Molecular and Biochemical Parasitology 102 (1999) 145–155

Purification and characterisation of a trypsin-like serineoligopeptidase from Trypanosoma congolense

Rory E. Morty a,1, Edith Authie b,2, Linda Troeberg a,3,John D. Lonsdale-Eccles c, Theresa H.T. Coetzer a,*

a Department of Biochemistry, Uni6ersity of Natal, Pri6ate Bag X01, 3209 Scotts6ille, South Africab International Li6estock Research Institute, P.O. Box 30709, Nairobi, Kenya

c Department of Biochemistry and Molecular Genetics, Uni6ersity of Alabama at Birmingham, 1918 Uni6ersity Boule6ard,Birmingham, AL 35294-0005, USA

Received 28 January 1998; received in revised form 27 April 1999; accepted 9 May 1999

Abstract

Trypanosoma brucei contain a serine oligopeptidase (OP-Tb) that is released into (and remains active in) the bloodof trypanosome-infected animals. Here a similar enzyme from Trypanosoma congolense is described. This oligopepti-dase, called OP-Tc, was purified using three-phase partitioning, and ion-exchange and affinity chromatography.OP-Tc is inhibited by alkylating agents, by serine peptidase-specific inhibitors including 3,4-dichloroisocoumarin,4-(2-aminoethyl)benzenesulfonylfluoride and diispropylfluoro-phosphate and by other peptidase inhibitors includingleupeptin, antipain and peptidyl chloromethyl ketones. Reducing agents such as dithiothreitol enhanced activity as didheparin, spermine and spermidine. The enzyme has trypsin-like specificity since it cleaved fluorogenic peptides thathave basic amino acid residues (Arg or Lys) in the P1 position. Potential substrates without a basic residue in P1 werenot hydrolysed. Although OP-Tc has weak arginine aminopeptidase activity, the enzyme clearly preferred substratesthat had amino acids in the P2 and P3 positions. Overall, OP-Tc appears to be less efficient than OP-Tb because itusually displayed lower kcat/Km values for the substrates tested. However, like OP-Tb, the best substrate for OP-Tcwas Cbz-Arg-Arg-AMC (Km=0.72 mM, kcat=96 s−1). OP-Tc preference for amino acids in the P2 position was(Gly,Lys,Arg)\Phe\Leu\Pro. The results also suggest that the P3-binding site has hydrophobic characteristics.

www.elsevier.com/locate/parasitology

Abbre6iations: AEBSF, 4-(2-aminoethyl)benzenesulfonylfluoride; AMC, 7-amino-4-methylcoumarin; Boc, t-butoxycarbonyl; Cbz,Na-carbobenzoxy; DCI, 3,4-dichloroisocoumarin; DFP, diisopropylfluorophosphate; DTT, dithiothreitol; IgY, chicken egg-yolkimmunoglobulin; OP-Tc, oligopeptidase from Trypanosoma congolense ; pCMB, para-chloromercuribenzoate; PMSF, phenyl-methanesulfonylfluoride; Tos, toluene-p-sulfonyl; TBS, Tris-buffered saline; TPP, three-phase partitioning.

* Corresponding author. Tel.: +27-331-2605463; fax: +27-331-2605462.E-mail address: coetzer@unpsun1.cc.unp.ac.za (T.H.T. Coetzer)1 Present address: Section of Microbial Pathogenesis, Boyer Center for Molecular Medicine, Yale University School of Medicine,

295 Congress Avenue, New Haven, CT 06536-0812, USA.2 Present address: Centre de Cooperation Internationale en Recherche Agronomique pour le Developpement, Departement

d’Elevage et Medecine Veterinaire (CIRAD-EMVT), BP5035, 34032 Montpellier, France.3 Present address: Department of Biochemistry, University of Cambridge, 80 Tennis Court Road, Cambridge CB2 1GA, UK.

0166-6851/99/$ - see front matter © 1999 Published by Elsevier Science B.V. All rights reserved.

PII: S 0166 -6851 (99 )00097 -3

R.E. Morty et al. / Molecular and Biochemical Parasitology 102 (1999) 145–155146

OP-Tc may not be a naturally immunodominant molecule because neither IgG nor IgM anti- OP-Tc antibodies weredetected in the blood of experimentally infected cattle. © 1999 Published by Elsevier Science B.V. All rights reserved.

Keywords: Oligopeptidase; Peptidase; Trypanotolerance; Trypanosoma (Nannomonas) congolense ; Enzyme inhibition; Substratespecificity; IgG; IgM; IgY

1. Introduction

Trypanosoma (Trypanozoon) brucei brucei, T.(Nannomas) congolense and T. (Duttonella) 6i6axare the etiological agents of bovine trypanosomia-sis, or nagana, in tropical Africa. These parasitesbelong to pathogenically distinct groups of try-panosomes. The T. brucei (‘humoral’) subgroupwhich causes an inflammatory degenerative dis-ease in humans characterised by necrosis, oedemaand hypotension [1], is of moderate pathogenicityin cattle [2]. Infections with T. congolense or T.6i6ax, which belong to the ‘haematic’ subgroup,are characterised primarily by extra-vascularhaemolytic anaemia [1].

The authors have previously reported thepurification and characterisation of a serineoligopeptidase from T. b. brucei [3,4], which theyhave called OP-Tb [4]. The authors have alsodemonstrated that this enzyme is released into thehost bloodstream, where it remains catalyticallyactive and may thus contribute to disease patho-genesis through the degradation of host peptidehormones [4]. Thus, OP-Tb has the potential to bean important pathogenic factor in T. b. bruceiinfections.

A related oligopeptidase (OP-Tc) from T. con-golense, a member of the ‘haematic’ subgroup hasnow been purified and characterised. While cys-teine peptidases have been reported from T. con-golense [5,6], this represents the first report on aserine peptidase from any member of the Nan-nomonas subgenus of African trypanosomes. Al-though OP-Tc seems to be slightly less active thanOP-Tb (about 3-fold), their properties are broadlysimilar. However, since T. brucei and T. congo-lense localise in different tissue in the infected host[7], the released oligopeptidase may have differentpathological effects in the host.

