Degradation of C1-inhibitor by plasmin: implications for the control of inflammatory processes.

BACKGROUND: A correct balance between protease and inhibitor activity is critical in the maintenance of homoeostasis; excessive activation of enzyme pathways is frequently associated with inflammatory disorders. Plasmin is an enzyme ubiquitously activated in inflammatory disorder, and C1-inhibitor (C1-Inh) is a pivotal inhibitor of protease activity, which is particularly important in the regulation of enzyme cascades generated in plasma. The nature of the interaction between plasmin and C1-Inh is poorly understood. MATERIALS AND METHODS: C1-Inh was immunoadsorbed from the plasma of normal individuals (n = 21), from that of patients with systemic lupus erythematosus (n = 18) or adult respiratory distress syndrome (n = 9), and from the plasma and synovial fluid of patients with rheumatoid arthritis (n = 18). As plasmin is a putative enzyme responsible for C1-Inh was examined using SDS-PAGE. In addition, peptides cleaved from C1-Inh by plasmin were isolated and sequenced and the precise cleavage sites determined from the known primary sequence of C1-Inh. Homology models of C1-Inh were then constructed. RESULTS: Increased levels of cleaved and inactivated C1-Inh were found in each of the inflammatory disorders examined. Through SDS-PAGE analysis it was shown that plasmin rapidly degraded C1-Inh in vitro. The pattern of C1-Inh cleavage seen in vivo in patients with inflammatory disorders and that produced in vitro following incubation with plasmin were very similar. Homology models of C1-Inh indicate that the majority of the plasmin cleavage sites are adjacent to the reactive site of the inhibitor. CONCLUSIONS: This study suggests that local C1-Inh degradation by plasmin may be a central and critical event in the loss of protease inhibition during inflammation. These findings have important implications for our understanding of pathogenic mechanisms in inflammation and for the development of more effectively targeted therapeutic regimes. These findings may also explain the efficacy of anti-plasmin agents in the treatment of C1-Inh deficiency states, as they may diminish plasmin-mediated C1-Inh degradation.     

Primary structure of the reactive site of human C1-inhibitor

Human C1-inhibitor (C1-Inh) forms an equimolar complex with complement proteinase C1s that is resistant to dissociation by sodium dodecyl sulfate. The formation of this stable complex results in the cleavage of a peptide bond near the carboxyl terminus of the inhibitor and, whereas the bulk of C1-Inh remains covalently bound to the light chain of C1s, the postcomplex inhibitor peptide can be isolated under denaturing conditions. We have sequenced the amino-terminal region of this peptide and deduced that it represents the carboxyl-terminal side of the reactive site of C1-Inh. Limited proteolysis of C1-Inh by Crotalus atrox protease results in an active derivative lacking an amino-terminal peptide of 36 residues. Further proteolysis of this derivative with Pseudomonas aeruginosa elastase inactivates the inhibitor and a peptide is released. The amino-terminal sequence of this peptide overlaps with that of the postcomplex peptide and indicates that the residue imparting primary specificity to the inhibitor is arginine.     

The functional integrity of the serpin domain of C1-inhibitor depends on the unique N-terminal domain,as revealed by a pathological mutant

C1-inhibitor (C1-Inh) is a serine protease inhibitor (serpin) with a unique, non-conserved N-terminal domain of unknown function. Genetic deficiency of C1-Inh causes hereditary angioedema. A novel type of mutation (Delta 3) in exon 3 of the C1-Inh gene, resulting in deletion of Asp62-Thr116 in this unique domain, was encountered in a hereditary angioedema pedigree. Because the domain is supposedly not essential for inhibitory activity, the unexpected loss-of-function of this deletion mutant was further investigated. The Delta 3 mutant and three additional mutants starting at Pro76, Gly98, and Ser115, lacking increasing parts of the N-terminal domain, were produced recombinantly. C1-Inh76 and C1-Inh98 retained normal conformation and interaction kinetics with target proteases. In contrast, C1-Inh115 and Delta 3, which both lack the connection between the serpin and the non-serpin domain via two disulfide bridges, were completely non-functional because of a complex-like and multimeric conformation, as demonstrated by several criteria. The Delta 3 mutant also circulated in multimeric form in plasma from affected family members. The C1-Inh mutant reported here is unique in that deletion of an entire amino acid stretch from a domain not shared by other serpins leads to a loss-of-function. The deletion in the unique N-terminal domain results in a "multimerization phenotype" of C1-Inh, because of diminished stability of the central beta-sheet. This phenotype, as well as the location of the disulfide bridges between the serpin and the non-serpin domain of C1-Inh, suggests that the function of the N-terminal region may be similar to one of the effects of heparin in antithrombin III, maintenance of the metastable serpin conformation.     

