The refined 2.15 A X-ray crystal structure of human liver cathepsin B: the structural basis for its specificity

Cyclodextrins as templates for the presentation of protease inhibitors

Mono(6-succinylamido-6-deoxy)-beta-cyclodextrin was synthesized by classical carbohydrate chemistry and used as a template mono-functionalized with the linear, fully flexible 4C-spacer carboxylate for covalent linkage of the calpain inhibitor leucyl-leucyl-norleucinal. Spectroscopic analyses of the conjugate do not support a self-inclusion of part of the hydrophobic peptide tail, but confirm its intra- or intermolecular interaction with the template moiety that leads to full water solubility. The inhibitory potency of the beta-cyclodextrin/peptide aldehyde construct was compared with that of the parent Ac-Leu-Leu-Nle-H against cathepsin B and calpain. Despite the large size of the template the inhibition of cathepsin B was only slightly reduced in full agreement with the X-ray structure of this enzyme which shows full accessibility of the S-subsites. For this enzyme the 4C-spacer is apparently sufficient to guarantee optimal interaction of the peptide tail with the binding cleft. Conversely, for mu-calpain a significantly decreased inhibitory potency was obtained with the conjugate suggesting steric interference of the template in the binding process. These results show that the beneficial properties of the cyclodextrin template can be retained in conjugates with bioactive peptides if attention is paid to optimize in each case the size and nature of the spacer for optimal recognition of the grafted biomolecule.

Tumor protease-activated, pore-forming toxins from a combinatorial library

We describe a library of two-chain molecular complementation mutants of staphylococcal alpha-hemolysin that features a combinatorial cassette encoding thousands of protease recognition sites in the central pore-forming domain. The cassette is flanked by a peptide extension that inactivates the protein. We screened the library to identify alpha-hemolysins that are highly susceptible to activation by cathepsin B, a protease that is secreted by certain metastatic tumor cells. Toxins obtained by this procedure should be useful for the permeabilization of malignant cells thereby leading directly to cell death or permitting destruction of the cells with drugs that are normally membrane impermeant.

Potency and selectivity of the cathepsin L propeptide as an inhibitor of cysteine proteases

The cathepsin L propeptide (phcl-2) was expressed in Saccharomyces cerevisiae using a human procathepsin L/alpha-factor fusion construct containing a stop codon at position -1 (the C-terminal amino acid of the proregion). Since the yield after purification was very low, the cathepsin L propeptide was also obtained by an alternate procedure through controlled processing of an inactive mutant of procathepsin L (Cys25Ser/Thrl10Ala) expressed in Pichia pastoris, by small amounts of cathepsin L. The peptide resulting from the cleavage of the proenzyme (phcl-1) was then purified by HPLC. The purified propeptides were characterized by N-terminal sequencing and mass spectrometry and correspond to incomplete forms of the proregion (87 and 81 aa for phcl-1 and phcl-2 respectively, compared to 96 aa for the complete cathepsin L propeptide). The two peptides were found to be potent and selective inhibitors of cathepsin L at pH 5.5, with Ki values of 0.088 nM for phcl-1 and 0.66 nM for phcl-2. The Ki for inhibition of cathepsin S was much higher (44.6 nM with phcl-1), and no inhibition of cathepsin B or papain could be detected at up to 1 microM of the propeptide. The inhibitory activity was also found to be strongly pH-dependent. Two synthetic peptides of 75 and 44 aa corresponding to N-terminal truncated versions of the propeptide were also prepared by solid phase synthesis and displayed Ki values of 11 nM and 2900 nM, respectively, against cathepsin L. The data obtained for the 4 propeptide derivatives of various lengths indicate that the first 20 residues in the N-terminal region of the propeptide are more important for inhibition than the C-terminal region which contributes little to the overall inhibitory activity.

Monoclonal antibodies against Dermatophagoides group I allergens as pseudo-cystatins blocking the catalytic site of cysteine proteinases

The enzyme allergens Der p I and Der f I produced by the house dust mites Dermatophagoides pteronyssinus and D. farinae display partial sequence homology with other members of the cysteine proteinase superfamily. We report that certain widely used mouse mAbs against these Group I allergens indeed crossreact with the plant enzymes papain, bromelain and ficin. The recognition sites of these anti Group I mAbs comprise conformational and thermolabile epitopes involved in molding the catalytic center of the proteinases. Thus, the mAbs inhibit the enzymatic hydrolysis of specific chromogenic substrates by the Group I allergens, while specific cysteine proteinase inhibitors abolish the recognition of the enzymes by the mAbs. Similarly, activation of the thiol-proteases with L-cysteine abrogates their binding in the two-site mAb system, indicating that the mAbs recognize a proenzyme conformational peptide epitope. It follows that mAb-based assays for mite Group I components can neither detect the allergens after inactivation, nor in their fully activated forms.

Crystal structures of human procathepsin B at 3.2 and 3.3 Angstroms resolution reveal an interaction motif between a papain-like cysteine protease and its propeptide

A wild-type human procathepsin B was expressed, crystallized in two crystal forms and its crystal structure determined at 3.2 and 3.3 Angstroms resolution. The structure reveals that the propeptide folds on the cathepsin B surface, shielding the enzyme active site from exposure to solvent. The structure of the enzymatically active domains is virtually identical to that of the native enzyme [Musil et al. (1991) EMBO J. 10, 2321-2330]: the main difference is that the occluding loop residues are lifted above the body of the mature enzyme, supporting the propeptide structure.

Structure of rat procathepsin B: model for inhibition of cysteine protease activity by the proregion

BACKGROUND

Cysteine proteases of the papain superfamily are synthesized as inactive precursors with a 60-110 residue N-terminal prosegment. The propeptides are potent inhibitors of their parent proteases. Although the proregion binding mode has been elucidated for all other protease classes, that of the cysteine proteases remained elusive.

RESULTS

We report the three-dimensional structure of rat procathepsin B, determined at 2.8 A resolution. The 62-residue proregion does not form a globular structure on its own, but folds along the surface of mature cathepsin B. The N-terminal part of the proregion packs against a surface loop, with Trp24p (p indicating the proregion) playing a pivotal role in these interactions. Inhibition occurs by blocking access to the active site: part of the proregion enters the substrate-binding cleft in a similar manner to a natural substrate, but in a reverse orientation.

