Chemical-enzymatic replacement of Ile64 in the reactive site of soybeen trypsin inhibitor (Kunitz).

All the reactive amino groups in soybean trypsin inhibitor (Kunitz) were protected by guanidination of 9 out of 10 lysyl residues with O-methylisourea and by carbamoylation of the NH2 terminal Asp with potassium cyanate. This derivative was converted to modified inhibitor (Arg63-Ile64 reactive site peptide bond hydrolyzed) by incubation with trypsin at pH 3. The NH2 terminal of Ile64 was allowed to react with phenyl isothiocyanate to produce inactive phenylthiocarbamoyl-modified inhibitor. Treatment with trifluoroacetic acid formed the anilinothiazolinone of Ile64 yielding des-Ile64-modified inhibitor. After renaturation and purification, this material coelectrophoresed with modified inhibitor but did not form a stable complex with trypsin. Incubation with tert-butyloxycarbonyl-(amino acid)-N-hydroxysuccinimide esters yielded [tert-butyloxycarbonyl-(amino acid64)]-modified inhibitor. The tert-butyloxycarbonyl protective group was removed in trifluoroacetic acid. After renaturation, active [amino acid64]-modified inhibitors were obtained for Ile64, Ala64, Leu64, and Gly64 replacements. The resynthesis of the reactive-site peptide bound by kinetic control dissociation of the trypsin-inhibitor complex yielded fully active [Ala64]-virgin inhibitor. Thus, soybean trypsin inhibitor (Kunitz) has been shown to tolerate the replacement of the P1' residue with retention of activity. The importance of P1' residues in the function of protein proteinase inhibitors is discussed.     

Conformational studies of [Abu(3, 11)]-SFTI-1, an analogue of the trypsin inhibitor isolated from sunflower seeds.

With only 14 amino acid residues, the trypsin inhibitor SFTI-1 is the smallest naturally occurring serine proteinase inhibitor. It consists of two cyclic fragments (with head-to-tail cyclization and a disulfide bridge). In our previous paper, we showed that the removal of the disulfide bridge produced 2.4-fold lower activity. Here, we present the total conformational analysis of the [Abu(3, 11)]-SFTI-1 analog by means of 2D NMR spectroscopy in conjunction with theoretical methods. The peptide was synthesized by Fmoc SPPS. It was cyclized with PyBop and DIPEA in DMF. The NMR studies were performed in DMSO-d(6) at 303 K. Conformations of the peptide studied were calculated by the following three approaches: distance geometry (DG), molecular dynamics (MD) and determination of the statistical weights of conformations. The first two algorithms use a CHARMM force field, whereas the last uses an ECEPP/3 force field. Our calculations resulted in three sets of conformers with 7, 9 and 6 representatives, respectively. All our results were compared with published ones. It was found that the peptide has an ill-defined structure. Despite its conformational flexibility, the binding loop (3-11 fragment) displayed geometry similar to the corresponding fragments of the other SFTI-1 analogs and to the inhibitor itself. Furthermore, the peptide bond between the Ile7 and Pro8 residues adopts cis geometry, which is essential for inhibitory activity. Copyright (c) 2008 European Peptide Society and John Wiley & Sons, Ltd.     

Reconstitution of new inhibitors through mutual exchange of Ile(64)-Met(84) peptides of soybean trypsin inhibitors, Tia and Tib

Singly modified soybean trypsin inhibitors (STIs), Tia* [Tia cleaved at Arg(63)-Ile(64)] and Tib* [Tib cleaved at Arg(63)-Ile(64)], were prepared by limited proteolysis with trypsin at pH 3.0. These singly modified inhibitors were further modified to yield doubly modified inhibitors, Tia** and Tib**, by limited proteolysis with subtilisin BPN', which cleaved the Met(84)-Leu(85) bonds of Tia* and Tib*, respectively. The doubly modified inhibitors could be separated into two parts: protein moiety A and peptide moiety a (IRFIAEGHPLSLKFDS-FAVIM) for Tia**, and protein moiety B and peptide moiety b (IRFIAEGNPLRLKFDS-FAVIM) for Tib**. These protein and peptide moieties showed no trypsin inhibitory activities alone. However, the inhibitors can be reconstituted through the mutual exchange of the protein and peptide moieties isolated from STIs. The reconstituted inhibitor which has tyrosine at position 62 and histidine at position 71 shows the highest inhibitory activity. Its Ki value for bovine trypsin is around 10(-10) M, which is almost the same as that of Tia for bovine trypsin. The inhibitor possessing either tyrosine at position 62 or histidine at position 71 exhibits a Ki value of around 10(-9) M, which is between those of Tia and Tib. The inhibitor having phenylalanine and asparagine at positions 62 and 71, respectively, shows the weakest inhibitory activity of around 10(-8) M similar to that of Tib for bovine trypsin.     

