Kinetics, subcellular localization, and contribution to parasite virulence of a Trypanosoma cruzi hybrid type A heme peroxidase (TcAPx-CcP)

Significance

Trypanosoma cruzi, the causative agent of Chagas disease, affects 8–10 million people in Latin America. Parasite antioxidant systems are essential for parasite survival and infectivity in the vertebrate host. Herein, we characterized the enzymic properties, subcellular localization, and contribution to parasite virulence of a T. cruzi hybrid type A member of class I heme peroxidases. The enzyme reacts fast with hydrogen peroxide and utilizes both ferrocytochrome c and ascorbate as reducing substrates [T. cruzi ascorbate peroxidase (TcAPx)-cytochrome c peroxidase (CcP)]. A unique subcellular distribution of TcAPx-CcP in the infective stages suggests a role during parasite–host interactions. Infection of macrophages and cardiomyocytes, as well as in mice, confirmed the involvement of TcAPx-CcP in pathogen virulence as part of the parasite antioxidant armamentarium.

Abstract

The Trypanosoma cruzi ascorbate peroxidase is, by sequence analysis, a hybrid type A member of class I heme peroxidases [TcAPx-cytochrome c peroxidase (CcP)], suggesting both ascorbate (Asc) and cytochrome c (Cc) peroxidase activity. Here, we show that the enzyme reacts fast with H₂O₂ (k = 2.9 × 10⁷ M⁻¹⋅s⁻¹) and catalytically decomposes H₂O₂ using Cc as the reducing substrate with higher efficiency than Asc (k_cat/K_m = 2.1 × 10⁵ versus 3.5 × 10⁴ M⁻¹⋅s⁻¹, respectively). Visible-absorption spectra of purified recombinant TcAPx-CcP after H₂O₂ reaction denote the formation of a compound I-like product, characteristic of the generation of a tryptophanyl radical-cation (Trp^233•+). Mutation of Trp²³³ to phenylalanine (W233F) completely abolishes the Cc-dependent peroxidase activity. In addition to Trp^233•+, a Cys²²²-derived radical was identified by electron paramagnetic resonance spin trapping, immunospin trapping, and MS analysis after equimolar H₂O₂ addition, supporting an alternative electron transfer (ET) pathway from the heme. Molecular dynamics studies revealed that ET between Trp²³³ and Cys²²² is possible and likely to participate in the catalytic cycle. Recognizing the ability of TcAPx-CcP to use alternative reducing substrates, we searched for its subcellular localization in the infective parasite stages (intracellular amastigotes and extracellular trypomastigotes). TcAPx-CcP was found closely associated with mitochondrial membranes and, most interestingly, with the outer leaflet of the plasma membrane, suggesting a role at the host–parasite interface. TcAPx-CcP overexpressers were significantly more infective to macrophages and cardiomyocytes, as well as in the mouse model of Chagas disease, supporting the involvement of TcAPx-CcP in pathogen virulence as part of the parasite antioxidant armamentarium.

The protozoan parasite Trypanosoma cruzi is the causative agent of Chagas disease (CD; also known as American trypanosomiasis). Up to 10 million people across Latin America are infected with this protozoan parasite, a distribution range that is expanding driven by migration of infected insects and hosts, with CD now emerging as a public health problem at nonendemic sites (1, 2). T. cruzi strains are heterogeneous, exhibiting a high degree of biochemical and genetic variability. Such differences are believed, at least in part, to be responsible for disease outcome, which ranges from being asymptomatic during the course of infection to fatal severe cardiac and digestive complications (3). It has been shown that the parasite antioxidant systems are essential for parasite survival and establishment of the infection in the vertebrate host (4–7). In contrast to most eukaryotes, T. cruzi lacks catalase and selenium-dependent glutathione peroxidases, which are enzymes capable of rapidly metabolizing high levels of H₂O₂ (8, 9). Instead, it expresses an array of complex enzyme-mediated mechanisms in which the trypanosomatid-specific thiol trypanothione ([T(SH)₂], N¹,N⁸-bisglutathionylspermidine) plays a central role in the funneling of reducing equivalents to the different peroxidase antioxidant systems (10). Two typical 2-Cys peroxiredoxins are located in the cytosol (CPX) and in the mitochondrial matrix (MPX), respectively, and efficiently scavenge H₂O₂, peroxynitrite, and small-chain organic hydroperoxides (11–13). Two glutathione-dependent peroxidases are located at the endoplasmic reticulum (ER) and in the cytosol, and seem to be important in the metabolization of lipid-derived hydroperoxides (14). Finally, a plant-like related heme peroxidase located at the ER displays ascorbate (Asc)-dependent peroxidase activity (APx) (15).

Because the T. cruzi antioxidant defense systems are distinct from its mammalian host, the trypanosomal activities are suitable targets for specific rationale pharmacological inhibition. During its life cycle, T. cruzi undergoes various morphological and biochemical changes. One of the most complex transformations occurs during metacyclogenesis, the process where a noninfective and replicative insect-derived epimastigote cell transforms to a highly infectious, nonreplicative, metacyclic trypomastigote. During this process, it has been shown that T. cruzi up-regulates several antioxidant enzymes to preadapt to the hostile environment of the vertebrate host (5, 16, 17). Enzymes known to be up-regulated during this differentiation process include peroxiredoxins, [T(SH)₂] synthase (5), Fe-containing superoxide dismutases, and APx (16).

T. cruzi ascorbate peroxidase (TcAPx) was first reported to reduce H₂O₂ catalytically in the presence of Asc as the reducing substrate (APx activity), but with a catalytic efficiency several fold lower than plant APx (18, 19). Its overexpression confers parasite resistance toward exogenously added H₂O₂ toxicity (13, 15, 20). Recent studies have shown that T. cruzi lacking TcAPx is still capable of infecting cultured mammalian cells, albeit at a reduced level, and can still establish an infection in the mouse model of CD (20). Although the enzyme does not seem to be essential for parasite infectivity, its enhanced expression may represent an additional skill for the parasite to deal with host-derived oxidant toxicity both in the acute and chronic stages of the disease (6, 21).

TcAPx shares significant homology (62% identity and 86% similarity) with its Leishmania major peroxidase (LmP) counterpart (18). Both parasite enzymes are related to the class I heme-peroxidase group of antioxidant enzymes that includes catalase-peroxidase (katG), APx, cytochrome c peroxidase (CcP), and the hybrid type (A and B) peroxidases (22). Phylogenetic studies have classified the T. cruzi and Leishmania enzymes as members of the hybrid type A, subfamily A1, heme peroxidases (23), with other sequences in this group displaying both Asc- and cytochrome c (Cc)-dependent peroxidase activities (22, 24). Based on this fact, we have renamed TcAPx as TcAPx-CcP. These hybrid type A peroxidases represent a real turning point in the evolution of ancient bifunctional catalase-peroxidase toward monofunctional specialization within this family, and their biochemical characterization may help to unravel the evolution of structure–function relationships of these abundant oxidoreductases (23).

LmP can use both Asc (as monofunctional APx) and Cc (as monofunctional CcP) as electron donors (25). H₂O₂ reacts with the heme generating the oxoferryl (Fe^{IV = O}) and the porphyrin-centered radical (P^•+) forming the classical compound I species. The presence of tryptophan located near the heme (Trp²⁰⁸ in LmP) rapidly reduces the P^•+, generating an amino acid-derived cation radical (Trp^208•+, compound I-like). Finally, using either Asc or Cc as the reducing substrate, the enzyme recovers the resting state (Fe-III-P). Classical compound I formation in these hybrid type A peroxidases was established by direct electron paramagnetic resonance (EPR) analysis (25) and/or by spectroscopic studies using a W208F mutant (25, 26). To date, LmP represents the only trypanosomatid hybrid type A peroxidase characterized as using Asc and Cc as electron donors, with its activity implicated in parasite differentiation and oxidative stress protection (27). Biochemical characterization of other members of this family of peroxidases may help to unravel the catalytic mechanism of these ancient enzymes as well as to define their role in pathogen survival, proliferation, and virulence.

