Amino acid substitutions in the N-terminal segment of cystatin C create selective protein inhibitors of lysosomal cysteine proteinases

We used site-directed mutagenesis to alter the specificity of human cystatin C, an inhibitor with a broad reactivity against cysteine proteinases. Nine cystatin C variants containing amino acid substitutions in the N-terminal (L9W, V10W, V10F and V10R) and/or the C-terminal (W106G) enzyme-binding regions were designed and produced in Escherichia coli. It was discovered that the inhibition profile of the cystatin could be altered by changing residues 9 and 10, which are proposed to bind in the S3 and S2 substrate-binding pockets respectively of the enzymes. All of the variants with substitutions in the N-terminal segment displayed decreased binding to cathepsins B and H, indicating that the S3 and S2 pockets of these enzymes cannot easily accommodate large aromatic residues. The introduction of a charged residue into S2 (variant V10R) created a more specific inhibitor to distinguish cathepsin B from cathepsin H. Cathepsin L showed a preference for larger aromatic residues in S2. In contrast, cathepsin S preferred phenylalanine to valine in S2, but bound less tightly to the V10W cystatin variant. The latter variant proved to be valuable for discriminating between cathepsin L and cathepsin S (Ki 2.4 and 190 pM respectively). The equilibrium dissociation constant of the complex between cathepsin L and variant L9W/W106G showed little difference in affinity from that of the cathepsin L complex with the singly substituted W106G variant. In contrast, the L9W/W106G variant displayed increased specificity for cathepsin S with a Ki of 10 pM. Our results clearly indicate differences in the specificity of interaction between the N-terminal region of cystatin C and cathepsins B, H, L and S, and that, although cystatin C has evolved to be a good inhibitor of all of the mammalian cysteine proteinases, more specific inhibitors of the individual enzymes can be engineered.

Control of breast tumor cell growth using a targeted cysteine protease inhibitor

The purpose of this study was to determine whether inhibition of lysosomal proteolysis could be used to selectively inhibit proliferation of tumor cells. The lysosomal cysteine protease inhibitor 9-fluorenylmethyloxycarbonyl-tyrosylalanyl-diazomethane was found to inhibit growth of the breast cancer cell lines SK-Br-3 and MCF-7. A humanized monoclonal antibody (huMAb 4D5) directed against the extracellular domain of p185HER2 specifically inhibited growth of the SK-Br-3 cells, which overexpress this antigen. The antibody and inhibitor together showed enhanced inhibition of growth of the SK-Br-3 cells only. When the protease inhibitor was radiolabeled and conjugated to the antibody (huMAb 4D5-125I-Tyr-Ala-CHN2) it was selectively bound to and taken up by the SK-Br-3 cell line. The conjugated inhibitor was delivered and targeted to cathepsin B and an unidentified protein of Mr 39,000 in the SK-Br-3 cells. Internalization of huMAb 4D5-125I-Tyr-Ala-CHN2 and inhibitor labeling of the proteins were temperature-dependent processes. huMAb 4D5-Tyr-Ala-CHN2 was significantly more effective in inhibiting proliferation of SK-Br-3 cells than the inhibitor-free analogue but was ineffective against MCF-7 cells. The results in this report show that targeting of cysteine protease inhibitors can selectively control tumor cell growth and that targeted cysteine protease inhibitors could prove valuable in the development of novel anticancer immunotherapies.

Radiation-induced sarcoma of the breast

A small but growing number of radiation-induced sarcomas after breast-conserving surgery for carcinoma have been reported. Because breast-conserving surgery followed by irradiation is becoming increasingly popular, the potential for the emergence of these sarcomas is growing. Unfortunately, because of postirradiation changes in the affected breast, detection of a new lesion can be difficult, resulting in a delay in diagnosis. This case history describes a 48-year-old woman in whom a malignant fibrous histiocytoma was discovered 5 years after she had had a lumpectomy followed by radiation therapy for infiltrating ductal carcinoma. Although the exact mechanism of radiation-induced carcinogenesis is unclear, one theory is discussed. Also discussed are the histologic composition of the malignant fibrous histiocytoma, its latency period, and the best treatment modality for radiation-induced sarcomas.

Patellar fatigue fractures

Three cases of patellar fatigue fracture are reviewed. Two fractures presented with acute displacement and were initially thought to represent pathological fractures. Histological appearances in one case, and the clinical course in another, subsequently indicated that these were fatigue fractures. A third patient developed a chronic undisplaced fracture and followed the typical clinical course of fatigue fractures at other sites. The importance of recognizing patellar fatigue fractures and of differentiating spontaneously displaced fatigue fractures from pathological fractures is emphasized.

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.

