Conserved cystatin segments as models for designing specific substrates and inhibitors of cysteine proteinases

Repurposing the McoTI-II Rigid Molecular Scaffold in to Inhibitor of 'Papain Superfamily' Cysteine Proteases

Clan C1A or ‘papain superfamily’ cysteine proteases are key players in many important physiological processes and diseases in most living systems. Novel approaches towards the development of their inhibitors can open new avenues in translational medicine. Here, we report a novel design of a re-engineered chimera inhibitor Mco-cysteine protease inhibitor (CPI) to inhibit the activity of C1A cysteine proteases. This was accomplished by grafting the cystatin first hairpin loop conserved motif (QVVAG) onto loop 1 of the ultrastable cyclic peptide scaffold McoTI-II. The recombinantly expressed Mco-CPI protein was able to bind with micromolar affinity to papain and showed remarkable thermostability owing to the formation of multi-disulphide bonds. Using an in silico approach based on homology modelling, protein–protein docking, the calculation of the free-energy of binding, the mechanism of inhibition of Mco-CPI against representative C1A cysteine proteases (papain and cathepsin L) was validated. Furthermore, molecular dynamics simulation of the Mco-CPI–papain complex validated the interaction as stable. To conclude, in this McoTI-II analogue, the specificity had been successfully redirected towards C1A cysteine proteases while retaining the moderate affinity. The outcomes of this study pave the way for further modifications of the Mco-CPI design for realizing its full potential in therapeutics. This study also demonstrates the relevance of ultrastable peptide-based scaffolds for the development of novel inhibitors via grafting.

Cyclic peptide engineered from phytocystatin inhibitory hairpin loop as an effective modulator of falcipains and potent antimalarial

Cystatins are classical competitive inhibitors of C1 family cysteine proteases (papain family). Phytocystatin superfamily shares high sequence homology and typical tertiary structure with conserved glutamine-valine-glycine (Q-X-V-X-G) loop blocking the active site of C1 proteases. Here, we develop a cysteine-bounded cyclic peptide (CYS-cIHL) and linear peptide (CYS-IHL), using the conserved inhibitory hairpin loop amino acid sequence. Using an in silico approach based on modeling, protein-peptide docking, molecular dynamics simulations and calculation of free energy of binding, we designed and validated inhibitory peptides against falcipain-2 (FP-2) and -3 (FP-3), cysteine proteases from the malarial parasite Plasmodium falciparum. Falcipains are critical hemoglobinases of P. falciparum that are validated targets for the development of antimalarial therapies. CYS-cIHL was able to bind with micromolar affinity to FP-2 and modulate its binding with its substrate, hemoglobin in in vitro and in vivo assays. CYS-cIHL could effectively block parasite growth and displayed antimalarial activity in culture assays with no cytotoxicity towards human cells. These results indicated that cyclization can substantially increase the peptide affinity to the target. Furthermore, this can be applied as an effective strategy for engineering peptide inhibitory potency against proteases.

Active site labeling of cysteine cathepsins by a straightforward diazomethylketone probe derived from the N-terminus of human cystatin C

We designed a straightforward biotinylated probe using the N-terminal substrate-like region of the inhibitory site of human cystatin C as a scaffold, linked to the thiol-specific reagent diazomethylketone group as a covalent warhead (i.e. Biot-(PEG)₂-Ahx-LeuValGly-DMK). The irreversible activity-based probe bound readily to cysteine cathepsins B, L, S and K. Moreover affinity labeling is sensitive since active cathepsins were detected in the nM range using an ExtrAvidin-peroxidase conjugate for disclosure. Biot-(PEG)₂-Ahx-LeuValGly-DMK allowed a slightly more pronounced labeling for cathepsin S with a compelling second-order rate constant for association (kass = 2,320,000 M(-1) s(-1)). Labeling of the active site is dose-dependent as observed using 6-cyclohexylamine-4-piperazinyl-1,3,5-triazine-2-carbonitrile, as competitive inhibitor of cathepsins. Finally we showed that Biot-(PEG)₂-Ahx-LeuValGly-DMK may be a simple and convenient tool to label secreted and intracellular active cathepsins using a myelomonocytic cell line (THP-1 cells) as model.