CONCLUSIONS

The structure of procathepsin B provides the first insight into the mode of interaction between a mature cysteine protease from the papain superfamily and its prosegment. Maturation results in only one loop of cathepsin B changing conformation significantly, replacing contacts lost by removal of the prosegment. Contrary to many other proproteases, no rearrangement of the N terminus occurs following activation. Binding of the prosegment involves interaction with regions of the enzyme remote from the substrate-binding cleft and suggests a novel strategy for inhibitor design. The region of the prosegment where the activating cleavage occurs makes little contact with the enzyme, leading to speculation on the activation mechanism.

Contribution to activity of histidine-aromatic, amide-aromatic, and aromatic-aromatic interactions in the extended catalytic site of cysteine proteinases

Within the papain family of cysteine proteinases few other residues in addition to the catalytic triad, Cys25-His159-Asn175 (papain numbering) are completely conserved [Berti & Storer (1995) J. Mol. Biol. 246, 273-283]. One such residue is tryptophan 177 which participates in a Trp-His-type interaction with the catalytic His159. In all enzymes of this class for which a three-dimensional structure has been reported, an additional highly conserved tryptophan, Trp181, also interacts with Trp177 via an aromatic-aromatic interaction in which the planes of the indole rings are essentially perpendicular. Also, both indole rings participate as pseudo-hydrogen bond acceptors in interactions with the two side chain amide protons of Asn175. Clearly, the proximity of Trp177 and Trp181 to the catalytic triad residues His159 and Asn175 and their network of interactions points to potential contributions of these aromatic residues to catalysis. In this paper, using cathepsin S, a naturally occurring variant that has a phenylalanine residue at position 181, we report the kinetic characterization of mutants of residues 175, 177, and 181. The results are interpreted in terms of the side chain contributions to catalytic activity and thiolate-imidazolium ion-pair stability. For example, the side chain of Asn175 has a major influence on the ion-pair stability presumably through its hydrogen bond to His159. The magnitude of this effect is modulated by Trp177, which shields the His159-Asn175 hydrogen bond from solvent. The His159-Trp177 interaction also contributes significantly to ion-pair stability; however, Trp181 and its interactions with Asn175 and Trp177 do not influence ion-pair stability to a significant degree. The observation that certain mutations at positions 177 and 181 result in a reduction of kcat/Km but do not appear to influence ion-pair stability probably reflects the contributions of these residues to substrate binding.

Structure of human beta-glucuronidase reveals candidate lysosomal targeting and active-site motifs

The X-ray structure of the homotetrameric lysosomal acid hydrolase, human beta-glucuronidase (332,000 Mr), has been determined at 2.6 A resolution. The tetramer has approximate dihedral symmetry and each promoter consists of three structural domains with topologies similar to a jelly roll barrel, an immunoglobulin constant domain and a TIM barrel respectively. Residues 179-204 form a beta-hairpin motif similar to the putative lysosomal targeting motif of cathepsin D, supporting the view that lysosomal targeting has a structural basis. The active site of the enzyme is formed from a large cleft at the interface of two monomers. Residues Glu 451 and Glu 540 are proposed to be important for catalysis. The structure establishes a framework for understanding mutations that lead to the human genetic disease mucopolysaccharidosis VII, and for using the enzyme in anti-cancer therapy.

Molecular modeling of substrate-enzyme reactions for the cysteine protease papain

AM1 quantum mechanical reaction coordinate (RC) calculations were run to simulate the rate-limiting deacylation (hydrolysis) reaction for a series of para-X-PhC(O)NHCH2-C(Y)-S-papain intermediates, where X = OCH3, CH3, H, Cl, NO2 and Y = O (thioester) or S (dithioester), for which a large body of structural, kinetic, and spectroscopic data is available. Several reaction zones, in particular the so-designated Large Zone and Small Zone, were extracted for these RC simulations from the fully solvated and energy-minimized X-ray crystal structure of papain (pdb9pap) bound to the appropriate substrate moiety. The major structural difference between these two zones was the absence of the oxyanion hole in the latter. For both the thioester and dithioester cases, the calculated Ea value associated with the parent (X = H) acyl-enzyme intermediate was lower by ca. 10 kcal/mol for the Large Zone than for the Small Zone. The magnitude of this difference suggests that the oxyanion hole plays a functional if not essential role in stabilizing the anionic tetrahedral intermediate with the cysteine proteases. The calculated Ea value was lower by ca. 10 kcal/mol for the thioester [-C(O)-S-] than for the corresponding dithioester [-C(S)-S-], in qualitative agreement with kinetic data for this series of substrates which reveal that the specific rate constant for deacylation k3 is ca. 60 times larger for the former. This difference is also consistent with both AM1 and 6-31G* calculations on model intermediates, which indicate that the weaker polarity of the dithioester compared with the thioester [i.e., -C(<--S)-S-versus-C(-->O)-S-] renders the former a much poorer site for nucleophilic attack. The anionic tetrahedral intermediate is energetically more stable for the dithioester than for the corresponding thioester, a finding that is discussed in terms of its kinetic and mechanistic implications. The mode of attack by the H2O nucleophile is "concerted" rather than "sequential" in terms of the order of proton abstraction by His-159 and nucleophilic attack on the acyl-enzyme intermediate. While the presumably key Sthiol . . . N nonbonded contact remained almost constant (ca. 2.90 A) up to formation of the [TS] structure, the substrate torsion angles phi and psi rotated significantly as the hybridization around the reaction site transforms from sp2 to sp3 during formation of the tetrahedral intermediate. The AM1-calculated frontier molecular orbitals for model thioester and dithioester acyl-enzyme intermediates generally associate the HOMOs with the reaction site and the LUMOs with the benzamide moiety. Computer graphics images corroborate our view that, in relation to the Sthiol . . . N interaction, the HOMOs and LUMOs should be identified, respectively, with Sthiol and N rather than the reverse, as suggested by other workers.