Chemical-enzymatic insertion of an amino acid residue in the reactive site of soybean trypsin inhibitor (Kunitz).

Modified (Arg63-Ile64 reactive-site peptide bond hydrolyzed) soybean trypsin inhibitor (Kunitz) with all reactive amino groups, except that of Ile64, protected was described in the preceding paper (Kowalski, D., and Laskowski, M., Jr. (1976), Biochemistry, preceding paper in this issue). Treatment of this inhibitor with tert-butyloxycarbonyl-Ala- and tert-butyloxycarbonyl-Ile-N-hydroxy-succinimide esters yields inactive endo-tert-butyloxycarbonyl-Ala63A-and endo-tert-butyloxycarbonyl-Ile63A-modified inhibitors. The tert-butyloxycarbonyl groups were removed by treatment of the proteins with trifluoroacetic acid. After renaturation and purification, the resultant endo-Ala63A- and endo-Ile63A-modified inhibitors co-electrophorese with modified inhibitor both on disc gels (pH 9.4) and sodium dodecyl sulfate gels (after reduction of disulfide bonds) and show end groups corresponding to the 63A residue. These derivatives fail to form stable complexes with trypsin, extending the previous observation (Kowalski, D., and Laskowski, M., Jr. (1972), Biochemistry 11, 3451) that acylation of the P1' residue in modified inhibitors leads to inactivation. However, the incubation of endo-Ala63A- and endo-Ile63A-modified inhibitors with trypsin at pH 6.5 leads to the synthesis of the Arg63-Ala63A and Arg63-Ile63A peptide bonds in 4% yield. This is very close to the yield anticipated from a semiquantitative theory for the value of the equilibrium constant for reactive-site peptide bond. An alternative chemical method of insertion is also described. Controlled treatment of modified inhibitor with the N-carboxyanhydride of Glu produced inactive endo-Glu63A-modified inhibitor. Incubation of this inactive derivative with trypsin at pH 6.5 leads to 16% synthesis of the Arg63-Glu63A peptide bond. The higher yield of single chain protein in this case is attributed to the influence of the negative charge of the Glu63A side chain. Thus, the insertion of an amino acid residue between the P1 and P1' residues in soybean trypsin inhibitor (Kunitz) converts a trypsin inhibitor into a trypsin substrate.     

Selection of low-molecular-mass trypsin and chymotrypsin inhibitors based on the binding loop of CMTI-III using combinatorial chemistry methods

Using a combinatorial chemistry approach, a decapaptide library containing the N-terminal fragment of trypsin inhibitor CMTI-III was synthesized by the solid-phase method. The peptide library was screened for trypsin and chymotrypsin inhibitory activity applying the iterative method in solution. Two decapeptides were selected and resynthesized for each enzyme. The association equilibrium constants ((1.1+/-0.2)x10(8) and (7.3+/-1.6)x10(7)) determined for peptides with trypsin inhibitory activity indicate that they are 3-4-fold less active than the CMTI inhibitors. On the other hand, they are significantly more effective as compared with the starting sequence. Two peptides selected as chymotrypsin inhibitors displayed about 10 times higher activity (1.7+/-0.4)x10(7) and (1.1+/-0.2)x10(7), respectively) than those monosubstituted in position P(1) of the CMTI-III analogue. Considering low molecular weight of peptides selected and the lack of conformational constraints in their structures, the results are promising. They are good templates as starting sequences for further selection of small, peptidomimetic proteinase inhibitors.     