It has been suggested that immune recognitionof parasite pathogenic factors may protect hosts

against the pathological effects of infection [8],and an effective immune response to invarianttrypanosome antigens is thought to be an impor-tant mechanism of resistance to trypanosomiasis[9,10]. Indeed, trypanotolerant N’Dama cattle(Bos taurus) elicit a strong IgG1 response againsta 33 kDa invariant antigen from T. congolense[11]. This response, which is weak in trypanosus-ceptible Boran cattle (Bos indicus ; [9]), is directedagainst the major lysosomal cysteine peptidase ofT. congolense, trypanopain-Tc [6] (also called con-gopain [11]). Thus, in this report the authors alsoinvestigated whether there is a differential recog-nition of OP-Tc, a possible pathogenic factor intrypanosome infections, by the humoral immunesystems of trypanotolerant and trypanosusceptiblecattle.

2. Materials and methods

2.1. Parasite culture

T. congolense (clone IL3000) [12] was culturedin adult male Sprague-Dawley rats, and purifiedfrom infected blood by a combination of Percoll®

isopycnic gradient centrifugation [13] and anion-exchange chromatography on DEAE-cellulose[14].

2.2. Enzyme and protein assays

OP-Tc activity was routinely measured against5 mM Na-carbobenzoxy (Cbz)-Arg-Arg-7-amino-4-methylcoumarin (AMC) at 37°C in assay buffer[50 mM Tris–HCl/10 mM dithiothreitol (DTT),pH 8.0] in an Hitachi F-2000 spectrofluorimeter(lex=370 nm, lem=460 nm). Trypanopain-Tcactivity was determined against 5 mM Cbz-Phe-Arg-AMC at 37°C in 170 mM CH3COONa/30mM CH3COOH/2 mM Na2EDTA/3 mM DTT,

R.E. Morty et al. / Molecular and Biochemical Parasitology 102 (1999) 145–155 147

pH 5.5. Protein assays were conducted accordingto the modified method of Bradford [15].

2.3. Gel electrophoresis and western blotting

Samples were resolved by reducing (5 mM 2-mercaptoethanol) or non-reducing Tris-TricineSDS-PAGE on 10% polyacrylamide gels [16] andproteins were visualised by silver staining [17].Western blotting was performed as described [18],except that electro-transfer from gels to nitrocellu-lose membranes was conducted in 10 mM 3-[cy-clohexylamino]-1-propanesulfonic acid/10% (v/v)methanol, pH 11 (30 V, 16 h; [19]). Unoccupiedsites on the nitrocellulose membranes wereblocked with 5% (m/v) non-fat milk in 20 mMTris–HCl/200 mM NaCl, pH 7.4 (Tris-bufferedsaline, TBS), followed by incubation with therelevant primary antibody diluted in 0.5% (m/v)BSA in TBS (2 h, room temperature). Membraneswere washed in TBS (3×5 min, room tempera-ture), followed by incubation in the relevant sec-ondary antibody diluted in 0.5% (m/v) BSA inTBS (1 h, room temperature). After a secondwash cycle (3×5 min, room temperature), im-mune complexes were visualised with 0.15 mgml−1 5-bromo-4-chloro-3-indolyl phosphate/0.3mg ml−1 nitroblue tetrazolium/100 mM Tris–HCl/5 mM MgCl2, pH 9.5. Rabbit anti-bovineIgG-alkaline phosphatase conjugate, monoclonalanti-bovine IgM clone BM-23-biotin conjugateand ExtrAvidin®-alkaline phosphatase conjugatewere from Sigma (St. Louis, MO, USA).

2.4. Generation of antibodies

OP-Tb was purified from T. b. brucei as de-scribed previously [4], and antibodies raised inchickens by intramuscular injection at two sites inthe breast muscle with a total of 20 mg of OP-Tbemulsified in Freund’s complete adjuvant [20].Chickens were boosted at 3, 7 and 11 weeks afterthe first immunisation, with 10 mg of OP-Tb perbooster, emulsified in Freund’s incomplete adju-vant. Chicken egg-yolk antibodies (IgY) were iso-lated [20], and anti-OP-Tb antibody productionmonitored by enzyme-linked immunosorbent as-say. Western blots were employed to demonstrate

the cross-reactivity of chicken anti-OP-Tb IgYwith OP-Tc, and to screen the serum of infectedbovine hosts for the presence of IgG and IgMantibodies directed against OP-Tc.

2.5. Preparation of immunoaffinity matrix

Anti-OP-Tb IgY (15 mg) was coupled tocyanogen bromide-activated Sepharose-4B (15 ml,packed volume) [21]. Coupling efficiency was esti-mated at 96.7%.

2.6. Purification of OP-Tc

Lysates of T. congolense (6.7×109 cell equiva-lents, 12 ml), prepared by cycles of freezing andthawing, were diluted to 25 ml with 0.1% (m/v)Brij 35. A 10–25% (m/v) three-phase partitioning(TPP) fraction was prepared as described previ-ously for T. brucei [4]. The pellet was resuspendedin buffer A [20 mM sodium acetate/1 mMNa2EDTA/0.02% (m/v) NaN3, pH 5.5] andclarified by centrifugation (15 000×g, 30 min,4°C). The resultant supernatant (15 ml) wasloaded onto a Q-Sepharose column (26×100mm, 1 ml min−1) equilibrated in buffer A con-taining 100 mM NaCl. Material bound to thecolumn that contained trypanopain-Tc activity,was eluted with a linear salt gradient (0.1–1 MNaCl in buffer A; 190 ml—:4—column vol-umes). The fractions that did not bind to thecolumn and which had activity against Cbz-Arg-Arg-AMC were dialysed (16 h, 4°C) againstbuffer B [50 mM Tris–HCl/0.02% (m/v) NaN3,pH 8]. After dialysis, the sample (44 ml) wasloaded onto a p-aminobenzamidine-Sepharosecolumn (120×15 mm, 1 ml min−1) pre-equili-brated in buffer B. The column was washed withbuffer B until the A280 reached baseline before thebound protein was eluted in a single step with 250mM NaCl in buffer B. The bound, active fractionfrom p-aminobenzamidine-Sepharose (22 ml) wasdialysed against buffer C [50 mM Tris–HCl/1 MNaCl/10 mM CaCl2/0.5% (m/v) Brij 35/0.02%(m/m) NaN3, pH 7.4; 16 h, 4°C] and applied tothe anti-OP-Tb-Sepharose immunoaffinity column(50×15 mm, 0.32 ml min−1) equilibrated inbuffer C. After washing of the column with buffer

R.E. Morty et al. / Molecular and Biochemical Parasitology 102 (1999) 145–155148

C until the A280 reached baseline, the boundOP-Tc was eluted with 3.5 M NaSCN in buffer C.NaSCN was removed by dialysis against buffer B,and protein concentrated by ultrafiltration in 3 mlpolysulfone concentrators (10 kDa cut-off; Mil-lipore, Bedford, USA).

