Leishmania major elongation factor 1B complex has trypanothione S-transferase and peroxidase activity.

In the Trypanosomatidae, trypanothione has subsumed many of the roles of glutathione in defense against chemical and oxidant stress. Crithidia fasciculata lacks glutathione S-transferase, but contains an unusual trypanothione S-transferase activity that is associated with eukaryotic translation elongation factor 1B (eEF1B). Here we describe the cloning, expression, and reconstitution of the purified alpha, beta, and gamma subunits of eEF1B from Leishmania major. Individual subunits lacked trypanothione S-transferase activity. Only eEF1B, formed by reconstitution or co-expression of the three subunits, was able to conjugate a variety of electrophilic substrates to trypanothione or glutathionylspermidine, but not glutathione. In contrast to the C. fasciculata eEF1B, the L. major enzyme also displayed peroxidase activity against a variety of organic hydroperoxides. The enzyme showed no activity with hydrogen peroxide and greatest activity with linoleic acid hydroperoxide (1 unit mg(-1)). Kinetic studies suggest a ternary complex mechanism, with Km values of 140 mum for trypanothione and 7.4 mm for cumene hydroperoxide and kcat=25 s(-1). Immunofluorescence studies indicate that the enzyme may be localized to the surface of the endoplasmic reticulum. These results suggest that, in addition to its role in protein synthesis, the Leishmania eEF1B may help protect the parasite from lipid peroxidation.

Parasitic protozoa of the family trypanosomatidae cause disease and death throughout the tropical and subtropical world. Trypanosoma brucei infections are estimated to cause ϳ400,000 cases of sleeping sickness per year; Trypanosoma cruzi, the cause of Chagas' disease, chronically infects ϳ17 million people and Leishmania spp. are thought to cause 2 million cases of leishmaniasis per year. 1 Current chemotherapies for these diseases are, on the whole, ineffective and toxic (2,3), whereas effective vaccines may never be developed. The development of effective treatments for these infections is therefore an urgent necessity.
New anti-parasitic drugs can be developed from inhibitors of biochemical pathways that are essential for parasite survival but absent from the host. One such target is the thiol metabo-lism of trypanosomatids. Uniquely, this is dependent upon trypanothione (N 1 ,N 8 -bis(glutathionyl)spermidine or T[SH] 2 ) 2 (4), whereas their human hosts use glutathione ((␥-L-glutamyl-L-cysteinylglycine or GSH). Trypanothione is involved in protective processes, defending against oxidative stress through peroxidase systems (5), against reactive aldehydes through the glyoxalase system (6,7), and against toxic xenobiotics through the trypanothione S-transferase (TST) (8). All these processes depend upon trypanothione being maintained in its dithiol form by trypanothione reductase. This NADPH-dependent flavoenzyme is essential to the trypanosomatids (9, 10) as, in these organisms, it is the only known route for the transfer of reducing equivalents from NADPH to low molecular mass thiols. In addition, because these parasites lack an equivalent to thioredoxin reductase, T[SH] 2 also reduces their orthologues of thioredoxins, the tryparedoxins (TryX) (5).
The role of T[SH] 2 in the defense against oxidative stress is particularly important, as reactive oxygen species are critical to the host-parasite interaction (11). Furthermore, several of the clinically used drugs, such as the antimonials used to treat leishmaniasis, and the nitroimidazoles and nitrofurans used to treat Chagas' disease, may all act through the induction of oxidative stress (12). Trypanosomatids lack catalase and selenium-dependent peroxidases and are unusually dependent upon a set of thiol-dependent peroxidases (5). These enzymes fall into two classes, both of which are thought to contain a redox-active cysteine residue in their active sites that is reduced by their first substrate and then oxidized by the hydroperoxide second substrate (13,14). The best characterized of these are the tryparedoxin peroxidases (13,(15)(16)(17). These are a group of peroxiredoxins that accept electrons from TryX and effectively reduce hydrogen peroxide. However, they vary in their abilities to metabolize aryl and alkyl hydroperoxides, such as cumene hydroperoxide (CuOOH). Recently, a second class of peroxidases has been discovered in T. cruzi (18,19) and T. brucei (14) with higher activity toward hydrophobic hydroperoxides. These are homologues of the selenoprotein glutathione peroxidase (GPX), with the active site selenocysteine being replaced by a cysteine residue. However, with the exception of the T. cruzi GPX II (20), these enzymes use TryX, rather than GSH as their preferred reductant (14,19).