A peptide from the eel pancreas with structural similarity to human pancreatic secretory trypsin inhibitor

The primary structure of a 61-amino-acid residue peptide from the pancreas of the European eel (Anguilla anguilla) has been established as E E K S G(5)L Y R K P(10)S C G E M(15)S A M H A(20)C P M N F(25)A P V C G(30)T D G N T(35)Y P N E C(40)S L C F Q(45)R Q N T K(50)T D I L I(55)T K D D R(60)C. There was no indication of microheterogeneity. This peptide shows structural similarity to pancreatic secretory trypsin inhibitors from several mammalian species and to a cholecystokinin-releasing peptide isolated from rat pancreatic juice. A comparison of the amino acid sequences of the peptides has identified a domain in the central region of the molecules that has been strongly conserved during evolution. In contrast, the amino acid sequence in the region corresponding to the reactive centre of the mammalian trypsin inhibitors is very poorly conserved in the eel peptide. The P1-P1' reactive site lysine-isoleucine (or arginine-isoleucine) bond in the mammalian trypsin inhibitors is replaced by a methionine-asparagine bond. This region does, however, show limited homology to the reactive centre of human alpha 1-protease inhibitor suggesting that the eel peptide may function as an inhibitor of other proteolytic enzymes in the pancreas.     

Evaluation of phage display system and leech-derived tryptase inhibitor as a tool for understanding the serine proteinase specificities

A small combinatorial library of LDTI mutants (5.2 x 10(4)) restricted to the P1-P4' positions of the reactive site was displayed on the pCANTAB 5E phagemid, and LDTI fusion phages were produced and selected for potent neutrophil elastase and plasmin inhibitors. Strong fusion phage binders were analyzed by ELISA on enzyme-coated microtiter plates and the positive phages had their DNA sequenced. The LDTI variants: 29E (K8A, I9A, L10F, and K11F) and 19E (K8A, K11Q, and P12Y) for elastase and 2Pl (K11W and P12N), 8Pl (I9V, K11W, and P12E), and 10Pl (I9T, K11L, and P12L) for plasmin were produced with a Saccharomyces cerevisiae expression system. New strong elastase and plasmin inhibitors were 29E and 2Pl, respectively. LDTI-29E was a potent and specific neutrophil elastase inhibitor K(i) =0.5 nM), affecting no other tested enzymes. LDTI-2Pl was the strongest plasmin inhibitor ( K(i) =1.7nM) in the LDTI mutant library. This approach allowed selection of new specific serine proteinase inhibitors for neutrophil elastase and plasmin (a thrombin inhibitor variant was previously described), from a unique template molecule, LDTI, a Kazal type one domain inhibitor, by only 2-4 amino acid replacements. Our data validate this small LDTI combinatorial library as a tool to generate specific serine proteinase inhibitors suitable for drug design and enzyme-inhibitor interaction studies.     

Design and chemical synthesis of a 32 residues chimeric microprotein inhibiting both trypsin and carboxypeptidase A

A chimeric peptide, 32 amino acids long, bearing two active sites, one inhibiting trypsin (Kd = 1.8 10(-9) M), one carboxypeptidase A (Kd = 3 10(-9) M), was designed and synthesized. It is a "squash inhibitor" (EETI II, 28 amino acids) elongated with the 4 amino acids from the C-terminus of the potato carboxypeptidase inhibitor. It has 3 disulfide bridges assembled in the particular knotted topology shared by the two inhibitors, by conotoxin omega, and possibly by E. coli enterotoxin ST1b.     

Identification of epitope-like consensus motifs using mRNA display

The mRNA display approach to in vitro protein selection is based upon the puromycin-mediated formation of a covalent bond between an mRNA and its gene product. This technique can be used to identify peptide sequences involved in macromolecular recognition, including those identical or homologous to natural ligand epitopes. To demonstrate this approach, we determined the peptide sequences recognized by the trypsin active site, and by the anti-c-Myc antibody, 9E10. Here we describe the use of two peptide libraries of different diversities, one a constrained library based on the trypsin inhibitor EETI-II, where only the six residues in the first loop were randomized (6.4 x 10(7) possible sequences, 6.0 x 10(11) sequences in the library), the other a linear-peptide library with 27 randomized amino acids (1.3 x 10(35) possible sequences, 2 x 10(13) sequences in the library). The constrained library was screened against the natural target of wild-type EETI, bovine trypsin, and the linear library was screened against the anti-c-myc antibody, 9E10. The analysis of selected sequences revealed minimal consensus sequences of PR(I,L,V)L for the first loop of EETI-II and LISE for the 9E10 epitope. The wild-type sequences, PRILMR for the first loop of EETI-II and QKLISE for the 9E10 epitope, were selected with the highest frequency, and in each case the complete wild-type epitope was selected from the library.     