Hybrids of chicken cystatin with human kininogen domain 2 sequences exhibit novel inhibition of calpain, improved inhibition of actinidin and impaired inhibition of papain, cathepsin L and cathepsin B

Chicken cystatin and human kininogen domain 2 are members of the cystatin superfamily of protein-type cysteine proteinase inhibitors. They show structural and functional similarities, but only human kininogen domain 2 inhibits calpain. Using recombinant chicken cystatin as a scaffold for hybrid cassette analysis, the known reactive-site regions (N-terminus, first hairpin loop and second hairpin loop) were substituted by the corresponding sequences of human kininogen domain 2 in a single and combined manner. Seven hybrids were expressed, purified to homogeneity, characterized protein-chemically, and their inhibition of papain, actinidin, human cathepsin B, human cathepsin L and calpain (80-kDa subunit of rabbit skeletal muscle calpain II and porcine erthrocyte calpain 1) was determined. Strong but temporary inhibition of calpain by chicken cystatin hybrids carrying the N-terminus alone (variant sc1-KD2) or the N-terminus together with the first hairpin loop (variant sc1/2-KD2) was observed; hybrids of the second hairpin loop (sc3-KD2, sc1/3-KD2, sc2/3-KD2, sc1/2/3-KD2) were less strong calpain inhibitors. These data indicate that the inhibiton of calpain by human kininogen domain 2 requires the correct conformation and combination of several contact sites, and suggest that the N-terminus and the first hairpin loop play a major role in this ensemble. Remarkably, hybrid sc2-KD2 exhibited 5 or 150 times stronger inhibition of actinidin compared to native chicken cystatin or to proteolytically isolated human kininogen domain 2, respectively. This indicates an important role of the first hairpin loop of cystatins in the interaction with actinidin. Along with the impaired inhibition of cathepsin L, papain, actinidin, cathepsin B and calpain by the hybrids sc1/3-KD2, sc2/3-KD2 and sc1/2/3-KD2, these results support our hypothesis that all three predicted contact regions of kininogen domain 2 contribute to binding in the active-site clefts of papain-like enzymes in a finely balanced manner.

Large-scale purification and preliminary x-ray diffraction studies of human aspartylglucosaminidase

Aspartylglucosaminidase (AGA) is a lysosomal asparaginase that takes part in the ordered degradation of glycoproteins and a deficiency of which results in a lysosomal accumulation disease aspartylglucosaminuria in human. The mature enzyme consists of 24-kDa and 17-kDa subunits, which are both heterogeneously glycosylated. Activation of the enzyme from a single precursor polypeptide into two subunits is accomplished in the endoplasmic reticulum (ER). The relative lack of this proteolytic capacity in several tested high-producing expression systems has complicated the production of active recombinant enzyme in high quantities, which would be an alternative for purification of this molecule for crystallization. Consequently, the AGA enzyme has to be purified directly from cellular or tissue sources for crystallographic analysis. Here we describe a large-scale purification method to produce milligram amounts of homogeneous AGA from human leukocytes. The purified AGA enzyme represents a heterogeneous pool of molecules not only due to glycosylation, but also heterogeneity at the polypeptide level, as demonstrated here. We were able to isolate a homogeneous peptide pool that was successfully crystallized and preliminary X-ray data collected from the crystals. The crystals diffract well to 2.0 angstroms and are thus suitable for determination of the crystal structure of AGA.

Studies on activation and inhibition of cathepsin B from buffalo liver

Cathepsin B (EC 3.4.22.1) was purified from buffalo liver. The enzyme activity against alpha-benzoyl-DL-arginine-naphthylamine (BANA) was substantially reduced by heat (above 37 degrees C) and by nondenaturing concentrations of urea (3 M) and guanidine hydrochloride (1 M). Cathepsin B was significantly activated by 1.5 mM EDTA alone. The activation of the enzyme was further enhanced in the presence of thiol compounds, e.g., cysteine thioglycolic acid, 2,3-dimercapto-1-propenol, and dithioerythritol (DTE). The minimum concentration of the thiol compound required for optimal activation of cathepsin B was found to be lowest (0.2 mM) for DTE. The BANA hydrolyzing activity of cathepsin B was substantially reduced by Cu2+ (20-200 microM) and Ca2+ (30-250 mM) as well as by thiol blocking reagents, e.g., iodoacetate, 5,5'-dithiobis(2-nitro-benzoic acid) (DTNB), and p-hydroxymercuribenzoate (pHMB). The enzyme activity was completely abolished when the molar ratio of the reagent: cathepsin B was close to 1. The number of free sulfhydryl groups in cathepsin B was determined to be 2 by titration against DTNB and pHMB. Modification of one free thiol group of cathepsin B resulted in complete loss of BANA hydrolyzing activity.

Identification of two novel Dictyostelium discoideum cysteine proteinases that carry N-acetylglucosamine-1-P-modification

Dictyostelium discoideum makes multiple developmentally regulated lysosomal cysteine proteinases. One of these, a lysosomal enzyme called proteinase I, contains a cluster of GlcNAc-alpha-1-P-Ser residues. We call this phosphoglycosylation. To study its function, a cDNA library from vegetative cells was screened, and two novel cysteine proteinase clones were characterized (cprD and cprE). Each of them has highly conserved regions expected for cysteine proteinases, but unlike any other, each has a serine-rich domain containing three distinct motifs, poly-S, SGSQ, and SGSG. cprD and cprE cDNAs were overexpressed in Dictyostelium and the active enzymes identified. cprD codes for a protein of approximately 36 kDa (CP4), which is recognized by monoclonal antibodies against GlcNAc-1-P and fucose. cprE corresponds to a 29-kDa protein, which is recognized by antibodies against GlcNAc-1-P. mRNA for both enzymes is present in the vegetative phase and increases during growth on bacteria but decreases throughout development. When the formation of the fruiting body is complete the mRNA for both messages is detected again but in very low levels. Having cloned cDNAs for proteins that carry GlcNAc-1-P should allow us to probe the function of the carbohydrate in these putative lysosomal enzymes.

Three-dimensional structure of human lysosomal aspartylglucosaminidase

The high resolution crystal structure of human lysosomal aspartylglucosaminidase (AGA) has been determined. This lysosomal enzyme is synthesized as a single polypeptide precursor, which is immediately post-translationally cleaved into alpha- and beta-subunits. Two alpha- and beta-chains are found to pack together forming the final heterotetrameric structure. The catalytically essential residue, the N-terminal threonine of the beta-chain is situated in the deep pocket of the funnel-shaped active site. On the basis of the structure of the enzyme-product complex we present a catalytic mechanism for this lysosomal enzyme with an exceptionally high pH optimum. The three-dimensional structure also allows the prediction of the structural consequences of human mutations resulting in aspartylglucosaminuria (AGU), a lysosomal storage disease.

