Design and Characterization of a Hybrid Miniprotein That Specifically Inhibits Porcine Pancreatic Elastase
2003; Elsevier BV; Volume: 278; Issue: 27 Linguagem: Inglês
10.1074/jbc.m212152200
ISSN1083-351X
AutoresKai Hilpert, Helga Wessner, Jens Schneider‐Mergener, Karin Welfle, Rolf Misselwitz, Heinz Welfle, Andreas C. Hocke, Stefan Hippenstiel, Wolfgang Höhne,
Tópico(s)Computational Drug Discovery Methods
ResumoStudying protease/peptide inhibitor interactions is a useful tool for understanding molecular recognition in general and is particularly relevant for the rational design of inhibitors with therapeutic potential. An inhibitory peptide (PMTLEYR) derived from the third domain of turkey ovomucoid inhibitor and optimized for specific porcine pancreatic elastase inhibition was introduced into an inhibitor scaffold to increase the proteolytic stability of the peptide. The trypsin-specific squash inhibitor EETI II from Ecballium elaterium was chosen as the scaffold. The resulting hybrid inhibitor HEI-TOE I (hybrid inhibitor from E. elaterium and the optimized binding loop of the third domain of turkey ovomucoid inhibitor) shows a specificity and affinity to porcine pancreatic elastase similar to the free inhibitory peptide but with significantly higher proteolytic stability. Isothermal titration calorimetry revealed that elastase binding of HEI-TOE I occurs with a small unfavorable positive enthalpy contribution, a large favorable positive entropy change, and a large negative heat capacity change. In addition, the inhibitory peptide and the hybrid inhibitor HEI-TOE I protected endothelial cells against degradation following treatment with porcine pancreatic elastase. Studying protease/peptide inhibitor interactions is a useful tool for understanding molecular recognition in general and is particularly relevant for the rational design of inhibitors with therapeutic potential. An inhibitory peptide (PMTLEYR) derived from the third domain of turkey ovomucoid inhibitor and optimized for specific porcine pancreatic elastase inhibition was introduced into an inhibitor scaffold to increase the proteolytic stability of the peptide. The trypsin-specific squash inhibitor EETI II from Ecballium elaterium was chosen as the scaffold. The resulting hybrid inhibitor HEI-TOE I (hybrid inhibitor from E. elaterium and the optimized binding loop of the third domain of turkey ovomucoid inhibitor) shows a specificity and affinity to porcine pancreatic elastase similar to the free inhibitory peptide but with significantly higher proteolytic stability. Isothermal titration calorimetry revealed that elastase binding of HEI-TOE I occurs with a small unfavorable positive enthalpy contribution, a large favorable positive entropy change, and a large negative heat capacity change. In addition, the inhibitory peptide and the hybrid inhibitor HEI-TOE I protected endothelial cells against degradation following treatment with porcine pancreatic elastase. Proteases can play a decisive role as indirect virulence factors promoting infection by viruses, bacteria, or parasites. Moreover, inappropriate or altered host protease activities are involved in many diseases, such as cancer, blood clotting, Alzheimer's disease, rheumatoid arthritis, arteriosclerosis, and pulmonary emphysema. Consequently protease inhibitors have a huge potential as therapeutic tools for treating any number of diverse diseases (1Helm K.v.d. Korant B.D. Cheronis J.C. Born G.V.R. Cuatrecasas P. Ganten D. Herken H. Starke K. Taylor P. Berlin Handbook of Experimental Pharmacology. Springer-Verlag, Heidelberg, Germany2000Google Scholar). A well known example is the generation of protease inhibitors against human immunodeficiency virus, type 1, protease (1Helm K.v.d. Korant