In the present work, we characterized the Cc-dependent peroxidase activity of TcAPx-CcP. After a fast reaction with H₂O₂, an intramolecular electron transfer (ET) pathway at the heme microenvironment necessary for the oxidation of Cc was revealed; indeed, in addition to Trp, the role of a Cys residue typically located in the active site of these hybrid type A peroxidases was disentangled. Also, in light of its CcP activity and the presence of a secretory motif, which predicts membrane localization in all other hybrid type A members studied (15, 22, 23), we explored the subcellular localization of the enzyme in the infective T. cruzi parasite stages. Finally, the role of TcAPx-CcP overexpression in parasite virulence was assessed through in vitro and in vivo infections in the cellular and acute murine model of CD.

Results

Visible Spectral Studies of TcAPx-CcP Reaction with H₂O₂.

Recombinant T. cruzi wild-type (wt) and W233F APx-CcP were purified to homogeneity (>99% pure) by affinity chromatography as assayed by SDS/PAGE (Fig. 1A and Fig. S1A). For all purifications ca. 70% of protein had Fe incorporated (as determined by the Soret peak absorbance in relation to the total protein content). To investigate the reaction of TcAPx-CcP with H₂O₂, the optical spectra of TcAPx-CcP (from 350 to 650 nm) were recorded before and after the addition of equimolar concentrations of H₂O₂ to assay the spectral shifts associated with the formation of a compound I-like product (Fig. 1B). The resting enzyme exhibited a Soret peak at 409 nm that shifted to 420 nm after reaction with H₂O₂, with the appearance of two humps at 539 and 560 nm. These changes in the optical spectra are indicative of the generation of a compound I-like product (25, 26, 28) (Fig. 1B). When the W233F mutant was analyzed, the Soret peak was found to be at 414 nm in the resting state, indicating some structural perturbation at the heme microenvironment due to the amino acid replacement (Fig. 1C). Upon reaction with H₂O₂, the Soret peak shifted to 418 nm, with only the 560-nm hump now present (Fig. 1C). Given the rapid generation of the compound I-like product, rapid kinetics analysis by stopped flow at 10 °C was performed to visualize the classical compound I in TcAPx-CcP. The maximal absorption (Soret peak) in the ferric state of heme peroxidases drops in intensity when the enzyme reacts with H₂O₂ to form compound I, so a decrease in absorbance at the Soret peak is expected after H₂O₂ reaction (26, 28, 29). A very fast decay in absorbance for the wt enzyme (Soret peak at 409 nm) was observed upon mixing with H₂O₂, indicating the formation of the Fe-^{IV = O}-P^•+ radical of compound I (Fig. 1D). Interestingly, data in Fig. 1E, show that in the T. cruzi W233F mutant, a very fast (5 ms) and minor decay in absorbance at the Soret peak (414 nm) was followed by a slower increase in absorbance, suggesting the generation of an “alternative compound I-like” species. The drop in absorbance at 414 nm for the W233F mutant is in agreement with the spectral data shift upon H₂O₂ addition shown in Fig. 1C. Because the compound I was not stabilized in the W233F mutant, the data support that other amino acids, in addition to Trp²³³, can participate in the reduction of the Fe-^{IV = O}-P^•+ radical in TcAPx-CcP.

Fig. 1.

Spectroscopic analysis of TcAPx-CcP compound I and compound I-like intermediates. (A) SDS/PAGE of purified, recombinant TcAPx-CcP (1) and its W233F mutant (2) visualized following Coomassie blue staining. Absorption spectra of TcAPx-CcP (2 μM) (B) and the W233F mutant (C) before (solid line) and after (dashed line) equimolar H₂O₂ addition (2 μM). (Inset) Spectra from 500 to 650 nm. Arrows indicate the Soret peak of the resting enzyme and after H₂O₂ reaction. Spectral shifts of TcAPx-CcP are indicative of compound I-like formation. TcAPx-CcP (5 μM) (D) or W233F (E) was mixed with (dashed line) or without (solid line) H₂O₂ (5 μM) at 10 °C. Changes in the absorbance at the Soret peak [409 and 414 nm for TcAPx-CcP (D) and W233F (E), respectively] were followed. The decrease in the absorbance at the Soret peak is indicative of compound I generation.

Fig. S1.

(A) Purification of wt and W233F TcAPx-CcP. Protein fractions (lanes 1–8 and 9–12 for wt and W233F TcAPx-CcP mutant, respectively) obtained during purification of the His-tagged proteins using nickel-nitrilotriacetic acid affinity chromatography and eluted with a linear imidazole gradient were resolved on 15% SDS/PAGE and visualized by Coomassie blue staining. (B) Inactivation of TcAPx-CcP activity by H₂O₂. TcAPx-CcP (2 μM) was incubated with H₂O₂ (8 μM) in phosphate buffer (100 mM, pH 7.4) at 25 °C. At time intervals, aliquots of TcAPx-CcP (0.1 μM) were assayed for CcP activity (Cc²⁺, 50 μM; H₂O₂, 10 μM) by following the decrease in absorbance at 550 nm due to oxidation of Cc²⁺ (ε value of 21,000 M⁻¹⋅cm⁻¹) in phosphate buffer. The data are expressed as a percentage of activity compared with the initial activity.

Kinetic Analysis of the Cc-Dependent Peroxidase Activity.

Both the formation of a compound I-like product (this work) and the phylogenetic analysis of TcAPx-CcP indicate that the T. cruzi enzyme might function using Asc and Cc as electron donors. The TcAPx-CcP activity with Cc as the reducing substrate was monitored spectrophotometrically at 550 nm following ferrocytochrome c (Cc²⁺) oxidation after H₂O₂ addition. As shown in Fig. 2A, Cc²⁺ oxidation was observed in the presence of variable concentrations of wt enzyme (0.05–0.1 μM) after H₂O₂ addition (50 μM). When the W233F enzyme was used, even at high concentrations (0.1–0.5 μM), no Cc²⁺ oxidation was observed, indicating the central role of Trp²³³ for the Cc-dependent enzyme activity (Fig. 2A). In this mutant, a 10-fold decrease in Asc-dependent activity was also measured. To characterize the Cc-dependent peroxidase activity further, steady-state kinetic analysis was performed in the presence of wt TcAPx-CcP (0.05 μM), and variable amounts of Cc²⁺ (0–60 μM) and H₂O₂ (0–50 μM) as shown in Fig. 2B. The activity of the wt enzyme obeys Michaelis–Menten kinetics as assessed by the dependency of the initial velocity (V₀) as a function of H₂O₂ concentration under different Cc²⁺ concentrations (Fig. 2B). A secondary plot of 1/V₀ vs. 1/[H₂O₂] generated linear parallel slopes, indicating that TcAPx-CcP displays a typical bisubstrate ping-pong kinetic mechanism (30, 31) (Fig. 2C). From the y-axis intercept, apparent V_max (V′_max) values for the different Cc²⁺ concentrations used were calculated. From the y- and x-axis intercepts of the secondary plot (1/V′_max vs. 1/[H₂O₂]; Fig. 2D), 1/V_max and 1/K_m values for Cc²⁺ were determined, respectively: K_m (Cc²⁺) = 23.5 μM, k_cat = 5.0 s⁻¹, and k_cat/K_m (Cc²⁺) = 2.1 × 10⁵ M⁻¹⋅s⁻¹. To determine the rate constant of the enzyme with H₂O₂, the absorbance at 390 nm (a λ value at which the resting and oxidized enzyme have significant absorbance differences; Fig. 1B) was followed by rapid stopped-flow techniques. The reaction of H₂O₂ (8 μM) with TcAPx-CcP (2 μM) led to a fast time-dependent decrease in absorbance (Fig. 2E). The time course was fitted to a double exponential decay, with observed rate constants for both phases dependent on H₂O₂ concentration (Fig. 2F). The slopes of the plot (Fig. 2F) yielded second-order rate constants for the H₂O₂-dependent oxidation of TcAPx-CcP of (2.9 ± 0.5) × 10⁷ M⁻¹⋅s⁻¹ and (5.4 ± 0.2) × 10⁵ M⁻¹⋅s⁻¹ for the fast and slow processes, respectively. The k value on the order of 10⁷ M⁻¹⋅s⁻¹ for the initial reaction of H₂O₂ with the resting ferric state of the TcAPx-CcP is in good agreement with previous reports for other heme peroxidases (26, 32). It was previously shown that APx enzymes (in the absence of reducing substrates) are inactivated by H₂O₂ (33). Thus, the slower process is likely to represent (because no reducing substrate was present) the reaction of a second molecule of H₂O₂ with the oxidized TcAPx-CcP species, leading to an inactive form of the enzyme. Indeed, under these experimental conditions, we also observed a time-dependent inactivation of the enzyme by H₂O₂ (SI Materials and Methods and Fig. S1). In Table 1, the parameters of the steady-state kinetic analysis at pH 7.4 of TcAPx-CcP when using Cc (Fig. 2) or Asc (Fig. S2) as the reducing substrates are shown. Overall, the efficiency of TcAPx-CcP when using Cc is one order of magnitude higher than with Asc.