Three-dimensional structure of the human 'protective protein': structure of the precursor form suggests a complex activation mechanism

BACKGROUND

The human 'protective protein' (HPP) forms a multi-enzyme complex with beta-galactosidase and neuraminidase in the lysosomes, protecting these two glycosidases from degradation. In humans, deficiency of HPP leads to the lysosomal storage disease galactosialidosis. Proteolytic cleavage of the precursor form of HPP involves removal of a 2 kDa excision peptide and results in a carboxypeptidase activity. The physiological relevance of this activity is, as yet, unknown.

RESULTS

The crystal structure of the 108 kDa dimer of the precursor HPP has been elucidated by making extensive use of twofold density averaging. The monomer consists of a 'core' domain and a 'cap' domain. Comparison with the distantly related wheat serine carboxypeptidase dimer shows that the two subunits in the HPP dimer differ by 15 degrees in mutual orientation. Also, the helical subdomain forming part of the cap domains is very different. In addition, the HPP precursor cap domain contains a 'maturation' subdomain of 49 residues which fills the active-site cleft. Merely removing the 'excision' peptide located in the maturation subdomain does not render the catalytic triad solvent accessible.

CONCLUSIONS

The activation mechanism of HPP is unique among proteases with known structure. It differs from the serine proteases in that the active site is performed in the zymogen, but is blocked by a maturation subdomain. In contrast to the zinc metalloproteases and aspartic proteases, the chain segment physically rendering the catalytic triad solvent inaccessible in HPP is not cleaved off to form the active enzyme. The activation must be a multi-step process involving removal of the excision peptide and major conformational changes of the maturation subdomain, whereas the conformation of the enzymatic machinery is probably almost, or completely, unaffected.

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

Peptide segments derived from consensus sequences of the inhibitory site of cystatins, the natural inhibitors of cysteine proteinases, were used to develop new substrates and inhibitors of papain and rat liver cathepsins B, H, and L. Papain hydrolyzed Abz-QVVAGA-EDDnp and Abz-LVGGA-EDDnp at about the same rate, with specificity constants in the 10(7) M-1 sec-1 range; cathepsin L also hydrolyzes both substrates with specificity constants in the 10(5) M-1 sec-1 range due to lower k(cat) values, with the Km's being identical to those with papain. Only Abz-LVGGA-EDDnp was rapidly hydrolyzed by cathepsin B, and to a lesser extent by cathepsin H. Peptide substrates that alternate these two building blocks (LVGGQVVAGAPWK and QVVAGALVGGAPWK) discriminate the activities of cathepsins B and L and papain. Cathepsin L was highly selective for cleavage at the G-G bond of the LVGG fragment in both peptides. Papain and cathepsin B cleaved either the LVGG fragment or the QVVAG fragment, depending on their position within the peptide. While papain was more specific for the segment located C-terminally, cathepsin B was specific for that in N-terminal position. Peptidyl diazomethylketone inhibitors based on these two sequences also reacted differently with papain and cathepsins. GlcA-QVVA-CHN2 was a potent inhibitor of papain and reacted with papain 60 times more rapidly (k + 0 = 1,100,000 M-1 sec-1) than with cathepsin L, and 220 times more rapidly than with cathepsin B. Cathepsins B and L were preferentially inhibited by Z-RLVG-CHN2. Thus cystatin-derived peptides provide a valuable frame-work for designing sensitive, selective substrates and inhibitors of cysteine proteinases.

The baculovirus cysteine protease has a cathepsin B-like S2-subsite specificity

Autographa californica nuclear polyhedrosis virus (AcNPV) encodes a functional cysteine protease of the papain family which is expressed after infection in Spodoptera fruglperda Sf9 cells. The protease displays an inhibition profile typical for cysteine proteases and is highly active against synthetic peptide substrates. The pH optimum of the bell-shaped pH-activity curve is between 5.0 and 5.5. The best substrate tested is Z-Arg-Arg-MCA which is specific for cathepsin B. The specificity constant (Kcat/Km) of AcNPV protease for this substrate is approximately two times higher than for human cathepsin B. In contrast to human cathepsins, AcNPV protease does not exhibit a discriminating specificity towards neutral hydrophobic residues in the P2 position. These substrates are hydrolysed at a ten-fold lower rate than the P2 arginine containing substrate. The pH activity profile against the Z-Arg-Arg-MCA substrate reveals a pK of 5.35 which can be assigned to a glutamate residue in the S2 subsite pocket. Like in cathepsin B, this residue facilitates the binding of positively charged P2 residues in the primary binding pocket. In this respect, the AcNPV protease resembles cathepsin B more than cathepsins L and S.

Isolation and expression pattern of a cDNA encoding a cathepsin B-like protease from Nicotiana rustica

Sequence analysis of a 1.33 kb clone from a root cDNA library of Nicotiana rustica revealed an open reading frame encoding a protein of 356 amino acids. The deduced protein has high levels of homology to human cathepsin B protease and a cathepsin B-like cysteine protease from wheat but much lower levels of homology with other plant cysteine proteinases. Southern blotting experiments suggest a limited number of cathepsin B-like genes are present in the genome of N. rustica and also that of N. tabacum. RNA analysis involving a range of tissues, harvested from both Nicotiana species 4-5 h after the beginning of a 16 h photoperiod, revealed the cathepsin B-like gene was being expressed strongly in roots, stem and developing flowers but weakly in mature leaves. Further analysis of RNA extracted from leaf tissue of N. tabacum revealed the gene showed rhythmic expression and also that its expression increased in response to wounding. Analysis of leaf tissues harvested during the latter part of a 16 h photoperiod (11 and 16 h after illumination commenced) showed that transcript levels were two three times higher than in leaf tissue harvested either towards the end of the dark period or 5 h after illumination commenced. When leaf tissue was wounded at 11:00 (5 h after plants were illuminated), and harvested for RNA extraction 6 h later, the level of cathepsin B-like transcript in mesophyll tissue was found to be increased ca. 2-fold relative to the level detected in unwounded controls.