Effect of reactive site loop elongation on the inhibitory activity of C1-inhibitor

The serine protease inhibitor C1-Inhibitor (C1-Inh) inhibits several complement- and contact-system proteases, which play an important role in inflammation. C1-Inh has a short reactive site loop (RSL) compared to other serpins. RSL length determines the inhibitory activity of serpins. We investigated the effect of RSL elongation on inhibitory activity of C1-Inh by insertion of one or two alanine residues in the RSL. One of five mutants had an increased association rate with kallikrein, but was nevertheless a poor inhibitor because of a simultaneous high stoichiometry of inhibition (>10). The association rate of the other variants was lower than that of wild-type C1-Inh. These data suggest that the relatively weak inhibitory activity of C1-Inh is not the result of its short RSL. The short RSL of C1-Inh has, surprisingly, the optimal length for inhibition.     

Changes in protein conformation and stability accompany complex formation between human C1 inhibitor and C1-s

The fluorescence spectrum of C1 inhibitor (C1-Inh) in aqueous buffer has a maximum at 324 nm which shifts to 358 nm in 6.0 M guanidinium chloride (GdmC1), indicating that fluorescent tryptophans are buried in the native protein. When titrated with GdmC1, the fluorescence intensity, polarization, and emission maximum of C1-Inh and C1-s exhibited clear transitions which were more prominent than those of the enzyme-inhibitor complex. Two of the variables (intensity and emission maximum) suggest biphasic unfolding of C1-Inh. Differential absorption measurements and sodium iodide quenching of intrinsic fluorescence were consistent with a net increase in the exposure of tryptophans and tyrosines upon complex formation. This reaction, i.e., complex formation, was also accompanied by an increase in the ability to enhance the fluorescence of the hydrophobic probe 8-anilino-1-naphthalenesulfonate. Fluorescence assays of heat denaturation showed transitions at 40 and 52 degrees C for C1-s and at 60 degrees C for C1-Inh whereas there was no detectable melting transition for the complex. Similarly, differential scanning calorimetric measurements revealed transitions at 42, 52, and 62 degrees C for C1-s and one transition at 60 degrees C for C1-Inh, with no major transitions detectable for the complex. The ratio of the calorimetric enthalpy to the apparent van't Hoff enthalpy for thermal unfolding of C1-Inh was 1.6. Taken together, these results suggest that C1-Inh and C1-s are each composed of at least two independently unfolding domains and that complex formation, which involves conformational change, yields a protein substantially more stable than either component alone.     

The structure of a Michaelis serpin-protease complex.

Serine protease inhibitors (serpins) regulate the activities of circulating proteases. Serpins inhibit proteases by acylating the serine hydroxyl at their active sites. Before deacylation and complete proteolysis of the serpin can occur, massive conformational changes are triggered in the serpin while maintaining the covalent linkage between the protease and serpin. Here we report the structure of a serpin-trypsin Michaelis complex, which we visualized by using the S195A trypsin mutant to prevent covalent complex formation. This encounter complex reveals a more extensive interaction surface than that present in small inhibitor-protease complexes and is a template for modeling other serpin-protease pairs. Mutations of several serpin residues at the interface reduced the inhibitory activity of the serpin. The serine residue C-terminal to the scissile peptide bond is found in a closer than usual interaction with His 57 at the active site of trypsin.     

The Croonian Lecture, 1980. The complex proteases of the complement system

The assembly and activation of the early components of complement, after their interaction with antibody-antigen complexes, are described in terms of the structures of the different proteins taking part. C1q, a molecule of unique half collagen--half globular structure, binds to the second constant domain of the antibody molecules through its six globular heads. A tetrameric complex of C1r2-C1s2 binds to the collagenous tails and leads to formation of the serine-type proteases C1r and C1s. C1s activates C4, which forms a covalent bond between its alpha' chain and the Fab section of the antibody. C2 is also activated by C1s and associates with the bound C4 molecule to form C42, a labile protease that activates C3, but which loses activity as the C2 peptide chains dissociate from C4. C2, by analogy with factor B, the equivalent component of the alternative pathway of activation, appears to be a novel type of serine protease with a similar catalytic site but different activation mechanism to the serine proteases that have been described previously.     

