A comparative study of methylglyoxal metabolism in trypanosomatids

The glyoxalase system, comprising the metalloenzymes glyoxalase I (GLO1) and glyoxalase II (GLO2), is an almost universal metabolic pathway involved in the detoxification of the glycolytic byproduct methylglyoxal to d-lactate. In contrast to the situation with the trypanosomatid parasites Leishmania major and Trypanosoma cruzi, this trypanothione-dependent pathway is less well understood in the African trypanosome, Trypanosoma brucei. Although this organism possesses a functional GLO2, no apparent GLO1 gene could be identified in the T. brucei genome. The absence of GLO1 in T. brucei was confirmed by the lack of GLO1 activity in whole cell extracts, failure to detect a GLO1-like protein on immunoblots of cell lysates, and lack of d-lactate formation from methylglyoxal as compared to L. major and T. cruzi. T. brucei procyclics were found to be 2.4-fold and 5.7-fold more sensitive to methylglyoxal toxicity than T. cruzi and L. major, respectively. T. brucei also proved to be the least adept of the ‘Tritryp’ parasites in metabolizing methylglyoxal, producing l-lactate rather than d-lactate. Restoration of a functional glyoxalase system by expression of T. cruzi GLO1 in T. brucei resulted in increased resistance to methylglyoxal and increased conversion of methylglyoxal to d-lactate, demonstrating that GLO2 is functional in vivo. Procyclic forms of T. brucei possess NADPH-dependent methylglyoxal reductase and NAD+-dependent l-lactaldehyde dehydrogenase activities sufficient to account for all of the methylglyoxal metabolized by these cells. We propose that the predominant mechanism for methylglyoxal detoxification in the African trypanosome is via the methylglyoxal reductase pathway to l-lactate.

The glyoxalase system, comprising the metalloenzymes glyoxalase I (GLO1) and glyoxalase II (GLO2), is an almost universal metabolic pathway involved in the detoxification of the glycolytic byproduct methylglyoxal to d-lactate. In contrast to the situation with the trypanosomatid parasites Leishmania major and Trypanosoma cruzi, this trypanothionedependent pathway is less well understood in the African trypanosome, Trypanosoma brucei. Although this organism possesses a functional GLO2, no apparent GLO1 gene could be identified in the T. brucei genome. The absence of GLO1 in T. brucei was confirmed by the lack of GLO1 activity in whole cell extracts, failure to detect a GLO1-like protein on immunoblots of cell lysates, and lack of d-lactate formation from methylglyoxal as compared to L. major and T. cruzi. T. brucei procyclics were found to be 2.4-fold and 5.7-fold more sensitive to methylglyoxal toxicity than T. cruzi and L. major, respectively. T. brucei also proved to be the least adept of the 'Tritryp' parasites in metabolizing methylglyoxal, producing l-lactate rather than d-lactate. Restoration of a functional glyoxalase system by expression of T. cruzi GLO1 in T. brucei resulted in increased resistance to methylglyoxal and increased conversion of methylglyoxal to d-lactate, demonstrating that GLO2 is functional in vivo. Procyclic forms of T. brucei possess NADPH-dependent methylglyoxal reductase and NAD + -dependent l-lactaldehyde dehydrogenase activities sufficient to account for all of the methylglyoxal metabolized by these cells. We propose that the predominant mechanism for methylglyoxal detoxification in the African trypanosome is via the methylglyoxal reductase pathway to l-lactate. mammalian hosts), which utilize glutathione (c-l-glutamyl-l-cysteinylglycine) [7]. This dithiol is primarily responsible for the maintenance of thiol-redox homeostasis within trypanosomatids, and is crucially involved in the protection of parasites from oxidative stress [8], heavy metals [9] and xenobiotics [10]. Several enzymes involved in trypanothione biosynthesis and its downstream metabolism have been genetically and chemically validated as essential for parasite survival [11]. Consequently, trypanothione-dependent enzymes have become the focus of much anti-trypanosomatid drug discovery.