B.D. Cheronis J.C. Born G.V.R. Cuatrecasas P. Ganten D. Herken H. Starke K. Taylor P. Berlin Handbook of Experimental Pharmacology. Springer-Verlag, Heidelberg, Germany2000Google Scholar). However, for each individual indication major efforts are required to find the appropriate specific, effective, stable, and non-toxic protease inhibitor. Apart from this, investigating inhibitor/protease interactions leads to a better understanding of protein molecular recognition and specificity in general. There are different methods and strategies to develop an inhibitor for a given protease. Previously, we described a method for characterizing and optimizing peptide inhibitor/protease interactions using cellulose-bound peptides (2Hilpert K. Hansen G. Wessner H. Schneider-Mergener J. Hohne W. J. Biochem. (Tokyo). 2000; 128: 1051-1057Google Scholar). We demonstrated an optimization process for a 9-mer peptide originating from the third domain of the turkey ovomucoid inhibitor, OMTKY3. 1The abbreviations used are: OMTKY3, third domain of turkey ovomucoid inhibitor; EETI, trypsin inhibitor from E. elaterium; HEI-TOE, hybrid inhibitor from E. elaterium and the optimized binding loop of the third domain of turkey ovomucoid inhibitor; PPE, porcine pancreatic elastase; MCEI, elastase inhibitor from M. charantia; ITC, isothermal titration calorimetry; HPLC, high pressure liquid chromatography; HLE, human leucozyte elastase; HUVEC, human umbilical cord vein endothelial cells; BPC, bovine pancreatic chymotrypsin; BPT, bovine pancreatic trypsin; ASA, accessible surface area; Suc, succinyl; pNA, p-nitroanilide; PIPES, 1,4-piperazinediethanesulfonic acid; MES, 4-morpholineethanesulfonic acid; MOPS, 4-morpholinepropanesulfonic acid. This 56-amino acid protein strongly inhibits a broad spectrum of serine proteases, e.g. bovine α-chymotrypsin with a Ki = 5.5 × 10–12m (3Bigler T.L. Lu W. Park S.J. Tashiro M. Wieczorek M. Wynn R. Laskowski Jr., M. Protein Sci. 1993; 2: 786-799Google Scholar). The peptide from the binding loop of OMTKY3 was optimized against porcine pancreatic elastase (PPE, EC 3.4.21.36). This 240-amino acid enzyme is a member of the S1 class of serine proteases. Its active site comprises eight subsites (S5–S3′; according to the nomenclature of Schechter and Berger (4Schechter I. Berger A. Biochem. Biophys. Res. Commun. 1967; 27: 157-162Google Scholar)) and cleaves the amide bond of amino acids with small hydrophobic side chains (5Bieth J.G. Barrett A.J. Rawlings N.D. Woessner Jr., F. Handbook of Proteolytic Enzymes. Academic Press, London1998: 42Google Scholar). Our optimization procedure led to a 7-mer peptide (PMTLEYR), which although showing a high specificity and affinity toward PPE, turned out to be unstable against proteolytic degradation by PPE. For more detailed investigation of the inhibitor/PPE interaction, e.g. by x-ray structural and biophysical analysis of corresponding complexes, as well as for medical applications, high proteolytic stability is an important prerequisite. One possible way to create a more stable inhibitor molecule is to introduce the inhibitory peptide sequence into a stable scaffold. Such scaffolds are provided by the squash-type inhibitors. Squash-type inhibitors are natural serine protease inhibitors occurring in plant seeds of Curcubitaceae. They consist of only ∼30 amino acids and have a rigid and stable structure cross-linked by three disulfide bridges. They are resistant to proteolytic cleavage and stable at elevated temperatures (for review, see Ref. 6Otlewski J. Krowarsch D. Biochemistry. 