Cysteine cathepsins: from structure, function and regulation to new frontiers

It is more than 50 years since the lysosome was discovered. Since then its hydrolytic machinery, including proteases and other hydrolases, has been fairly well identified and characterized. Among these are the cysteine cathepsins, members of the family of papain-like cysteine proteases. They have unique reactive-site properties and an uneven tissue-specific expression pattern. In living organisms their activity is a delicate balance of expression, targeting, zymogen activation, inhibition by protein inhibitors and degradation. The specificity of their substrate binding sites, small-molecule inhibitor repertoire and crystal structures are providing new tools for research and development. Their unique reactive-site properties have made it possible to confine the targets simply by the use of appropriate reactive groups. The epoxysuccinyls still dominate the field, but now nitriles seem to be the most appropriate “warhead”. The view of cysteine cathepsins as lysosomal proteases is changing as there is now clear evidence of their localization in other cellular compartments. Besides being involved in protein turnover, they build an important part of the endosomal antigen presentation. Together with the growing number of non-endosomal roles of cysteine cathepsins is growing also the knowledge of their involvement in diseases such as cancer and rheumatoid arthritis, among others. Finally, cysteine cathepsins are important regulators and signaling molecules of an unimaginable number of biological processes. The current challenge is to identify their endogenous substrates, in order to gain an insight into the mechanisms of substrate degradation and processing. In this review, some of the remarkable advances that have taken place in the past decade are presented. This article is part of a Special Issue entitled: Proteolysis 50 years after the discovery of lysosome.

Kininogen-derived peptides for investigating the putative vasoactive properties of human cathepsins K and L

Macrophages at an inflammatory site release massive amounts of proteolytic enzymes, including lysosomal cysteine proteases, which colocalize with their circulating, tight-binding inhibitors (cystatins, kininogens), so modifying the protease/antiprotease equilibrium in favor of enhanced proteolysis. We have explored the ability of human cathepsins B, K and L to participate in the production of kinins, using kininogens and synthetic peptides that mimic the insertion sites of bradykinin on human kininogens. Although both cathepsins processed high-molecular weight kininogen under stoichiometric conditions, only cathepsin L generated significant amounts of immunoreactive kinins. Cathepsin L exhibited higher specificity constants (kcat/Km) than tissue kallikrein (hK1), and similar Michaelis constants towards kininogen-derived synthetic substrates. A 20-mer peptide, whose sequence encompassed kininogen residues Ile376 to Ile393, released bradykinin (BK; 80%) and Lys-bradykinin (20%) when incubated with cathepsin L. By contrast, cathepsin K did not release any kinin, but a truncated kinin metabolite BK(5-9) [FSPFR(385-389)]. Accordingly cathepsin K rapidly produced BK(5-9) from bradykinin and Lys-bradykinin, and BK(5-8) from des-Arg9-bradykinin, by cleaving the Gly384-Phe385 bond. Data suggest that extracellular cysteine proteases may participate in the regulation of kinin levels at inflammatory sites, and clearly support that cathepsin K may act as a potent kininase.

Congopain from Trypanosoma congolense: drug target and vaccine candidate

Trypanosomes are the etiological agents of human sleeping sickness and livestock trypanosomosis (nagana), which are major diseases in Africa. Their cysteine proteases (CPs), which are members of the papain family, are expressed during the infective stages of the parasites' life cycle. They are suspected to act as pathogenic factors in the mammalian host, where they also trigger prominent immune responses. Trypanosoma congolense, a major pathogenic species in livestock, possesses at least two families of closely related CPs, named CP1 and CP2. Congopain, a CP2-type of enzyme, shares structural and functional resemblances with cruzipain from T. cruzi and with mammalian cathepsin L. Like CPs from other Trypanosomatids, congopain might be an attractive target for trypanocidal drugs. Here we summarise the current knowledge in the two main areas of research on congopain: first, the biochemical properties of congopain were characterised and its substrate specificity was determined, as a first step towards drug design; second, the possibility was being explored that inhibition of congopain by host-specific antibodies may mitigate the pathology associated with trypanosome infection.