Crystal structure of a conserved protease that binds DNA: the bleomycin hydrolase, Gal6

Bleomycin hydrolase is a cysteine protease that hydrolyzes the anticancer drug bleomycin. The homolog in yeast, Gal6, has recently been identified and found to bind DNA and to act as a repressor in the Gal4 regulatory system. The crystal structure of Gal6 at 2.2 A resolution reveals a hexameric structure with a prominent central channel. The papain-like active sites are situated within the central channel, in a manner resembling the organization of active sites in the proteasome. The Gal6 channel is lined with 60 lysine residues from the six subunits, suggesting a role in DNA binding. The carboxyl-terminal arm of Gal6 extends into the active site cleft and may serve a regulatory function. Rather than each residing in distinct, separable domains, the protease and DNA-binding activities appear structurally intertwined in the hexamer, implying a coupling of these two activities.

Immunolocalization of cathepsin B in human glioma: implications for tumor invasion and angiogenesis

The poor prognosis of patients with malignant gliomas is at least partially due to the invasive nature of these tumors. In this study, the authors investigated the possibility that the cysteine protease cathepsin B (CB) is a participant in the process of glial tumor cell invasion. To accomplish this, an immunohistochemical analysis was made of the localization of antibodies to CB in biopsies of five specimens of normal brain, 16 astrocytomas, 33 anaplastic astrocytomas, and 33 glioblastomas multiforme. Staining was scored according to the percentage of positive cells and the intensity of the stain, graded from 0 to 3+. Staining for CB was not seen in any of five samples of normal brain except for occasional neuronal cell bodies and microglia. Only five (31%) of 16 astrocytomas showed a small percentage of positive cells (0.01%-3%) that were stained in a light, diffuse cytoplasmic pattern (1+). Twenty-nine (87.8%) of 33 anaplastic astrocytomas showed positive light, granular staining in 2% to 40% of cells. In anaplastic astrocytoma, the staining within a tumor was heterogeneous with intensities of 1+ (17%), 1+ to 2+ (29%), or 2+ (55%). In contrast, all 33 (100%) glioblastomas were positive in 10% to 90% of cells. The staining was present in a coarse, granular pattern with an intensity of 2+ (12%) or 3+ (88%). Tumor cells infiltrating into brain adjacent to malignant gliomas stained positively in 26 cases that could be evaluated for glioblastoma multiforme; these invading cells frequently followed penetrating blood vessels as typical "secondary structures of Scherer." Moderate to intense CB staining associated with endothelial proliferation in high-grade tumors was also observed, especially in regions of tumor infiltration into adjacent normal brain. These results provide evidence consistent with the hypothesis that CB is functionally significant in the process of tumor invasion and angiogenesis in the clinical progression of the malignant phenotype in astrocytes.

Expression of full-length human procathepsin L cDNA in Escherichia coli and refolding of the expression product

From human embrional lung fibroblasts mRNA was obtained and converted to cDNA. The procathepsin L coding region was amplified by PCR, inserted into pALTER and, after checking the nucleotide sequence, transferred into pET81F1+. Procathepsin L was expressed by induction of recombinant E. coli strain BL21[DE3](pLysS) with IPTG and was found to be deposited into inclusion bodies. These were isolated and solubilized in guanidinium hydrochloride. The soluble proteins were sulphonated and procathepsin L was obtained after gel filtration. Purified proenzyme was refolded by dialysis and autoactivated into a form of the expected size and enzymatic activity against a fluorogenic substrate.

Characterization and localization of cathepsin B proteinases expressed by adult Ancylostoma caninum hookworms

The hookworm Ancylostoma caninum induces human eosinophilic enteritis (EE), probably via allergic responses to its secretions. Cysteine and metallo-proteinases may be the components of these secretions that elicit hypersensitivity reactions. In order to characterize genes encoding cysteine proteinases (CP) secreted by A. caninum, an adult hookworm cDNA library was constructed and screened with a cloned fragment of a hookworm CP gene. This fragment was obtained using consensus oligonucleotide, CP-gene-specific primers in the polymerase chain reaction. cDNAs encoding two CPs were obtained from the library and sequenced. The first gene, AcCP-1, encoded a cathepsin B-like zymogen CP of 343 amino acids (aa), predicted to be processed in vivo into a mature CP of 255 aa. Closest nucleotide identities were to Haemonchus contortus cysteine protease (61%) and to human cathepsin B (60%). The second gene, AcCP-2, encoded a mature CP of 254 aa, that showed 86% identity to AcCP-1, and 58% and 47% identity to bovine cathepsin B and human cathepsin B, respectively. Rabbit antisera raised against recombinant AcCP-1 reacted with esophageal, amphidial and excretory glands in formalin-fixed, paraffin embedded sections of both male and female adult hookworms, and with an antigen of approx. 40 kDa in Western blot analysis of excretory/secretory products from adult hookworms. Together, these immuno-hybridization results strongly suggest that the CP encoded by the AcCP-1 gene is secreted by hookworms. These are the first reported CP genes from hookworms. Proteinases encoded by these genes may be responsible for the CP activity that we have shown previously to be secreted by adult A. caninum.

The preparation of catalytically active human cathepsin B from its precursor expressed in Escherichia coli in the form of inclusion bodies

A cDNA clone encoding human procathepsin B was expressed at a high level in Escherichia coli using a T7 polymerase expression system, resulting in the formation of insoluble cytoplasmic protein aggregates (inclusion bodies). The recombinant product was solubilized and renatured by refolding and reoxidation. The proenzyme was subsequently processed with pepsin to produce an enzymically active enzyme. By systematic variation of the parameters influencing the folding, formation of disulphide bonds, and processing of procathepsin B to the catalytically active mature form, a simple renaturation procedure was designed, allowing the production of about 3 mg purified active cathepsin B/l E. coli culture broth. The enzyme obtained in this way consists of a single chain and, as a consequence of pepsin treatment, possesses a three-amino-acid extension at its N-terminus. The enzyme has similar kinetic and immunological properties to native human cathepsin B.

Regulation of the activity of lysosomal cysteine proteinases by pH-induced inactivation and/or endogenous protein inhibitors, cystatins

The kinetics of pH-induced inactivation of human cathepsins B and L was studied by conventional and stopped-flow methods. The inactivation of both enzymes was found to be an irreversible, first-order process. The inactivation rate constants increased exponentially with pH for both enzymes. From log kinac vs pH plots, 3.0 and 1.7 protons were calculated to be desorbed for pH-induced inactivation of cathepsins L and B. Cathepsin B was thus substantially more stable than cathepsin L (approximately 15-fold at pH 7.0 and 37 degrees C). Cathepsin B was efficiently inhibited by cystatin C at pH 7.4, whereas the inhibition by stefin B and high molecular weight kininogen was only moderate. In contrast, cathepsin L was efficiently inhibited by both chicken cystatin and stefin B at this pH kass approximately 3.3 x 10(7) m-1 s-1).