Pancreatic proteolytic enzymes of ostrich purified on immobilized protein inhibitors. Characterization of a new form of chymotrypsin (Chtr1)

Four forms of chymotrypsin (Chtr1, Chtr2, Chtr3, Chtr4), one form of trypsin and one form of elastase were purified from a slightly alkaline extract of ostrich (Struthio camelus) pancreas. The zymogens in the crude extract were activated with immobilized trypsin and then separated by affinity chromatography using immobilized inhibitors and ion exchange chromatography. One of the purified forms of chymotrypsin (Chtr1) exhibited an unusual interaction with the highly selective protein trypsin inhibitor from Cucurbita maxima (CMTI). Interactions with other protein trypsin inhibitors such as basic pancreatic trypsin inhibitor (BPTI), soybean trypsin inhibitor (STI), trypsin inhibitors from Cyclanthera pedata (CyPTI), Cucurbita pepo (CPTI), Cucurbita pepo var. giramontia (CPGTI) and Linum usitatissimum (LUTI) were also investigated. This study demonstrated the affinity of Chtr1 to inhibitors containing Arg at P1 position. Studies of substrate specificity of Chtr1 using oxidized B-chain of insulin revealed four susceptible bonds: Tyr15-Leu16, Phe24-Phe25, Phe25-Tyr26 and, surprisingly, Arg22-Gly23. The amino acid composition, as well as the first 13 residues of the N-terminal amino acid sequence, was determined. Studies of ostrich elastase showed that it can interact with immobilized CMTI in the presence of 5 M NaCl. This unusual characteristic is reported for the first time and suggests that elastase specificity depends on ionic strength. The kinetic constants K(M), k(cat) and k(cat)/K(M) for purified ostrich trypsin, chymotrypsin 4 and elastase were also determined.     

Synthesis and evaluation of the sunflower derived trypsin inhibitor as a potent inhibitor of the type II transmembrane serine protease, matriptase.

We report here the synthesis of a 14-amino acid long bicyclic peptide, previously isolated from sunflower seeds. This peptide, termed sunflower trypsin inhibitor (SFTI-1), is one of the most potent naturally occurring small-molecule trypsin inhibitors. In addition to inhibiting trypsin, the synthetic SFTI-1 is also a very potent inhibitor, with a K(i) of 0.92nM, of the recently identified epithelial serine protease, termed 'matriptase'.     

Identification of chymotrypsin inhibitors from a second-generation template assisted combinatorial peptide library.

In an earlier study (McBride JD, Freeman N, Domingo GJ, Leatherbarrow RJ. Selection of chymotrypsin inhibitors from a conformationally-constrained combinatorial peptide library. J. Mol. Biol. 1996; 259: 819-827) we described a resin-bound cyclic peptide library, constructed based on the sequence of the anti-tryptic reactive site loop of Bowman Birk Inhibitor (BBI), a proteinase inhibitor protein. This library was used to identify re-directed chymotrypsin inhibitors with Ki values as low as 17 nM. We have now extended this work by constructing an enhanced library in which a further position, at the P4 site of the inhibitor, has been randomized. This new library has variation at three target locations (P4, P1 and P2) within the inhibitory loop region, producing 8,000 variants. Screening this library allowed selection of new inhibitor sequences with Ki values as low as 3.4 nM. The success of this approach is reflected by the fact that the inhibition constant given by the selected peptide sequence is slightly lower than that reported against chymotrypsin for the most studied full length BBI protein, Soybean BBI 2-IV.     

Subcellular trafficking abnormalities of a prion protein with a disrupted disulfide loop

The recombinant phage antibody system pCANTAB 5E has been used to display functionally active leech-derived tryptase inhibitor (LDTI) on the tip of the filamentous M13 phage. A limited combinatorial library of 5.2 x 10(4) mutants was created with a synthetic LDTI gene, using a degenerated oligonucleotide and the pCANTAB 5E phagemid. The mutations were restricted to the P1-P4' positions of the reactive site. Fusion phages and appropriate host strains containing the phagemids were selected after binding to thrombin and DNA sequencing. The variants LDTI-2T (K8R, I9V, S10, K11W, P12A), LDTI-5T (K8R, I9V, S10, K11S, P12L) and LDTI-10T (K8R, I9L, S10, K11D, P12I) were produced with a Saccharomyces cerevisiae expression system. The new inhibitors, LDTI-2T and -5T, prolong the blood clotting time, inhibit thrombin (Ki 302 nM and 28 nM) and trypsin (Ki 6.4 nM and 2.1 nM) but not factor Xa, plasma kallikrein or neutrophil elastase. The variant LDTI-10T binds to thrombin but does not inhibit it. The relevant reactive site sequences of the thrombin inhibiting variants showed a strong preference for arginine in position P1 (K8R) and for valine in P1' (I9V). The data indicate further that LDTI-5T might be a model candidate for generation of active-site directed thrombin inhibitors and that LDTI in general may be useful to generate specific inhibitors suitable for a better understanding of enzyme-inhibitor interactions.     