Highly potent inhibitors of human cathepsin L identified by screening combinatorial pentapeptide amide collections

By screening a combinatorial pentapeptide amide collection in an inhibition assay, we systematically evaluated the potential of 19 proteinogenic amino acids and seven nonproteinogenic amino acids to serve as building blocks for inhibitors of human cathepsin L. Particularly efficient were aromatic, bulky, hydrophobic amino-acid residues, especially leucine, and positively charged residues, especially arginine. Building blocks for potential inhibitory peptides were combined by random selection from their activity pattern. This random approach for the design of inhibitors was introduced to compensate for the inaccuracy induced by shifted docking of combinatorial compound collections at the active center of cathepsin L. Thereby, we obtained structurally defined pentapeptide amides which inhibited human cathepsin L at nanomolar concentrations. Among the most potent novel inhibitors, one peptide, RKLLW-NH2, shares the amphiphilic character of the nonamer fragment VMNGLQNRK of the autoinhibitory, substrate-like, but reverse-binding prosegment of human cathepsin L which blocks the active center of the enzyme. Obviously, RKLLW-NH2 carries the functions that are important for enzyme-peptide interaction in a condensed form. This hypothesis was confirmed by structure-activity studies using truncated and modified pentapeptides.

Fluorescence-quenched solid phase combinatorial libraries in the characterization of cysteine protease substrate specificity

To map the substrate specificity of cysteine proteases, two combinatorial peptide libraries were synthesized and screened using the archetypal protease, papain. The use of PEGA resin as the solid support for library synthesis facilitated the application of an on-resin fluorescence-quenched assay. Results from the screening of library 2 indicated a preference for Pro or Val in the S3 subsite and hydrophobic residues in S2; the most prevalent residue not being Phe but Val. The S1 subsite exhibited a dual specificity for both small, nonpolar residues, Ala or Gly, as well as larger, Gln, and charged residues, Arg. Small residues predominated in the S1'-S4' subsites. Active peptides from the libraries and variations thereof were resynthesized and their kinetics of hydrolysis by papain assessed in solution phase assays. Generally, there was a good correlation between the extent of substrate cleavage on solid phase and the kcat/KM's obtained in solution phase assays. Several good substrates for papain were obtained, the best substrates being Y(NO2)PMPPLCTSMK(Abz) (kcat/KM = 2109 (mM s)-1), Y(NO2)PYAVQSPQK(Abz) (kcat/KM = 1524 (mM s)-1), and Y(NO2)PVLRQQRSK(Abz) (kcat/KM = 1450 (mM s)-1). These results were interpreted in structural terms by the use of molecular dynamics (MD). These MD calculations indicated two different modes for the binding of substrates in the narrow enzyme cleft.

Enzymatic cleavage of peptide-linked radiolabels from immunoconjugates

We have incorporated peptides selected by combinatorial library [Peterson, J. J., and Meares, C. F. (1998) Bioconjugate Chem. 9, 618-626) into peptide-linked radiolabeled immunoconjugates of the form DOTA-peptide-antibody. Decapeptide linkers -GFQGVQFAGF- and -GFGSVQFAGF-, selected for cleavage by human liver cathepsin B, were rapidly digested in vitro when compared to the simple model tetrapeptide motif of the prototype -GGGF- [Li, M., and Meares, C. F. (1993) Bioconjugate Chem. 4, 275-283]. Cleavage properties of these library-selected substrates for cathepsin B compared favorably with decapeptide linkers -GLVGGAGAGF- and -GGFLGLGAGF-, which incorporate two of the most labile extended cathepsin B substrates from the literature. The decapeptide linker -GFGSTFFAGF-, selected from the library for cleavage by human liver cathepsin D, was rapidly digested by cathepsin D while the others were not.

Interdependency of sequence and positional specificities for cysteine proteases of the papain family

The specificity of cysteine proteases is characterized by the nature of the amino acid sequence recognized by the enzymes (sequence specificity) as well as by the position of the scissile peptide bond (positional specificity, i.e., endopeptidase, aminopeptidase, or carboxypeptidase). In this paper, the interdependency of sequence and positional specificities for selected members of this class of enzymes has been investigated using fluorogenic substrates where both the position of the cleavable peptide bond and the nature of the sequence of residues in P2-P1 are varied. The results show that cathepsins K and L and papain, typically considered to act strictly as endopeptidases, can also display dipeptidyl carboxypeptidase activity against the substrate Abz-FRF(4NO2)A and dipeptidyl aminopeptidase activity against FR-MCA. In some cases the activity is even equal to or greater than that observed with cathepsin B and DPP-I (dipeptidyl peptidase I), which have been characterized previously as exopeptidases. In contrast, the exopeptidase activities of cathepsins K and L and papain are extremely low when the P2-P1 residues are A-A, indicating that, as observed for the normal endopeptidase activity, the exopeptidase activities rely heavily on interactions in subsite S2 (and possibly S1). However, cathepsin B and DPP-I are able to hydrolyze substrates through the exopeptidase route even in absence of preferred interactions in subsites S2 and S1. This is attributed to the presence in cathepsin B and DPP-I of specific structural elements which serve as an anchor for the C- or N-terminus of a substrate, thereby allowing favorable enzyme-substrate interaction independently of the P2-P1 sequence. As a consequence, the nature of the residue at position P2 of a substrate, which is usually the main factor determining the specificity for cysteine proteases of the papain family, does not have the same contribution for the exopeptidase activities of cathepsin B and DPP-I.