Hairpin loop mutations of chicken cystatin have different effects on the inhibition of cathepsin B, cathepsin L and papain

Five recombinant hairpin loop variants of chicken cystatin (delta V55, delta V55-S56, delta P103-L105, delta I102-Q107, loop2-KD2) were constructed by cassette mutagenesis, expressed in E. coli, purified to homogeneity, characterized by protein-chemical means and by their inhibitory properties. The variant forms, modified in two of the three postulated cysteine proteinase binding regions, were inhibitorily active. However, the equilibrium dissociation constants of the complexes between papain as well as human cathepsin B or L and the cystatin variants show a weaker affinity for all three enzymes compared with recombinant chicken cystatin. These results prove the contribution of both hairpin loops to complex formation with the three enzymes. Furthermore, the kinetic constants indicate discrete differences in the molecular mechanism of interaction between chicken cystatin and papain, cathepsin B, and cathepsin L. Inhibition of cathepsin L was much less affected than inhibition of papain or cathepsin B by the modifications achieved in the five variants. Remarkably, at high enzyme concentration (above 0.5 nM) inhibition of papain by these variants was 'temporary', that means, active papain was released from the enzyme-inhibitor complex within minutes to hours (compare [1]).

Structural basis for the biological specificity of cystatin C. Identification of leucine 9 in the N-terminal binding region as a selectivity-conferring residue in the inhibition of mammalian cysteine peptidases

The structural basis for the biological specificity of human cystatin C has been investigated. Cystatin C and other inhibitors belonging to family 2 of the cystatin superfamily interact reversibly with target peptidases, seemingly by independent affinity contributions from a wedge-shaped binding region built from two loop-forming inhibitor segments and a binding region corresponding to the N-terminal segment of the inhibitor. Human cystatin C variants with Gly substitutions for residues Arg-8, Leu-9, and/or Val-10 of the N-terminal binding region, and/or the evolutionarily conserved Trp-106 in the wedge-shaped binding region, were produced by site-directed mutagenesis and Escherichia coli expression. A total of 10 variants were isolated, structurally verified, and compared to wild-type cystatin C with respect to inhibition of the mammalian cysteine peptidases, cathepsins B, H, L, and S. Varying contributions from the N-terminal binding region and the wedge-shaped binding region to cystatin C affinity for the four target peptidases were observed. Interactions from the side chains of residues in the N-terminal binding region and Trp-106 are jointly responsible for the major part of cystatin C affinity for cathepsin L and are also of considerable importance for cathepsin B and H affinity. In contrast, for cathepsin S inhibition these interactions are of lesser significance, as reflected by a Ki value of 10(-8) M for the cystatin C variant devoid of Arg-8, Leu-9, Val-10, and Trp-106 side chains. The side chain of Val-10 is responsible for most of the affinity contribution from the N-terminal binding region, for all four enzymes. The contribution of the Arg-8 side chain is minor, but significant for cystatin C interaction with cathepsin B. The Leu-9 side chain confers selectivity to the inhibition of the target peptidases; it contributes to cathepsin B and L affinity by factors of 200 and 50, respectively, to cathepsin S binding by a factor of 5 only, and results in a 10-fold decreased affinity between cystatin C and cathepsin H.

Crystal structures of recombinant rat cathepsin B and a cathepsin B-inhibitor complex. Implications for structure-based inhibitor design

The lysosomal cysteine proteinase cathepsin B (EC 3.4.22.1) plays an important role in protein catabolism and has also been implicated in various disease states. The crystal structures of two forms of native recombinant rat cathepsin B have been determined. The overall folding of rat cathepsin B was shown to be very similar to that of the human liver enzyme. The structure of the native enzyme containing an underivatized active site cysteine (Cys29) showed the active enzyme conformation to be similar to that determined previously for the oxidized form. In a second structure Cys29 was derivatized with the reversible blocking reagent pyridyl disulfide. In this structure large side chain conformational changes were observed for the two key catalytic residues Cys29 and His199, demonstrating the potential flexibility of these side chains. In addition the structure of the complex between rat cathepsin B and the inhibitor benzyloxycarbonyl-Arg-Ser(O-Bzl) chloromethylketone was determined. The complex structure showed that very little conformational change occurs in the enzyme upon inhibitor binding. It also allowed visualization of the interaction between the enzyme and inhibitor. In particular the interaction between Glu245 and the P2 Arg residue was clearly demonstrated, and it was found that the benzyl group of the P1 substrate residue occupies a large hydrophobic pocket thought to represent the S'1 subsite. This may have important implications for structure-based design of cathepsin B inhibitors.

Cathepsin B in angiogenesis of human prostate: an immunohistochemical and immunoelectron microscopic analysis

BACKGROUND

Angiogenesis (or neovascularization) is required for the growth of solid organ tumors and precedes invasion of the adjacent stroma by neoplastic cells. We investigated the relative density and distribution of cathepsin B (CB) immunostained microvessels (i.e., small blood vessels and capillaries) in benign prostatic hyperplasia (BPH), prostatic intraepithelial neoplasia (PIN), and prostatic adenocarcinoma (CAP) by immunocytochemical localization of an antibody directed against a cathepsin B-derived synthetic peptide (Syn-CB).

METHODS

We studied 16 formalin-fixed, prostatectomy specimens that were embedded in paraffin/paraplast for histological examination by hematoxylin and eosin and immuno-localization of the Syn-CB antibody. Selected paraformaldehyde-fixed specimens were embedded in K4M Lowicryl or LRWhite resins. We localized the antibody in thin sections using immunoelectron microscopy techniques.

RESULTS

Eight patients had BPH [4 patients with BPH alone, 2 with BPH and PIN, and 2 with BPH and CAP]. Ten cancer cases included one with Gleason histologic score 4, two with score 6, four with score 7, and three with score 8. In CAP cases, Gleason score 6 and 7 tumors had more microvessels than the score 4 or 8 tumors. In both BPH and CAP cases, the antibody was localized chiefly in the endothelial cells of microvessels, but occasionally in ductal and glandular epithelial cells. Ultrastructurally, CB-immunoreactive gold particles were markedly increased at the luminal and basal plasma membrane surfaces and folds of endothelial cells in neoplastic prostate, but not in the endothelial cells of BPH. Furthermore, the presence of CB localizing gold particles in collagen and smooth muscle fibers near the microvessels indicated leakage of the enzyme in prostatic stroma of neoplastic prostate. Similar leakage was not observed in BPH. Morphometric analysis showed that the relative density of microvessels increased two to three times in cancer patients when compared to patients with BPH alone. Our study also indicated that BPH associated with PIN or CAP had an increased density of microvessels when compared to BPH alone.