Fig. 2.

Investigating the Cc-dependent peroxidase activity of TcAPx-CcP. (A) Oxidation of Cc²⁺ (50 μM) was followed at 550 nm in the presence of H₂O₂ (10 μM) before and after the addition of TcAPx-CcP (wt; solid and dashed black lines, respectively) or its W233F mutant (solid and dashed gray lines, respectively). Proteins concentrations are noted, whereas the arrow indicates when enzyme was added to each reaction. Abs, absorbance. (B) Activity assay described in A was performed in the presence of TcAPx-CcP (0.05 μM) using different concentrations of H₂O₂ (0–50 μM) and Cc²⁺ (4, 10, 25, and 60 μM). The continuous lines show the best fit to a rectangular hyperbola for the experimental data. Double reciprocal (C) and secondary (D) plots from the data obtained in B. (E) TcAPx-CcP (2 μM) was mixed in the stopped-flow spectrophotometer with H₂O₂ (8 μM) in sodium phosphate buffer (100 mM, pH 7.4) containing DTPA (0.1 mM) at 25 °C. The time-dependent change in absorbance at 390 nm (maximal absorbance difference observed for resting and oxidized enzymes; Fig. 1B) was followed. (F) Effect of H₂O₂ concentrations on the observed rate constants of the biphasic decay (at 390 nm). Filled (●) and empty (○) circles represent fast and slow phases of decay, respectively. The continuous lines represent the best linear fit of experimental data. Data shown are from a representative experiment reproduced three times.

Table 1.

Steady-state kinetic parameters for TcAPx-CcP

Substrate	Km, μM	kcat/Km, M−1⋅s−1	pH	Source
Asc	192 ± 19	5.7 × 104	6.5	(15)
	190 ± 2	3.5 × 104	7.4	This work
Cc	29 ± 8	2.1 × 105	7.4	This work

Fig. S2.

Steady-state kinetic analysis of TcAPx-CcP with Asc. (A) Oxidation of Asc (0–600 μM) was monitored at 290 nm (ε₂₉₀ = 2,800 M⁻¹⋅cm⁻¹) using TcAPx-CcP (0.01 μM) in the presence of H₂O₂ (50 μM) in phosphate buffer (100 mM, pH 7.4). (B) From the double reciprocal plot (1/V₀ versus 1/[Asc]), a K_m value of 190 ± 2 μM for Asc was determined. (C) Asc-dependent peroxidase activity was assayed in the presence of Asc (600 μM) and different concentrations of H₂O₂ (0–100 μM) in phosphate buffer (100 mM, pH 7.4). (D) From the double reciprocal plot (1/V₀ versus 1/[H₂O₂]), a K_m value of 3 ± 0.5 μM for H₂O₂ was determined. The k_cat (s⁻¹) for the reaction of the reducing substrate was obtained from the V_max value of substrate oxidation of B and considering all of the enzyme as an oxoheme complex intermediate, k_cat = V_max/[E]. The k_cat/K_m for Asc was calculated at pH 7.4 (3.5 × 10⁴ M⁻¹⋅s⁻¹) being similar to the k_cat/K_m for Asc previously reported at pH 6.5 (Table 1).

Detection of TcAPx-CcP Trp^233•+ Radical and Role of Cys²²².

To determine the amino acid-derived radicals participating in the TcAPx-CcP catalytic cycle after reaction with H₂O₂, EPR spin trapping studies were carried out. Upon addition of equimolar concentrations of H₂O₂ to the wt enzyme (60 μM) in the presence of the spin trap 3,5-dibromo-4-nitroso benzene sulfonate (DBNBS) (10 mM), an immobilized carbon-centered radical adduct signal that has a nitrogen hyperfine splitting constant (a_N) value of 32 G (or 2a_N = 64 G); this signal was absent in the W233F mutant (Fig. 3). Overall, these results strongly support the participation of a Trp²³³-derived radical in the generation of the compound I-like product. Detailed inspection of the heme microenvironment in hybrid type A and CcP peroxidases revealed the presence of a Cys residue (Cys²²² for T. cruzi enzyme) located near Trp²³³ (25, 34); this Cys residue is absent in the APx family (23). It has been proposed that the sulfur atom of this Cys residue favors stabilization of the LmP compound I-like product. Indeed, mutation of Cys¹⁹⁷ to Thr in LmP decreased more than 100-fold the Cc²⁺-dependent peroxidase activity without affecting the Asc-dependent function (25, 26). We thus explored the involvement of Cys²²² in TcAPx-CcP peroxidase activity by means of alkylation of enzyme thiols with N-ethyl-maleimide (NEM). Of the six Cys residues present in TcAPx-CcP, two are buried in the protein structure and four (including Cys²²²) are solvent-accessible. As shown in Fig. 4, NEM-treated wt enzyme reacts with H₂O₂ (as wt nontreated enzyme) with the generation of a compound I-like product (Fig. 4A). Although the NEM-treated enzyme reacted with H₂O₂, Cc²⁺-dependent peroxidase activity was decreased by more than 70% compared with wt enzyme, although the Asc-dependent peroxidase activity was not affected (25, 26) (Fig. 4D). Interestingly, when the NEM-treated W233F mutant was exposed to H₂O₂, no changes in the Soret peak at 414 nm of the resting Fe-III was observed (Fig. 4C). In this condition, a significant drop in Asc-dependent peroxidase activity was also evident for the W233F-NEM enzyme (Fig. 4D). Overall, the above results indicate that Cys²²² (located near the heme and Trp²³³) is necessary for compound I-like generation/stability.

Fig. 3.

EPR spectra of DBNBS radical adducts obtained for oxidized TcAPx-CcP. The spectra were obtained after reaction of TcAPx-CcP (wt) and the W233F enzyme (60 μM) in DBNBS (10 mM) in the absence (A) or presence (B) of H₂O₂ (60 μM). 2a_N, the distance between the external peaks in G corresponds to twice the value of the hyperfine splitting constant of the nitrogen atom of the radical adduct (2a_N = 64 G).

Fig. 4.

Peroxidase activity of TcAPx-CcP following cysteine alkylation. (A) Spectral change of resting, NEM-treated (solid line; Soret peak at 409 nm), or H₂O₂-treated (2 μM; dashed line; Soret peak at 420 nm) TcAPx-CcP (2 μM). (B) Cc²⁺-dependent peroxidase activity of untreated (solid line) and NEM-treated (dashed line) TcAPx-CcP (0.01 μM) toward H₂O₂ (20 μM) using Cc²⁺ (20 μM) as a reductant. Activity was followed by monitoring the change in absorbance at 550 nm. The arrow indicates the addition of enzyme. (C) Spectral change of resting, NEM-treated (solid line), or H₂O₂-treated (2 μM; dashed line) TcAPx-CcP W233F (2 μM; Soret peak at 414 nm). (D) Asc-dependent peroxidase activity of TcAPx-CcP (0.13 μµM) and the W233F mutant (1 μM), with (+) or without (−) NEM treatment, toward H₂O₂ (40 μM). Assays were conducted in the presence of Asc (200 μM), and activity was monitored by following the change in absorbance at 290 nm. Data represent the mean ± SEM from five independent determinations and are expressed as absorbance units per second. ***P < 0.0001 by one-way ANOVA and post-Tukey’s multiple comparison test.