2.7. Substrate specificity of OP-Tc

OP-Tc activity against fluorogenic substrateswas determined by addition of substrate afterpreincubation of OP-Tc (1 ng, 37°C, 5 min) inassay buffer (Section 2.2). The initial steady-statevelocity (60) was determined by continuous assayfor a range of substrate concentrations (45 nM–75 mM). Km and Vmax were determined by hyper-bolic regression of the kinetic data using thesoftware package Hyper 1.01 (obtained from J.S.Easterby, University of Liverpool, UK). The kcat

was determined from kcat=Vmax/[E ]0, where [E ]0represents the active enzyme concentration deter-mined with 4-methylumbelliferyl-p-guanidoben-zoate [22].

2.8. Enzymatic characterisation of OP-Tc

The pH profile for OP-Tc was conducted asdescribed above, except that buffers of constantionic strength (100 mM acetic acid/200 mM Tris–HCl/100 mM Mes; I=0.1 [23]) containing 1 mMDTT and 4 mM Na2EDTA over the pH range4–12 were substituted for assay buffer. Reductiveactivation of OP-Tc was investigated by pre-incu-bating OP-Tc (2.5 ng, 37°C, 10 min) in assaybuffer containing DTT, reduced glutathione orL-cysteine (1–25 mM, 37°C, 5 min) prior to theaddition of Cbz-Arg-Arg-AMC (5 mM final con-centration). Effects of divalent metal ions, nucle-otides, polyamines and heparin were examined bypreincubating OP-Tc (5 ng, 37°C, 10 min) in assaybuffer (without DTT) containing the respectivecompound prior to the addition of 5 mM Cbz-Arg-Arg-AMC.

Effects of irreversible peptidase inhibitors wereinvestigated by adding an aliquot of inhibitor (10ml) to the enzyme (140 ml, containing 50–100 ngOP-Tc in 50 mM Tris–HCl, pH 8.0, 37°C).Aliquots were removed at timed intervals and

residual activity (6t) determined using Cbz-Arg-Arg-AMC as described above. Pseudo-first-orderinhibition rate constants (kobs) were obtainedfrom plots of ln 6t/60 versus time where 60 repre-sents the activity prior to addition of inhibitor.Apparent second-order inhibition rate constants(kass) were obtained from the relationship kass=kobs/[I ] where [I ] represents the inhibitor concen-tration [24]. The time required for the free enzymeconcentration to decrease by 50% (half-life, t1

2) is

given by t12=0.693/kass[I ] [25], and kass values

were determined in the presence and absence ofDTT.

For reversible inhibitors, inhibition mechanismwas determined from the effect of inhibitors onthe Km and Vmax for the hydrolysis of Cbz-Arg-Arg-AMC by OP-Tc.

For competitive inhibitors, the inhibition con-stant, Ki, was determined by continuously moni-toring the enzyme-catalysed hydrolysis ofCbz-Arg-Arg-AMC [to establish an uninhibitedrate of substrate hydrolysis (60)], after which a20-fold molar excess of inhibitor over enzyme wasadded, and the new steady-state velocity in thepresence of the inhibitor (6i) determined [25]. Theapparent inhibition constant in the presence ofsubstrate [Ki(app)] was given by 60/6i=1+ [I ]/Ki(app). The true Ki for competitive inhibitors,adjusting for the presence of substrate, was calcu-lated from the relationship Ki=Ki(app)/1+ [S ]/Km

where [S ] denotes substrate concentration. Thekass between OP-Tc and reversible inhibitors wasdetermined as described for irreversible inhibitors,except that the kass was corrected for the presenceof substrate by multiplying kobs/[I ] by 1+ ([S ]/Km)[25]. The rate constant for complex dissociation(kdiss) was determined from the relationship Ki=kdiss/kass [26].

2.9. Infection of bo6ine hosts with T. congolense

Five N’Dama (trypanotolerant) and five Boran(trypanosusceptible) cattle, with a history of pre-vious exposure to T. congolense infections [27],were challenged with T. congolense IL13E-3 [2,28]via the bites of 5–10 infected tsetse flies (Glossinamorsitans centralis), as described [9,29]. All fiveN’Dama cattle exhibited transient parasitaemia

R.E. Morty et al. / Molecular and Biochemical Parasitology 102 (1999) 145–155 149

Table 1Purification table for the isolation of OP-Tc from T. congolense lysates

Total proteinVolume (ml)Fraction Total activity Yield (%)Specific activity Purification (fold)(pmol s−1 mg−1)(pmol s−1)(mg)

196Lysate 222812 11 1 10015TPP 7.4 1473 198 18 66

6.8 1109 16344 15Q-Sepharose 5022 0.413 1012p-Aminobenza- 2450 221 45

midine0.054 754Immunoaffinity 13 9624 1 269 34

and no anaemia. Three Boran cattle developedmoderate anaemia during 6 months of patentparasitaemia and eventually self-cured. The re-maining two Boran cattle became severelyanaemic, their haematocrit fell below 15% anddrug therapy was required to prevent death [9].Sera were collected at intervals during infectionand immediately stored at −70°C. For the cur-rent study, sera taken 3 months post infectionfrom the five N’Dama cattle and from the threeBoran cattle that recovered, were pooled in twoseparate samples, later referred to as ‘N’Dama’and ‘Boran’.