The glyoxalase system, comprising the metalloenzymes glyoxalase I (GLO1, EC 4.4.1.5) and glyoxalase II (GLO2, EC 3.1.2.6), together with glutathione as cofactor, is a widely distributed pathway involved in metabolism of the toxic and mutagenic glycolytic byproduct methylglyoxal [12,13]. A unique trypanothione-dependent glyoxalase system has been identified in Leishmania spp. and T. cruzi [14][15][16]. In the first step, GLO1 isomerizes the spontaneous hemithioacetal adduct formed between trypanothione and methylglyoxal to S-d-lactoyltrypanothione [14]. In the second step, GLO2 catalyses hydrolysis of this ester, releasing d-lactate and regenerating trypanothione. The trypanothione-dependent glyoxalase system in these parasites differs significantly from that employed by their mammalian hosts, which depends entirely on glutathione as a thiol cofactor. These differences in substrate specificity may provide an opportunity for the specific chemotherapeutic targeting of these enzymes in the trypanosomatids. As inhibitors of the glyoxalase system have already been shown to possess both anticancer [17] and antimalarial [18] activities, it is possible that inhibition of the trypanothione-dependent glyoxalase pathway may prove toxic to trypanosomatids.
Although glyoxalase metabolism has been well defined in both Leishmania major and T. cruzi, this pathway is less well understood in T. brucei. Intriguingly, the recently completed T. brucei genome revealed that although this organism possesses a functional GLO2 [19], no apparent GLO1 gene or homologue could be identified [20]. This was unexpected, as the bloodstream form of T. brucei has an extremely high glycolytic flux and relies solely on substrate-level phosphorylation for ATP production [21]. Triose phosphates are a major source of methylglyoxal [12,13], and thus the reported antiproliferative effects of exogenous dihydroxyacetone [22] or endogenous modulation of triose phosphate isomerase in T. brucei [23] could be due to methylglyoxal toxicity. Should the absence of GLO1 from this pathogen be confirmed, it may have important implications for the viability of the glyoxalase system as a target for antitrypanosomatid chemotherapy. In this study, we attempted to further characterize the unusual methylglyoxal metabolism of T. brucei and directly compare it to that of T. cruzi and L. major.

Results and Discussion
Analysis of methylglyoxal-catabolizing enzymes in trypanosomatid cell extracts Sequencing of the 'Tritryp' genomes has revealed several interesting distinctions between the cellular metabolism of T. brucei, T. cruzi and L. major [20]. In our current study, we sought to examine the apparent absence of a gene encoding a GLO1 homologue from the T. brucei genome, GLO1 being a ubiquitous enzyme required for the metabolism of methylglyoxal. Initially, the relative activities of enzymes involved in methylglyoxal metabolism were compared in these medically significant trypanosomatids. Whole cell extracts of T. cruzi epimastigotes, L. major promastigotes and T. brucei (bloodstream and procyclic forms) were prepared, and the activities of methylglyoxalcatabolizing enzymes were determined (Table 1). In keeping with previously published data [14,15], trypanothione-dependent GLO1 activity was detected in both L. major and T. cruzi extracts with specific activities of 85 and 42 nmolAEmin )1 AEmg )1 , respectively. However, GLO1 activity could not be detected in extracts of T. brucei procyclic or bloodstream forms, with either trypanothione or glutathione hemithioacetals as substrate. In contrast, trypanothione-dependent GLO2 activity was detected in all cell lysates. With S-d-lactoyltrypanothione as a substrate, L. major extracts demonstrated GLO2 activity of 62.8 nmolAEmin )1 AEmg )1 , over sixfold higher than that of T. cruzi extracts (8.8 nmolAEmin )1 AEmg )1 ). Despite the apparent lack of GLO1 activity, both T. brucei bloodstream form and procyclic extracts effectively metabolized S-d-lactoyltrypanothione, with specific activities of 18 and 23 nmolAEmin )1 AEmg )1 , respectively. Trypanothione reductase activities were also assayed in each lysate to ensure adequate extraction of the parasites, and were in line with previously published data [24].