To determine the generation of a Cys²²² radical accurately after reaction of TcAPx-CcP with H₂O₂, immunospin trapping assays using 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) were performed. Untreated or NEM-treated wt and W233F enzymes (10 μM) were exposed to H₂O₂ (0–30 μM) in the presence of DMPO (100 mM). DMPO-protein adducts were evaluated by Western blot using the anti–DMPO-nitrone antibody as previously reported (35). In the wt enzyme, DMPO-nitrone adducts were observed with a signal intensity dependent on H₂O₂ concentration (Fig. 5A). These immunoreactive signals were completely abrogated in the NEM-treated enzyme, strongly suggesting the participation of Cys in the detected protein adduct (Fig. 5A). Moreover, the signal was greatly enhanced in the W233F mutant, suggesting that in the absence of Trp²³³, Cys²²² could be involved in the recovery of compound I, leading to an alternative ET pathway, as suggested from data in Fig. 1D. These results were corroborated by EPR spin trapping analysis (Fig. 5B). Reaction of the wt enzyme (60 μM) with equimolar concentrations of H₂O₂ in the presence of DMPO (100 mM) yielded a detectable paramagnetic signal; notably, the EPR signal was significantly enhanced in the W233F mutant (Fig. 5B, line B). One hyperfine splitting constant of the beta hydrogen (a^H_β) of the adduct signal (a^H_β = 16 G) is fully consistent with the trapping of an APx-CcP-cysteinyl radical, whereas another (a^H_β = 9.3 G) suggests that a Tyr radical may be also trapped (36, 37). Furthermore, the formation of the DMPO radical adduct was almost fully inhibited in the NEM-treated wt and W233F enzymes (Fig. 5B, lines C). The results indicate the generation of a protein-derived radical located at the active site Cys²²² (alternative compound I-like) that is important for the TcAPx-CcP Cc-dependent peroxidase activity (Fig. 4B). To obtain full confirmation of the formation of Cys²²² radical upon H₂O₂ reaction, peptide mapping-MS analysis of TcAPx-CcP was performed. A manual search for the peptide of interest revealed both L²⁰³-K²³⁷ and DMPO-adduct peptides coeluting at 60.2 min. Most abundant ions were those ions tetra-charged for both peptides although triple-charged ions were also found. Enhanced resolution analysis was performed for best accuracy and mass-to-charge ratio (m/z) were obtained for each peptide: 1,258.9 ([M + 3H]³⁺); 944.4 ([M + 4H]⁴⁺); 1,294.9 ([M + DMPO + 3H]³⁺); 971.4 ([M + DMPO + 4H]⁴⁺). The fragmentation pattern of the DMPO-peptide adduct was obtained by an enhanced product ion of m/z 971.4 (Fig. 5C). Several b and y ions were found, ensuring identity of the peptide sequence. Interestingly, the y-series (y₉–y₁₅ ions) suggests that Tyr²²⁹ is unmodified after treatment with H₂O₂ in the presence of DMPO. The assignment of the DMPO-modified amino acid was finally obtained with the data of the b₁₇ fragment (unmodified peptide) and the b₂₂ fragment (modified peptide) in full support of Cys²²² as the preferential site of adduct formation by DMPO in the W233F enzyme (Fig. 5C). Overall, EPR spin trapping, immunospin trapping, and peptide mapping-MS data support the critical role of Cys²²² in the generation of the alternative compound I-like in TcAPx-CcP.

Fig. 5.

Immunospin trapping, EPR spectra, and mass spectrometry analysis of TcAPx-CcP protein adducts in the presence of DMPO. (A) Blot containing TcAPx-CcP and W233F mutant (10 μg) protein, with (+) or without (−) NEM treatment, treated with H₂O₂ (0, 10, and 30 μM) in the presence of DMPO (100 mM) was probed with anti-DMPO antibody (green) to detect protein-DMPO-nitrone adducts and anti–TcAPx-CcP antisera (red). Merging (TcAPx-CcP plus DMPO) is shown in yellow. (B) EPR spectra of TcAPx-CcP (wt, Left) and the W233F mutant (Right, 60 μM each) in the presence of DMPO (100 mM) before (line A) and after (line B) the addition of H₂O₂ (60 μM). Line C is the same as line B, except the enzymes were NEM-treated. A composite signal with values of the hyperfine splitting constant of the hydrogen atom (a^H_β) of adducts of 9.4 G and 16 G is consistent with trapping of tyrosyl and cysteinyl radicals, respectively. (C) Tandem MS analysis of the TcAPx-CcP-DMPO adduct. The spectrum of the tetra-charged ion at m/z 971.1 (retention time = 60.2 min) from a tryptic digestion of the TcAPx-CcP W233F mutant (16 μM) after treatment with equimolar H₂O₂ in the presence of DMPO (100 mM) is shown. The major N-terminal (b, red-labeled) and C-terminal (y, blue-labeled) fragment ions that allowed the sequence L203-K237 assignment containing a cysteine-linked DMPO residue are shown. The upper part of C shows the amino acid sequence of peptide L203-K237, indicating major b and y ions detected by EPI.

Molecular Dynamics of ET Pathways at the Heme Microenvironment Involving Trp²³³ and Cys²²².

Using the previously reported crystallographic structure of LmP [Protein Data Bank (PDB) ID code 3RIV], we performed homology models for the TcAPx-CcP (25) (Fig. S3), from which 100-ns-long molecular dynamics (MD) simulations were conducted. The proximal site of the heme in TcAPx-CcP contains the conserved residues Trp²³³, Cys²²², and His²¹⁷ present in all hybrid type A peroxidases (23). The critical Trp²³³ side chain is maintained nearby the His²¹⁷ that coordinates the heme iron via π-stacking interactions (Fig. 6). This structural property determines the capability of Trp²³³ of being able to reduce the heme after its reaction with H₂O₂. Particularly for TcAPx-CcP, Trp²³³ is located close to Cys²²². MD simulations indicate that the interactions between the Trp²³³ and Cys²²² side chains are influenced by the thiol or thiolate character of Cys²²² (i.e., Cys²²² pK_a). The thiol form shows a direct interaction with Trp²³³, but also with the heme porphyrin aromatic system mainly in the form of thiolate (Fig. 6 and Fig. S4A). To understand the redox interplay at the heme microenvironment involving the heme, Trp²³³, and Cys²²² better, we performed ET probability calculations by using the “pathways” formalism. By means of this method, the possible paths for ET from Trp²³³ and Cys²²² to the oxoferryl heme were computed throughout the MD trajectories. The best possible structural path (i.e., the higher electronic coupling matrix) connecting the electron donor and acceptor was obtained, allowing us to analyze and compare key structural features governing the ET process. In Fig. S4B, the average values are shown. The data support that the preferred ET pathway involved the heme-dependent oxidation of Trp²³³ (Fig. 7, route A) to the Trp^233•+ intermediate; this intermediate may lose a proton with the generation of the neutral indolyl radical (NTrp^233•) to be reduced by Cys²²² to yield the corresponding cysteinyl radical and establishing a redox equilibrium between the two amino acid-derived radical species. On the other hand, in the W233F mutant, the oxidized heme will promote the direct one-electron oxidation of the thiolate form of Cys²²² (Fig. 7, route B), thus maintaining the Asc peroxidase activity of the enzyme in the absence of Trp²³³ (Fig. 4D).

Fig. S3.

(Left) Alignment of the LmP crystallographic structure (cyan) with the TcAPx-CcP homology model (gray). Before superimposition, both systems were optimized, solvated, and equilibrated to 300 K, as described in Materials and Methods. (Right) Close-up image of the heme binding site. Side chains of His²¹⁷, Trp²³³, and Cys²²² (numbering in bold uppercase refers to TcAPx-CcP) are depicted by means of a “licorice” representation and colored by atom types, with the exception of carbon atoms, which are colored matched to the corresponding backbone (cyan for LmP and gray for TcAPx-CcP).

Fig. 6.

Predicted ET pathway across the active site of TcAPx-CcP. The Trp²³³→heme, Cys²²²→heme, and Cys²²²→Trp²³³ paths are shown in green, blue, and red, respectively.

Fig. S4.