In mammalian cells, another class of enzymes that are im-portant in the metabolism of hydrophobic hydroperoxides are the glutathione S-transferases or GSTs. The GSTs are detoxification enzymes that catalyze the nucleophilic attack of GSH on a wide variety of hydrophobic substrates (23). These enzymes can therefore not only reduce the phospholipid hydroperoxides generated during lipid peroxidation (24,25), but they can also detoxify the reactive aldehydes such as 4-hydroxynon-2-enal produced by the breakdown of these oxidized lipids (26).
Recently, a trypanothione S-transferase (TST) activity was detected in several members of the trypanosomatidae and purified from Crithidia fasciculata. The TST complex was identified as the eukaryotic translation elongation factor 1B (eEF1B), which is involved in ribosomal protein synthesis, and the TST active site localized to the ␥ subunit (eEF1B␥) (8). Previous studies in other organisms had shown that the other two subunits (eEF1B␣ and -␤) of eEF1B act to recycle translation elongation factor 1A complexed with GDP back to the active GTP-complexed form (27,28). The S-transferase activity of eEF1B␥ suggests possible roles of this subunit in translational responses to oxidative and xenobiotic stress. In this paper we describe the expression of the three subunits of the Leishmania major eEF1B and the reconstitution and enzymatic characterization of the complex. The eEF1B holocomplex, but not the isolated subunits, was able to catalyze the conjugation of T[SH] 2 to a variety of electrophiles, being most active with hydrophobic hydroperoxides.
Cloning of L. major eEF1B Subunits-The genes encoding the L. major eEF1B subunits were identified in the L. major genome data base (www.genedb.org). The eEF1B␣ gene was amplified by PCR from genomic DNA, using the sense primer 5Ј-CATATGTCCACCCTCAAG-GAAGTCAAC-3Ј and the antisense primer 5Ј-GGATCCTTAAATCTT-GTTCCAGGCGACGA-3Ј. The PCR product was then cloned into the pCR-Blunt II-TOPO plasmid using the Zero Blunt TOPO PCR cloning kit (Invitrogen) and fully sequenced. Similarly, an eEF1B␤ gene was amplified and cloned using the sense primer 5Ј-CATATGTCTCTGAA-GGACGTGAGCAAGAAG-3Ј and the antisense primer 5Ј-GGATCCTC-AGATCTTGTTCCAGGCGACGA-3Ј and the eEF1B␥ gene using the sense primer 5Ј-CATATGACCTACAAGCTCCTCGCCC-3Ј and the antisense primer 5Ј-GGATCCTTACTTGAAGCAGCGGCCCTC-3Ј. These primers added 5Ј NdeI sites and 3Ј BamHI sites (underlined), allowing subcloning of these three genes into the expression plasmid pET3aTr (33) and, in addition, the eEF1B␤ gene into pET15b (Novagen).
The polycistronic expression construct pST39.eEF1B was assembled by the insertion of these three genes into translation cassettes 1, 2, and 4 of the polycistronic expression vector pST39 (33). First, the eEF1B␤ gene was excised from pET15b.eEF1B␤ by digestion with XbaI and BamHI, this removed a fragment containing the plasmid 5Ј-untranslated region and the eEF1B␤ gene as an in-frame fusion with the vector-derived N-terminal hexahistidine tag. This fragment was then ligated into the corresponding sites of pST39, to create the plasmid pST39.eEF1B␤. Similarly, the eEF1B␣ gene was excised by EcoRI and HindIII digestion from pET3aTr.eEF1B␣ and ligated into these sites in the pST39.eEF1B␤ plasmid, to create pST39.eEF1B␣␤. The eEF1B␥ gene was then re-amplified, using the pET3aTr.eEF1B␥ plasmid as template, the sense primer 5Ј-CCCGGGTTGGAATTCGCTAGCTCTA-GAAATAAT-3Ј, and the antisense primer 5Ј-CTCGAGTTACTTGAAG-CAGCGGCCCTC-3Ј. These primers amplified a fragment containing the pET3aTr 5Ј-untranslated region and added a 5Ј XmaI site and a 3Ј XhoI site (underlined), allowing subcloning into the BspEI and XhoI sites of the pST39.eEF1B␣␤ plasmid, to create pST39.eEF1B.