CONCLUSIONS

Our study showed that the relative density and distribution of microvessels are the most important features of neovascularization in prostatic tumors. The relative density of microvessels increased in PIN and CAP when compared to BPH alone. Although the localization of CB is associated with lysosomes of endothelial cells in both BPH and CAP, there is a greater association of CB with the plasma membranes of endothelial cells in CAP than BPH. Immunoelectron microscopy provided evidence that CB might be involved in dissolution of basement membranes in neoplastic tumors during angiogenesis. CB localization has the potential of defining a role for this protease in degradation of extracellular matrix constituents during early steps of angiogenesis.

Identification of bovine stefin A, a novel protein inhibitor of cysteine proteinases

For the first time, three different stefins, A, B and C, have been isolated from a single species. The complete amino acid sequence of bovine stefin A was determined. The inhibitor, with a calculated M(r) of 11,123, consists of 98 amino acid residues. Although it exhibits considerable similarity to human and rat stefin A, some significant differences in inhibition kinetics were found. Bovine stefin A bound tightly and rapidly to cathepsin L (kass = 9.6 x 10(6) M-1.s-1, Ki = 29 pM). The binding to cathepsin H was also rapid (kass = 2.1 x 10(6) M-1.s-1), but weaker (Ki = 0.4 nM) due to a higher dissociation rate constant. In contrast, the binding to cathepsin B was much slower (kass = 1.4 x 10(5) M-1.s-1), but still tight (Ki = 1.9 nM).

The role of phagocyte proteinases and proteinase inhibitors in multiple organ failure

Although numerous other inflammatory mediators are important, the following review of our research and that of other authors reveals a prominent role for the phagocyte proteinases, polymorphonuclear (PMN) elastase and cathepsin B, in the development of multiple organ failure. The release of these enzymes in relation to the severity of trauma- and/or infection-induced inflammation was clearly verified in a variety of clinical studies. The amounts of the extracellularly discharged phagocyte proteinases were highly predictive of forthcoming organ failure and ultimate patient outcome. Moreover, the consumption of important proteinase inhibitors (e.g., alpha 1-proteinase inhibitor, antithrombin III) and other plasma proteins (e.g., fibrinogen), which are highly susceptible to proteolytic degradation, coincided with the occurrence of proteolytic activity, especially that of PMN elastase. Therefore, the therapeutic use of specific PMN elastase and/or thrombin inhibitors should prevent multiple organ failure or at least reduce severe signs of inflammation.

The role of thiol proteases in tissue injury and remodeling

Human lung macrophages express all four of the known lysosomal thiol proteases: cathepsins B, H, L, and S. These enzymes share a similar size and targeting mechanism for lysosomal accumulation and all have relatively indiscriminate substrate specificity in comparison with such highly selective serine proteases as urokinase or thrombin. These enzymes do have distinctive properties: only cathepsin B has C-terminal dipeptidase activity, only cathepsin H has potent aminopeptidase activity, and only cathepsin L and S are elastolytic. Cathepsin S is unique in that it is stable at neutral pH; indeed, at neutral pH it has elastolytic activity roughly comparable with that of neutrophil elastase. Recent studies of the differential expression of these cathepsins suggest they not only cooperate in terminal degradation of endocytized protein but also have specific functions such as proenzyme activation, antigen processing, and tissue remodeling, especially bone matrix resorption. Lysates of lung macrophages degrade elastin at neutral pH, suggesting that necrosis of macrophages at sites of macrophage accumulation, e.g., caseation necrosis, could contribute to tissue destruction. Tissue destruction and remodeling by thiol proteases expressed by live macrophages, however, is limited by tight compartmentalization of cathepsins to lysosomes. Nonetheless, macrophages accumulate at sites of known injury in cigarette smokers. Because these cells contain potent elastases, and because lysosomal enzyme release and cell surface acidification are regulated events, dysregulation of thiol protease expression in stimulated macrophages may contribute to the injury observed in cigarette smokers with non-alpha-1-protease inhibitor-type emphysema.

Elongation on the amino-terminal part of stefin B decreases inhibition of cathepsin H

Two mutants of the cysteine proteinase inhibitor, stefin B, were prepared by ligating the amino-terminal region from cystatin C and kininogen, members of two other families of cystatin superfamily. The mutant proteins were expressed in Escherichia coli and purified to homogeneity. Inhibition and kinetic constants were determined for authentic and mutated stefins against the four different cysteine proteinases, papain and human cathepsins B, L and H. Inhibition of both amino-terminal elongated stefin B mutants was decreased particularly for cathepsin H. A model of the tertiary structure of cathepsin H and its complex with stefin B was constructed. The framework for the model of cathepsin H consisted of structurally conserved regions from tertiary structures of three cysteine proteinases. Variable regions were selected from fragments of other proteins from the protein data base. We suggest that reduced binding of stefins with elongated amino termini is caused by the mini chain of cathepsin H which is probably in close proximity to the amino termini in the complexes. This mini chain is bridged to Cys214 and has already been proposed to be responsible for the aminopeptidase activity of cathepsin H. We conclude that the amino-terminal region of stefin B plays an important role in determining the strength of inhibition of cathepsin H.

Sequence and structure similarities of cathepsin B from the parasite Schistosoma mansoni and human liver

A three-dimensional structure of Schistosoma mansoni cathepsin B was modelled using the coordinates of the crystal structure of the human liver enzyme. Both enzymes appear to share remarkable structural similarity. However, an examination of the models complexed with two synthetic inhibitors revealed differences in inhibitor binding, as confirmed by differences in the 50% inhibitory concentration of the same inhibitor.