(A) Superimposition of representative structures of TcAPx-CcP with Cys²²² in its two protonation states, obtained from MD simulations. The backbone of the structure, where Cys²²² [as a thiolate (R-S⁻)], colored in orange, and the corresponding thiol form (R-SH), colored in gray, is shown. Cys²²² is depicted with a balls and sticks representation in both cases, and distances (in angstroms) between the S atom of Cys²²² and the closer atom from Trp²³³ are shown as dashed lines. (B) ET coupling matrix average values (<T_DA>) and SDs in electron volts, obtained with the pathways algorithm.

Fig. 7.

Representation of the heme microenvironment redox pathways after H₂O₂ reaction with TcAPx-CcP. Pathway A shows heme reduction via Trp²³³ with the formation of the tryptophanyl radical (thermodynamically favored pathway), whereas pathway B involves alternative heme reduction via Cys²²² (in its thiolate form), yielding the corresponding cysteinyl radical. ET between Trp²³³ and Cys²²² in both pathways is shown. Arrows widths and lengths are indicative of relative flow rates.

Subcellular Localization of TcAPx-CcP in the Infective Parasite Stages.

Here, the kinetic data shown above indicate that the TcAPx-CcP–preferred source of reducing equivalents is Cc. In addition to the previously reported ER localization in the noninfective epimastigote stage (15), we searched for the presence of TcAPx-CcP in the mitochondrial compartment (where Cc is present) in the different parasite stages (noninfective epimastigotes, infective intracellular amastigotes, and extracellular trypomastigotes) using specific antibodies (Fig. S5). First, and using TcAPx-CcP overexpressers (epimastigote), we conducted controlled-digitonin permeabilization studies to evaluate the coelution of the enzyme with mitochondrial markers (Fig. S6A). TcAPx-CcP was found to coelute with Cc and Fe-superoxide dismutase A (SODA) (mitochondrial markers), and it was also present in the membrane fraction (Fig. S6A). This result indicates that at least a fraction of the enzyme is associated with mitochondria. Immunofluorescence microscopy of T. cruzi epimastigotes evidences a nonhomogeneous distribution of the enzyme, compatible with ER localization (15). The enzyme is also present in close proximity to mitochondrial DNA and Fe-SODA (Fig. S6B). Finally, to localize proteins at the ultrastructural level precisely, providing simultaneous visualization of antigen epitopes together with the ultrastructure of membrane compartments, we conducted immunoelectron microscopy (immuno-EM) (38) of TcAPx-CcP in the infective T. cruzi stages. TcAPx-CcP enzyme was found to be located in mitochondria, particularly associated with membranes of the cristae (Fig. 8B and Figs. S7 and S8). The localization of TcAPx-CcP in mitochondria was different from the localization observed for MPX located at the mitochondrial matrix (11) (Fig. 8 C–E). Intriguingly, the enzyme was found to be associated with the plasma membrane of the intracellular amastigote and extracellular trypomastigote stages, potentially forming a “shield” at the host–parasite interface (Fig. 8A and Figs. S7 and S8). These observations were confirmed in T. cruzi pTEX-APx-9E10 overexpressers using the monoclonal anti–c-Myc antibody in both the trypomastigote and amastigote stages (Fig. S7). The presence of TcAPx-CcP at the host–parasite interface strongly suggests its participation in the parasite defense mechanisms toward host-derived toxicity. To date, the physiological reducing substrate (i.e., Asc, Cc, or both) for this ancient enzyme in T. cruzi is still elusive (Discussion).

Fig. S5.

Evaluation of the specificity of the T. cruzi antibodies used for immunolocalization studies. (A) Parasites extracts from T. cruzi amastigotes (Am), trypomastigotes (Try), and epimastigotes (Epi) (10 μg of total protein) run on 12% SDS/PAGE were probed with anti–TcAPx-CcP (Left, 1:2,000) or anti-TcMPX (Right, 1:2,000) antibodies. For all parasites, stages analyzed in only one immunoreactive band were observed. (B) Western blot analysis to verify that there are no cross-reactive epitopes between T. cruzi antibodies and cardiomyocytes. Protein extracts from cardiomyocytes (lane 1, 100 μg) and T. cruzi epimastigotes (lane 2, 100 μg) were run on 15% SDS/PAGE and probed with anti-GAPDH (1:2,000; Sigma), anti-TcMPX (1:2,000), and anti-TcAPx-CcP (1:2,000). T. cruzi antibodies did not cross-react with cardiomyocyte proteins. MM, molecular mass.

Fig. S6.

(A) T. cruzi epimastigotes were permeabilized with sequential additions of digitonin (0.1 mg) by means of stepwise addition until reaching a final concentration of 1 mg of total digitonin (SI Material and Methods), and proteins released were analyzed by Western blot for the detection of TcAPx-CcP, mitochondrial Fe-SODA, anti-T. brucei Cc, and anti-CPX. P, pellet extract (membrane-enriched) fraction after the sequential addition of 0.1 mg of digitonin reaching a final concentration of 1 mg. Note that proteins immunodetected under each digitonin value represent a released fraction from the remaining protein left from the previous permeabilization steps. (B, Upper) Immunofluorescence microscopy of T. cruzi epimastigotes showing DNA staining (DAPI colored blue), Fe-SODA-PE (red), TcAPx-CcP (green-FITC), and a merged image. All panels correspond to the same image. (B, Lower) Colocalization (white arrows, merge colored yellow) of kinetoplastid (k) DNA (DAPI colored red) and Fe-SODA-PE (colored green). (B, Lower Right) Bright field of the image. All images were taken at 400× magnification.

Fig. 8.

TcAPx-CcP is localized to the plasma membrane and mitochondrion of intracellular T. cruzi amastigotes. Cultured cardiomyocytes infected with T. cruzi (wt) were analyzed with antisera raised against TcAPx-CcP (A and B) or TcMPX (C–E) (APX¹⁰ and MPX¹⁰ nm of gold, respectively). The white outlined rectangle shows a high magnification of the amastigote plasma membrane, and black arrowheads indicate amastigote mitochondrial membranes. Am, amistigote; k, kinetoplast; m, mitochondrion; n, nucleus of infected cell. (Scale bars: A–D, 500 nm; E, 200 nm.)

Fig. S7.

TcAPx-CcP is confined to the plasma and mitochondrial membranes in the infective amastigote and trypomastigote stages. Immuno-EM localization of TcAPx-CcP in a culture-derived amastigote (A; Am), trypomastigote (B; Try), and axenic epimastigote (C; Epi) using a monoclonal antibody against the c-Myc9E10–tagged enzyme in T. cruzi overexpressers (CL-Brener-pTEX-APX-9E10). Labeling is abundant on the cell surface of amastigotes (A) and trypomastigotes (B) and is associated with limiting mitochondrial membrane and luminal membranes of the cristae (arrowheads) in all parasite stages. g, Golgi, k, kinetoplast; m, mitochondrion. (Scale bars: 200 nm.)

Fig. S8.

TcAPx-CcP is localized to the plasma membrane and mitochondrion of intracellular T. cruzi amastigotes. Cultured cardiomyocytes infected with T. cruzi (wt) were analyzed with antisera raised against TcAPx-CcP by immuno-EM. The figure shows three intracellular amastigotes with abundant labeling at the plasma membrane (arrowheads) and associated with mitochondrial membranes (m, arrowheads). n, parasite nucleus; pm, plasma membrane. (Scale bar: 500 nm.)

Enhanced Virulence of TcAPx-CcP Overexpressers in Cellular and Mice Infections.