Identification of the minimum region of Bordetella pertussis Vag8 required for interaction with C1 inhibitor

An autotransporter of Bordetella pertussis, virulence-associated gene 8 (Vag8), binds and inactivates the complement regulator, C1 inhibitor (C1-Inh), and plays a role in evasion of the complement system. However, the molecular interaction between Vag8 and C1-Inh remains unclear. Here, we localized the minimum region of Vag8 required for interaction with C1-Inh by examining the differently truncated Vag8 derivatives for the ability to bind and inactivate C1-Inh. The truncated Vag8 containing amino-acid residues 102–548, but not 102–479 and 202–648, showed the full activity of intact Vag8, suggesting that the separate 102–202 and 548–648 amino-acid regions of Vag8 mediate the interaction with C1-Inh.     

Mechanism of action of anti-C1-inhibitor autoantibodies: prevention of the formation of stable C1s-C1-inh complexes.

BACKGROUND: Acquired C1-inhibitor (C1-inh) deficiency is usually associated with the presence of circulating C1-inh autoantibodies. These autoantibodies have been shown previously to bind to two synthetic peptides corresponding to C1-inh amino acid residues 438-449 (peptide 2) and 448-459 (peptide 3) but not to peptide 1 (residues 428-440). MATERIALS AND METHODS: Affinity-purified C1-inh autoantibodies from two patients with acquired C1-inh deficiency were studied for their effects on the inhibition of C1s activity by C1-inh using SDS-PAGE and hydrolysis of a synthetic ester. RESULTS: Functional studies confirmed that the anti-C1-inh autoantibodies abrogated C1-inh activity, and their maximum effect was produced when the concentrations of C1-inh and autoantibody were approximately equimolar. The autoantibodies prevent the formation of the C1s-C1-inh complex, but they do not dissociate the preformed complex, suggesting that the autoantibodies act prior to the formation of the enzyme-inhibitor complex. In the presence of autoantibodies, C1s cleaves C1-inh, and a stable covalent bond between C1s and C1-inh does not form. Peptides 2 and 3, but not peptide 1 inhibited autoantibody activity, thus C1-inh inhibitory activity for C1s was expressed fully. CONCLUSIONS: Our data indicate that the anti-C1-inh autoantibodies convert C1-inh to a substrate by preventing the formation of the stable covalent protease-serpin complex. The data also suggest a possible therapeutic use for peptides 2 and 3 or their derivatives in the management of patients with type II acquired angioedema (AAE).     

Sulfated galactan is a catalyst of antithrombin-mediated inactivation of alpha-thrombin

Novel compounds presenting anticoagulant activity, such as sulfated polysaccharides, open new perspectives in medicine. Elucidation of the molecular mechanism behind this activity is desirable by itself, as well as because it allows for the design of novel compounds. In the present study, we investigated the action of an algal sulfated galactan, which potentiates alpha-thrombin inactivation by antithrombin. Our results indicate the following: 1) both the sulfated galactan and heparin potentiate protease inactivation by antithrombin at similar molar concentrations, however they differ markedly in the molecular size required for their activities; 2) this galactan interacts predominantly with exosite II on alpha-thrombin and, similar to heparin, catalyzes the formation of a covalent complex between antithrombin and the protease; 3) the sulfated galactan has a higher affinity for alpha-thrombin than for antithrombin. We propose that the preferred pathway of sulfated galactan-induced inactivation of alpha-thrombin by antithrombin starts with the polysaccharide binding to the protease through a high-affinity interaction. Antithrombin is then added to the complex and the protease is inactivated by covalent interactions. Finally, the antithrombin-alpha-thrombin covalent complex dissociates from the polysaccharide chain. This mechanism resembles the action of heparin with low affinity for antithrombin, as opposed to heparin with high affinity for serpin.     

Quantitative analysis of the interaction between immune complex and C1q complement subcomponent. The role of interdomain interactions in rabbit IgG in binding of C1q to immune precipitates.