Primary structure and functional properties of cobra (Naja naja naja) venom Kunitz-type trypsin inhibitor.

A trypsin inhibitor from the venom of the cobra Naja naja naja has been isolated by a single step of reverse-phase high-performance liquid chromatography. The protein strongly inhibits trypsin (Ki = 3.5 pM). The primary structure was determined by peptide analysis of the [14C]carboxymethylated inhibitor. The 57-residue polypeptide chain belongs to the family of Kunitz-type inhibitors, and exhibits 42% residue identity with bovine pancreatic trypsin inhibitor. The structure shows only 70% identity with the corresponding peptide from the Capa cobra (Naja nevia), establishing that the inhibitor molecule exhibits extensive variations. Functionally, a basic residue at position P3' correlates with strong inhibition.     

Amino acid sequence of winged bean (Psophocarpus tetragonolobus (L.) DC.) chymotrypsin inhibitor, WCI-3

The complete amino acid sequence of winged bean chymotrypsin inhibitor 3 (WCI-3) was determined by the conventional methods. WCI-3 consisted of 183 amino acid residues, but was heterogeneous in the carboxyl terminal region owing to the loss of one to four carboxyl terminal amino acid residues. The sequence of WCI-3 was highly homologous with those of soybean trypsin inhibitor Tia, winged bean trypsin inhibitor WTI-1, and Erythrina latissima trypsin inhibitor DE-3. One of the reactive site peptide bonds of WCI-3 was identified as Leu(65)-Ser(66), which was located at the same position as those of the other Kunitz-family leguminous proteinase inhibitors.     

Phage display selection of P1 mutants of BPTI directed against five different serine proteinases

The P1 position of protein inhibitors and oligopeptide substrates determines, to a large extent, association energy with many serine proteinases. To test the agreement of phage display selection with the existing thermodynamic data, a small library of all 20 P1 mutants of basic pancreatic trypsin inhibitor (BPTI) was created, fused to protein III, and displayed on the surface of M13 phage. The wild type of displayed inhibitor monovalently and strongly inhibited trypsin with an association constant of Ka = 3 x 10(11) M(-1). The library was applied to select BPTI variants active against five serine proteinases of different specificity (bovine trypsin and chymotrypsin, human leukocyte and porcine pancreatic elastases, human azurocidin). The results of enrichment with four proteinases agreed well with the available thermodynamic data. In the case of azurocidin, the phage display selection allowed determination of the P1 specificity of this protein with the following frequencies for selected P1 variants: 43% Lys, 36% Leu, 7% Met, 7% Thr, 7% Gln.     

Epidermal growth factor-urogastrone receptor: selective alteration in simian virus 40 transformed mouse fibroblasts

Comparative data on the properties of four thiol proteinase inhibitors, and of four serine proteinase inhibitors (two subtilisin and two trypsin inhibitors) isolated from seeds of Vigna are presented. They were similar in their molecular weights (5000–15,000) and dissociation constants (10^?8–10^?9m). The range of isoelectric points of the thiol proteinase inhibitors was 6.5 to 10.6, and of the serine proteinase inhibitors was 5.0 to 5.9. The amino acid compositions of one papain isoinhibitor, one of subtilisin, and one of trypsin are presented. Papain inhibitor A1 and subtilisin inhibitor 2a were low in cystine. All of the inhibitors were stable upon heating to 80 °C for 5 min at low pH. The subtilisin inhibitor did not bind to catalytically inactive subtilisin derivatives, whereas the papain inhibitor was stoichiometrically bound to the Hg or thioacetamide derivatives of papain. Incubation of the subtilisin inhibitor with catalytic amounts of subtilisin led to the formation of a modified form with the same inhibitor activity as the native inhibitor but with a different electrophoretic mobility. There was no indication of a similar modification of the papain inhibitor by papain. Separate sites are present on the trypsin-chymotrypsin inhibitors for trypsin and chymotrypsin. The papain inhibitors have the same binding sites for papain and ficin.     

