Michael J. North

Cysteine proteinases of parasitic protozoa

Proteinases are involved with many processes in living organisms. In recent years, there has been increasing interest in elucidating the functions the enzymes perform in parasites. These studies have revealed that one class of proteinases, the cysteine proteinases, predominates in many parasitic protozoa. In this article Mick North, Jeremy Mottram and Graham Coombs review what is known about the cysteine proteinases of parasitic protozoa and discuss the approaches being pursued in attempts to design antiparasite drugs based on inhibitors or substrates of these enzymes.

A cysteine proteinase cDNA from Trypanosoma brucei predicts an enzyme with an unusual C-terminal extension.

A cDNA for a Trypanosoma brucei cysteine proteinase has been cloned and sequenced. The deduced protein can be divided into four domains, based on homologies with other cysteine proteinases: the pre‐, pro‐ and central regions show considerable homology to the cathepsin L class of mammalian enzymes, whilst the long C‐terminal extension distinguishes the trypanosome enzyme from all mammalian cysteine proteinases reported. This 108 amino acid extension, which includes 9 contiguous prolines near the junction with the central domain, appears likely to be processed in part to produce the mature enzyme, and may be involved in targetting the protein within the cell. The trypanosome genome contains more than 20 copies of the cysteine proteinase gene arranged in a long tandem array.

The use of a highly sensitive electrophoretic method to compare the proteinases of trichomonads

A highly sensitive electrophoretic method involving gelatin-containing polyacrylamide gels has been used to analyse trichomonad proteinases. Multiple forms, optimally active at pH 5-6, were present in all four species examined, but the species could be distinguished from one another by both quantitative and qualitative differences. The intestinal parasites, Trichomitus batrachorum and Pentatrichomonas hominis, had lower specific activities than the urogenital parasites, Trichomonas vaginalis and Tritrichomonas foetus, and, in the case of P. hominis, there were fewer enzyme forms. The high activity proteinases of Tritrichomonas foetus had low apparent molecular weights (less than 25 kDa), while the predominant enzymes of Trichomonas vaginalis were of high apparent molecular weight (68-110 kDa). Distinct differences were also observed between the proteinase patterns of various isolates of T. vaginalis. All of the enzymes were stimulated by dithiothreitol, suggesting that they were cysteine proteinases. This was confirmed for the T. vaginalis and Tritrichomonas foetus proteinases from their inhibition by antipain, leupeptin, TLCK and iodoacetic acid. The method allows the detection of proteinases in samples of Trichomonas vaginalis containing as few as 10(4) cells or as little as 1 microgram protein. It was also possible to detect proteinase activity released into the medium. For both T. vaginalis and Tritrichomonas foetus, the extracellular enzymes present during early log phase were qualitatively different from the intracellular proteinases, although the latter were present in samples of media obtained from later cultures (cell densities greater than 1 X 10(5) parasites ml-1). The results show the potential of this technique for detecting proteinases in trichomonad samples in studies aimed at determining proteinase function in pathogenesis and host-parasite relationships.

Trichomonas vaginalis, Tritrichomonas foetus, and Trichomitus batrachorum: Comparative proteolytic activity

At least four proteolytic activities were detected in the lysates of each of Trichomonas vaginalis, Tritrichomonas foetus, and Trichomitus batrachorum. These were HPAase, a dithiothreitol-dependent activity on hide powder azure; AZCase, a dithiothreitol-dependent activity on azocasein; and two distinct activities towards peptide nitroanilide derivatives--one was optimally active at pH 7 and stimulated by dithiothreitol; the other had no dithiothreitol requirement and was highly active at pH 5. HPAase and AZCase were active over a broad pH range. Overall, with respect to these four activities, T. batrachorum and T. vaginalis were quite similar. In contrast, T. vaginalis and T. foetus differed from one another in several respects, notably the level of HPAase activity and the properties of the dithiothreitol-independent activity. Multiple bands of proteinase activity were demonstrated with each species after electrophoresis of parasite extracts on polyacrylamide gels containing denatured haemoglobin. They appeared optimally at acid pH and in the presence of dithiothreitol. The proteinase band patterns of T. foetus were similarly complex (at least six bands), whereas T. batrachorum gave a much simpler pattern (three bands). The sensitivities to proteinase inhibitors suggested that all the activities were due to cysteine proteinases. The results show that there are some similarities in the proteolytic activities of all three trichomonad species, and that the two parasites of the urinogenital tracts of mammals possess additional features in common.