1996; 43: 431-444Google Scholar). EETI II from the squirting cucumber Ecballium elaterium is a well characterized squash-type inhibitor with a high affinity and specificity for trypsin (Kd of 1.2 × 10–12m) (7Favel A. Mattras H. Coletti-Previero M.A. Zwilling R. Robinson E.A. Castro B. Int. J. Pept. Protein Res. 1989; 33: 202-208Google Scholar). It was shown to fold into its native structure even if considerable amino acid exchanges and length variations are incorporated within its binding loop (8Wentzel A. Christmann A. Kratzner R. Kolmar H. J. Biol. Chem. 1999; 274: 21037-21043Google Scholar). Natural squash inhibitors that inhibit porcine pancreatic elastase are also known from the literature. The inhibitors MCEI I (9Hara S. Makino J. Ikenaka T. J. Biochem. (Tokyo). 1989; 105: 88-91Google Scholar) and MCEI II, III, and IV (10Hamato N. Koshiba T. Pham T.N. Tatsumi Y. Nakamura D. Takano R. Hayashi K. Hong Y.M. Hara S. J. Biochem. (Tokyo). 1995; 117: 432-437Google Scholar) were isolated from bitter gourd (Momordica charantia) and inhibit PPE with Ki values between 9.7 × 10–7 and 4.7 × 10–9m (9Hara S. Makino J. Ikenaka T. J. Biochem. (Tokyo). 1989; 105: 88-91Google Scholar, 10Hamato N. Koshiba T. Pham T.N. Tatsumi Y. Nakamura D. Takano R. Hayashi K. Hong Y.M. Hara S. J. Biochem. (Tokyo). 1995; 117: 432-437Google Scholar). We describe here the introduction of sequences, derived from the inhibitory peptide PMTLEYR, into the squash inhibitor EETI II, and the characterization of several hybrid squash inhibitor variants with respect to inhibition strength, specificity, and stability. The energetics of the inhibitor-elastase complex formation at different temperatures and in various buffer systems was studied using isothermal titration calorimetry (ITC). Furthermore, the effects of the optimized inhibitory peptide PMTLEYR and of one of the hybrid squash inhibitors (HEI-TOE I) on elastase-treated human endothelial cells were compared. Peptide Synthesis—The peptides and squash inhibitors were synthesized in reduced form on solid phase according to standard Fmoc (N-(9-fluorenyl)methoxycarbonyl) machine protocols using a multiple peptide synthesizer (Abimed, Langenfeld, Germany). After a purification step by reverse phase HPLC, the disulfide bridges were formed using charcoal according to a recently described cyclization protocol (11Volkmer-Engert R. Landgraf C. Schneider-Mergener J. J. Pept. Res. 1998; 51: 365-369Google Scholar). The degree of disulfide formation was monitored by measuring the decrease of free cysteine content with 5,5′-dithiobis(2-nitrobenzoic acid). The cyclization procedure was applied until no free cysteine was detectable. After disulfide formation another reverse phase HPLC purification step was performed. The purity of the final products was greater than 98%. Kiand Proteolytic Stability Determination—All enzymes and substrates were purchased from Serva (Heidelberg, Germany). The activity of the enzymes was measured by monitoring hydrolysis of the corresponding substrates on a recording spectrophotometer model UV-160A (Shimadzu, Duisburg, Germany). Varying the substrate (63 μm to 2.5 mm) and inhibitor concentrations (55 nm to 1 mm) the Ki values were determined in a 200-μl assay at 25 °C. For PPE, human leukocyte elastase (HLE), and trypsin a 0.1 m Tris buffer, pH 8.5, was used; in the case of chymotrypsin, a 0.1 m Tris buffer, pH 7.5, was used. The activity of chymotrypsin (Serva, Heidelberg, Germany) was determined using Suc-Ala-Ala-Phe-pNA (Serva, Heidelberg, Germany); for PPE, Suc-Ala-Ala-Ala-pNA; for HLE, Suc-Ala-Ala-Val-pNA; and for trypsin, Nα-benzoyl-dl-arginine-pNA, all solubilized in Me2SO. The proteolytic stability of the peptides was measured (determined via PPE activity) by preincubation of different inhibitor concentrations (10–500 μm) with PPE (15–150 units in 150-μl assay volume) for between 2 min and 8 days in 0.1 m Tris buffer, pH 8.5, at 25 °C. Isothermal Titration Experiments—Isothermal titrations were performed at 6.8, 11.5, 16.0 and 25 °C using a MicroCal MCS isothermal titration calorimeter (MicroCal Inc., Northampton, MA) as described recently (12Hahn M. Winkler D. Welfle K. Misselwitz R. Welfle H. Wessner H. Zahn G. Scholz C. Seifert M. Harkins R. Schneider-Mergener J. Hohne W. J. Mol. Biol. 2001; 314: 293-309Google Scholar, 13Welfle K. Misselwitz R. Sabat R. Volk H.D. Schneider-Mergener J. Reineke U. Welfle H. J. Mol. Recognit. 