Structural and functional characterization of two allelic variants of human cystatin D sharing a characteristic inhibition spectrum against mammalian cysteine proteinases

Human cystatin D is a novel member of the cystatin superfamily of cysteine proteinase inhibitors present in saliva and tears. Two alleles of the cystatin D gene (CST5), encoding protein variants with either Cys or Arg as residue 26 in their 122-residue polypeptide chains, are present in the population. Expression of the two alleles was investigated by immunochemical analyses of the secreted cystatin D in saliva from individuals homozygous for each of the two alleles, with results demonstrating that both are expressed at similar levels. The inhibitory characteristics of the two cystatin D variants were studied, by determination of dissociation equilibrium constants (Ki) for their complexes with papain and with the mammalian cysteine proteinases, cathepsins B, H, L, and S. The results demonstrate that 1) cystatin D has a characteristic inhibition profile since it does not inhibit cathepsin B (Ki > 1 microM), and when compared to cystatin C and all other known cystatins it is a much poorer inhibitor of cathepsin L (mean Ki 25 nM) but binds cathepsin H and S relatively tightly (mean Ki values of 8.5 and 0.24 nM, respectively); and 2) the inhibitory activities of the two cystatin D variants are not significantly different, demonstrating that the presence of an extra cysteine residue in the cystatin D molecule affects neither the stability nor the functional activity of the inhibitor, thus explaining the widespread distribution of the Cys26-cystatin D encoding allele in the population. The inhibitory properties displayed by cystatin D suggest that it has a function in saliva as inhibitor of either endogenous or exogenous enzymes with cathepsin S- or H-like properties.

Surface activation of pro-cathepsin L

Pro-cathepsin L is an inactive zymogen that has been shown previously to undergo autolysis at pH 3.0 to give mature forms of the enzyme. We have now been able to demonstrate that this enzyme can undergo activation at pH 5.5 in the presence of negatively charged surfaces. Activation could also be measured at pH 6.0, but no activation occurred at pH 6.5 or higher. The initiation of activation depends upon the presence of a small percentage of active pro-enzyme, and this is then followed by a more rapid activation to give mature forms of the enzyme. No significant intermediate molecular forms of the enzyme were seen. The time taken for processing of the pro-enzyme to single-chain mature enzyme is comparable to that seen in biosynthetic pulse-chase experiments.

The specificity and elastinolytic activities of bovine cathepsins S and H

Cathepsins S and H were purified from bovine spleen and their catalytic properties compared. The enzymes were shown to be similar by chromatographic properties and by the ability to hydrolyze Bz-Phe-Val-Arg-NHMec. They could however be distinguished by the fact that cathepsin S reacted with Z-[125I]Tyr-Ala-CHN2 and hydrolyzed Z-Phe-Arg-NHMec whereas cathepsin H did not. The substrate and inhibitor specificities of cathepsin H suggest that unlike cathepsins B, L, and S, it cannot accommodate peptides with aromatic side chains in P2. Cathepsins L and S can accommodate the aromatic side chain of tyrosine in P1 readily, whereas cathepsins H and B cannot. The specificities of each enzyme for synthetic substrates and inhibitors have enabled the construction of models of the architecture of the active sites of the mammalian cysteine proteinases which clearly show the differences between the four enzymes. A significant characteristic of cathepsin S is that it can hydrolyze insoluble elastin at both acidic and neutral pH; this distinguishes it from all of the other lysosomal proteinases.

Inhibition of cysteine proteinases in lysosomes and whole cells

Inhibitors of cysteine proteinases have been used extensively to dissect the roles of these proteinases in cells. Surprisingly though, little work has been performed to demonstrate unequivocally that the inhibitors reach and inactivate their target proteinases in cell culture or in vivo. In the present study, the permeability of lysosomes and whole cells has been studied. Benzyloxycarbonyl (Z)-[125I]iodo-Tyr-Ala-diazomethane (CHN2), an inhibitor of cathepsins L and B, has been shown to label active forms of these enzymes in lysosomes and whole cells. The ability of other cysteine proteinase inhibitors to block this labelling has been used to indicate the permeation of these compounds. All the inhibitors were able to block labelling by Z-[125I]iodo-Tyr-Ala-CHN2 in lysosomal extracts. In intact lysosomes or cells, however, only N-[N-(L-3-trans-ethoxycarbonyloxirane-2-carbonyl)-L-leucyl]-3- methylbutylamine ('E-64d') Z-Tyr-Ala-CHN2, Z-Phe-Ala-CHN2 and Z-Phe-Phe-CHN2 were able to block labelling by Z-[125I]iodo-Tyr-Ala-CHN2. N-[N-(L-3-trans-Carboxyoxirane-2-carbonyl)-L-leucyl]amino-4-gua nidinobutane (E-64) and leupeptin were unable to block labelling by Z-[125I]iodo-Tyr-Ala-CHN2 in lysosomes or in cells. The ability to block labelling in lysosomes is an indication of the ability of the inhibitor to diffuse across membranes. Thus E-64 and leupeptin do not readily permeate membranes and therefore their uptake into cells probably only occurs via pinocytosis.

