex with subtilisin Novo [McPhalen, C.A., Svendsen, I., Jonassen, I., & James, M.N.G. (1985) Proc. Natl. Acad. Sci. U.S.A. 82, 7242-7246]. However, the reactive site loop has some significant conformational differences between the free inhibitor and its complexed form. The residues in this segment of polypeptide exhibit relatively large thermal motion parameters and some disorder in the uncomplexed form of the inhibitor. The reactive site bond is between Met-59I and Glu-60I in the consecutive sequential numbering of CI-2 (Met-60-Glu-61 according to the alignment of Svendsen et al. [Svendsen, I., Hejgaard, J., & Chavan, J.K. (1984) Carlsberg Res. Commun. 49, 493-502]). The network of hydrogen bonds and electrostatic interactions stabilizing the conformation of the reactive site loop is much less extensive in the free than in the complexed inhibitor.

No abstract