Expression and Purification of eEF1B Subunits-The pET3aTr. eEF1B␣, -␤, and -␥ plasmids were transformed into BL21(DE3)pLys-S Escherichia coli. These were grown at 37°C in LB media containing 100 g ml Ϫ1 carbenicillin and 12.5 g ml Ϫ1 chloramphenicol to an OD of 0.6. After cooling to 25°C, expression was induced for 4 h with 1 mM isopropyl ␤-D-galactopyranoside and the cells then harvested and frozen.
All three eEF1B subunits were expressed as untagged proteins from pET3aTr and purified by anion-exchange and size-exclusion chromatography. Unless otherwise specified, all procedures were carried out at 4°C. Cells were resuspended in a final volume of 30 ml of lysis buffer (75 mM (Na ϩ ) phosphate, pH 7.5, 1 mM benzamidine, 3 g ml Ϫ1 leupeptin, 250 M 4-(2-aminoethyl)benzenesulfonyl fluoride, 1 M pepstatin A) to which dithiothreitol and EDTA were added to a final concentration of 1 mM. The cells were then lysed by sonication (4 ϫ 30-s bursts), with cooling to Ͻ4°C between pulses. After centrifugation (45,000 ϫ g, 1 h), the supernatant was applied at 2 ml min Ϫ1 to a 25-ml (1.6 ϫ 8 cm) Q-Sepharose HP anion-exchange column equilibrated in either buffer A (20 mM (Na ϩ ) histidine, pH 6) for eEF1B␣ and -␤, or buffer B (25 mM (Na ϩ ) bis-Tris, pH 6.5) for eEF1B␥. The column was washed for 30 min and bound proteins were then eluted with a linear gradient of 0 -1 M NaCl, in the same buffer. Fractions containing the recombinant protein were pooled and concentrated to 5 ml. This sample was then applied to a 319-ml (2.6 ϫ 60 cm) Superdex 200 26/60 size-exclusion column equilibrated with buffer C (50 mM (Na ϩ ) HEPES, pH 7.5, 300 mM NaCl, 0.01% (w/v) NaN 3 ) and eluted at a flow rate of 2 ml min Ϫ1 .
Complex Reconstitution-Purified eEF1B subunits were mixed at a final concentration of 50 M in 400 l of buffer C containing 1 mM dithiothreitol and incubated for 10 min at 25°C. The mixtures were either separated by size-exclusion chromatography, as described below, or dialyzed for 1 h against 1 liter of buffer B and applied at 2 ml min Ϫ1 to a 1-ml (0.64 ϫ 3 cm) Resource Q anion-exchange column (Amersham Biosciences) equilibrated in buffer B. Bound proteins were then eluted with a linear gradient of 0 -0.5 M NaCl in buffer B.
Polycistrionic Expression and Complex Purification-BL21(DE3) pLys-S E. coli, transformed with pST39.eEF1B, were grown and protein expression was induced, as before. The cells were lysed, as described above, in lysis buffer modified by the addition of 500 mM NaCl and 1 mM 2-mercaptoethanol. After centrifugation (45,000 ϫ g, 1 h), the supernatant was applied at 2 ml min Ϫ1 to a nickel-charged 5-ml HiTrap chelating Sepharose column (Amersham Biosciences) in buffer D (50 mM (Na ϩ ) phosphate, pH 7.5, 200 mM NaCl) and bound proteins were eluted with a linear gradient of 0 -500 mM imidazole in buffer D. Fractions containing eEF1B were pooled and dialyzed overnight against 2 liters of buffer B containing 1 mM dithiothreitol, and then further purified by anion-exchange chromatography using a 25-ml Q-Sepharose HP column in buffer B, as before.
Production of Antiserum Against C. fasciculata eEF1B and Immunoblot Analysis-A polyclonal antiserum was raised in BALB/c mice against the purified C. fasciculata eEF1B complex (34). Cells (5 ϫ 10 6 ) from C. fasciculata and L. major cultures were harvested and resuspended in SDS-PAGE sample buffer. The proteins were then separated by SDS-PAGE and transferred onto Sequi-Blot polyvinylidene difluoride membranes (Bio-Rad). Immunoblotting was then performed, essentially as described (35), using a 1:10,000 dilution of the primary polyclonal antiserum and a 1:5000 dilution of secondary antibody (horseradish peroxidase-conjugated anti-mouse IgG, Sigma). Bound antibodies were detected using the ECL detection kit (Amersham Biosciences), according to the manufacturer's instructions.