Blindness associated with toxoplasmosis in canaries

Seven of 30 canaries in an aviary in New Zealand developed ophthalmic problems. Clinically, 5 birds had unilateral and 2 birds had bilateral lesions characterized by conjunctivitis, crusty exudates on eyelids, and collapse of the eyeball. Microscopic lesions in 12 of 14 eyes examined included inflammation of the choroid and retina, with osseous replacement of the globe in some. Numerous Toxoplasma gondii tachyzoites were seen in the detached retina and vitreous humor of acutely affected birds. The diagnosis of toxoplasmosis was confirmed by immunohistochemical staining with T gondii antiserum. Affected birds had encephalitis, and T gondii was localized in the brains of these by immunohistochemical examination and by use of bioassays in mice. Toxoplasmosis should be considered in differential diagnosis of ophthalmitis in canaries.

Cigarette smoking induces an elastolytic cysteine proteinase in macrophages distinct from cathepsin L

Degradation of the interstitium of the lung by elastolytic enzymes is thought to be a critical component of the pathogenesis of emphysema. Alveolar macrophages are increased in numbers in cigarette smokers and contain the elastolytic cysteine proteinase cathepsin L. We sought to determine if cigarette smoking induces a change in cathepsin L levels in alveolar macrophages which would, in turn, alter the expression of elastolytic activity. Lysates of smokers' macrophages, assayed at pH 5.50, degraded more than seven times as much [3H]elastin as did lysates from nonsmokers' macrophages (44 +/- 20.8 vs. 6 +/- 1.6 micrograms.10(6) cells-1.24 h-1). Little or no activity was demonstrable at neutral pH. Immunoblots of macrophage lysates demonstrated that smokers' cells contain 3.7 +/- 1.1 times as much 25-kDa cathepsin L antigen as nonsmokers' cells. However, as judged by active site labeling, levels of active cathepsin L in smokers and nonsmokers are indistinguishable, suggesting that most of the 25-kDa antigen found in smokers' macrophages is inactive. Inhibitors of cathepsin L had little effect on lysate elastolytic activity, confirming that an enzyme other than cathepsin L is responsible for the increased elastolytic activity seen in smokers' macrophages. Further experiments demonstrated that this second enzyme(s) has a profile of inhibition indicating that it is a cysteine proteinase with optimal activity at pH 5.50. It is this second elastolytic cysteine proteinase(s) that is induced by exposure to cigarette smoke and is responsible for the sevenfold increase in elastolytic activity found in smokers' macrophage lysates.

Lethal toxoplasmosis in a little penguin (Eudyptula minor) from Tasmania

A penguin (Eudyptula minor) from Tasmania, Australia, died following a 3-day illness. The liver and spleen parenchyma contained grossly visible foci of discoloration. Numerous tachyzoites of Toxoplasma gondii were seen associated with necrosis in sections of liver and spleen. The diagnosis was confirmed by immunohistochemical staining of parasites. This apparently is the first report of fatal toxoplasmosis in a penguin.

Uptake of extracellular enzyme by a novel pathway is a major determinant of cathepsin L levels in human macrophages

The phorbol myristate acetate (PMA)-differentiated myelomonocytic cell line, THP-1, and human alveolar macrophages contain the cysteine proteinase cathepsin L. This enzyme is synthesized as a 43-kD proenzyme and processed to the active 25-kD form. Differentiation of THP-1 cells in the presence of human serum resulted in an increase in the size of the vacuolar compartment and the accumulation of more 25-kD cathepsin L antigen, as compared with THP-1 cells differentiated in the presence of fetal calf serum. Cells cultured in both types of sera have equivalent levels of cathepsin L mRNA. Metabolic labeling experiments demonstrated equivalent rates of synthesis, processing to the active form, and persistence in both culture conditions. An extracellular source of enzyme was documented by immunoblotting human serum which demonstrated 25-kD cathepsin L antigen; furthermore, we demonstrated that both THP-1 cells, differentiated in human serum, and human alveolar macrophages take up the 43-kD proenzyme and process it to the 25-kD form. Thus, human serum contains a factor(s) that induces both a marked increase in the size of the vacuolar compartment in differentiated THP-1 cells and a novel pathway that is responsible for the uptake and processing of extracellular cathepsin L. The activity of this inducible pathway is a major determinant of levels of intracellular cathepsin L. Cathepsin L is a potent elastase and the regulation of its uptake and processing may play a role in the pathogenesis of disease processes characterized by the destruction of elastin, such as pulmonary emphysema.