3. Results and discussion

3.1. Purification of OP-Tc

The purification of OP-Tc involves a 4-stepprocedure, which is two steps fewer than that usedfor the purification of OP-Tb [4] and results in aslightly higher yield of enzyme (34 vs. 28%, Table1). Although the major lysosomal cysteine pepti-dase of T. congolense, trypanopain-Tc, co-precipi-tated with OP-Tc in the 10-25% (m/v)(NH4)2SO4-cut, complete separation was effectedby anion-exchange chromatography on Q-Sep-harose, thus facilitating their simultaneous purifi-cation from the same starting material.

Para-aminobenzamidine-Sepharose chromatog-raphy removed 94% of the remaining contaminat-ing protein with only a 5% loss in yield (Table 1).The cross-reactivity of chicken anti-OP-Tb anti-bodies with OP-Tc in T. congolense lysates (Fig.1) prompted their use for immunoaffinity purifica-

tion of OP-Tc. Cbz-Arg-Arg-AMC-hydrolysingactivity was eluted in a single, well-resolved peakfrom the immunoaffinity column. OP-Tc was ho-mogenous on reducing Tris-Tricine SDS-PAGE,producing a single band at 80 kDa (Fig. 1C).Elution of OP-Tc from the immunoaffinitycolumn with 50 mM glycine–HCl, pH 2.3 insteadof 3.5 M NaSCN resulted in a lower final yield(24%, results not shown).

The purified OP-Tc was 5494% active by ac-tive-site titration. The specific activity of thepurified OP-Tc is very similar to that obtained forOP-Tb (13 368 pmol s−1 mg−1) using a longer

Fig. 1. Cross-reactivity of anti-OP-Tb antibodies with Op-Tc.T. congolense lysates (A; 42 mg) and T. b. brucei lysates (B; 40mg) were resolved by non-reducing SDS-PAGE, transferred toa nitrocellulose membrane and probed with chicken anti-OP-Tb IgY (75 mg ml−1). Antigen-antibody complexes were de-tected with a rabbit anti-IgY-alkaline-phosphatase conjugate(40 mg ml−1) as described in the text. (C) Purified OP-Tc (200ng) from the bound, active fraction eluted from the anti-OP-Tb-Sepharose column, resolved by reducing Tris-Tricine SDS-PAGE and visualised by silver staining.

R.E. Morty et al. / Molecular and Biochemical Parasitology 102 (1999) 145–155150

Table 2Amidolytic activity of OP-Tca

Km (mM) kcat (s−1)Substrate number (Sc ) kcat/Km (M−1 s−1)Substrate

P4��P3��P2��P1��P1%0.72 96.0Cbz-Arg-Arg-AMC 133 300 00011.19 114.02 95 800 000Cbz-Gly-Gly-Arg-AMC1.09 67.0Cbz-Phe-Arg-AMC 61 500 0003

4 Cbz-Ala-Arg-Arg-AMC 2.07 92.0 44 400 0001.11 44.0Boc-Leu-Gly-Arg-AMC 39 600 00051.04 37.06 35 600 000Boc-Leu-Lys-Arg-AMC1.69 54.0Boc-Leu-Arg-Arg-AMC 32 000 0007

Boc-Gly-Arg-Arg-AMC8 2.21 49.0 22 200 0002.10 40.0Boc-Val-Gly-Arg-AMC 19 100 0009

H-Ala-Phe-Lys-AMC10 3.00 27.0 9 000 0002.29 16.0 7 000 00011 Boc-Gly-Lys-Arg-AMC4.34 29.0Boc-Val-Leu-Lys-AMC 6 700 000123.03 12.013 4 000 000Cbz-Arg-AMC5.79 22.0Boc-Ala-Gly-Pro-Arg-AMC 3 800 000149.89 17.0 1 700 00015 Boc-Val-Pro-Arg-AMC

54.40 2.7H-Arg-AMC 50 00016

a No activity was detected against Ac-Ala-Ala-Pro-Ala-AMC, H-Gly-AMC; H-Leu-AMC; MeoSuc-Gly-Trp-Met-AMC; Suc-Leu-Tyr-AMC; Glt-Gly-Gly-Phe-AMC or H-Gly-Pro-AMC after 1 h.

purification method [4]. The specific activity of13 962 pmol−1 s−1 mg−1 is :10 times thatreported for a potentially related trypsin-like ser-ine oligopeptidase (1379 pmol p-nitroanilide s−1

mg−1 [30]) and a chymotrypsin-like serineoligopeptidase (1620 pmol p-nitroanilide s−1

mg−1 [31]) isolated from the pollen of Prosophis6elutina. However, these activities were deter-mined at 25°C (as opposed to 37°C in this study)and employed the less-sensitive photometric p-ni-troanilide substrates [25] in contrast to the fluori-metric substrates employed in this study, whichmay account for the higher specific activity re-ported here.

3.2. Substrate specificity of OP-Tc

A variety of commercially available fluorogenicpeptides were tested for their potential as sub-strates for OP-Tc and clearly showed that OP-Tchas a trypsin-like activity (Table 2). A basicamino acid residue (Arg or Lys) in the P1 position(nomenclature of Schechter and Berger [32]) isnecessary for OP-Tc to cleave the potential sub-strate. The low second-order rate constant (kcat/Km) for H-Arg-AMC (Sc16; 50 000 M−1 s−1)

and lack of activity against H-Gly-AMC andH-Leu-AMC indicates that OP-Tc has pooraminopeptidase activity. In contrast, peptideswith a blocked N-terminus are much better sub-strates. For example, the kcat/Km for the hydroly-sis of Cbz-Arg-AMC (Sc13) by OP-Tc is 80-foldgreater than that for H-Arg-AMC (Sc16). Bothsubstrate binding and catalytic rate are improvedwhen the P1- and the P2-binding sites are occu-pied. The failure of amastatin and bestatin toinhibit OP-Tc (see later) also points to the lack ofaminopeptidase activity in OP-Tc.