Western blot analyses of cell extracts
To confirm the absence of GLO1 from T. brucei at the protein level, immunoblots of trypanosomatid whole cell lysates were probed with L. major GLO1-specific polyclonal antiserum (Fig. 1). As expected, a protein of 16 kDa, which is equivalent to the predicted molec-ular mass of GLO1, reacted strongly with the antiserum in both the L. major and the T. cruzi lysates. No GLO1-like protein was detected in whole cell lysates of T. brucei procyclics, despite overexposure of the blot. In combination with our enzymatic analysis of cell extracts, these data confirm the absence of a functional GLO1 enzyme within T. brucei. This situation is not entirely without precedence. Cestode and digenean parasitic helminths have been studied that lack GLO1 while maintaining high levels of GLO2 activity [25]. One explanation for the retention of this enzyme is that T. brucei GLO2 has methylglyoxal-independent functions. Indeed, human GLO2 has demonstrated substrate promiscuity in efficiently hydrolysing thiol esters of simple acids such as formic acid, succinic acid and mandelic acid [13]. The identification of the true physiological substrate of T. brucei GLO2 will form the basis of our future studies.

Effects of methylglyoxal on trypanosomatid growth
The absence of GLO1 from T. brucei suggested that these parasites may be particularly susceptible to the toxic effects of methylglyoxal. With this in mind, T. cruzi, L. major and T. brucei were grown in the presence of increasing methylglyoxal concentrations, and the rela-tive growth of each culture was determined after 72 h (Fig. 2). To allow the direct comparison of the methylglyoxal sensitivity of these parasites, each cell line was adapted for growth in SDM-79 medium prior to analysis. T. brucei procyclics were the most sensitive to methylglyoxal toxicity, with an EC 50 of 70 ± 2 lm, whereas T. cruzi epimastigotes and L. major promastigotes were 2.4-fold and 5.7-fold less sensitive, with EC 50 values of 171 ± 11 and 397 ± 27 lm, respectively.    Bloodstream trypanosomes could not be adapted for growth in SDM-79 medium, and attempts to determine the methylglyoxal sensitivity of these cells in HMI-9 medium proved unsuccessful, due to the propensity of methylglyoxal to react with thiols in this culture medium. In a previous study on the curative effect of methylglyoxal in cancer-bearing mice [26], Ghosh et al. established the pharmacokinetic properties of methylglyoxal in blood following oral dosing. Using this methodology, we examined the effects of methylglyoxal on an in vivo T. brucei infection. The maximum achievable methylglyoxal concentration in blood following oral dosing of mice was 20 lm, and at this level there was no discernible effect on the progression of the parasite infection (data not shown). These results suggest that the methylglyoxal EC 50 for bloodstream T. brucei in vivo is in excess of 20 lm.

Trypanosomatid metabolism of methylglyoxal
The rate of exogenous methylglyoxal metabolism by T. cruzi, L. major and T. brucei (bloodstream and procyclic forms) was determined (Fig. 3). Each cell line was resuspended in a minimal medium that had been preincubated with 1.5 mm methylglyoxal for 90 min. At defined intervals, culture supernatants were removed and analysed for residual methylglyoxal. In keeping with both our enzymatic analysis of whole cell lysates and EC 50 data, L. major promastigotes dealt with exogenous methylglyoxal most efficiently, with an initial rate of 67 nmolAEmin )1 AEmL )1 . In comparison, T. cruzi epimastigotes were considerably less effective at metabolizing methylglyoxal (47.6 nmolAEmin )1 AEmL )1 ). However, T. brucei procyclics and bloodstream forms proved to be the least adept at dealing with this toxic oxoaldehyde, metabolizing methylglyoxal with initial rates of 7.4 nmolAE min )1 AEmL )1 and 9.8 nmolAEmin )1 AEmL )1 , respectively. These results suggest that although T. brucei is predicted to be the most vulnerable of the 'Tritryp' trypanosomatids to methylglyoxal toxicity, it can effectively metabolize methylglyoxal despite the absence of a complete glyoxalase pathway.