A novel method was developed for the analysis of the interaction of large multivalent ligands with surfaces (matrices) to analyse the binding of complement subcomponent C1q to immune precipitates. Our new evaluation method provides quantitative data characteristic of the C1q-immune-complex interaction and of the structure of the immune complex as well. To reveal the functional role of domain-domain interactions in the Fc part of IgG the binding of C1q to different anti-ovalbumin IgG-ovalbumin immune complexes was studied. Immune-complex precipitates composed of rabbit IgG in which the non-covalent or covalent bonds between the heavy chains had been eliminated were used. Non-covalent bonds were abolished by splitting off the CH3 domains, i.e. by using Facb fragments, and the covalent contact was broken by reduction and alkylation of the single inter-heavy-chain disulphide bond. The quantitative analysis of the binding curves provides a dissociation constant (K) of 200 nM for the interaction between C1q and immune precipitate formed from native IgG. Surprisingly, for immune precipitates composed of Facb fragments or IgG in which the inter-heavy-chain disulphide bond had been selectively reduced and alkylated, stronger binding (K = 30 nM) was observed. In this case, however, changes in the structure of the immune-complex matrix were also detected. These structural changes may account for the strengthening of the C1q-immune-complex interaction, which can be strongly influenced by the flexibility and the binding-site pattern of the immune-complex precipitates. These results suggest that domain-domain interactions in the Fc part of IgG affect the segmental mobility of IgG molecules and the spatial arrangement of the immune-complex matrix rather than the affinity of individual C1q-binding sites on IgG.     

The binding of human complement component C4 to antibody-antigen aggregates.

The binding of human complement component C4 to antibody-antigen aggregates and the nature of the interaction have been investigated. When antibody-antigen aggregates with optimal C1 bound are incubated with C4, the C4 is rapidly cleaved to C4b, but only a small fraction (1-2%) is bound to the aggregates, the rest remaining in the fluid phase as inactive C4b. It has been found that C4b and th antibody form a very stable complex, due probably to the formation of a covalent bond. On reduction of the C4b-immunoglobulin G (IgG) complex, the beta and gamma chains, but not the alpha' chain, of C4b are released together with all the light chain, but only about half of the heavy chain of IgG. The reduced aggregates contain two main higher-molecular-weight complexes, one shown by the use of radioactive components to contain both IgG and C4b and probably therefore the alpha' chain of C4b and the heavy chain of IgG, and the other only C4b and probably an alpha' chain dimer. The aggregates with bound C1 and C4b show maximal C3 convertase activity, in the presence of excess C2, when the alpha'-H chain component is in relatively highest amounts. When C4 is incubated with C1s in the absence of aggregates, up to 15% of a C4b dimer is formed, which on reduction gives an alpha' chain complex, probably a dimer. The apparent covalent interaction between C4b and IgG and between C4b and other C4b molecules cannot be inhibited by iodoacetamide and hence cannot be catalysed by transglutaminase (factor XIII). The reaction is, however, inhibited by cadaverine and putrescine and 14C-labelled putrescine is incorporated into C4, again by a strong, probably covalent, bond. It is suggested that a reactive group, possibly an acyl group, is generated when C4 is activated by C1 and that this reactive group can react with IgG, with another C4 molecule, or with water.     

Recombinant human C1-inhibitor produced in Pichia pastoris has the same inhibitory capacity as plasma C1-inhibitor

Therapeutic application of the serpin C1-inhibitor (C1-Inh) in inflammatory diseases like sepsis, acute myocardial infarction and vascular leakage syndrome seems promising, but large doses may be required. Therefore, a high-yield recombinant expression system for C1-Inh is very interesting. Earlier attempts to produce high levels of C1-Inh resulted in predominantly inactive C1-Inh. We describe the high yield expression of rhC1-Inh in Pichia pastoris, with 180 mg/l active C1-Inh at maximum. On average, 30 mg/l of 80-100% active C1-Inh was obtained. Progress curves were used to study the interaction with C1s, kallikrein, coagulation factor XIIa and XIa, and demonstrated that rhC1-Inh had the same inhibitory capacity as plasma C1-Inh. Structural integrity, as monitored via heat stability, was comparable despite differences in extent and nature of glycosylation. We conclude that the P. pastoris system is capable of high-level production of functionally and structurally intact human C1 inhibitor.     

Disulfide Bond Formation Between Nerve Growth Factor and the Nerve Growth Factor Receptor from Embryonic Sensory Neurons

Recent studies with sympathetic neurons using radiolabeled nerve growth factor have indicated that a high-molecular-weight covalent complex is formed. This complex is between the nerve growth factor and the high-affinity (type I) receptor and occurs through the formation of a disulfide bond. Studies presented in the present article demonstrate a similar complex is formed on chicken embryonic sensory neurons. The formation of this complex is inhibited by the addition of unlabeled nerve growth factor, metabolic energy inhibitors (dinitrophenol and NaF), and of sulfhydryl reagents. On the other hand, formation of this complex is not inhibited by temperature, or by the addition of insulin or epidermal growth factor. The receptor involved in the covalent complex formation is the high-affinity (type I) receptor. The molecular weight of this complex is approximately 232,000 daltons. Evidence indicates that this covalent complex may be required for the biological activity of the nerve growth factor.     