Definition of peptide inhibitors from a synthetic peptide library by targeting gelatinase B/matrix metalloproteinase-9 (MMP-9) and TNF-α converting enzyme (TACE/ADAM-17)

Gelatinase B/matrix metalloproteinase-9 (MMP-9) is a regulatory and effector metalloproteinase in inflammation. TNF-α is an important proinflammatory cytokine and is released by the action of a Zn(2+)-containing converting enzyme (TACE/ADAM-17). Both metallo-enzymes play important roles during the development of shock syndromes. Combinatorial chemical synthesis and subsequent library deconvolution were previously used to define a peptide inhibitor (Regasepin1) acting, almost to the same degree, on neutrophil collagenase/MMP-8 and MMP-9 in vitro, and protecting mice against lethal endotoxinemia in vivo. We have now extended this approach by incorporating D-form amino acids and residues preferred by TACE. A new peptide library was designed and synthesized, and by deconvolution new peptide inhibitors were defined. These included a TACE-specific inhibitor, an MMP-9- specific inhibitor, and inhibitors for both enzymes.     

A synthetic 13-residue peptide designed to resemble the primary binding site of the basic pancreatic trypsin inhibitor

A peptide, AC-Pro-Cys-Lys-Ala-Arg-Ile-DPhe-Pro-Tyr-Gly-Gly-Cys-Arg-NH2, which resembles the binding site of the basic pancreatic trypsin inhibitor, has been prepared by solid-phase peptide synthesis. A partially protected peptide was first obtained from the solid-phase product by removal of all side-chain protecting groups except the acetamidomethyl (Acm) groups on the cysteines. This di-Acm-peptide was deprotected, with concomitant formation of the cyclic product, by treatment with I2 in AcOH. The cyclic 13-residue peptide is a reversible, competitive inhibitor of trypsin with a Ki (app) of 2 . 10(-6) M, but loses its inhibitory activity upon incubation with trypsin. The di-Acm-peptide precursor has a Ki (app) of 5 . 10(-5) M and is deactivated more rapidly by trypsin. The effectiveness of the 13-residue peptides as inhibitors is in part attributed to the conformation induced by the beta-turn directing the -DPhe-Pro portion of the sequence.     

What can the structures of enzyme-inhibitor complexes tell us about the structures of enzyme substrate complexes?

Proteinases perform many beneficial functions that are essential to life, but they are also dangerous and must be controlled. Here we focus on one of the control mechanisms: the ubiquitous presence of protein proteinase inhibitors. We deal only with a subset of these: the standard mechanism, canonical protein inhibitors of serine proteinases. Each of the inhibitory domains of such inhibitors has one reactive site peptide bond, which serves all the cognate enzymes as a substrate. The reactive site peptide bond is in a combining loop which has an identical conformation in all inhibitors and in all enzyme-inhibitor complexes. There are at least 18 families of such inhibitors. They all share the conformation of the combining loops but each has its own global three-dimensional structure. Many three-dimensional structures of enzyme-inhibitor complexes were determined. They are frequently used to predict the conformation of substrates in very short-lived enzyme-substrate transition state complexes. Turkey ovomucoid third domain and eglin c have a Leu residue at P(1). In complexes with chymotrypsin, these P(1) Leu residues assume the same conformation. The relative free energies of binding of P(1) Leu (relative to either P(1) Gly or P(1) Ala) are within experimental error, the same for complexes of turkey ovomucoid third domain, eglin c, P(1) Leu variant of bovine pancreatic trypsin inhibitor and of a substrate with chymotrypsin. Therefore, the P(1) Leu conformation in transition state complexes is predictable. In contrast, the conformation of P(1) Lys(+) is strikingly different in the complexes of Lys(18) turkey ovomucoid third domain and of bovine pancreatic trypsin inhibitor with chymotrypsin. The relative free energies of binding are also quite different. Yet, the relative free energies of binding are nearly identical for Lys(+) in turkey ovomucoid third domain and in a substrate, thus allowing us to know the structure of the latter. Similar reasoning is applied to a few other systems.