Products of trypanosomatid metabolism of methylglyoxal
In all studies to date, the principal product of thioldependent metabolism of methylglyoxal has been d-lactate [27][28][29]. Consequently, methylglyoxal-treated parasites were monitored for the production of lactate, using d-lactate and l-lactate dehydrogenase-based assays ( Table 2). As expected, both L. major and T. cruzi cells produced considerable amounts of d-lactate following exposure to methylglyoxal, accounting for approximately 30% of free methylglyoxal in the medium. In contrast, T. brucei (procyclics and bloodstream forms) produced only trace amounts of d-lactate. Instead, methylglyoxal-treated T. brucei procyclics and bloodstream forms produced significant quantities of the stereoisomer l-lactate (120 and 221 lm in 2 h, respectively). The sixfold higher rate of l-lactate production by bloodstream parasites in the absence of exogenous methylglyoxal reflects the extremely high glycolytic rate in this developmental form of the African trypanosome [30]. The addition of methylglyoxal marginally decreased the amount of l-lactate detected in the supernatants of both L. major and T. cruzi cultures. These data suggest that T. brucei may metabolize methylglyoxal by an alternative pathway.
In a previous study [31], Ghoshal et al. identified NADPH-dependent methylglyoxal reductase activity in Leishmania donovani promastigotes. These parasites were shown to metabolize approximately 1.2% of the exogenous methylglyoxal added to cultures via this reductase, generating l-lactaldehyde as an end-product. In view of the generation of considerably higher levels of l-lactate by methylglyoxal-treated T. brucei, we hypothesized that methylglyoxal reductase activity may be elevated in T. brucei to compensate for the absence of GLO1. Indeed, when NADPH-dependent methylglyoxal reductase activity was measured in all three trypanosomatid cell lysates, a twofold higher reductase activity was observed in T. brucei procyclic and bloodstream extracts, respectively, than that seen in L. major and T. cruzi cells (Table 1). If we consider that procyclics metabolize exogenous methylglyoxal at a rate of 7.4 nmolAEmin )1 per 10 8 cells (Fig. 3), and assuming that 10 8 cells is equivalent to 1 mg of protein [32], this elevated methylglyoxal reductase activity could conceivably account for all methylglyoxal metabolism in T. brucei. Although a T. brucei methylglyoxal reductase has yet to be identified, two putative aldoketo reductase genes (Tb927.2.5180 and Tb11.02.3040), whose protein products are members of the same aldoketo reductase superfamily as methylglyoxal reductase, have been annotated in the genome. To date, attempts to express these genes as soluble recombinant proteins have proved unsuccessful. In mammalian cells, methylglyoxal can also be detoxified by two methylglyoxal dehydrogenase enzymes (oxoaldehyde dehydrogenase and betaine aldehyde dehydrogenase) [33]. No homologues of these enzymes were identified in the T. brucei genome, and neither NAD + -dependent nor NADP +dependent methylglyoxal dehydrogenase activities were detected in T. brucei extracts (data not shown).
To complete the metabolism of l-lactaldehyde to lactate, T. brucei would require a functional l-lactaldehyde dehydrogenase. Although lactaldehyde dehydrogenase activity has previously been detected in L. donovani cell lysates [31], it has yet to be identified in either T. cruzi or in T. brucei. Using l-lactaldehyde as a substrate, l-lactaldehyde dehydrogenase activity was measured in all three insect-stage trypanosomatid cell lysates (Table 1), and was found to be relatively similar in L. major and T. cruzi lysates (0.51 ± 0.004 and 0.48 ± 0.02 nmolAEmin )1 AEmg )1 , respectively). In comparison, l-lactaldehyde dehydrogenase activity was found to be elevated approximately 2.4-fold in T. brucei procyclic cell lysates (1.24 ± 0.11 nmolAE min )1 AEmg )1 ). However, activity could not be detected in the bloodstream stage of the parasite. These studies confirm that procyclic T. brucei organisms are capable of metabolizing methylglyoxal, via a methylglyoxal reductase-dependent pathway, to l-lactate; however, it remains to be seen whether this is the predominant pathway for methylglyoxal detoxification in these cells. Our failure to detect NAD + -dependent l-lactaldehyde dehydrogenase activity in T. brucei bloodstream forms may be due to technical reasons, such as NADH oxidation via the glycerophosphate oxidase system masking the formation of NADH.