Sulfated galactan is a catalyst of antithrombin-mediated inactivation of α-thrombin

Novel compounds presenting anticoagulant activity, such as sulfated polysaccharides, open new perspectives in medicine. Elucidation of the molecular mechanism behind this activity is desirable by itself, as well as because it allows for the design of novel compounds. In the present study, we investigated the action of an algal sulfated galactan, which potentiates α-thrombin inactivation by antithrombin. Our results indicate the following: 1) both the sulfated galactan and heparin potentiate protease inactivation by antithrombin at similar molar concentrations, however they differ markedly in the molecular size required for their activities; 2) this galactan interacts predominantly with exosite II on α-thrombin and, similar to heparin, catalyzes the formation of a covalent complex between antithrombin and the protease; 3) the sulfated galactan has a higher affinity for α-thrombin than for antithrombin. We propose that the preferred pathway of sulfated galactan-induced inactivation of α-thrombin by antithrombin starts with the polysaccharide binding to the protease through a high-affinity interaction. Antithrombin is then added to the complex and the protease is inactivated by covalent interactions. Finally, the antithrombin–α-thrombin covalent complex dissociates from the polysaccharide chain. This mechanism resembles the action of heparin with low affinity for antithrombin, as opposed to heparin with high affinity for serpin.     

Inhibition of Staphylococcus aureus cysteine proteases by human serpin potentially limits staphylococcal virulence

Bacterial proteases are considered virulence factors and it is presumed that by abrogating their activity, host endogenous protease inhibitors play a role in host defense against invading pathogens. Here we present data showing that Staphylococcus aureus cysteine proteases (staphopains) are efficiently inhibited by Squamous Cell Carcinoma Antigen 1 (SCCA1), an epithelial-derived serpin. The high association rate constant (k(ass)) for inhibitory complex formation (1.9×10(4) m/s and 5.8×10(4) m/s for staphopain A and staphopain B interaction with SCCA1, respectively), strongly suggests that SCCA1 can regulate staphopain activity in vivo at epithelial surfaces infected/colonized by S. aureus. The mechanism of staphopain inhibition by SCCA1 is apparently the same for serpin interaction with target serine proteases whereby the formation of a covalent complex result in cleavage of the inhibitory reactive site peptide bond and associated release of the C-terminal serpin fragment. Interestingly, the SCCA1 reactive site closely resembles a motif in the reactive site loop of native S. aureus-derived inhibitors of the staphopains (staphostatins). Given that S. aureus is a major pathogen of epithelial surfaces, we suggest that SCCA1 functions to temper the virulence of this bacterium by inhibiting the staphopains.     

Molecular mechanism analysis of Gloydius shedaoensis venom gloshedobin interaction with inhibitors by homology modeling

Snake venom thrombin-like enzymes (SVTLEs) are widely applied in the treatment of thrombotic diseases, however, the molecular mechanism of its inhibition by synthetic and natural proteinaceous inhibitors is not yet understood. Here we investigated effects of protease inhibitors including phenylmethylsulfonil fluoride (PMSF), benzamidine (BMD) and its derivates on the activity of recombinant gloshedobin, a SVTLE from the snake Gloydius shedaoensis. The molecular inhibition mechanism was postulated by separately docking inhibitors into three-dimensional model of gloshedobin using protein C activator from Agkistrodon contortrix contortrix venom (ACC-C, which bear 78% identity with gloshedobin) as template. The analysis indicated that the strongest inhibitor, PMSF, was via a covalent bond with the catalytic Ser195, while other inhibitors showing weaker inhibitory activity were via hydrogen bond with Ser195 or non-catalytic residues.     

beta-NGF-endopeptidase: structure and activity of a kallikrein encoded by the gene mGK-22

