Identification of dipeptidyl nitriles as potent and selective inhibitors of cathepsin B through structure-based drug design

Cathepsin B is a member of the papain superfamily of cysteine proteases and has been implicated in the pathology of numerous diseases, including arthritis and cancer. As part of an effort to identify potent, reversible inhibitors of this protease, we examined a series of dipeptidyl nitriles, starting with the previously reported Cbz-Phe-NH-CH(2)CN (19, IC(50) = 62 microM). High-resolution X-ray crystallographic data and molecular modeling were used to optimize the P(1), P(2), and P(3) substituents of this template. Cathepsin B is unique in its class in that it contains a carboxylate recognition site in the S(2)' pocket of the active site. Inhibitor potency and selectivity were enhanced by tethering a carboxylate functionality from the carbon alpha to the nitrile to interact with this region of the enzyme. This resulted in the identification of compound 10, a 7 nM inhibitor of cathepsin B, with excellent selectivity over other cysteine cathepsins.

Effect of the selective and non-selective cysteine protease inhibitors on the intracellular processing of interleukin 6 by HEPG2 cells

The effects were measured and compared of three nonselective cysteine cathepsin inhibitors (leupeptin, trans-Epoxy-succinyl-L-Leucylamido(4-guanidino)-butane (E-64), and Z-Phe-Ala-CH2F) and a selective cathepsin B inhibitor, CA074Me, on the intracellular processing of 125I-labeled human recombinant Interleukin 6 (IL-6) by HepG2 cells. The uptake and processing of 125I-IL-6 by cells treated with inhibitors was followed over a 7-h period. All inhibitors caused an increased residence time of IL-6 inside the cell and a corresponding decrease in the output of non-trichloroacetic acid-precipitable fragments of radiolabeled protein. Maximal effect was achieved with leupeptin at 200 microM, with which the rate of IL-6 digestion was reduced to 50% that of control cells. The specific inhibitor CA074Me was the least effective in slowing the intracellular processing of IL-6. The effects of all of the inhibitors on the production of haptoglobin, either stimulated by IL-6 or basal, was negligible over a similar time period, indicating continued cell viability. The data from this model suggest that cathepsin inhibitors would not interfere with lysosomal processing to an extent which would prohibit the development of selective and potent cathepsin inhibitors for the treatment of diseases in which individual cysteine cathepsins play clearly pathophysiological roles.

Characterization of recombinant human cathepsin B expressed at high levels in baculovirus

The lysosomal cysteine protease cathepsin B has been studied intensely for many years because of its unique characteristics and its potential involvement in disease states. A reproducible, high yield expression system for active recombinant protein is key to biochemical and biophysical studies as well as rational drug design. Although several microbial and mammalian expression systems for recombinant human cathepsin B have been described, these have been limited by low or variable yields. Further, in some of these systems hyper-glycosylation of the enzyme near the active site affects its activity. We describe a baculovirus expression system and purification scheme that solve all of these problems. Yields of active, protected enzyme were reproducibly in excess of 25 mg/L. Since this protein was not hyper-glycosylated, it had greater activity than cathepsin B produced in yeast systems as indicated by a threefold increase in Kcat. In addition, the biophysical properties of the baculovirus-expressed cathepsin B, as measured by dynamic light scattering, were more amenable to crystallographic study since the data indicated proteins of more uniform size. Therefore, this system for the production of recombinant human cathepsin B constitutes a major improvement in both quantity and quality over those previously reported. Further, we demonstrate that the manner of expression and purification of this enzyme has profound effects on its kinetic and physical parameters.

Characterization of the S3 subsite specificity of cathepsin B

Five synthetic substrates containing different amino acid residues at the P3 position (acetyl-X-Arg-Arg-AMC, where X is Gly, Glu, Arg, Val, and Tyr and where AMC represents 7-amindo-4-methylcoumarin) were used to investigate the S3 subsite specificity of cathepsin B. At pH 6.0, the specificity constant, kcat/Km, for tripeptide substrate hydrolysis was observed to increase in the order Glu < Gly < Arg < Val < Tyr. Molecular modeling studies of substrates containing a P3 Glu, Arg, or Tyr covalently bound as the tetrahedral intermediate to the enzyme suggest that the specificity for a P3 Tyr is because of a favorable aromatic-aromatic interaction with Tyr75 on the enzyme as well as a possible H bond between the P3 Tyr hydroxyl and the side chain carboxyl of Asp69.

Identification of a novel aspartyl endothelin converting enzyme in porcine aortic endothelial cells

An enzyme which catalyzes the conversion of proendothelin-1 to the potent vasoconstrictor peptide endothelin-1 has been identified in the detergent extract of primary porcine aortic endothelial cell membranes. Partial purification was accomplished by anion exchange and Con A affinity chromatography. The enzyme was active at pH 4 and was inhibited by 100 nM peptstatin A. Hydrolysis products of proendothelin-1 were characterized by bioassay, RIA, HPLC and molecular mass analysis. Comparisons to cathepsin D and renin demonstrated that the endothelin converting enzyme activity from the porcine aortic endothelial cells was unrelated to the known enzymes. These results suggest that the processing of proendothelin-1 by endothelial cells involves a novel pepstatin-sensitive aspartyl protease.