Immunolocalization of L. major eEF1B-L. major promastigotes were washed twice in PBS before being dried onto polylysine-coated microscope slides. The cells were then fixed in 4% (w/v) paraformaldehyde in PBS for 10 min, followed by methanol at Ϫ20°C for 2 min. Following a 30-min incubation in PBS containing 1% (w/v) saponin and 1 mg ml Ϫ1 heat-treated RNase, the fixed cells were incubated in PBS containing 5% (v/v) fetal calf serum for 5 min. The cells were then co-labeled with anti-C. fasciculata eEF1B antiserum (1:500 in PBS) and anti-T. brucei BiP antiserum (1:200 in PBS; raised in rabbits against the T. brucei homologue of immunoglobulin heavy chain-binding protein (BiP); a gift from James D. Bangs, University of Wisconsin-Madison Medical School, Madison, WI). Following incubation for 1 h in a dark humid chamber, the slides were washed with PBS and incubated for 1 h in fluorescein isothiocyanate-conjugated goat anti-mouse IgG antiserum (1:500 in PBS, Sigma) and tetramethylrhodamine isothiocyanate-conjugated goat anti-rabbit IgG antiserum (1:500 in PBS, Sigma). The slides were washed with PBS before being mounted using the SlowFade Light Antifade kit with 4Ј,6-diamidino-2-phenylindole (Molecular Probes), according to the manufacturer's instructions. Images were collected using a Zeiss Axiovert 200 M fluorescence microscope.
Enzyme Assay-TST was assayed at 25°C in 100 mM (Na ϩ ) phosphate, pH 6.5, using either a Shimadzu UV-2401 PC or a Beckman DU640 spectrophotometer. In the standard assay T[SH] 2 was produced immediately before use by mixing T[S] 2 (Bachem) with a 2-fold excess of tris(2-carboxyethyl)phosphine (TCEP) and a 5-fold excess of NaOH. One unit of TST activity corresponds to 1 mol of sulfhydryl group conjugated per min. Unless otherwise noted, the activities with these substrates were calculated using the published absorbance coefficients for the corresponding GSH conjugates. In the case of the peroxidase, dehydroascorbate reductase, and thioltransferase assays, TCEP was omitted and the required concentration of T[S] 2 , glutathionylspermidine disulfide (Bachem), or GSSG was reduced by 10 g ml Ϫ1 trypanothione or glutathione reductase and a 150 M excess of NADPH, before addition of the second substrate. Linoleic acid hydroperoxide assays contained 0.05% (v/v) Tween 20 in the final volume. The rate of these reactions was followed by observing the consumption of NADPH at 340 nm. Hydroperoxide concentrations were standardized using a modified TCEP assay for peroxide (36). Briefly, samples of hydroperoxide were reduced by 75 M TCEP in 50 mM (Na ϩ ) Tris, pH 7.5. After a 5-min incubation at 25°C, 5,5Ј-dithiobis(2-nitrobenzoic acid) was added to 200 M and the absorbance at 412 nm was used to calculate residual TCEP, using the extinction coefficient of 14.14 mM cm Ϫ1 for the TNB Ϫ anion (37). When this procedure was used to assay linoleic acid hydroperoxide, 0.1% (v/v) Triton X-100 was included in the buffer.
Purification and Characterization of the Subunits of L. major eEF1B-All three subunits of L. major eEF1B were expressed at a high level as soluble proteins from pET3aTr. This allowed the simple purification of these proteins by anion-exchange and size-exclusion chromatography to more than 95% purity (Fig.  1). The masses of eEF1B␣ and -␤ were significantly different from their apparent mass by SDS-PAGE (Table I). This has previously been noted for the C. fasciculata eEF1B␣ and -␤ proteins and may relate to the low isoelectric points of these proteins (predicted pI values of 4.56 and 5.07 for L. major eEF1B␣ and -␤, respectively), as the anomalous migration in SDS-PAGE of other highly acidic proteins has been noted previously (38). All three subunits self-associated in solution, with the eEF1B␣ and -␤ forming complexes of similar molecular mass. These may represent a eEF1B␣ heptamer (predicted M r 162,000) and a eEF1B␤ hexamer (predicted M r 153,000). The eEF1B␥ subunit appears to form a tetramer (predicted M r 180,000).