Discrimination of cruzipain, the major cysteine proteinase of Trypanosoma cruzi, and mammalian cathepsins B and L, by a pH-inducible fluorogenic substrate of trypanosomal cysteine proteinases

The substrate specificity of cruzipain, the major cysteine proteinase of Trypanosoma cruzi, was investigated using a series of dansyl-peptides based on the putative autoproteolytic sequence of the proteinase (VVG-GP) located at the hinge region between the catalytic domain and the C-terminal extension. Replacing Val with Pro at P2 in this sequence greatly improved the rate of cleavage by cruzipain. Tyr and Val residues are preferred at P3 by all cysteine proteinases whatever their origin, whereas only cruzipain and cathepsin L cleaved substrate with a His at that position. The combination of a Pro at P2 and His at P3 abolished cleavage by cathepsin L, so that only cruzipain was able to cleave the HPGGP peptide at the GG bond. A substrate with intramolecularly quenched fluorescence was raised on this sequence (Abz-HPGGPQ-EDDnp) which was also specifically cleaved by cruzipain (kcat/Km of 157 000 m-1. s-1) and by a homologous proteinase from Trypanosoma congolense. The pH activity profile of cruzipain on Abz-HPGGPQ-EDDnp showed a narrow peak with a maximum at pH 5.5 and no cleavage above pH 6.8, although trypanosomal cysteine proteinases remain active at basic pH. The lack of activity at neutral and basic pH was due to a decrease in kcat, while the Km remained essentially unchanged, demonstrating that the substrate still binds to the enzyme and therefore behaves as an inhibitor. Changing the substrate into an inhibitor depended on the deprotonation of the His residue in the substrate, as deduced from a comparison of the pH activity profile with that of a related, but uncharged, substrate. Abz-HPGGPQ-EDDnp also inhibited mammalian cathepsins B and L but was not cleaved by these proteinases at any pH. The importance of the His residue at P3 for cleavage by cruzipain was confirmed by substituting Lys for His at that position. The resulting peptide was not cleaved by cruzipain in spite of the presence of a positively charged group at P3, but still interacted with the enzyme. It was concluded that the presence of an imidazolium group at P3 was essential to endow the HPGGPQ sequence with the properties of a cruzipain substrate.

Cathepsin substrates as cleavable peptide linkers in bioconjugates, selected from a fluorescence quench combinatorial library

Several extended peptide substrates for the human liver enzymes cathepsin B and cathepsin D have been selected as cleavable linkers for lysosomal proteolysis of bioconjugates. A one-bead-one-peptide combinatorial library of 9(4) fluorogenic substrates was employed. We designed this library to explore a set of substrates containing nonionizable/nonoxidizable groups to meet the requirements of prelabeling [Li et al. (1994) Bioconjugate Chem. 5, 101-104] as well as to yield stable conjugates whose preparation is straightforward.

Substrate specificity of barley cysteine endoproteases EP-A and EP-B

The cysteine endoproteases (EP)-A and EP-B were purified from green barley (Hordeum vulgare L.) malt, and their identity was confirmed by N-terminal amino acid sequencing. EP-B cleavage sites in recombinant type-C hordein were determined by N-terminal amino acid sequencing of the cleavage products, and were used to design internally quenched, fluorogenic peptide substrates. Tetrapeptide substrates of the general formula 2-aminobenzoyl-P2-P1-P1'-P2'-tyrosine(NO2)-aspartic acid, in which cleavage occurs between P1 and P1', showed that the cysteine EPs preferred phenylalanine, leucine, or valine at P2. Arginine was preferred to glutamine at P1, whereas proline at P2, P1, or P1' greatly reduced substrate kinetic specificity. Enzyme cleavage of C hordein was mainly determined by the primary sequence at the cleavage site, because elongation of substrates, based on the C hordein sequence, did not make them more suitable substrates. Site-directed mutagenesis of C hordein, in which serine or proline replaced leucine, destroyed primary cleavage sites. EP-A and EP-B were both more active than papain, mostly because of their much lower Km values.