Proteinases of Leishmania mexicana amastigotes and promastigotes: analysis by gel electrophoresis.

The proteinases of Leishmania mexicana mexicana amastigotes and promastigotes have been analysed by electrophoresis on polyacrylamide gels containing denatured haemoglobin. Eleven bands of activity were detected indicating multiple proteinases. These were significant quantitative and qualitative differences between the proteinases of the two developmental forms. Four, B-E, were present in both forms but were of much higher activity in the amastigote. There were two major activities in promastigotes, A and D. The other proteinases, F-K, were of lower activity; I and K were not detected in promastigotes. All proteinases were active optimally at pH 4.0. Most of them, including the major proteinases A-E, were thiol proteinases since they were stimulated by 1 mM dithiothreitol and were sensitive to inhibitors such as HgCl2, leupeptin, antipain and iodoacetic acid.

The release of hydrolases from Trichomonas vaginalis and Tritrichomonas foetus.

Trichomonas vaginalis and Tritrichomonas foetus were found to release large amounts of beta-N-acetylglucosaminidase (EC 3.2.1.30), alpha-mannosidase (EC 3.2.1.24), beta-glucosidase (EC 3.2.1.21), acid phosphatase (EC 3.1.3.2) and proteinases during axenic growth in vitro. The enzymes were released continually throughout the growth phase, with the extracellular activity being of the same order as that within the cells. There was differential release of proteinases from Trichomonas vaginalis. The subcellular localization of the hydrolases was determined by differential and isopycnic centrifugation. The intracellular enzymes were shown to be mostly located within particle populations. Centrifugation on Percoll gradients allowed the separation of sub-populations of the particles in T. vaginalis; two distinct sub-populations were apparent with equilibrium densities in 20% (v/v) Percoll of 1.035 and 1.050 g cm-3 respectively. The higher density particles were rich in the hydrolases released most abundantly, suggesting a possible link between enzyme release and these organelles. Distinct subpopulations of hydrolase-containing particles were not detected in Tritrichomonas foetus. The results demonstrate that hydrolytic enzyme release represents a major activity during trichomonad growth.

The specificity of trichomonad cysteine proteinases analysed using fluorogenic substrates and specific inhibitors.

The multiple cysteine proteinases of Trichomonas vaginalis and Tritrichomonas foetus, both those retained intracellularly and those released, were separated using gelatin-SDS-PAGE, and their activity towards a range of 15 fluorogenic peptidyl aminomethylcoumarins determined together with their relative sensitivity to inhibitors. Three types of enzyme were apparent in T. vaginalis: (i) an 86-kDa enzyme active only on Z-Arg-Arg-NHMec; (ii) a 54-kDa proteinase which was most active on Z-Phe-Arg-NHMec but also able to hydrolyse N-t-Boc-Val-Leu-Lys-NHMec, Suc-Ala-Phe-Lys-NHMec, H-Pro-Phe-Arg-NHMec and Z-Arg-Arg-NHMec; and (iii) a group of six enzymes which preferentially hydrolysed substrates with bulky residues at the P2 and P3 positions. N-t-Boc-Val-Leu-Lys-NHMec and H-Leu-Val-Tyr-NHMec were the best substrates for the latter group. The 86-kDa proteinase was inactivated by E-64, but only at high concentrations, and was relatively insensitive to the peptidyl diazomethanes. The other proteinases were inhibited by low concentrations of E-64 and by Z-Phe-Ala-CHN2, and to a lesser extent by Z-Phe-Phe-CHN2. Differences between the proteinases of T. foetus were also demonstrated. All of them were active on Z-Arg-Arg-NHMec, but their activity towards other substrates varied. Three predominantly extracellular proteinases (25, 27 and 34 kDa), hydrolysed Z-Arg-Arg-NHMec specifically. Other proteinases (apparent Mr of 20,000 and 32,000) hydrolysed a number of other substrates, with the 32-kDa enzyme having greater activity towards N-t-Boc-Val-Leu-Lys-NHMec and H-Leu-Val-Tyr-NHMec than towards Z-Arg-Arg-NHMec. At a high concentration (270 microM), E-64 inhibited all of the T. foetus enzymes, but lower concentrations were less effective, with the 18-kDa proteinase being particularly insensitive. Z-Phe-Ala-CHN2 and Z-Phe-Phe-CHN2 were relatively poor inhibitors. The results demonstrate that the proteinases of both species are a heterogeneous group with respect to specificity, and have highlighted significant differences between the enzymes of T. vaginalis and T. foetus. The information on the specificities will be useful for assessing the features required in proteinase inhibitors if they are to be of potential value as antitrichomonal agents.