Although efficient hydrolysis by OP-Tc requiresthat the P2 position be occupied, not all aminoacids are equally acceptable in the P2-positionsince the nature of the residues occupying thesesites clearly influences the rate of hydrolysis. Twosubstrates (Sc14, Sc15) were tested that hadP2-Pro, and both were hydrolysed only slowly.Thus, Pro does not appear to be readily acceptedin the OP-Tc P2-binding site. In contrast, sub-strates that had Gly, Lys, Arg, Phe or Leu in P2

appeared to be more acceptable. The second-or-der rate constants for the hydrolysis of the Boc-Leu-Xaa-Arg-AMC family of peptides (Sc5,Sc6 and Sc7) were very similar indicating that

R.E. Morty et al. / Molecular and Biochemical Parasitology 102 (1999) 145–155 151

the residues Gly, Lys, and Arg were almostequally acceptable in the P2-position. Comparisonof Sc1 and Sc3 shows that P2-Arg is preferredover P2-Phe. Thus, the overall P2-binding sitepreference of OP-Tc seems to be (Gly,Lys,Arg)\Phe\Leu\Pro. However, it is also clear thatamino acid and other residues in P3 and P4 caninfluence the enzyme interactions with P2. Forexample, a comparison of the substrates (Sc8and Sc11) versus (Sc6 and Sc7) shows that,in the pair Boc-Leu-Xaa-Arg-AMC, the substratewith a P2-Lys is hydrolysed about 10% faster thanthat with a P2-Arg, whereas, in the pair Boc-Gly-Xaa-Arg-AMC, the situation is reversed and thesubstrate with a P2-Arg is hydrolysed about 200%better than that with a P2-Lys. Despite such com-plications, the same general P2 preference trend(Gly\Phe\Leu\Pro) is also observed in thereactivity of OP-Tc with different peptidyl-in-hibitors of OP-Tc (see later).

Because of the apparent cross-talk between thedifferent binding-subsites, it is even more difficultto define the properties of the P3 and P4 bindingsites of OP-Tc. However, it appears that bothsubsites are hydrophobic since, for example, thesecond-order rate constant for the hydrolysis ofthe substrate Boc-Leu-Lys-Arg-AMC (Sc6) is5-fold greater than that for Boc-Gly-Lys-Arg-AMC (Sc11). In this regard it is interesting thatthe best substrate tested (Sc1; Cbz-Arg-Arg-AMC) also has a hydrophobic (benzyloxycar-bonyl) group in the P3 position.

3.3. Enzymatic characterisation of OP-Tc

OP-Tc activity against Cbz-Arg-Arg-AMCpeaked at pH 9. However, OP-Tc remained sub-stantially active (73% of maximal activity) atphysiological pH (7.4). OP-Tc is primarily a cyto-solic enzyme (Morty, unpublished observations)as also is OP-Tb [33]. As the cytosol is a reducingenvironment, it is possible that intracellular re-ducing agents such as glutathione and trypanoth-ione may act as in vivo regulators of activity ofthese oligopeptidases. Indeed, cysteine, reducedglutathione and dithiothreitol, all enhanced OP-Tc activity at 10 mM by 4, 5 and 7-fold, respec-tively (Table 3). Similarly, the polyamines

spermine and spermidine (50 mM) enhanced theactivity of OP-Tc against Cbz-Arg-Arg-AMC by79 and 77%, respectively when compared to un-treated OP-Tc. Since trypanosomes containpolyamines, including putrescine and spermidine[34], polyamines may also be involved in theregulation of OP-Tc in vivo. Interestingly, hep-arin, which carries an opposite charge topolyamines, also enhanced OP-Tc activity by 66%at 30 mg ml−1. Nucleotides (ATP, GTP) andornithine were without effect. Similarly, chloridesalts of Mn2+, Mg2+, Ba2+, Ca2+, which do notreact with cysteine residues, were without effect(results not shown). In contrast, OP-Tc activitywas completely inhibited by the chloride salts ofZn2+, Hg2+, Cd2+ and Fe2+ at 1 mM (resultsnot shown), and this probably occurs by theformation of mercaptides between heavy metalions and a reactive thiol group [35]. These resultsare consistent with OP-Tc having a cysteineresidue which must be maintained in its reducedstate to preserve catalytic activity (see also below).

Classical reversible aspartic- and metallo-pepti-dase inhibitors were without effect against OP-Tc.The lack of inhibitory activity of 1 mM EDTA,ethyleneglycol-bis(b-aminoethylether)-N,N,N %,N %-tetraacetic acid and 1,10-phenanthroline rein-forces the observation that OP-Tc has no metal-ion dependence.

Table 3Effect of potential activators on OP-Tc activitya

Relative activity (%)Compound Concentration(mM)

100–None10L-cysteine 43410 762DTT

Reduced glu- 54510tathione

30 mg ml−1Heparin 166Spermine 1790.05Spermidine 1770.05

108Putrescine 0.05

a No effect on activity was observed in the presence of 50mM ornithine, 1 mM ATP or 1 mM GTP. S.D.s for theactivities were B95%.

R.E. Morty et al. / Molecular and Biochemical Parasitology 102 (1999) 145–155152

Table 4Inhibition of OP-Tc activity by irreversible inhibitors of cys-teine and serine proteinases

t12

b (s)kassa (M−1 s−1)Inhibitor

AEBSF 22.0294.05 126DCI 16167.00923.60

21313.0992.17DFP1.1190.71Iodoacetamidec 2 497

1 265Iodoacetic acidc 2.1990.081 6111.7290.11N-ethylmaleimidec

29.1094.47pCMBc 953 046PMSF 0.9190.02

a Data reflect the mean kass9S.D. (n=3).b t1

2at 250 mM inhibitor concentration.

c Assays conducted in the absence of dithiothreitol. In thepresence of 10 mM dithiothreitol, the t1

2for the inhibition of

OP-Tb by thiol-reactive agents was elevated as follows:iodoacetamide (2520 s), iodoacetic acid (2559 s), N-ethyl-maleimide (1900 s) and pCMB (1854 s).