Structure/function implications for the aminopeptidase specificity of aleurain

The cysteine protease aleurain, a member of the papain superfamily, was characterized by its specificity constants, kcat/Km, for the hydrolysis of different substrates of the type H-P1-NH-Mec (NH-Mec, 4-methylcoumaryl-7-amide). The determined constants for the different substrates decrease in the order citrulline > Arg = Phe >> Ala. A 75-fold loss of specificity was observed when the substrate Bz-Arg-NH-Mec (Bz, benzoyl), with a blocked N-terminus, was used instead of H-Arg-NH-Mec. The pH dependence of kcat/Km for H-Arg-NH-Mec was bell-shaped with pKa1 and pKa2 values of 5.81 and 7.27, respectively, at 25 degrees C. The residue corresponding to a pKa1 value of 5.81 could be identified by its ionisation enthalpy, delta Hion, of 15 kJ/mol as a carboxylate group of the enzyme interacting electrostatically with the residue with pKa2 7.27, attributed to the alpha-amino group of the substrate by its delta Hion value of 48 kJ/mol. Aleurain can be titrated at the active site with L-trans-epoxy-succinylleucylamido(4-guanidino)butane, and the reaction was characterized by its association rate constant of 19,000 M-1.s-1. Native chicken cystatin inhibited aleurain competitively with Ki 133 nM. Recombinant chicken cystatin variants Ala-Glu-Phe-[Met1, Ile29, Leu89] chicken egg-white cystatin, (variant 1) and the N-terminally truncated form des-(S1-P11)-[Ala12, Glu12, Phe14, Met15, Leu89]-chicken egg-white cystatin, (variant 2), inhibited aleurain competitively with Ki values of 125 nM and 5 microM, respectively. Implications for the aminopeptidase activity of aleurain are discussed using cathepsin H for comparison.

Immunological significances of invariant chain from the aspect of its structural homology with the cystatin family

The primary structure of p31 of invariant chain (Ii-chain) shows about 50% homology with those of the cystatin family which are endogenous cysteine protease inhibitors. The binding domains between Ii-chain and HLA-DR-7 were estimated from the structural homology between cystatin and Ii-chain and also between cathepsins and DR-7, respectively. The QL64-71 and GS76-88 of Ii-chain were estimated to be the binding domains with GG45-51 and VS57-63 of HLA-DR7, respectively. The purified human Ii-chain from spleen is capable of forming four molecular forms from monomer to tetramer by redox-potential dependent disulfide bond formation. The Ii-chain inhibits cathepsin L and H competitively as a dimer and the K(i) value for cathepsin L was 4.1 x 10(-8) M, but cathepsin B was not inhibited at all. The Ii-chain showed mainly a dimer (60 kDa) under the assay condition of cathepsins with cysteine and was not degraded by these cathepsins. The Ii-chain may play an important role in the regulation of antigenic peptide presentation to MHC class II.

Crystal structure of the cysteine protease interleukin-1 beta-converting enzyme: a (p20/p10)2 homodimer

Interleukin-1 beta-converting enzyme (ICE) proteolytically cleaves pro-IL-1 beta to its mature, active form. The crystal structure at 2.5 A resolution of a recombinant human ICE-tetrapeptide chloromethylketone complex reveals that the holoenzyme is a homodimer of catalytic domains, each of which contains a p20 and a p10 subunit. The spatial separation of the C-terminus of p20 and the N-terminus of p10 in each domain suggests two alternative pathways of assembly and activation in vivo. ICE is homologous to the C. elegans cell death gene product, CED-3, and these may represent a novel class of cytoplasmic cysteine proteases that are important in programmed cell death (apoptosis). Conservation among members of the ICE/CED-3 family of the amino acids that form the active site region of ICE supports the hypothesis that they share functional similarities.

Families of cysteine peptidases

This chapter presents families of cysteine peptidases. The activity of all cysteine peptidases depends on a catalytic dyad of cysteine and histidine. The order of the cysteine and histidine residues (Cys/His or His/Cys) in the linear sequence differs between families and this is among the lines of evidence suggesting that cysteine peptidases have had many separate evolutionary origins. The families C1, C2, and C10 can be described as “papainlike,” and form clan CA. The papain family contains peptidases with a wide variety of activities, including endopeptidases with broad specificity, endopeptidases with narrow specificity, aminopeptidases, and peptidases with both endopeptidase and exopeptidase activities. Papain homologs are generally either lysosomal or secreted proteins. The calpain family includes the calcium-dependent cytosolic endopeptidase calpain, which is known from birds and mammals, and the product of the sol gene in Drosophila. Calpain is a complex of two peptide chains. Picornains are a family of polyprotein-processing endopeptidases from single-stranded RNA viruses. Each picornavirus has two picornains (2A and 3C).

Mechanism and regulation of antigen processing by cathepsin B

Cellular and humoral immune responses to vaccines of hepatitis B and rabies as antigens were suppressed by specific inhibitors of cathepsin B, anti-cathepsin B antibody and the specific substrate of cathepsin B. The antigenic peptides of these vaccines are processed by cathepsin B and the fragments are capable of binding with the desetope of MHC class II, beta-chain, because one of the active sites of cathepsin B (14, 15) VN217-222 shares high homology with a part of the desetope, VN57-62, of MHC class II, beta-chain. Rechallenge of the synthesized antigenic peptides of these vaccine molecules shows a strong proliferative response to the splenocyte primed by these vaccines. However, the response to these antigenic peptides was not inhibited by cathepsin B inhibitors. These findings suggest that cathepsin B inhibitors do not inhibit any other processes of immune responses than the proteolytic processing of antigens. Some investigators reported recently that the Ii-chain is degraded by purified cathepsin B in vitro (23-25). However, we showed that the suppression of these immune responses by cathepsin B inhibitors is not due to the inhibition of invariant chain degradation. We found that the invariant chain shares about 40% homology with the cystatin family which are the endogenous inhibitors of cysteine proteases (23, 24). Therefore, the Ii-chain is one of the members of the cystatin superfamily and may participate in the regulation of presentation of antigenic peptides and also antigen processing by cathepsin B.

Molecular cloning of cDNA for the 29 kDa proteinase participating in decomposition of the larval fat body during metamorphosis of Sarcophaga peregrina (flesh fly)

A cDNA clone for the 29 kDa proteinase participating in tissue disintegration during metamorphosis of Sarcophaga was isolated. This proteinase, named Sarcophaga cathepsin B, consisted of 256 amino acid residues, and contained three putative N-glycosylation sites. By comparison with other cathepsins B, its unique substrate specificity was partly explained by Ala at position 248.