2001; 14: 89-98Google Scholar). PPE and HEI-TOE I were dialyzed against the following buffers at pH 6.0 containing 0.15 m NaCl, 50 mm PIPES, 50 mm MES, and 50 mm MOPS, which have ionization enthalpies at 7.0 °C of 11.1 kJ·mol–1, 15.2 kJ·mol–1, and 21.1 kJ·mol–1, respectively (14Fukada H. Takahashi K. Proteins. 1998; 33: 159-166Google Scholar). The ionization enthalpies at different temperatures were calculated using a heat capacity change, ΔCp0, of 0.019 kJ·mol–1·K–1 for PIPES (14Fukada H. Takahashi K. Proteins. 1998; 33: 159-166Google Scholar). The concentrations of PPE and HEI-TOE I were determined spectrophotometrically at 280 and 276 nm, respectively, using extinction coefficients of 5.2 × 104m–1·cm–1 (15Shotton D.M. Methods Enzymol. 1970; 19: 113-140Google Scholar) and 1.885 × 103m–1·cm–1 (calculated from the amino acid composition according to Gill and von Hippel (16Gill S.C. von Hippel P.H. Anal. Biochem. 1989; 182: 319-326Google Scholar)). In a typical titration experiment 5 μl of HEI-TOE I solution (around 560 μm) were titrated into 1.4 ml of PPE solution (around 12 μm). The experimental raw data were corrected by subtraction of the heat of dilution of the peptide, transformed into kJ·mol–1 of injectant, then fitted to a model of identical and independent binding sites using the ORIGIN software package provided with the calorimeter (MicroCal Inc.). Enthalpy changes, ΔHbind0, of complex formation were determined by correction of experimentally estimated enthalpy changes, ΔHexp0, for ΔHion0 of the respective buffer. The total entropy ΔStot0 can be dissected according to Equation 1, ΔStot0=ΔSsolv0+ΔSconf0+ΔSrt0(Eq. 1) in terms of the main contributions of solvation entropy, ΔSsolv0, conformational entropy, ΔSconf0, and translational and rotational entropy, ΔSrt0. The value of ΔSrt0 is debated in the literature and here ΔSrt0=−33 J·mol–1·K–1 was taken, which is close to the cratic entropy (17Lavigne P. Bagu J.R. Boyko R. Willard L. Holmes C.F. Sykes B.D. Protein Sci. 2000; 9: 252-264Google Scholar). Change of the solvation entropy, ΔSsolv0, was calculated according to Equation 2, ΔSsolv0(T)=ΔCp·ln(T/T0*)(Eq. 2) where T is the temperature in Kelvin, ΔT0=385 K is the reference temperature where ΔSsolv0(ΔT0) is zero (18Baker B.M. Murphy K.P. Methods Enzymol. 1998; 295: 294-315Google Scholar) and ΔCp is the heat capacity change. Entropic contributions of the hydrophobic effect ( ΔSHE0) to HEI-TOE I-PPE binding were calculated according to Equation 3 (19Spolar R.S. Record Jr., M.T. Science. 1994; 263: 777-784Google Scholar), ΔSHE0(T)=0.32·ΔAnp·ln(T/386)(Eq. 3) where ΔAnp is in Å2, T is in degrees Kelvin, and ΔSHE0(T) is in entropy units (e.u. or cal·mol–1·K–1). Ts is the temperature where ΔS0 is zero and was calculated according to Equation 4, ln Ts=−ΔS0/ΔCp+ln T (eq. 4) with the experimental values ΔStot0=150 J·mol–1·K–1 at 298 K and ΔCp,exp = –1.275 kJ·mol–1·K–1. At Ts the entropy changes from the hydrophobic effect ΔSHE0 and from any other entropic contribution ΔSother0 despite ΔSHE0 add to zero. ΔS0(Ts)=0=ΔSHE0(Ts)+ΔSother0(Eq. 5) Polar (ΔASAp) and nonpolar (ΔASAnp) buried surface areas were calculated according to Refs. 20Murphy K.P. Freire E. Adv. Protein Chem. 1992; 43: 313-361Google Scholar and 21Xie D. Freire E. Proteins. 