T. cruzi interaction with macrophages leads to the assembly of the membrane-bound NADPH oxidase with the generation of sustained (90 min) and large amounts of superoxide radical, and thus hydrogen peroxide, directed toward the internalized parasite (7). Once internalized, T. cruzi has to deal with the oxidative environment of the phagosome to survive and escape to the host cytosol, where it proliferates as amastigotes. Thus, we first searched for parasite survival in naive macrophage infections. Macrophages were infected with wt or TcAPx-CcP–overexpressing trypomastigotes, and infection yields were evaluated after 24 h by intracellular amastigote counting. Trypomastigotes overexpressing TcAPx-CcP were more resistant to macrophage toxicity compared with wt parasites, indicating the participation of TcAPx-CcP in the H₂O₂ detoxification generated during parasite internalization (Fig. 9A). Finally, the infectivity of TcAPx-CcP overexpressers was evaluated in cardiomyocytes, cells where amastigotes can reside during chronic infection. It has been shown that T. cruzi cardiomyocyte infection leads to the establishment of host–mitochondrial dysfunction, with an increase in H₂O₂ and inflammatory cytokine (IL-1β and TNF-α) production (39). We evaluated the infectivity of TcAPx-CcP overexpressers following 96 h of infection, an incubation time sufficient to cause mitochondrial dysfunction in cardiomyocytes and to achieve amastigote replication in the host cell cytoplasm. Rat-derived ventricular cardiomyocytes (H9c2) were infected as above and incubated for 96 h. As was observed for macrophages, TcAPx-CcP overexpressers were more infective than wt parasites, supporting the participation of the enzyme in enhancing virulence (Fig. 9A). These results correlate well with in vitro studies where TcAPx-CcP–null parasites infect myoblast cells at a reduced level (20). Finally, and to validate the enhanced virulence of the TcAPx-CcP overexpressers in vivo, we conducted mouse (BALB/c susceptible and C57BL/6 resistance lineages; Fig. 9B) infections with culture-derived trypomastigotes from wt (CL-Brener) and TcAPx-CcP overexpressers. As shown in Fig. 9B, TcAPx-CcP overexpressers produced higher blood parasitemias in the acute infection model of CD, indicating the role of this enzyme in parasite virulence and consistent with its function as part of the pathogen defense against host-derived cytotoxic oxidants.

Fig. 9.

TcAPx-CcP overexpression enhances parasite virulence. (A) Culture-derived T. cruzi trypomastigotes (CL-Brener and TcAPx-CcP overexpressers) were used to infect macrophages at a ratio of five trypanosomes per cell. Following 24 h of infection, cells were stained with DAPI and the number of intracellular amastigotes was counted. Cardiomyocyte infection was performed as above, except that the time of infection was increased to 96 h. Results are expressed as the number of intracellular amastigotes per 100 cells. Data represent the mean ± SEM of at least three independent experiments. (B) BALB/c or C57BL/6 mice (five per group) were infected with 1.5 × 10⁷ culture-derived T. cruzi trypomastigotes (CL-Brener and TcAPx-CcP overexpressers). Blood taken from each mouse at daily intervals from day 3 postinfection onward was microscopically examined for the presence of the bloodstream form of trypomastigotes. Data represent the mean ± SEM for each day. *P < 0.005 as assayed by the Student’s t test.

SI Materials and Methods

Purification of Recombinant TcAPx-CcP wt and W233F Mutant Proteins.

Escherichia coli BL21 DE3 cells containing the plasmid were grown at 37 °C in NZCYM broth containing ampicillin (50 μg/mL). Expression of recombinant TcAPx-CcP was induced by isopropyl-β-d-thiogalacto-pyranoside (0.8 mM) addition when the culture reached A₆₀₀ = 0.6 in the presence of g-amino-levulinic acid (0.5 mM), and the temperature lowered to 22 °C for overnight protein expression. The purification was performed in a 5-mL HiTrap affinity column (Amersham Biosciences) charged with Ni²⁺ and equilibrated with binding buffer [50 mM sodium phosphate (pH 7.4), 10 mM imidazole, and 500 mM NaCl] at a flow rate of 3 mL⋅min⁻¹. TcAPx-CcP was eluted with a linear imidazole (10–500 mM) gradient in sodium phosphate [50 mM (pH 7.6) containing 500 mM NaCl]. Imidazole was removed immediately by buffer exchange in sodium phosphate buffer (50 mM, pH 7.4) using HiTrap desalting columns (Amersham Biosciences).

Controlled Digitonin-Mediated Cell Permeabilization and Western Blot Analysis.

Trypanosoma cruzi epimastigotes overexpressing TcAPx-CcP (3 × 10⁸ cells per milliliter, 1 mg of total parasite protein) were washed three times with 1 mL of Dulbecco’s PBS (dPBS) at pH 7.4. The parasite pellet was resuspended in dPBS (1 mL) containing digitonin (0.1 mg/mL) and allowed to set for 5 min at 4 °C. Parasites were collected by centrifugation at 800 × g for 10 min, and the supernatant was recovered and kept on ice. The parasite pellet was washed three times (1 mL) with dPBS and resuspended in dPBS containing digitonin (0.1 mg/ mL). The procedure was repeated 10 times, reaching a final digitonin concentration of 1 mg. Proteins from the supernatants of each sequential extraction step, whole-parasite extract (total parasite extract), and the membrane-enriched fraction [remaining pellet after the last 0.1 mg/mL digitonin addition (total pellet extract)] were subjected to 15% SDS/PAGE, transferred to nitrocellulose membranes, and probed with anti-T. brucei Cc [kindly provided by Andre Schneider, University of Berne, Department of Chemistry and Biochemistry, Berne, Switzerland (63)], anti-TcAPx-CcP (1:2,000), anti-FeSOD-A, and anti-CPX (1:2,000).

Immunolocalization Studies.

For fluorescence microscopy, wt epimastigotes (CL-Brener) were washed three times by centrifugation at 800 × g in PBS at pH 7.4 for 10 min and fixed in 4% (vol/vol) fresh PFA for 30 min at room temperature. Parasites were washed as above and permeabilized in PBS at pH 7.4 containing Triton X-100 (0.5% vol/vol) and BSA (3% wt/vol) for 30 min. Permeabilized parasites were incubated with rabbit anti-TcAPx-CcP and/or mouse anti–Fe-SODA polyclonal antibody (1:200) in the same solution and incubated overnight at 4 °C. Parasites were washed five times in PBS at pH 7.4 containing Triton X-100 (0.5% vol/vol) and incubated with anti-rabbit IgG-488 antibody and/or anti-mouse–PE (1/100; Invitrogen) for 1 h at room temperature. After staining, slides were mounted with antifade mounting solution plus DNA staining (DAPI) and visualized by epifluorescence microscopy (Eclipse TE-2000-E2-Nikon).

Discussion

By phylogenetic (23) and biochemical studies (this work), TcAPx-CcP is a hybrid type A peroxidase sharing both Asc- and Cc-dependent peroxidase activity. Although initially described to use Asc as a reducing substrate (15), the catalytic efficiency with Cc is higher, suggesting that, in vivo, the enzyme may also function as a CcP, so long as a component of TcAPx-CcP is present where this “alternative” reducing equivalent is found, such as the mitochondrion. Spectral analysis during the reaction of the enzyme with H₂O₂ showed the formation of a compound-I like (Fig. 1B) characteristic of the generation of the Trp^233•+ in these enzymes (25). The classical compound I was transiently detected following fast absorption spectra of the resting enzyme at the Soret peak (409 nm) after the H₂O₂ reaction (Fig. 1D). The participation in the reaction of the critical Trp²³³ located in close proximity to the heme was supported by EPR-spin trapping experiments (Fig. 3) and by the site-directed enzyme mutant (W233F) in experiments that included the lack of detection of the EPR signal in parallel with a great drop in the Cc-dependent activity and a 10-fold decrease in Asc-dependent peroxidase activity (Figs. 2 and 4). Rapid UV-visible scanning of the W233F enzyme (Soret peak at 414 nm) permitted detection of an alternative compound I-like species (Fig. 1 C and E), indicating that another redox-active amino acid located in the heme microenvironment was able to participate rapidly in the one-electron reduction of the porphyrin cation radical present in classical compound I. Protist hybrid type A heme peroxidases contain a Cys residue located near the heme (Cys²²² and Cys¹⁹⁷ for T. cruzi and LmP, respectively) (22). Immunospin trapping, EPR-spin trapping, and peptide mapping-mass spectrometry analysis using DMPO revealed the presence of a cysteinyl radical in wt TcAPx-CcP upon H₂O₂ reaction that was enhanced in the W233F mutant (Fig. 5). Enzyme thiol alkylation dramatically lowered the Cc-dependent peroxidase activity of TcAPx-CcP in line with the participation of this residue in enzyme activity. In this sense, MD simulations revealed that in fact, once the Trp^233•+ is formed to yield the alternative compound I, a subsequent ET can occur with the participation of Cys²²². For this ET to occur, once the Trp^233•+ is formed, it must lose a proton to yield the neutral indolyl radical (NTrp^233•) that, in turn, can abstract a hydrogen atom for the neutral HS-Cys²²², generating a cysteinyl radical (Cys²²²-S^•) (Eq. 1 and Fig. 7), establishing the following equilibrium:

$${\text{NTrp}}^{233•}+{\text{Cys}}^{222}-\text{SH}\rightleftharpoons {\text{HNTrp}}^{233}+{\text{Cys}}^{222}{-\text{S}}^{•}.$$ [1]