Table 5Peptidyl chloromethyl ketone inhibitors of OP-Tca

Inhibitor kassb (×10−5 M−1 t1

2

c (s)s−1)

Tos-LysCH2Cl 7.974.4990.23biotin-ArgCH2Cl 9.354.0590.49

2.2990.36 12.83Boc-Val-Leu-Gly-LysCHN2

Leu-Glu-Gly-ArgCH2Cl 2.9490.27 13.59Cbz-Phe-LysCH2Cl 15.572.0290.08

16.821.7790.16Asp-Val-Phe-LysCH2Cl2.0790.44Asp-Val-Leu-LysCH2Cl 20.08

32.69Asp-Phe-Pro-ArgCH2Cl 0.9990.39

a No inhibition was detected after 30 min pre-incubationwith 100 mM Ac-Ala-Ala-Ala-AlaCH2Cl; Cbz-Gly-Gly-PheCH2Cl; Cbz-Gly-Leu-PheCH2Cl; Cbz-Leu-Leu-MetCHN2;biotin-PheCH2Cl; Tos-PheCH2Cl; Cbz-Ala-AlaCHN2; Cbz-Ile-LeuCHN2; Cbz-Phe-AlaCH2Cl; Cbz-Phe-AlaCHN2; Cbz-Phe-Gly-TyrCHN2; Cbz-Phe-PheCH2Cl; Cbz-Phe-PheCHN2;Cbz-Phe-Tyr(OBut)CHN2 or MeoSuc-Ala-Ala-Pro-ValCH2Cl.

b Data reflect the mean kass9S.D. (n=3).c t1

2at 10 mM inhibitor concentration.

OP-Tc was inactivated by non-peptide irre-versible inhibitors of serine peptidases (Table 4) atrates comparable to those observed for other ser-ine proteases, including OP-Tb ([3]; Morty et al.,manuscript in preparation). These inhibitors in-cluded 3,4-dichloroisocoumarin (DCI), 4-(2-aminoethyl)benzenesulfonylfluoride (AEBSF),diisopropylfluorophosphate (DFP) and phenyl-methanesulfonylfluoride (PMSF). The inhibitionof OP-Tc activity by thiol-blocking agents, includ-ing N-ethylmaleimide, iodoacetate, iodoacetamideand para-chloromercuribenzoate (pCMB), sup-ports the reductive activation data, and suggeststhat a reactive cysteine residue is at or near theactive site of the enzyme. Modification of such aresidue would impair substrate binding or cataly-sis, as has been described for porcine prolyloligopeptidase [36,37].

OP-Tc was inhibited rapidly (kass\105 M−1

s−1) by peptidyl chloromethyl ketones (and adiazomethane) with basic residues in P1, but notby related molecules lacking Arg or Lys in P1

(Table 5). This supports the contention that OP-Tc has trypsin-like activity. The inhibition resultsalso show that the OP-Tc P2-subsite can accom-modate small, uncharged and hydrophobic (Gly,Phe, Leu), as well as basic residues (Arg, Lys).This is consistent with the trends observed for the

hydrolysis of fluorogenic peptide substrates byOP-Tc (Table 2). Again, proline in P2 appeared tobe detrimental to reactivity. OP-Tc was reversiblyinhibited by the peptide aldehydes antipain andleupeptin (Table 6), which have the aldehyde onthe P1-Arg residue However, OP-Tc was not in-hibited by chymostatin, thus providing furtherevidence that Phe is not acceptable in P1 (Tables 2and 5).

Table 6Inhibition of OP-Tc activity by competitive reversibleinhibitorsa

Ki (mM)Inhibitor kass (M−1 s−1) kdiss (s−1)

1.1×1030.54×10521.1×10−3Leupeptin2.1×10−3Antipain 8.9×105 1.9×103

Aprotinin 0.9×10−3 n.d. n.d247.0 n.d. n.d.Benzamidine

73.7 n.d.E-64 n.d.

a No inhibition was observed with amastatin (125 mM),bestatin (125 mM), chicken ovomucoid (100 mg ml−1), chymo-statin (1–100 mM), EDTA (1 mM), ethyleneglycol-bis(b-aminoethylether)-N,N,N %,N %-tetraacetic acid (1 mM), elastinal(150 mM), lima bean trypsin inhibitor (100 mg ml−1), pepstatin(1 mM), 1,10-phenanthroline (1 mM) or soybean trypsin in-hibitor (100 mg ml−1). n.d., not determined. In these cases, thekass was too fast to be measured experimentally.

R.E. Morty et al. / Molecular and Biochemical Parasitology 102 (1999) 145–155 153

Curiously, OP-Tc was inhibited competitivelyby E-64 with a Ki of 73.7 mM (Table 6). Al-though this molecule is generally considered toinhibit only cysteine peptidases [38], other work-ers have demonstrated that E-64 is an effectiveinhibitor of trypsin [39]. Another unexpected ob-servation was that benzamidine was a compara-tively poor inhibitor of OP-Tc (Ki 247 mM). Incontrast, bovine b-trypsin (Ki 36 mM) and mast-cell tryptase (Ki 12 mM) [40] are inhibited wellby benzamidine. Whether this reflects a morerestricted P1-binding site in OP-Tc, or the needfor additional subsite binding by OP-Tc, is notclear. The differential inhibition of OP-Tc byaprotinin (Ki 0.89 nM; Table 6) and soybeantrypsin inhibitor (no inhibition) indicates thatOP-Tc (like OP-Tb) can accommodate only lowmolecular mass polypeptides, but not proteins,in its active-site. Such limited reactivity willclearly affect the pattern of host molecule hy-drolysis that may be expected should OP-Tc bereleased into the host bloodstream. The authorshave previously illustrated that a trypsin-likeserine oligopeptidase is present in the blood-stream of T. b. brucei-infected rats [4], and theyhypothesise that this activity is released into thehost bloodstream by dying parasites. Overall,the enzymatic and physical properties of OP-Tcare quite similar to those for endopeptidasesidentified in other protozoans including OP-Tbfrom T. brucei [4] although, on average, OP-Tcappears to be about 3-fold less active than OP-Tb. For example, the Km (0.72 mM, Table 2) forthe hydrolysis of Cbz-Arg-Arg-AMC (Sc1) byOP-Tc is 3.5-fold above that obtained for OP-Tb from T. b. brucei (210 nM; Morty et al.,manuscript in preparation), and closely approxi-mates those obtained for similar peptidasespurified from Crithidia fasciculata (1 mM [41])and recombinant oligopeptidase B from T. cruzi(5 mM [42,43]). However, the enzyme kineticparameters with different substrates and in-hibitors are not the same. For example, the sec-ond-order rate constant for the hydrolysis ofCbz-Arg-Arg-AMC (Sc1) by OP-Tc is abouttwo-fold greater than that for Cbz-Phe-Arg-AMC (Sc3), but for OP-Tb this difference is

6-fold ([4]; Morty et al., manuscript in prepara-tion).