Characterization of the active site of human multicatalytic proteinase

The activity of multicatalytic proteinase against synthetic substrates and the kinetics of its inhibition by a range of class-specific inhibitors have been investigated. The enzyme was found to have a broader pH activity profile than previously noted, being active against succinyl-Ala-Ala-Phe-7-amino-4-methylcoumarin optimally at pH 4.5 and against benzyloxycarbonyl-Gly-Gly-Arg-7-amino-4-methylcoumarin optimally at pH 10.5. Neither activity was inhibited by the class-specific inhibitors 1,10-phenanthroline, EDTA, pepstatin, di-isopropyl fluorophosphate, peptidyl chloromethanes, peptidyl diazomethanes or L-3-carboxy-2,3-trans-epoxypropionyl-leucylamido-(4-guanidin o)butane (E-64), indicating that the enzyme is not a typical metallo-, aspartic, serine or cysteine proteinase. Inhibition by HgCl2, iodoacetamide and N-ethylmaleimide suggests that free thiols are necessary for the enzyme to maintain activity, but that these thiols are not particularly reactive as is the case for cysteine proteinases of the papain superfamily. The peptidyl aldehydes chymostatin and leupeptin were found to be reversible inhibitors of multicatalytic proteinase. Chymostatin inhibited activity against succinyl-Ala-Ala-Phe-7-amino-4-methylcoumarin at pH 4.5 (Ki 160 +/- 22 microM) whereas leupeptin (200 microM) was not inhibitory. Inhibition of activity against benzyloxycarbonyl-Gly-Gly-Arg-7-amino-4-methylcoumarin by these compounds was more complex, in that they behaved as slow tight-binding inhibitors. kon values were determined to be 12 +/- 2 M-1.s-1 and 1290 +/- 125 M-1.s-1 for chymostatin and leupeptin, respectively. The upper limit for Ki values for these two inhibitors was estimated as 5 +/- 1.5 microM and 25 +/- 5 nM, respectively. The different inhibition characteristics for each substrate were also apparent at an intermediate pH of 8.5, showing that the two activities are distinct. Dichloroisocoumarin, a mechanism-based inhibitor of serine proteinases, did inhibit activity against succinyl-Ala-Ala-Phe-7-amino-4-methylcoumarin with a rate constant of 250 M-1.s-1, suggesting that multicatalytic proteinase is an atypical serine proteinase.

Studies on the pharmacokinetics and mutagenic potential of rhodamine B

Rhodamine B is used as a marker dye in herbicide sprays. There is evidence that spray operators and others may absorb rhodamine B through the skin. This study was undertaken to investigate the in vivo mutagenicity of rhodamine B, to compare the in vitro mutagenicity of two commercial preparations of the dye with that of known mutagens including rhodamine 6G and to elucidate the pharmacokinetics of rhodamine B in the rabbit. Following the i.v. treatment of adult female New Zealand White rabbits with rhodamine B (1 mg/kg body wt), the plasma concentration of rhodamine B decreased rapidly and was accompanied by the appearance of at least four fluorescent polar metabolites. These metabolites, as well as a very small amount of rhodamine B, were also present in the urine which, when tested in the Ames assay was not significantly mutagenic against Salmonella typhimurium strains TA98 and TA100 either with or without metabolic activation. Urine from a human subject who had been contaminated with marker dye was also non-mutagenic. Both commercial preparations of rhodamine B were found to be weakly mutagenic, using the same assay system. It is concluded that while appropriate hygiene measures should be exercised by users of products containing this dye, the results do not support the hypothesis that rhodamine B is a genotoxic hazard in the mammalian organism.

The use of benzyloxycarbonyl[125I]iodotyrosylalanyldiazomethane as a probe for active cysteine proteinases in human tissues

The ability of benzyloxycarbonyl-(125I)Tyr-Ala-CHN2 to label cysteine proteinases in a variety of human tissues was investigated. The inhibitor bound only to cathepsin B in tissues homogenized at pH 5.0. When liver was autolysed at pH 4.0 for up to 4 h, the inhibitor also bound to a protein of Mr 25,000. This was identified immunologically and chromatographically as cathepsin L. Both cathepsins B and L were found primarily in kidney, liver and spleen. In spleen, an additional protein of Mr 25,000 was also labelled. This protein could not be precipitated by antibodies to any of cathepsins B, H and L. This protein has tentatively been identified as human cathepsin S by its tissue distribution, chromatographic properties and molecular size. This work clearly shows that peptidyldiazomethanes are specific probes for cysteine proteinases, and that benzyloxycarbonyl-(125I)Tyr-Ala-CHN2 binds to three such enzymes in human tissues.