Mouse nerve growth factor (NGF) is cleaved at a histidine-methionine bond to release an NH2-terminal octapeptide (NGF1-8). The enzyme responsible, beta-NGF-endopeptidase, is structurally and functionally similar to gamma-NGF and epidermal growth factor-binding protein (EGF-BP) and cleaves mouse low molecular weight kininogen to produce bradykinin-like activity. These data have suggested that, like gamma-NGF and EGF-BP, beta-NGF-endopeptidase is a mouse glandular kallikrein. Evidence for a physiological role for NGF1-8 encouraged studies to further characterize the structure and function of this enzyme. Purified beta-NGF-endopeptidase migrated as a single band on isoelectric focusing and reducing SDS-polyacrylamide gels. As was expected, it removed NGF1-8 from NGF. Interestingly, enzymatic activity on an artificial substrate, and on NGF, was inhibited by NGF1-8 and by bradykinin. These studies further supported the view that beta-NGF-endopeptidase acts on both NGF and kininogen. The first 30 NH2-terminal amino acids of beta-NGF-endopeptidase were sequenced. This analysis demonstrated that the enzyme is encoded by the gene designated mGK-22 (Evans et al., 1987). The sequence of this gene corresponds to that of EGF-BP type A (Anundi et al., 1982; Drinkwater et al., 1987), and so studies were performed to determine whether or not beta-NGF-endopeptidase participates in EGF complex formation. Chromatographic and kinetic data gave no evidence that beta-NGF-endopeptidase is an EGF-binding protein. Our studies suggest that contamination of high molecular weight (HMW) EGF preparations with beta-NGF-endopeptidase erroneously led to earlier designation of the product of mGK-22 as an EGF-BP.(ABSTRACT TRUNCATED AT 250 WORDS)     

Interaction of thymidylate synthetase with 5-nitro-2'-deoxyuridylate

5-Nitro-2'-deoxyuridylate (NO2dUMP) is a potent mechanism-based inhibitor of dTMP synthetase. After formation of a reversible enzymeìnhibitor complex, there is a rapid first order loss of enzyme activity which can be protected against by the nucleotide substrate dUMP. From studies of model chemical counterparts and the NO2dUMPdTMP synthetase complex, it has been demonstrated that a covalent bond is formed between a nucleophile of the enzyme and carbon 6 of NO2dUMP. The covalent NO2dUMPènzyme complex is sufficiently stable to permit isolation on nitrocellulose membranes, and dissociates to give unchanged NO2-dUMP with a first order rate constant of 8.9 x 10(-3) min-1. Dissociation of the complex formed with [6-3H]NO2dUMP shows a large alpha-secondary isotope effect of 19%, verifying that within the covalent complex, carbon 6 of the heterocycle is sp3-hybridized. The spectral changes which accompany formation of the NO2dUMPènzyme complex support the structural assignment and, when used to tritrate the binding sites, demonstrate that 2 mol of NO2dUMP are bound/mol of dimeric enzyme. The interaction of NO2dUMP with dTMP synthetase is quite different than that of other mechanism-based inhibitors such as 5-fluoro-2'-deoxyuridylate in that it neither requires nor is facilitated by the concomitant interaction of the folate cofactor, 5,10-CH2-H4folate, and that the covalent complex formed is unstable to protein denaturants.     

Refined crystal structure of the complex of subtilisin BPN' and Streptomyces subtilisin inhibitor at 1.8 A resolution

The crystal structure of subtilisin BPN' complexed with a proteinaceous inhibitor SSI (Streptomyces subtilisin inhibitor) was refined at 1.8 A resolution to an R-factor of 0.177 with a root-mean-square deviation from ideal bond lengths of 0.014 A. The work finally established that the SSI-subtilisin complex is a Michaelis complex with a distance between the O gamma of active Ser221 and the carbonyl carbon of the scissile peptide bond being an intermediate value between a covalent bond and a van der Waals' contact, 2.7 A. This feature, as well as the geometry of the catalytic triad and the oxyanion hole, is coincident with that found in other highly refined crystal structures of the complex of subtilisin Novo, subtilisin Carlsberg, bovine trypsin or Streptomyces griseus protease B with their proteinaceous inhibitors. The enzyme-inhibitor beta-sheet interaction is composed of two separate parts: that between the P1-P3 residues of SSI and the 125-127 chain segment (the "S1-3 site") of subtilisin and that between the P4-P6 residues of SSI and th 102-104 chain segment (the "S4-6 site") of subtilisin. The latter beta-interaction is unique to subtilisin. In contrast, the beta-sheet interaction previously found in the complex of subtilisin Novo and chymotrypsin inhibitor 2 or in the complex of subtilisin Carlsberg and Eglin C is distinct from the present complex in that the two types of beta-interactions are not separate. As for the flexibility of the molecules comprising the present complex, the following observations were made by comparing the B-factors for free and complexed SSI and comparing those for free and complexed subtilisin BPN'. The rigidification of the component molecules upon complex formation occurs in a very localized region: in SSI, the "primary" and "secondary" contact regions and the flanking region; in subtilisin BPN', the S1-3 and S4-6 sites and the flanking region.