3.4. Immunological studies

Both N’Dama and Boran cattle produced IgGprimarily against T. congolense antigens of 30,38 and 50 kDa (results not shown). The 38 kDaantigen is possibly trypanopain-Tc [11,44], andits higher immunogenicity in N’Dama cattle isconsistent with a previous report that, followinga T. congolense re-infection, IgG to congopainwas detected in all N’Dama cattle and in theleast affected of the Boran cattle [9]. OP-Tc didnot appear to elicit IgG antibodies in N’Damaor Boran cattle.

IgM from both N’Dama and Boran seratargeted T. congolense antigens primarily at 38and 80 kDa (results not shown). Boran seraproduced a stronger response, supporting earlierobservations that susceptible Boran cattlehave higher IgM responses, but weaker IgG re-sponses when compared with trypanotolerantN’Dama cattle [9]. Despite an 80 kDa bandtargeted in T. congolense lysates, purified OP-Tcwas not targeted by IgM from N’Dama or Bo-ran sera.

Taken together, these data show that OP-Tcis not an especially immuno-dominant epitope inT. congolense infections and that it is not differ-entially recognised by the humoral immune sys-tems of trypanotolerant and trypanosusceptiblecattle. However, because trypanosome-infectedanimals have uninhibited endopeptidase activityin their blood [4], it will be interesting to exam-ine whether active immunisation with OP-Tc hasany effect on the progression of a subsequentchallenge infection with T. congolense.

Acknowledgements

This study received financial support fromthe South African Foundation for Research De-velopment and the University of Natal Re-search Fund. The authors thank Drs. Oma-lokoho Tosomba and Robert Pike for helpfuldiscussions.

R.E. Morty et al. / Molecular and Biochemical Parasitology 102 (1999) 145–155154

References

[1] Facer CA, Crossley JM, Clarkson MJ, Jenkins GC.Haemolytic anaemia in bovine trypanosomiasis. J CompPathol 1982;92:393–401.

[2] Nantulya VM, Musoke AJ, Rurangirwa FR, Moloo SK.Resistance of cattle to tsetse-transmitted challenge withTrypanosoma brucei or T. congolense after spontaneousrecovery from syringe-passaged infections. Infect Immun1984;43:735–8.

[3] Kornblatt MJ, Mpimbaza GWN, Lonsdale-Eccles JD.Characterisation of an endopeptidase of Trypanosomabrucei brucei. Arch Biochem Biophys 1992;293:25–31.

[4] Troeberg L, Pike RN, Morty RE, Berry RK, CoetzerTHT, Lonsdale-Eccles JD. Proteases from Trypanosomabrucei brucei : purification, characterisation and interac-tions with host regulatory molecules. Eur J Biochem1996;238:728–36.

[5] Rautenberg P, Schadler R, Reinwald E, Risse H-J. Studyon a proteolytic enzyme from Trypanosoma congolense.Purification and some biochemical properties. Mol CellBiochem 1982;47:151–9.

[6] Mbawa ZR, Gumm ID, Shaw E, Lonsdale-Eccles JD.Characterisation of a cysteine protease from bloodstreamforms of Trypanosoma congolense. Eur J Biochem1992;204:371–9.

[7] Banks KL. Injury induced by Trypanosoma congolenseadhesion to cell membranes. J Parasitol 1980;66:34–7.

[8] Playfair JHL, Taverne J, Bate CAW, De Souza JB. Themalaria vaccine: anti-parasite or anti-disease. ImmunolToday 1990;11:25–9.

[9] Authie E, Muteti DK, Williams DJL. Antibody responseto invariant antigens of Trypanosoma congolense in cattleof differing susceptibility to trypanosomiasis. Parasite Im-munol 1993;15:101–11.

[10] Taylor KA. Immune responses of cattle to African try-panosomes: protective or pathogenic? Int J Parasitol1998;28:219–40.

[11] Authie E, Muteti DK, Mbawa Z, Lonsdale-Eccles JD,Webster P, Wells C. Identification of a major antigen ofTrypanosoma congolense as a cysteine proteinase. MolBiochem Parasitol 1992;56:103–16.

[12] Fish WR, Muriuki CW, Muthiani AM, Grab DJ, Lons-dale-Eccles JD. Disulfide bond involvement in the mainte-nance of the cryptic nature of the cross reactingdeterminant of metacyclic forms of Trypanosoma congo-lense. Biochemistry 1989;28:5415–21.

[13] Grab DJ, Bwayo JJ. Isopycnic isolation of African try-panosomes on Percoll gradients formed in situ. Acta Trop1982;39:363–6.

[14] Lanham SM, Godfrey DG. Separation of trypanosomesfrom the blood of infected rats and mice by anion-ex-changers. Exp Parasitol 1970;28:521–34.

[15] Read SM, Northcote DH. Minimisation of variation inthe response to different proteins of the Coomassie Bluedye-binding assay for protein. Anal Biochem1981;116:53–64.

[16] Schagger H, von Jagow G. Tricine-sodium dodecyl sul-fate-polyacrylamide gel electrophoresis for the separationof proteins in the range from 1 to 100 kDa. Anal Biochem1987;166:368–79.

[17] Blum H, Beier H, Gross HJ. Improved silver staining ofplant proteins, RNA and DNA in polyacrylamide gels.Electrophoresis 1988;8:93–9.