A comparative study of the proteolytic enzymes of Trypanosoma brucei, T. equiperdum, T. evansi, T. vivax, Leishmania tarentolae and Crithidia fasciculata

Four types of proteolytic activity were detected in the bloodstream form of each of the four Trypanosoma species: (i) HPAase, active on hide powder azure and detected on polyacrylamide gels containing denatured haemoglobin; (ii) AZCase, active on azocasein; (iii) type 1, active on the chromogenic peptide N-benzoyl-L-prolyl-L-phenylalanyl-L-arginine p-nitroanilide in the presence of dithiothreitol, and (iv) type 2, active against several nitroanilide derivatives in the absence of dithiothreitol. Studies of the pH optimum, dithiothreitol requirement and inhibitor sensitivities of the proteolytic activities suggested that: (a) HPAase and type 1 activities could be due to the same enzymes, probably a family of cysteine proteinases; (b) AZCase had some characteristics of a cysteine proteinase, but was not identical to HPAase, and (c) type 2 activity could be due to a serine proteinase. Procyclic T. brucei contained relatively low cysteine proteinase activities (HPAase, AZCase and type 1) but high type 2 activity. Their proteolytic enzymes thus were apparently more similar to those in Crithidia fasciculata and Leishmania tarentolae promastigotes than those in T. brucei bloodstream forms.

Parasite cysteine proteinases

Cysteine proteinases of protozoan and helminth parasites are considered to have a high potential as targets for novel antiparasite agents. This has stimulated research on the enzymes and a large body of data has now been accumulated. This Perspective provides an overview of the current situation, with recent advances being highlighted. Emphasis is given to the Type I cysteine proteinases of trypanosomatid protozoa, which are atypical in having an extra C-terminal domain, and the asparaginyl endopeptidase ofSchistosoma, for which the only homologues known are those in plants. The locations of parasite cysteine proteinases are described, with emphasis on the extralysosomal sites, and the putative roles of the enzymes in host-parasite interactions, including parasite nutrition and host invasion, are discussed. The major advances being made in elucidating cysteine proteinase function in protozoan parasites through genetic manipulation, including targeted gene deletion, are presented, with the Type I cysteine proteinases ofLeishmania being used as an example. The importance of this approach in the validation of the enzymes as drug targets is discussed. The current status of attempts to exploit parasite cysteine proteinases with drugs is presented, and future prospects are outlined.

Analysis of the proteinases of Trypanosoma brucei.

A method comprising enzyme separation by SDS-PAGE and subsequent use of peptidyl aminomethylcoumarins as substrates has been used to study proteinases of the protozoan parasite Trypanosoma brucei. The application of this method has allowed investigation of the substrate specificities of individual proteinases in cell lysates without the need for enzyme purification. The results show that T. brucei contains a group of cysteine proteinases, probably four in number, with substrate and inhibitor specificities similar to those of cathepsin L. A second group of proteinases, larger enzymes with significantly different substrate specificities and sensitivity to inhibitors, was also detected. Peptidyl diazomethanes inhibited the cysteine proteinases and also parasite growth, offering promise that peculiarities in the substrate specificity of trypanosomal cysteine proteinases could be exploited by compounds of this type.