1994; 19: 291-301Google Scholar by means of empirical Equations 6, 7, 8, ΔCp=a·ΔASAnp+b·ΔASAp(Eq. 1) ΔH0(T)=ΔH0(60 °C)+ΔCp(T−60 °C)(Eq. 7) ΔH0(60 °C)=c·ΔASAnp+d·ΔASAp(Eq. 8) where T is in °C, a = 1.88 J·K–1·mol–1·Å–2, b = –1.09 J·K–1·mol–1·Å–2, c = –35.3 J·mol–1·Å–2, and d = 131 J·mol–1·Å–2. Buried surface areas used for thermodynamic calculations were obtained from atomic coordinates of elastase-HEI-TOE I complex (Protein Data Bank code 1mcv, adapted for the calculations) with program CALCASA of the program suite STC (structure-based thermodynamic calculations) (17Lavigne P. Bagu J.R. Boyko R. Willard L. Holmes C.F. Sykes B.D. Protein Sci. 2000; 9: 252-264Google Scholar). In CALCASA is implemented the Lee and Richards algorithm (22Lee B. Richards F.M. J. Mol. Biol. 1971; 55: 379-400Google Scholar); probe radius and slice widths are 1.4 and 0.25 Å, respectively. Identical structures for both PPE and HEI-TOE I were assumed in the complex and in free form. The program suite STC sums up to ΔSconf0 the individual contributions of all residues buried in the complex. Program THERMO of the STC suite calculates thermodynamic parameters ΔCp,calc, ΔHcalc0, ΔGcalc0, and dissociation constant Kd (17Lavigne P. Bagu J.R. Boyko R. Willard L. Holmes C.F. Sykes B.D. Protein Sci. 2000; 9: 252-264Google Scholar). THERMO provides the possibility to vary arbitrarily the value for ΔSconf0 and thus to achieve apparent but meaningless agreement of experimental and calculated ΔG0. Preparation of Human Endothelial Cells and F-actin Staining— Human umbilical cord vein endothelial cells (HUVEC) were isolated from umbilical cord veins and identified as described previously (23Hippenstiel S. Soeth S. Kellas B. Fuhrmann O. Seybold J. Krull M. Eichel-Streiber C. Goebeler M. Ludwig S. Suttorp N. Blood. 2000; 95: 3044-3051Google Scholar). Briefly, cells obtained from collagenase digestion were washed, resuspended, cultivated in MCDB 131, 10% fetal calf serum, and seeded onto 24-well Thermanox® slides (Nunc, Wiesbaden, Germany) (23Hippenstiel S. Soeth S. Kellas B. Fuhrmann O. Seybold J. Krull M. Eichel-Streiber C. Goebeler M. Ludwig S. Suttorp N. Blood. 2000; 95: 3044-3051Google Scholar, 24Hippenstiel S. Witzenrath M. Schmeck B. Hocke A. Krisp M. Krull M. Seybold J. Seeger W. Rascher W. Schutte H. Suttorp N. Circ. Res. 2002; 91: 618-625Google Scholar). Studies were performed using confluent endothelial cell monolayers in their second passage. For the investigations with the protease inhibitors human umbilical cord vein endothelial cells were incubated with 0.5 unit of PPE for 1 h. HUVEC were grown on Thermanox® slides. After stimulation, slides were fixed for 1 h in 3% paraformaldehyde at room temperature and rinsed three times in phosphate-buffered saline. Permeabilization of cell membranes was performed using 0.1% Triton X-100 for 5 min followed by three wash steps with phosphate-buffered saline. F-actin was stained with phalloidin Alexa 488 (Molecular Probes, Eugene, OR) (1:400) as described previously (24Hippenstiel S. Witzenrath M. Schmeck B. Hocke A. Krisp M. Krull M. Seybold J. Seeger W. Rascher W. Schutte H. Suttorp N. Circ. Res. 2002; 91: 618-625Google Scholar). Thermanox® slides were placed on glass slides and embedded with Gelmount mounting media (Biomeda). Slides were analyzed using a Pascal 5 confocal scanning laser microscope (Zeiss, Jena, Germany) equipped with an air-cooled argon laser (Axioskop 2 Mot microscope, Zeiss). Alexa 488 fluorescence was excited with a 488-nm argon-ion laser beam and imaged using a NT80/20/488 beamsplitter and a 505-nm longpass emission filter. All images were recorded with a 40 × 1.3 NA Plan-Apochromat III DIC objective with image resolution 1024 × 1024 pixels. Rational Design of Hybrid Squash Inhibitors—The binding loop of squash inhibitor EETI II is located at the N terminus of the miniprotein and is stabilized by two cysteine bonds (see Fig. 1 for the corresponding sequences). Two different strategies were used to insert the inhibitory peptide into this binding loop. First, the inhibitory peptide was introduced in such a way that its P1 position (according to the nomenclature of Schechter and Berger (4Schechter I. Berger A. Biochem. Biophys. Res. Commun. 