Reconstitution of eEF1B␣␤ and eEF1B Complexes-The purified subunits were co-incubated at 1:1 ratios in three pairwise combinations and one 1:1:1 eEF1B␣, -␤, and -␥ combination. Initially, these mixtures were analyzed by size-exclusion chromatography (Fig. 2). Only the ternary mixture formed a novel complex of M r 540,000 (Fig. 2, closed squares), whereas no additional high molecular mass peaks were observed in any of the binary combinations. However, as separation of the individual subunits was not possible because of the similar M r values of their homocomplexes, these data do not exclude the formation of other heterocomplexes. The four protein mixtures were therefore also separated by anion-exchange chromatography and this showed that an ␣␤ heterocomplex was formed (data not shown). The subunit stoichiometries of the purified ␣␤ and ␣␤␥ complexes were determined by SDS-PAGE and densitometry. This indicated that the ␣␤ species contained the subunits in a 1:1 ratio and that the subunits in the ␣␤␥ complex were at a 1:1:2 ratio (Fig. 1). In combination with the M r values of these complexes, this suggested that the species formed were an (␣␤) 4 octamer and an [␣␤(␥ 2 )] 4 hexadecamer ( Table I). The latter agrees with our previous analysis of eEF1B from C. fasciculata (8). Only the [␣␤(␥ 2 )] 4 holocomplex showed TST activity with CDNB or CuOOH substrates (Table I).
Polycistronic Expression of Recombinant L. major eEF1B Complex-As these reconstitution experiments indicated that all three subunits were required to form a complex with Stransferase activity, the three subunits were cloned into the pST39 polycistronic expression vector. The fusion of an Nterminal hexahistidine tag to eEF1B␤ allowed the simple purification of the eEF1B holocomplex by metal-affinity chromatography followed by anion-exchange chromatography (Fig.  3A). The purified complex is more than 95% pure but appears to show some proteolytic degradation of the eEF1B␣ subunit. The M r of the recombinant complex was 570,000 and is identical, within the precision of the method, with the M r of the holocomplex formed by reconstitution from the isolated subunits.
Subcellular Localization of eEF1B-An antiserum was produced that detected all three subunits of the C. fasciculata eEF1B (Fig. 3B, lane 1). Immunoblotting of L. major extracts (Fig. 3B, lane 2) and recombinant L. major eEF1B (data not shown) showed that this antiserum cross-reacted with the eEF1B␤ and -␥ subunits, but not the eEF1B␣ subunit. This antiserum was therefore used to determine the subcellular localization of the L. major eEF1B by immunofluorescence microscopy. Initial studies showed the complex in the anterior of the parasite, particularly around the nucleus and kinetoplast. This pattern is characteristic of proteins located to the endoplasmic reticulum (ER) (20,39). Double labeling of parasites with anti-C. fasciculata eEF1B antiserum and an antiserum raised against the T. brucei homologue of the ER-localized chaperone BiP (39) was therefore performed. The resulting images showed labeling of these proteins that was again con- centrated at the anterior end of the parasite, around the nucleus and kinetoplast (Fig. 4, panels A-D). No labeling of the nucleus, kinetoplast, or flagellum was observed. When the images were superimposed, the labeling of the two proteins was shown to be largely identical (Fig. 4, panels E and F, yellow  staining). These data indicate that the L. major eEF1B is localized on or in the ER, but do not exclude the possibility that the complex may be present in other structures in this region, such as the Golgi apparatus or flagellar pocket.