Expression of T. cruzi GLO1 (TcGLO1) in T. brucei
Can T. brucei utilize a complete glyoxalase system? To address this question, a tetracycline-inducible pLew100-TcGLO1 construct was generated and transfected into both bloodstream and procyclic cells. Western blot analysis of transgenic parasites, following induction with tetracycline, confirmed the expression of a 16-kDa protein that reacted strongly with GLO1specific antiserum ( Fig. 4; bloodstream data not shown). This protein was not evident in cells transfected with an unrelated vector (pLew100-luciferase). Antiserum against T. brucei pteridine reductase 1 was used to establish equal loading of samples. The expression of recombinant TcGLO1 in procyclics and bloodstream forms was confirmed when GLO1 activity (23.0 ± 1.9 and 38.2 ± 1.9 nmolAEmin )1 AEmg )1 , respectively) was detected in cell extracts. Indeed, the rate of exogenous methylglyoxal metabolism in these transgenic T. brucei cell lines increased markedly, with GLO1-expressing procyclic and bloodstream cells metabolizing the toxic oxoaldehyde 1.7-fold and 2.7-fold more effectively, respectively (Table 3). Most importantly, TcGLO1-expressing T. brucei procyclics were almost 3.5-fold less sensitive to methylglyoxal than wild-type or luciferase-expressing cells.
To confirm that enhanced methylglyoxal tolerance in GLO1-expressing T. brucei was due to complementation of the glyoxalase system, lactate production in the supernatants of methylglyoxal-treated wild-type and transgenic cells was measured (Table 4). Whereas l-lactate levels in the supernatants of GLO1-expressing T. brucei (bloodstream forms and procyclics) were very similar to those of the wild-type, d-lactate production was found to be significantly higher ( 3-fold, P < 0.0001). d-Lactate levels failed to reach those seen in the supernatants of methylglyoxal-treated L. major and T. cruzi, but were sufficient to suggest that GLO1 expression in T. brucei procyclic and bloodstream parasites results in a complete glyoxalase system.

Implications for parasite chemotherapy
Mammalian cells maintain a repertoire of four pathways for metabolism of methylglyoxal [33], whereas our studies suggest that the African trypanosome may be solely dependent upon methylglyoxal reductase (Fig. 5). The absence of a functioning glyoxalase system within T. brucei, recognized as the principal route of oxoaldehyde detoxification in almost all cells, is especially perplexing. As methylglyoxal is generated primarily as a byproduct of glycolysis, and African trypanosomes are entirely dependent upon glycolysis for energy, it would be reasonable to assume that T. brucei would preserve robust methylglyoxal-metabolizing systems. Without an TcGLO1   Fig. 4. TcGLO1 expression in T. brucei procyclics. Immunoblots of cell extracts of T. brucei procyclics transfected with either pLew100-luciferase or pLew100-TcGLO1 were probed with antiserum to L. major GLO1 and T. brucei pteridine reductase 1 (PTR1) (1 · 10 7 parasites in each lane). Cells were induced with tetracycline for 24 h prior to analysis. intact glyoxalase pathway, these cells should be particularly vulnerable to methylglyoxal toxicity, and our current studies appear to confirm this. These findings have broad implications for the targeting of methylglyoxal metabolism for antitrypanosomatid chemotherapy. Previous studies have suggested that the contrasting substrate specificities of the human and trypanosomatid glyoxalase enzymes (GLO1 and GLO2) make them attractive targets for rational drug design [14,15,19]. Whereas this may still be the case in T. cruzi and Leishmania spp., methylglyoxal reductase is clearly a more promising drug target in the African trypanosome. Identification of the genes that encode this enzyme in T. brucei should now be a priority.
In order to directly compare the effects of methylglyoxal on the growth of these trypanosomatids, triplicate cultures containing methylglyoxal were seeded at 5 · 10 5 parasites per mL. As methylglyoxal interferes with the Alamar blue assay for viable cells, cell densities were determined using the CASY Model TT cell counter (Scha¨rfe, Renlingen, Germany) after culture for 72 h. Concentrations of inhibitor causing a 50% reduction in growth (EC 50 ) were determined using the following two-parameter equation by nonlinear regression using grafit: where the experimental data were corrected for background cell density and expressed as percentages of the uninhibited control cell density. In this equation, [I] represents inhibitor concentration, and m is the slope factor.