A comparison of endothelin-converting enzyme activity from primary porcine aortic endothelial cells with activities in cultured human cell lines

We examined four commercially available human cell lines for endothelin-converting-enzyme-(ECE) like activity and compared the results with primary porcine aortic endothelial cell enzymes. The cells that were investigated were 293 (transformed primary human embryonal kidney), Hep G2 (human hepatocellular carcinoma), HUVECs (human umbilical vein endothelial cells), and U937 (human histiocytic lymphoma). The relative ECE-like activities were determined in cytosolic and particulate extracts of each cell type. Enzyme activity against pro-ET-1 was measured at pH 4 and 7, using a C-terminal Trp-specific antibody to ET-1 radioimmunoassay and by high-performance liquid chromatography analysis of the enzyme hydrolysis products of pro-ET-1. Inhibition by EDTA at pH 7 or pepstatin at pH 4 was used to classify the pro-ET-1 processing enzymes from the human cell lines as either metallo- or aspartylproteinases. The particulate extract of the primary porcine aortic endothelial cells contained both aspartyl and metallo ECE-like enzymes. No ECE-like activity was present in either the cytosolic or particulate extracts of the U937 nor 293 cells. Neither the particulate extract of the Hep G2 cells nor the cytosolic extract of the HUVECs had any ECE-like activity. The cytosolic extract of the Hep G2 cells and the particulate extract of the HUVECs had an ECE-like activity at pH 7 that was inhibited by 10 mM EDTA, qualifying these enzymes as members of the metalloproteinase family.

Protection from chemical modification of nucleotides in complexes of M1 RNA, the catalytic subunit of RNase P from E coli, and tRNA precursors

Certain nucleotides in M1 RNA, the catalytic RNA subunit of RNase P from E coli, are protected from chemical modification when M1 RNA forms complexes with tRNA precursor molecules (ES complexes). Many of these nucleotides are important in the formation of the Michaelis complex. In the presence of tRNA precursor molecules, the pattern of protection from chemical modification of a region in M1 RNA that resembles the E site in 23S rRNA is similar to the pattern of protection of the E site in the presence of deacylated tRNA. In the complex with the RNA enzyme, more nucleotides in the substrate become accessible to modification, an indication that the substrate is in an unfolded conformation under these conditions.

Chemical modification of the RTEM-1 thiol beta-lactamase by thiol-selective reagents: evidence for activation of the primary nucleophile of the beta-lactamase active site by adjacent functional groups

The RTEM-1 thiol beta-lactamase (Sigal, I.S., Harwood, B.G., Arentzen, R., Proc. Natl. Acad. Sci. U.S.A. 79:7157-7160, 1982) is inactivated by thiol-selective reagents such as iodoacetamide, methyl methanethiosulfonate, and 4,4'-dipyridyldisulfide, which modify the active site thiol group. The pH-rate profiles of these inactivation reactions show that there are two nucleophilic forms of the enzyme, EH2 and EH, both of which, by analogy with the situation with cysteine proteinases, probably contain the active site nucleophile in the thiolate form. The pKa of the active site thiol is therefore shown by the data to be below 4.0. This low pKa is thought to reflect the presence of adjacent functionality which stabilizes the thiolate anion. The low nucleophilicity of the thiolate in both EH2 and EH, with respect to that of cysteine proteinases and model compounds, suggests that the thiolate of the thiol beta-lactamase is stabilized by two hydrogen-bond donors. One of these, of pKa greater than 9.0, is suggested to be the conserved and essential Lys-73 ammonium group, while the identity of the other group, of pKa around 6.7, is less clear, but may be the conserved Glu-166 carboxylic acid. beta-Lactamase activity is associated with the EH2 form, and thus the beta-lactamase active site is proposed to contain one basic or nucleophilic group (the thiolate in the thiol beta-lactamase) and two acidic (hydrogen-bond donor) groups (one of which is likely to be the above-mentioned lysine ammonium group).

Inactivation of the thiol RTEM-1 beta-lactamase by 6-beta-bromopenicillanic acid. Identity of the primary active-site nucleophile

The thiol RTEM-1 beta-lactamase [Sigal, Harwood & Arentzen (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 7157-7160] is inactivated by 6-beta-bromopenicillanic acid with formation of a characteristic chromophore, absorbing maximally at 350 nm, which is covalently bound to the enzyme. Model studies suggest that the chromophore is that of a 6-carboxylate thiol ester of 2,3-dihydro-2,2-dimethyl-1,4-thiazine-3,6-dicarboxylate, which can arise by rearrangement of the thiol-penicilloate obtained by thiolysis of the beta-lactam of 6-beta-bromopenicillanate. Loss of activity of the enzyme is also concerted with disappearance of its single (cysteine) thiol group. These results indicate that the thiol group of this enzyme is indeed a nucleophilic catalyst in beta-lactam turnover. The thiol beta-lactamase is also inactivated by clavulanic acid with formation of a chromophore, presumably a 3-aminoacrylate thiol ester, at 308 nm. Both 6-beta-bromopenicillanate and clavulanate are hydrolysed more slowly by the thiol enzyme than by the native serine beta-lactamase, but, probably as a consequence of this, both compounds inactivate the former enzyme more efficiently. Cefoxitin, a poor substrate of the native enzyme, does not appear to interact covalently with the thiol beta-lactamase.