Major increase in endopeptidase activity of human cathepsin B upon removal of occluding loop contacts

The main feature distinguishing cathepsin B from other cysteine proteases of the papain family is the presence of a large insertion loop, termed the occluding loop, which occupies the S' subsites of the enzyme. The loop is held in place mainly by two contacts with the rest of the enzyme, involving residues His110 and Arg116 on the loop that form salt bridges with Asp22 and Asp224, respectively. The influence of this loop on the endopeptidase activity of cathepsin B has been investigated using site-directed mutagenesis and internally quenched fluorogenic (IQF) substrates. Wild-type cathepsin B displays poor activity against the substrates Abz-AFRSAAQ-EDDnp and Abz-QVVAGA-EDDnp as compared to cathepsin L and papain. Appreciable increases in kcat/KM were observed for cathepsin B containing the single mutations D22A, H110A, R116A, and D224A. The highest activity however is observed for mutants where both loop to enzyme contacts are disrupted. For the triple-mutant D22A/H110A/R116A, an optimum kcat/KM value of 12 x 10(5) M-1 s-1 was obtained for hydrolysis of Abz-AFRSAAQ-EDDnp, which corresponds to a 600-fold increase relative to wild-type cathepsin B and approaches the level of activity observed with cathepsin L or papain. By comparison, the mutations have little effect on the hydrolysis of Cbz-FR-MCA. The influence of the mutations on the pH dependency of activity also indicates that the complexity of pH activity profiles normally observed for cathepsin B is related to the presence of the occluding loop. The major increase in endopeptidase activity is attributed to an increase in loop "flexibility" and suggests that the occluding loop might move when an endopeptidase substrate binds to the enzyme. The possible contribution of these interactions in regulating endopeptidase activity and the implications for cathepsin B activity in physiological or pathological conditions are discussed.

A comparison of the enzymatic properties of the major cysteine proteinases from Trypanosoma congolense and Trypanosoma cruzi

Congopain and cruzipain, the major cysteine proteinases from Trypanosoma congolense and Trypanosoma cruzi, were compared for their activities towards a series of new, sensitive fluorogenic substrates of the papain family of cysteine proteinases and for their sensitivity to inhibition by cystatins and related biotinylated peptidyl diazomethanes. Low Ki values, in the 10 pM range, were found for the interaction of both proteinases with natural cystatin inhibitors. The kinetic constants for the hydrolysis of cystatin-derived substrates, and the inhibition by related diazomethanes were essentially identical. Unlike cathepsins B and L, the related mammal papain family proteinases, congopain and cruzipain accomodate a prolyl residue in P2'. Substrates having the sequence VGGP from P2 to P2' were hydrolysed by both congopain and cruzipain with a k(cat)/Km greater than 4.10(3) mM(-1) s(-1). Irreversible diazomethane inhibitors, deduced from the unprime sequence of cystatin-derived substrates, inhibited the two parasite proteinases. N-terminal labelling of diazomethanes with a biotin group did not alter the rate of inhibition significantly, which provides a useful tool for examining the distribution of these enzymes in the parasite and in the host. Despite their similar activities on cystatin-derived substrates, congopain and cruzipain had significantly different pH-activity profiles when assayed with a cystatin-derived substrate. They were correlated with structural differences, especially at the presumed S2 subsites.

Inhibition of cathepsin B by its propeptide: use of overlapping peptides to identify a critical segment

Ten overlapping 15-mer peptides (peptidyl amides) spanning the proregion of rat cathepsin B (residues 1p-60p) were constructed to identify minimal segments having inhibitory activity towards the mature enzyme, that could be used to develop a new generation of peptide-derived inhibitors specifically targeting the active site of the corresponding proteinase. Three synthetic peptides, containing the pentapeptide Leu-Cys-Gly-Thr-Val (residues 41p-45p) in their sequence, inhibited cathepsin B with Ki values in the micromolar range. Alkylation of the thiol group of Cys-42p of peptide PB8 (36p-50p) resulted in its rapid proteolytic degradation, suggesting that this residue is essential for inhibition. The inhibition constant was slightly improved (Ki = 2 microM) using a longer peptide (26p-50p) which was completely resistant to cleavage even after a prolonged incubation. Alkylation of its cysteinyl residue also resulted in rapid cleavage of the peptide chain. Peptides derived from the rat cathepsin B prosequence also inhibited human cathepsin B with similar Ki values. Unlike rat cathepsin B, which cleaves peptide PB8 at the G47p-G48p bond after prolonged incubation, the human enzyme cleaved both PB8 and PB11 at the Lys-40p-Leu-41p bond, in agreement with the different kinetic properties of these two proteinases. New probes with improved specificity for cysteine proteinases may therefore be designed based on the sequences of their propeptides.