(800 g, 10 min, 4°C), washed once in PSG buffer [NaCl ⁄ P i , pH 8.0, 1.5% (w ⁄ v) glucose and 0.5 mgAEmL )1 BSA], resuspended in ice-cold de-ionized dH 2 O, (500 lL) and vortexed. Lysed bloodstream trypanosomes were then incubated on ice for 10 min prior to the addition of 2· lysis buffer (500 lL) and further vortexing. From this point, all lysates were treated in an identical manner. Following centrifugation (800 g, 20 min, 4°C), cell supernatants were collected and dialysed against 50 mm Hepes (pH 7.0) with 25 mm NaCl and 150 lm 2-mercaptoethanol at 4°C to remove components of less than 3.5 kDa. The protein concentration of each lysate was determined using Bradford reagent (Bio-Rad, Hemel Hempstead, UK). GLO1 activity in the trypanosomatid cell lysates was measured by monitoring the formation of S-d-lactoyltrypanothione spectrophotometrically at 240 nm [14]. Trypanothione and methylglyoxal were preincubated at 25°C for 10 min in 50 mm Hepes (pH 7.0) plus 25 mm NaCl, 50 lm adduct, and 100 lm free thiol. Reactions were initiated with enzyme extract. Methylglyoxal reductase and GLO2 activities were determined as previously described [14,31,36]. The activity of trypanothione reductase, used as a control enzyme, was assayed as previously described [37].
L-Lactaldehyde dehydrogenase activity in trypanosomatids l-Lactaldehyde, the substrate of l-lactaldehyde dehydrogenase, was prepared from d-threonine, as previously described [38]. Briefly, 25 mmol of d-threonine, 9.1 g of ninhydrin and 600 mL of 0.05 m sodium citrate buffer (pH 5.4) were combined and boiled for 15 min with continual stirring. After being cooled to room temperature, the mixture was filtered and treated with sufficient Dowex 1-X8 resin (bicarbonate form) to raise the pH to 6.5. After stirring for a further 2-3 h, the resin was again filtered, and the filtrate was adjusted to pH 4.0 by the addition of Dowex 50 resin (hydrogen ion form). Following filtration, the filtrate was concentrated down to 50-100 mL using a rotary evaporator. The resulting concentrate was then sequentially treated with Dowex 1-X8 and Dowex 50 resins, as previously described, and further concentrated to 20-30 mL. Dowex 1-X8 resin was then added to the concentrated filtrate in batches until the solution was colourless, and the pH was adjusted to 7.5. The l-lactaldehyde yield from this reaction was determined by monitoring NADH production at 340 nm following incubation with aldehyde dehydrogenase from baker's yeast (Fluka, Gillingham, UK). Reactions were performed in 100 mm Tris ⁄ HCl (pH 8.5), 3 mm NAD + and 10 units of aldehyde dehydrogenase. The purity of the synthetic l-lactaldehyde was analysed by liquid chromatography-MS. Samples were derivatized with excess 2,4-dinitrophenylhydrazine (Fluka) in 5 mm HCl, diluted with acetonitrile ⁄ water (1 : 1), and analysed by liquid chromatography-MS (Phenomenex Gemini C18 column, 50 · 3.0 mm, 5 lm particle size; mobile phase, water ⁄ acetonitrile + 0.1% HCOOH 80 : 20 to 5 : 95 over 3.5 min, and then held for 1.5 min; flow rate 0.5 mLAEmin )1 ). LC-MS analysis detected the expected lactaldehyde hydrazone plus an additional hydrazone (the contaminant was not present in the underivatized l-lactaldehyde preparation, or the 2,4-dinitrophenylhydrazine reagent). The mass and retention time of the contaminating hydrazone was consistent with the impurity in the l-lactaldehyde preparation being acetone or propionaldehyde (as shown by comparison with the hydrazones of acetone and propionaldehyde synthesized as described above). Biochemical assays on L. major cell-free extracts indicated that neither acetone nor propionaldehyde was responsible for the observed activity.