1967; 27: 157-162Google Scholar), the main specificity-determining amino acid residue adjacent to the scissile peptide bond of a peptide or polypeptide substrate) Leu-4 resembled that of the original squash inhibitor EETI II. Also, to maintain the important disulfide bridges, the sequence of the inhibitory peptide PMTLEYR was modified in position 2 to PCTLEYR. Furthermore, to avoid problems with an adjacent positive charge, and since a substitution for methionine is possible without decreasing inhibition (2Hilpert K. Hansen G. Wessner H. Schneider-Mergener J. Hohne W. J. Biochem. (Tokyo). 2000; 128: 1051-1057Google Scholar, 8Wentzel A. Christmann A. Kratzner R. Kolmar H. J. Biol. Chem. 1999; 274: 21037-21043Google Scholar), the Arg-7 was changed to methionine, a conserved position in most squash inhibitors. This inhibitor with the N-terminal sequence now reading PCTLEYM (see Fig. 1) was called HEI-TOE I: hybrid elastase inhibitor-turkey ovomucoid E. elaterium. In the second strategy, the optimized peptide was inserted between the two cysteine residues of the squash inhibitor N-terminal binding loop. Christmann et al. (25Christmann A. Walter K. Wentzel A. Kratzner R. Kolmar H. Protein Eng. 1999; 12: 797-806Google Scholar) showed that it is possible to replace the six residues of this loop by a 17-residue epitope from human bone Gla protein. This EETI II variant was fully oxidized and correctly folded, and the epitope inserted into the binding loop was readily recognized by a monoclonal antibody (25Christmann A. Walter K. Wentzel A. Kratzner R. Kolmar H. Protein Eng. 1999; 12: 797-806Google Scholar). We inserted the optimized peptide into the binding loop in the same way, creating the variant HEI-TOE II (Fig. 1). Both hybrid squash inhibitors were synthesized, purified, and oxidized as described under "Experimental Procedures." Inhibition of PPE activity was measured during the course of disulfide bridge formation. Interestingly, HEI-TOE I and II showed opposing behaviors. Whereas PPE inhibition by HEI-TOE I increases during disulfide formation it decreases for HEI-TOE II (Fig. 2A). To verify these data, the Ki values for HEI-TOE I and II were determined before disulfide formation, after disulfide formation and after a subsequent second purification step to eliminate improperly folded or cross-reacting species. For comparison, the Ki values for the free peptides corresponding to the binding loop of the two hybrid squash variants were included (Fig. 2B and Table I). The data confirmed that only the inhibitor HEI-TOE I, with the P1 site position conserved within the structure, inhibits PPE strongly after folding and disulfide formation is finished. We cannot exclude that non-native disulfide bonds are predominantly formed with HEI-TOE II, resulting in the decreased inhibition strength of this inhibitor variant, but more likely, the decreasing inhibitory strength during the formation of disulfide bridges reflects structural restrictions imposed on the PPE-interacting loop region.Table 1Specificity and pH dependence of ovomucoid derived and squash-type inhibitorsInhibitorpHKiPPEHLEBPCBPTMOMTKY3aThe data for OMTKY3 are taken from Ref. 43 and for EETI II in one case from Ref. 44. The values given in parentheses are ranges.8.52.4 × 