Biotin-labelled peptidyl diazomethane inhibitors derived from the substrate-like sequence of cystatin: targeting of the active site of cruzipain, the major cysteine proteinase of Trypanosoma cruzi

Biotin-labelled peptidyl diazomethane inhibitors of cysteine proteinases, based on the N-terminal substrate-like segment of human cystatin C, a natural inhibitor of cysteine proteinases, were synthesized. These synthetic derivatives were tested as irreversible inhibitors of cruzipain, the major cysteine proteinase of Trypanosoma cruzi, to compare the kinetics of the inhibition of the parasite proteinase with that of the mammalian cathepsins B and L. The accessibility of the active sites of these proteinases to these probes was also investigated. The inhibition of cruzipain by Biot-LVG-CHN2 (where Biot represents biotinyl and L,V and G are single-letter amino acid residue abbreviations) and Biot-Ahx-LVG-CHN2 (where Ahx represents 6-aminohexanoic acid) was similar to that of unlabelled inhibitor. Biotin labelling of the inhibitor slowed the inhibition of both cathepsin B and cathepsin L. Adding a spacer arm (Ahx) between the biotin and the peptide moiety of the derivative increased the inhibition of cathepsin B but not that of cathepsin L. The discrimination provided by this spacer is probably due to differences in the topologies of the binding sites of proteinases, a feature that can be exploited to improve targeting of individual cysteine proteinases. Analysis of the blotted proteinases revealed marked differences in the accessibility of extravidin-peroxidase conjugate to the proteinase-bound biotinylated inhibitor. Cruzipain molecules exposed to Biot-LVG-CHN2 or Biot-Ahx-LVG-CHN2 were readily identified, but the reaction was much stronger when the enzyme was treated with the spacer-containing inhibitor. In contrast with the parasite enzyme, rat cathepsin B and cathepsin L treated with either Biot-LVG-CHN2 or Biot-Ahx-LVG-CHN2 produced no detectable bands. Papain, the archetype of this family of proteinases, was poorly labelled with Biot-LVG-CHN2, but strong staining was obtained with Biot-Ahx-LVG-CHN2. These findings suggest that optimized biotinylated diazomethanes might considerably improve their selectivity for the T. cruzi target enzyme.

Investigation of the substrate specificity of cruzipain, the major cysteine proteinase of Trypanosoma cruzi, through the use of cystatin-derived substrates and inhibitors

A panel of intramolecularly quenched fluorogenic substrates containing the conserved QVVA and LVG inhibitory sequences of cystatin inhibitors was used to describe the specificity of the major cysteine proteinase of Trypanosoma cruzi (cruzipain or cruzain). This approach was based on the observations that: (1) cruzipain is strongly inhibited by chicken cystatin and rat T-kininogen, two representative members of cystatin families 2 and 3; (2) the QVVA- and LVG-containing substrates are specifically hydrolysed by papain-like proteinases; and (3) the cystatin-like motifs are similar to the proteolytically sensitive sequences in cruzipain that separate the pro-region and/or the C-terminal extension from the catalytic domain. Specificity constants (kcat/Km) were determined and compared with those of mammalian cathepsins B and L from rat liver lysosomes. Cruzipain and the mammalian proteinases cleaved cystatin-derived substrates at the same site, but their specificities differed significantly. Increased specificity for cruzipain was obtained by replacing amino acids at critical positions on both sides of the cleavage sites, especially at position P2'. The specificity constants (k(cat)/Km) obtained for the two substrates with a prolyl residue at P2' (O-aminobenzoyl-QVVAGP-ethylenediamine 2-4-dinitrophenyl and O-aminobenzoyl-VVGGP-ethylenediamine 2-4-dinitrophenyl) were about 50 times higher for cruzipain than for rat cathepsin L and about 100 times higher than for cathepsin B. Diazomethylketone derivatives, based on the non-prime sequence of cystatin-derived substrates, inhibited cruzipain irreversibly, but their inactivation rate constants were considerably lower than those for mammalian cathepsins B and L, confirming the importance of P' residues for cruzipain specificity.