l-Lactaldehyde dehydrogenase activity was measured in soluble trypanosomatid extracts, prepared as above, except that a further centrifugation (100 000 g, 1 h, 4°C) and dialysis step was introduced prior to analysis. Activity was measured at 27°C in 100 mm Tris ⁄ HCl (pH 8.5) with 3 mm NAD + and 1 mm l-lactaldehyde [39]. Trypanosomatid extracts were incubated with NAD + at 27°C for 5 min, prior to the initiation of the reaction with l-lactaldehyde.
Reactions were monitored at 340 nm for the formation of NADH.

Western blot analyses of trypanosomatid cell extracts
Polyclonal antisera against L. major GLO1 were raised in adult male Wistar rats. An initial injection of 100 lg of purified antigen, emulsified in complete Freund's adjuvant, was followed by two identical booster injections of antigen emulsified in Freund's incomplete adjuvant at 2 week intervals.
Trypanosomatid whole cell extracts (30 lg) were separated by SDS ⁄ PAGE and subsequently transferred onto nitrocellulose. After blocking with 7% skimmed milk in NaCl ⁄ P i for 1 h, blots were incubated with L. major GLO1 polyclonal antiserum (1 : 700 dilution) for 1 h, washed in NaCl ⁄ P i containing 0.1% (v ⁄ v) Tween-20, and then incubated with a secondary antibody [rabbit anti-(rat IgG)] (Dako, Ely, UK; 1 : 10 000 dilution). Immunoblots were developed using the ECL plus (enhanced chemiluminescence) system from Amersham Biosciences (Piscataway, NJ, USA). maintenance medium that had been preincubated with 1.5 mm methylglyoxal for 90 min prior to resuspension. In the case of T. brucei bloodstream forms, cells were pelleted by centrifugation (800 g, 10 min, 4°C), and washed in PSG buffer with 0.5 mgAEmL )1 BSA. Bloodstream trypanosomes were then resuspended at 1 · 10 8 mL )1 in PSG buffer with 0.5 mgAEmL )1 BSA that had been preincubated with 1.5 mm methylglyoxal for 90 min prior to resuspension. In all cases, cell viability was monitored by visibly checking motility throughout the experiment. Metabolism of methylglyoxal by these cells was determined by measuring the methylglyoxal concentration in cell-free assay buffer. At defined intervals, aliquots were removed, cells were pelleted at 16 000 g for 5 min, and the supernatants were analysed for residual methylglyoxal by the semicarbizide assay [14].
The production of lactate by methylglyoxal-treated mid-log L. major promastigotes, T. cruzi epimastigotes and both T. brucei procyclic and bloodstream trypanosomes (2 · 10 8 cells) was determined. Cells were incubated with 1.5 mm methylglyoxal in an identical manner to that previously described for the methylglyoxal metabolism studies. Following a 2 h incubation, cells were pelleted (16 000 g, 5 min), and supernatants were assayed without further treatment by the addition of either d-lactaldehyde dehydrogenase or l-lactaldehyde dehydrogenase, as per the manufacturer's instructions. The amount of NADH formed was measured at 340 nm, and the limit of detection for these assays was determined to be 1 lm.

Cloning and expression of recombinant TcGLO1 in T. brucei
The T. cruzi GLO1 gene (Tc00.1047053510659.240) was amplified by PCR from genomic DNA using the sense primer 5¢-AAGCTTATGTCAACACGACGACTTATGCAC A-3¢ and the antisense primer 5¢-GGATCCGGATCCTT AAGCCGTTCCCTGTTC-3¢ with additional HindIII and BamHI restriction sites (italicised), respectively. The PCR product was then cloned into pCR-Blunt II-TOPO (Invitrogen) and sequenced. The pCR-Blunt II-TOPO-TcGLO1 construct was then digested with HindIII and BamHI, and the fragment was ligated into the tetracycline-inducible expression vector pLew100 [40], resulting in a pLew100-TcGLO1 construct.