Using CuOOH as a model substrate, the mechanism of recombinant L. major eEF1B peroxidase reaction was studied in more detail. Substrate-mediated inactivation of the enzyme does not appear to occur, as linear reaction rates were observed at high concentrations of CuOOH, with assays containing 2 mM peroxide giving linear rates until NADPH was completely consumed (data not shown). Initial rates were measured over four fixed concentrations of CuOOH with varying concentrations of T[SH] 2 or its metabolic precursor, N 1 -glutathionylspermidine. Lineweaver-Burk plots of individual fits of the resulting data to the Michaelis-Menten equation were clearly not parallel. Rather, they intersected at or near the x axis, indicating that a ternary complex is formed (Fig. 5). Both random and ordered ternary complex mechanisms were fitted to the data, and a rapid equilibrium random order mechanism was selected for this preliminary kinetic analysis, as it gave the best fit. Further studies on this enzyme will be required to distinguish between the range of possible ternary complex mechanisms.
The kinetic constants obtained with these two thiol substrates were identical, within experimental error (Table III). DISCUSSION The L. major eEF1B complex possesses trypanothione S-transferase activity against a range of xenobiotic compounds and is most active with hydrophobic hydroperoxides. The inability of the isolated L. major eEF1B␥ subunit to catalyze these reactions is consistent with the complete loss of TST activity after separation of the subunits of the C. fasciculata eEF1B (8). These results indicate that the enzymatically active conformation of eEF1B␥ is only achieved in the eEF1B complex. Interactions between the eEF1B subunits can affect their enzymatic activities, with the nucleotide-exchange activity of Artemia salina eEF1B␣ being increased 2-fold by the addition of equimolar amounts of eEF1B␥ (28). However, the absence of TST activity in C. fasciculata and L. major eEF1B␥ contrasts with the GST activity of the recombinant rice eEF1B␥ (40). This eEF1B␥ protein has a GST activity with 1-chloro-2,4-dinitrobenzene that is comparable with that of the native eEF1B␤␤Ј␥ complex. Unfortunately, although the rice eEF1B␥ was shown not to be active with t-butyl hydroperoxide as a substrate (in common with the L. major eEF1B), no assays were performed with other hydrophobic hydroperoxides that are substrates for the L. major eEF1B. Further studies to test the specificity of the rice eEF1B␥ and determine whether the mammalian eEF1B␥ proteins have GST activity might therefore be desirable.  (8), whereas the published value for GS-DNB of 9.6 mM cm Ϫ1 at 340 nm was used for GSH (62).
b The published absorbance coefficient of 9.6 mM cm Ϫ1 at 340 nm for the GSH adduct was used for both thiol substrates (1). c An absorbance coefficient of 5.12 mM cm Ϫ1 at 400 nm for p-nitrophenol at pH 6.5 was calculated from the published pK a and absorbance coefficient of the p-nitrophenolate anion (22). d The published absorbance coefficient of 13.75 mM cm Ϫ1 at 224 nm for the GSH adduct was used for both thiol substrates (21). Interestingly, the assembly process of the L. major eEF1B complex appears to be significantly different from that of the mammalian complex. The ␣ and ␤ subunits of the rabbit eEF1B do not interact in the absence of the eEF1B␥ subunit (41). However, rabbit eEF1B␥ can bind to either of these subunits to form eEF1B␣␥ and eEF1B␤␥ complexes that may be intermediates in the assembly process. This contrasts with the L. major proteins, where the eEF1B␣ and -␤ subunits appear to be capable of binding to eEF1B␥ only after forming an octameric eEF1B␣␤ complex.