10-111.6 × 10-10NI7.55.5 × 10-12PACTLEYRC8.51.7 (±0.4) × 10-42.5 (±0.3) × 10-4ND7.51.0 (±0.7) × 10-4EETI II8.58.2 (±2.2) × 10-44.0 (±0.7) × 10-41.3 × 10-12a7.51.9 (±0.3) × 10-5MCEI III8.51.3 (±0.21) × 10-82.1 (±1.5) × 10-83.1 (±0.5) × 10-57.52.9 (±0.3) × 10-7PMTLEYR8.56.7 (±0.4) × 10-71.1 (±0.1) × 10-4ND7.56.3 (±2.1) × 10-71.3 (±0.1) × 10-4ND6.51.3 (±0.7) × 10-6ND5.54.9 (±1.3) × 10-5NDHEI-TOE I8.59.8 (±2.2) × 10-84.5 × 10-42.2 × 10-47.53.8 (±0.3) × 10-71.3 × 10-46.51.3 (±0.1) × 10-75.53.3 (±0.9) × 10-7HEI-TOE II8.57.0 (±0.7) × 10-53.6 × 10-56.0 × 10-57.52.9 × 10-5HEI-TOE III8.55.6 (±0.3) × 10-8NDNDND7.52.9 (±0.2) × 10-7NDNDND6.53.1 (±0.1) × 10-7NDNDND5.02.4 (±0.1) × 10-7NDNDNDa The data for OMTKY3 are taken from Ref. 43 and for EETI II in one case from Ref. 44. The values given in parentheses are ranges. Open table in a new tab The natural elastase-specific squash inhibitors MCEI I–IV differ only in N-terminal additions of glutamic acid residues: MCEI I has no Glu at the N terminus, but MCEI IV has three (see Fig. 1). This correlates directly with considerable differences in their affinity, where the inhibition strength of MCEI III is 240-fold higher than that of MCEI I (10Hamato N. Koshiba T. Pham T.N. Tatsumi Y. Nakamura D. Takano R. Hayashi K. Hong Y.M. Hara S. J. Biochem. (Tokyo). 1995; 117: 432-437Google Scholar). To investigate the influence of those Glu residues on our hybrid squash inhibitor, we added the three N-terminal residues of MCEI III (EER) to our sequence (Fig. 1, HEI-TOE III). This variant indeed showed increased inhibition of PPE, but only by a factor of 1.8 at pH 8.5 (Table I). Thus, the mode of interaction with PPE is apparently different for HEI-TOE III and MCEI III. Specificity and Stability of Native and Hybrid Squash Inhibitors—For comparison, Ki values for the two hybrid miniproteins HEI-TOE I and HEI-TOE II were determined with PPE and three other proteases: HLE, chymotrypsin (BPC), and trypsin (BPT) from bovine pancreas (Table I). A specificity similar to PPE has been reported for HLE (26Bode W. Meyer E.J. Powers J.C. Biochemistry. 1989; 28: 1951-1963Google Scholar). BPT was chosen because EETI II, the scaffold donor, shows a high specificity toward trypsin. BPC is included because of its deviating specificity and strong inhibition by OMTKY3. HEI-TOE II did not significantly distinguish between the tested serine proteases, in contrast to HEI-TOE I, which showed high specificity for PPE. The observed PPE specificity was similar to that of the optimized peptide PMTLEYR. In fact HEI-TOE I affinity was about seven times higher than that of PMTLEYR, but 3–4 orders of magnitude lower than that of OMTKY3, which in turn showed little difference between PPE, HLE, and BPC but did not inhibit BPT at all (27Kato I. Schrode J. Kohr W.J. Laskowski Jr., M. Biochemistry. 1987; 26: 193-201Google Scholar). The natural squash inhibitor EETI II is highly specific for trypsin due to the Arg in position P1 of the binding loop, but, at a low affinity level, it does not distinguish between the elastases PPE and HLE. The natural inhibitor MCEI III described as an inhibitor for PPE has a low affinity against trypsin and, at high affinity level, does not distinguish between both elastases (PPE and HLE) and also chymotrypsin (see Table I). In contrast to the free peptide, the hybrid inhibitor variant HEI-TOE II containing the optimized peptide PMTLEYR directly inserted into the EETI II binding loop shows no much difference in affinity toward all proteases tested (Table I). Obviously, the distortion of peptide conformation imposed by the framework constraint occurs in such a way that important contacts, especially to PPE, cannot be established. The data also show that the substitutions leading from EETI