It is important to note that the estimated one-electron redox potentials of the iron/oxo complex and the tryptophanyl and cysteinyl radicals are approximately: 1,156, 1,025 and 970 mV, respectively, at pH 7.4 (40, 41), all of which are thermodynamically consistent with preferential route “A” of the ET shown in Fig. 7. In light of these data, we propose that the equilibrium involving the presence of Cys²²² (Eq. 1) allows the stabilization of the protein-derived radicals, permitting a more selective reactivity of Trp^233•+ toward Cc²⁺ after binding of the reducing substrate to the active site. Overall, these observations indicate distinctive characteristics of the LmP enzyme compared with the T. cruzi enzyme (e.g., formation of compound-like I in Tc-APx-CcP) and provide a detailed atomistic picture of the redox processes occurring at the heme microenvironment during catalysis in this family of peroxidases. Future studies involving crystallization of TcAPx-CcP and comparative analysis with LmP should help to rationalize the subtle differences observed herein in the reaction mechanism.

The hybrid type APx-CcP peroxidases have been proposed to represent an evolutionary link between catalase-peroxidase and Cc-peroxidases, although their physiological roles are still under discussion (23). LmP, located at the parasite mitochondria, is involved in H₂O₂ detoxification during parasite differentiation and infectivity (27, 42). In light of its CcP activity, we searched for TcAPx-CcP localization in the different T. cruzi parasite stages. In addition to the early report of ER localization in the noninfective epimastigote stage, immuno-EM experiments revealed that the enzyme was also associated with the mitochondria cristae in all parasite stages (Fig. 8 and Fig. S7), suggesting that the mitochondrial-associated fraction of the enzyme may truly work as a CcP. In the intracellular amastigotes and extracellular trypomastigotes, TcAPx-CcP was also found to be noticeably located at the plasma membrane of the parasite in close interaction with the host cytosol, creating an “antioxidant shield” at the host–parasite interface (Fig. 8). This membrane localization was not evident in the T. cruzi epimastigote stage (Fig. S7). In this subcellular localization, TcAPx-CcP may principally use host cell-derived Asc as a reductant to protect the parasite from H₂O₂. Overall, we speculate that the infective stage of the parasite uses the reducing substrates alternatively depending on their availability. Although micromolar Asc is found in both the host cytoplasm (43) and T. cruzi trypomastigotes (17), micromolar Cc is also present in the mitochondrial intermembrane space (44), supporting that TcAPx-CcP primarily uses Asc in its plasma membrane localization and Cc in the mitochondria (Table 1).

The cysteine-protease cruzipain is an example of enzyme subcellular relocalization along the T. cruzi life cycle, where restriction to vesicles of the endosomal/lysosomal system in the epimastigote stage is relocalized to the parasite surface in the intracellular amastigote stage (45). The C-terminal extension in cruzipain was suggested to participate in this relocalization (46). TcAPx-CcP contains a positively charged 17-aa sequence of unknown function close to the carboxyl terminus (15) that could be involved in such enzyme relocalization.

To evaluate the significance of TcAPx-CcP in infectivity, studies using cultured cells and animals were performed. Two cellular models (macrophages and cardiomyocytes) were used for the in vitro infection studies. First, T. cruzi infection of naive macrophages leads to activation of the plasma membrane NADPH oxidase, yielding large amounts of superoxide radical, and subsequently H₂O₂ (due to superoxide dismutase-catalyzed dismutation), toward the internalized parasite (7, 47). Second, T. cruzi cardiomyocyte infection leads to the establishment of host–mitochondrial dysfunction with increased H₂O₂ production (39) that can diffuse and reach the intracellular amastigote. In both cellular models, parasites overexpressing TcAPx-CcP were more infective than controls, thus supporting the hypothesis that this peroxidase plays a role in protecting T. cruzi from the cytotoxic effects of host cell-derived H₂O₂ (Fig. 9A). Recent proteomic data indicate that virulent T. cruzi strains express high levels of TcAPx-CcP, with this expression being further elevated in the infective stage (16, 48). However, it has now been shown that although TcAPx-CcP–null parasites are more susceptible to oxidant killing in vitro, the peroxidase activity is not a determinant for establishment of an infection in the animal model of CD, and is thus not essential for parasite survival within the mammalian host (20). As an ancient parasite, T. cruzi is well adapted to infect and persist in the mammalian host, developing redundant activities that have allowed it to succeed in hostile oxidative environments. Trypanosomes null for TcAPx-CcP contain other oxidant-metabolizing systems, including CPX, that are able to detoxify H₂O₂ efficiently and make parasite infection possible. For comparative purposes, the reaction rate values of H₂O₂ with TcAPx-CcP are 2.9 × 10⁷ M⁻¹⋅s⁻¹ (this work) and 3 × 10 ⁷ M⁻¹⋅s⁻¹ for CPX (49), supporting that both systems are, in principle, readily capable of eliminating cytotoxic levels of host cell-derived H₂O₂. Moreover, two other putative heme peroxidases are present in the parasite genome (TcCLB.507011.130 and TcCLB.511143.30), and the relevance of their expression under different biological conditions remains to be defined. Finally, and to assay the infectivity of TcAPx-CcP-overexpressing parasites, infection trials in two different mice lineages (BALB/c and C57BL/6) were conducted. Higher and significantly different parasitemias were observed in both mouse models of acute CD for the TcAPx-CcP overexpressers with respect to wt parasites, confirming its role in parasite virulence (Fig. 9B).

Overall, the data presented herein provide conclusive biochemical evidence of the hybrid nature of the type A heme peroxidase in T. cruzi (TcAPx-CcP); unravel mechanistic aspects of its redox enzymology with an atomic level of detail; and contribute to define its biological function, particularly its role as a virulence factor in CD.

Materials and Methods

Expression, Purification, and Site-Directed Mutagenesis of Recombinant TcAPx-CcP.

The plasmid for heterologous expression of TcAPx-CcP (pTrcHis-APX) was kindly provided by Shane Wilkinson, Queen Mary University, London (15). Purification of recombinant enzymes was performed as done previously with minor modifications, as noted in SI Materials and Methods (15). Purity of TcAPx-CcP preparations was evaluated by 15% SDS/PAGE (Fig. S1), and the Soret peak at 409 nm for wt enzyme was used to quantitate the heme protein content (ε = 101 mM⁻¹⋅cm⁻¹) (18). Site-directed mutagenesis of TcAPx-CcP (W233F) was performed using a site-directed mutagenesis kit with the following primers: forward, 5′-GGGCTACGTGGGTCCGTTCACGCACGACAAG-3′, and reverse, 5′-CTTGTCGTGCGTGAACGGACCCACGTAGCCC-3′. Sequence fidelity was confirmed by DNA sequencing (Institut Pasteur, Montevideo, Uruguay). The TcAPx-CcP W233F mutant was purified as described above for the wt enzyme, and the Soret peak at 414 nm was used to quantitate the heme protein content as described above. Protein concentration was measured by the Bradford protein assay.

Thiol Alkylation.

Alkylation of wt and W233F TcAPx-CcP was carried out incubation of the enzyme (60 μM) with NEM (10 mM) for 2 h in phosphate buffer (100 mM, pH 7.4) at 4 °C. Excess NEM was immediately removed using HiTrap desalting columns (Amersham Biosciences) in PBS (100 mM, pH 7.4). Thiol alkylation was confirmed by the 5,5′-dithiobis-2-nitrobenzoic acid assay (50).

Spectroscopic Analysis.

UV-visible spectra were recorded at 25 °C from Asc-free wt and W233F enzymes (2 μM) in the presence or absence of an equimolar H₂O₂ concentration in PBS (100 mM, pH 7.4).

Stopped-Flow Analysis.