[18] Towbin H, Staehelin T, Gordon J. Electrophoretic trans-fer of proteins from polyacrylamide gels to nitrocellulosesheets: procedure and applications. Proc Natl Acad SciUSA 1979;76:4350–4.

[19] Matsudaira P. Sequence from picomole quantities ofproteins electroblotted onto polyvinylene difluoride mem-branes. J Biol Chem 1987;262:10035–8.

[20] Polson A, Coetzer THT, Kruger J, von Maltzahn E, vander Merwe KJ. Improvements in the isolation of IgYfrom the yolks of eggs laid by immunized hens. ImmunolInvest 1985;14:323–7.

[21] Hermanson GT, Mallia AK, Smith PK. ImmobilisedAffinity Ligand Techniques. San Diego: Academic Press,1992:223.

[22] Jameson GW, Roberts DV, Adams RW, Kyle WS, El-more DT. Determination of the operational molarity ofsolutions of bovine alpha-chymotrypsin, trypsin,thrombin and factor Xa by spectrofluorometric titration.Biochem J 1973;131:107–17.

[23] Ellis KJ, Morrison JF. Buffers of constant ionic strengthfor studying pH-dependant processes. Methods Enzymol1982;87:405–26.

[24] Potempa J, Pike RN, Travis J. Titration and mapping ofthe active site of cysteine proteinases from Porphyromonasgingi6alis (gingipains) using peptidyl chloromethanes. BiolChem 1997;378:223–30.

[25] Salvesen G, Nagase H. Inhibition of proteolytic enzymes.In: Benyon RJ, Bond, JS, editors. Proteolytic Enzymes: aPractical Approach. Oxford: IRL Press, 1989;83–104.

[26] Beith JG. Theoretical and practical aspects of proteinaseinhibition kinetics. Methods Enzymol 1995;248:59–84.

[27] Paling RW, Moloo SK, Scott JR, et al. Susceptibility ofN’Dama and Boran cattle to sequential challenges withtsetse-transmitted clones of Trypanosoma congolense. Par-asite Immunol 1991;13:413–20.

[28] Morrison WI, Roelants GE, Mayor-Withey KS, MurrayM. Susceptibility of inbred strains of mice to Try-panosoma congolense : correlation with changes in spleenlymphocyte population. Clin Exp Immunol 1978;32:25–40.

[29] Williams DJL, Naessens J, Scott JR, McOdimba FA.Analysis of peripheral blood leukocyte populations inN’Dama and Boran cattle following rechallenge infectionwith Trypanosoma congolense. Parasite Immunol1991;13:171–7.

[30] Matheson N, Schmidt J, Travis J. Isolation and proper-ties of an angiotensin II-cleaving peptidase from mesquitepollen. Am J Resp Cell Mol Biol 1995;12:228–441.

[31] Matheson N, Travis J. Purification and characterisationof a novel peptidase (Dimes) from mesquite (Prosophis6elutina) pollen. J Biol Chem 1998;273:16771–7.

R.E. Morty et al. / Molecular and Biochemical Parasitology 102 (1999) 145–155 155

[32] Schechter I, Berger A. On the size of the active-site inproteases. I. Papain. Biochem Biophys Res Comm1967;27:157–62.

[33] Lonsdale-Eccles JD, Grab DJ. Lysosomal and non-lyso-somal peptide hydrolases of the bloodstream forms ofTrypanosoma brucei brucei. Eur J Biochem 1987;169:467–75.

[34] Fairlamb A, Le Quesne SA. Polyamine metabolism intrypanosomes. In: Hide G, Mottram JC, Coombs GH,Holmes, PH, editors. Trypanosomiasis and Leishmania-sis: Biology and Control. Oxford: CAB International,1997;9:149–161.

[35] Barron ESG. Thiol groups of biological importance. AdvEnzymol 1995;11:201–53.

[36] Polgar L. pH-dependant mechanism in the catalysis ofprolyl oligopeptidase from pig muscle. Eur J Biochem1991;197:441–7.

[37] Fulop V, Bocskei Z, Polgar L. Prolyl oligopeptidase: anunusual b-propeller domain regulates proteolysis. Cell1998;94:161–70.

[38] Barrett AJ, Kambhavi AA, Brown MA, et al. L-trans-epoxysuccinyl-leucylamido(4-guanido)butane (E-64) andits analogues as inhibitors of cysteine proteases, includingcathepsins B, H and L. Biochem J 1992;201:189–98.

[39] Sreedharan SK, Verma SC, Caves LS, et al. Demonstra-tion that 1-trans-epoxysuccinyl-L-leucylamido-(4-

guanidino) butane (E-64) is one of the most effective lowMr inhibitors of trypsin-catalysed hydrolysis. Characteri-zation by kinetic analysis and by energy minimization andmolecular dynamics simulation of the E-64-beta-trypsincomplex). Biochem J 1996;316:777–86.

[40] Caughey GH, Raymond WW, Bacci E, Lombardy RJ,Tidwell RR. Bis(5-amidino-2-benzamizazolyl)methaneand related amidines are potent, reversible inhibitors ofmast cell tryptases. J Pharmacol Exp Therap1993;264:676–82.

[41] Ashall F. Characterisation of an alkaline peptidase ofTrypanosoma cruzi and other trypanostomatids. MolBiochem Parasitol 1990;38:77–88.

[42] Burleigh BA, Andrews NW. A 120 kDa alkaline pepti-dase from Trypanosoma cruzi is involved in the generationof a novel Ca2+-signalling factor for mammalian cells. JBiol Chem 1995;270:5172–80.

[43] Burleigh BA, Caler EV, Webster P, Andrews NW. Acytosolic serine endopeptidase from Trypanosoma cruzi isrequired for the generation of Ca2+ signalling in mam-malian cells. J Cell Biol 1997;136:609–20.

[44] Authie E, Duvallet G, Robertson C, Williams DJL. Anti-body response to a 33 kDa cysteine proteinase of Try-panosoma congolense : relationship to ‘trypanotolerance’in cattle. Parasite Immunol 1993;15:465–74.

.

.