II to the hybrid variant HEI-TOE I transform a squash inhibitor previously highly specific for trypsin into an specific inhibitor for porcine pancreatic elastase. Another feature of inhibitors is important: their stability against proteolytic cleavage, here either at the P1 position or somewhere else within structurally exposed sequence parts. OMTKY3 itself, the starting protein for our previous optimization experiments, is very stable against proteolytic attack (28Ardelt W. Laskowski M.J. Acta Biochim. Pol. 1983; 30: 115-126Google Scholar). The peptide PACTLEYRP representing the binding loop of OMTKY3 was synthesized, and its stability against proteolytic cleavage was measured as a change in the degree of PPE inhibition (data not shown). The peptide undergoes rapid degradation by PPE with a half-life ranging below 1 h. This suggests that the binding sequence itself is not responsible for the high hydrolytic stability of OMTKY3 but rather the rigid structure imposed by the scaffold. The time scale for degradation of the optimized peptide PMTLEYR enters the same range as for PACTLEYRP if 10-fold more enzyme is used (data not shown). The hybrid miniprotein HEI-TOE I is clearly more stable against proteolytic attack than the free inhibiting peptides (24% residual inhibitory activity after 145 h of incubation with PPE). The time-dependent reduction of the inhibitor peptides was also ascertained by HPLC measurements (data not shown). In the case of the natural squash inhibitor MCEI III no cleavage is detectable even over 8 days. The difference between the peptides and the squash inhibitors could be explained by a lower efficiency of the acyl enzym formation for the complete miniproteins because of the rigid structure of the binding loop. In addition, it was reported that protein inhibitors with the scissile bond already cleaved can bind to the enzyme as well and thus maintain inhibition or even resynthesize the cleaved peptide bond (29Otlewski J. Zbyryt T. Dryjanski M. Bulaj G. Wilusz T. Biochemistry. 1994; 33: 208-213Google Scholar). HEI-TOE I binding to PPE causes small shifts of loops in the vicinity of the binding region of the protease (30Aÿ J. Hilpert K. Krauss N. Schneider-Mergener J. Höhne W. Acta Crystallogr. Sect. D Biol. Crystallogr. 2003; 59: 247-254Google Scholar). It is possible that these shifts impose a higher degree of flexibility on the binding loop, thus resulting in a faster cleavage of the hybrid inhibitor compared with the natural squash inhibitor MCEI III. The free peptides are even more flexible, and the acyl enzyme may form faster. The pH dependence of Ki of the hybrid squash inhibitors HEI-TOE I and III is not very pronounced (Table I) and similar to that of the inhibitory peptide PMTLEYR. At pH 6.0, where the proteolytic activity of PPE is diminished, the affinity of HEI-TOE I for PPE is large enough for the co-crystallization of the complex (31Hilpert K. Schneider-Mergener J. Ay J. Acta Crystallogr. Sect. D Biol. Crystallogr. 2002; 58: 672-674Google Scholar). The complex structure shows a complete HEI-TOE I inhibitor with clear electron density for the scissile Leu-Glu bond in the active site of PPE (30Aÿ J. Hilpert K. Krauss N. Schneider-Mergener J. Höhne W. Acta Crystallogr. Sect. D Biol. Crystallogr. 2003; 59: 247-254Google Scholar). Comparison of the structure of HEI-TOE I in a complex with PPE and a structure of EETI II variant (for sequence see Fig. 1) in a complex with trypsin 2I. Uson, personal communication. showed a good match in the peptide backbone (data not shown). This means that the overall structure of HEI-TOE I is similar to the EETI II donor framework. There are only small deviations in the binding loop and also in two other loop regions. Intere
Referência(s)