The ability of the L. major eEF1B to utilize hydrophobic hydroperoxides is highly significant, as they are not substrates of the previously characterized C. fasciculata TST (8). Moreover, these compounds may not be effectively detoxified by the TryX/tryparedoxin peroxidases pathway, which, in L. major, prefers hydrogen peroxide as substrate and does not use cumene hydroperoxide at all (42). This may be because hydrophobic hydroperoxides can rapidly inactivate tryparedoxin peroxidases by causing the overoxidation of a redox-active cysteine in the active site (43,44). This inactivation reaction is common to most eukaryotic peroxiredoxins and has been proposed to function in oxidative stress signaling (45). The observed insensitivity of the TST to inactivation by hydroperoxides could result from the enzyme catalyzing the direct attack of T[SH] 2 on hydroperoxides, with no formation of a sulfenic acid intermediate on an active-site cysteine. The TST activity of the eEF1B complex may therefore be particularly important as a backup system under conditions of severe oxidative stress. However, other peroxidases may also complement the tryparedoxin peroxidases system in trypanosomatids, with non-selenium GPX homologues recently being characterized in T. cruzi (18) and T. brucei (14). Interestingly, the T. cruzi GPX II was reported to be reduced directly by GSH and not by TryX or T[SH] 2 (20). Moreover, GPX II appeared to be specific for fatty acid and lipid hydroperoxides and to be localized to the ER. However, the activity of this enzyme was 70-fold less with linoleic acid hydroperoxide than that of the L. major eEF1B. The relative roles of GPX II and any T. cruzi eEF1B S-transferase activity in antioxidant metabolism are therefore unclear. Recently, the T. cruzi GPX I (19) and the T. brucei GPX (14) were shown to be most active with TryX as a electron donor, but the activity of these enzymes with fatty acid and lipid peroxides was not reported. The physiological role and substrate specificity of the L. major GPX homologue LmjF36.3010 cannot therefore be predicted. Significantly, the activity of all TryX-dependent peroxidases may be limited under even moderate levels of oxidative stress by the slow rate of reduction of TryX by T[SH] 2 (13,19). As the TST activity of eEF1B does not require TryX, its peroxidase activity will not be limited by this reaction. Moreover, eEF1B may contribute a significant proportion of the L. major cellular lipid peroxidase activity even under nonstress conditions, as its relatively low activity could be compensated for by its abundance, with this complex forming 0.4% of total soluble protein in C. fasciculata (8). The activity of the L. major eEF1B of 0.95 unit mg Ϫ1 with linoleic acid hydroperoxide is, however, comparable with that of the mammalian GSTs, with the specific activities of the rat isoenzymes ranging from 0.06 to 5.3 units mg Ϫ1 with this substrate (46).
A Leishmania TST has been previously proposed to be the uncharacterized T[SH] 2 -dependent activity required for highlevel antimony resistance (47). The characterization of a Leishmania TST will therefore allow studies to directly address any function of this enzyme in the resistance phenotype. Interestingly, overexpression in T. cruzi of their eEF1B␥ gene confers resistance to the trypanothione reductase inhibitor clomipramine (48), a compound that might therefore be expected to induce oxidative stress in the parasite. However, the sensitivity of T. cruzi to other chemotherapeutic compounds that induce oxidative stress was not affected.
The localization of the eEF1B complex to the outer surface of the ER has previously been observed in human fibroblasts (49) and Xenopus laevis oocytes (50). This localization may be the result of an interaction between the ER-resident integral membrane protein kinectin and eEF1B␤ (51). Interestingly, the association of eEF1B to the ER membrane has also been proposed to involve a hydrophobic domain of eEF1B␥ (28). This affinity for membranes would be consistent with a role for eEF1B␥ in the metabolism of the lipid hydroperoxides and cytotoxic aldehydes produced during lipid peroxidation. The targeting of peroxidases to the ER might also relate to a reduced ability of antioxidant systems in the ER lumen to protect these membranes from oxidative damage, resulting from the oxidizing environment that is maintained within this cellular compartment (52,53). Moreover, the ER membranes may be exposed to oxidative species produced in the ER lumen during the oxidation of cysteine residues during the folding of secreted proteins. This is catalyzed by the flavoprotein Ero1p (54) and has been proposed to be a major source of reactive oxygen species (55).
The eEF1B complex may also mediate translational control in response to cellular oxidative stress. Protein synthesis is potently and reversibly inhibited by glutathione disulfide (56) and an in vivo study has indicated that translation elongation is the step in this process that is most sensitive to oxidative stress (57). Interestingly, the eEF1B␤ subunit is glutathionylated in response to oxidative stress (58) and the eEF1B␥ subunit contains a redox-active pair of cysteine residues (59). These post-translational modifications might affect protein synthesis either by direct alterations in the nucleotide-exchange activity of eEF1B, or through the allosteric effects that interactions with this complex have on tRNA synthetases (60,61).
The identification of a TST activity in the L. major eEF1B complex has implications for translational control and the mechanisms of resistance to chemical and oxidative stress in these pathogens. Future studies on the activities of host and parasite eEF1B␥ proteins may therefore reveal further important roles of these elongation factors in cellular physiology.

TABLE III
Kinetic parameters of recombinant L. major eEF1B trypanothione-dependent peroxidase activity Peroxidase activity with cumene hydroperoxide was assayed as described, and the resulting data fitted to a rapid-equilibrium random order mechanism.