Rapid acquisition spectra and absorbance time courses were obtained following Asc-free wt and W233F enzymes before and after H₂O₂ addition by mixing equal amounts (4 μM each) in sodium phosphate buffer (100 mM, pH 7.4) containing diethylenetriaminepentaacetic acid (DTPA) (0.1 mM) at 10 °C, reaching a final concentration of 2 μM reagents. Absorbance was recorded with an Applied Photophysics SX-20 stopped-flow spectrofluorometer (mixing time of ≤2 ms) equipped with a rapid-scanning diode array. Formation of compound I species was followed at a Soret peak of 409 nm for wt and 414 nm for W233F enzyme.

Steady-State Activity Assays and Rate Constant Determinations.

The steady-state activity assays for wt TcAPx-CcP and W233F mutant were performed using Asc and Cc²⁺ as reducing substrates, as previously described for LmP (25, 26). The Cc was prereduced with dithionite, and the excess reductant was removed immediately using HiTrap desalting columns in phosphate buffer (100 mM, pH 7.4). The concentration of Cc²⁺ was calculated spectrophotometrically following absorbance at 550 nm (ε_{550 nm} = 8,900 and 29,900 M⁻¹⋅cm⁻¹ for reduced and oxidized Cc, respectively). Assays were performed on a Cary UV-visible spectrophotometer at room temperature in sodium phosphate buffer (100 mM, pH 7.4). Both Asc- and Cc²⁺-dependent peroxidase activities were assayed for the comparative analysis of wt TcAPx-CcP and W233F mutant. The oxidation of Asc was monitored at 290 nm (ε₂₉₀ = 2,800 M⁻¹⋅cm⁻¹) using TcAPx-CcP and W233F (0–1 μM) in the presence of H₂O₂ (40 μM) and Asc (200 μM). The Cc²⁺ oxidation was measured at 550 nm (ε₅₅₀ = 21,000 M⁻¹⋅cm⁻¹) using TcAPx-CcP (0.05–0.3 μM) and W233F (0.1–1 μM) in the presence of H₂O₂ (0–40 μM) and Cc²⁺ (50 μM). For the inactivation studies, wt TcAPx-CcP (2 μM) was mixed at 25 °C with H₂O₂ (8 μM), and aliquots were taken at different times (0–40 min). The activity was assayed as above using TcAPx-CcP (0.1 μM), Cc²⁺ (50 μM), and H₂O₂ (20 μM). For the bisubstrate kinetics analysis of the Cc²⁺-dependent peroxidase activity of TcAPx-CcP, the assay was performed with increasing concentrations of H₂O₂ (0–50 μM) at four different concentrations of Cc²⁺ (4, 10, 25, and 60 μM). From a double reciprocal plot (1/V₀ versus 1/[Cc²⁺]), the apparent maximum velocity (V′_max) was determined at each concentration of Cc²⁺. From the secondary plot 1/V′_max versus 1/[Cc²⁺], real values for V_max and K_m (Cc²⁺), respectively, were determined. The k (s⁻¹) value was obtained from the V_max value of substrate oxidation and considering the entire enzyme as the oxoheme complex intermediate, k_cat = V_max/[E]. A similar experimental approach was used for the determination of kinetic parameters for Asc (SI Materials and Methods and Fig. S2). For determination of the rate constant of the H₂O₂-dependent oxidation of TcAPx-CcP, the enzyme (2 μM) was rapidly mixed (using a stopped-flow spectrofluorometer) with H₂O₂ (0–12 μM), and the change in absorbance at 390 nm due to enzyme oxidation was recorded. Data were fitted to biexponential curves. The observed rate constant values (k_obs) for the fast and slow processes were plotted as a function of H₂O₂ concentration, and the second-order rate constants were determined from the slope of the plots.

EPR-Spin Trapping and Immunospin Trapping Studies of wt and W233F TcAPx-CcP.

EPR spectra of Asc-free TcAPx-CcP, W233F mutant, and NEM-treated enzymes were recorded at room temperature (25 °C) on a MiniScope MS400 (Magnetech Instruments). For EPR experiments, the reaction mixture contained: wt or W233F (60 μM), DBNBS (10 mM) or DMPO (100 mM), H₂O₂ (60 μM), and DTPA (0.1 mM) in phosphate buffer (100 mM, pH 7.4). Immediately after oxidant addition, samples were transferred to a 100-μL flat cell and the spectra recorded within 1 min (one-spectrum acquisition). Instrumental conditions were as follows: microwave power, 20 mW; modulation amplitude, 2.5 G; time constant, 0.2 s; and scan rate, 1.67 G/s. For immunospin trapping, wt, W233F (10 μM), and NEM-treated enzymes were exposed to H₂O₂ (0–30 μM) in the presence of DMPO (100 mM). After treatment, proteins were subjected to SDS/PAGE, transferred to nitrocellulose membranes, and blocked in PBS (50 mM, pH 7.4) containing dry milk (5% wt/vol) for 1 h. Protein-DMPO adducts were detected using a chicken-derived polyclonal anti–DMPO-nitrone primary antibody (kindly provided by Ronald Mason, National Institutes of Environmental Health Sciences, Research Triangle Park, NC) and rabbit polyclonal anti-TcAPx-CcP (1:2,000 dilution in PBS) containing Tween-20 (0.1% vol/vol) and BSA (4% vol/vol) as previously described (5). Membranes were washed and then probed for 1 h with anti-chicken IgG (IR Dye-800; Licor Bioscience) and anti-rabbit IgG (IR Dye-680; Licor Bioscience) in PBS containing Tween 20 (0.1% vol/vol). Immunoreactive proteins were visualized with an infrared fluorescence detection system (Odyssey; Licor Bioscience).

MD Simulations.

MD simulations for wt and W233F mutant TcAPx-CcP were performed. Both initial structures were generated by homology modeling using the Swiss-Model package (51), using the structure of LmP solved at 1.76 Å resolution as a template, complexed with a heme prosthetic group and two cations: calcium and potassium (PDB ID code 3RIV). Both cations were kept in the systems, and the heme group was considered in the oxoferryl state. A pK_a prediction of ionizable protein residues was made with PROPKA (52). Considering the PROPKA-computed values and examining the structures, protonation states were assigned for ionizable residues, promoting hydrogen bond formation. Because thiol/thiolate equilibrium may result critical in ET processes involving Cys residues, we also performed independent MD simulations for both protonation states of Cys²²². The systems were placed into a truncated octahedral box of three point transferable intermolecular potential (TIP3P) water model and optimized with two steps of energy minimization, each consisting of 400 cycles with a force constant of 500 kcal⋅mol⁻¹⋅Å⁻² applied over all atoms, excluding water molecules in the former and over backbone atoms only in the latter. Temperature was increased from 0 to 10 K with 10-ps constant-volume MD, a time step of 0.1 fs, and a harmonic restraint potential of 50 kcal⋅mol⁻¹⋅Å⁻² applied over all protein residues, ions, and ferryl-oxoheme. Thereafter, the temperature was raised from 10 to 300 K in 90-ps constant-volume MD with a 0.5-fs time step, applying a force constant of 10 kcal⋅mol⁻¹⋅Å⁻² to the protein backbone atoms, ions, and ferryl-oxoheme. After heating, density was equilibrated with 400-ps MD simulation at a constant temperature and pressure with a time step of 1 ps and applying a force constant of 1 kcal kcal⋅mol⁻¹⋅Å⁻² to the protein backbone atoms, ions, and ferryl-oxoheme. A Langevin thermostat was used for temperature control, whereas a Berendsen barostat was chosen to adjust the pressure to 1 bar (both regulated every 1 ps). Then, 100-ns-long MD simulations within the NTP ensemble were carried out, with a time step of 2 fs. All of the simulations were performed under periodic boundary conditions (53) using the SHAKE algorithm (54) to keep hydrogen atoms at equilibria bond lengths, and long-range electrostatics interactions were handled with Ewald sums, setting a cutoff distance of 10 Å. Parameters used for ferryl-oxoheme were taken from a study by Capece et al. (55).

ET Pathway Calculations.

The probability of ET toward the ferryl-oxoheme group was calculated using the pathways algorithm (56, 57). This method looks for the best possible path connecting the electron donor and the acceptor, and it was successfully used by our group previously (58–62). Briefly, according to the Marcus theory (63), the ET rate constant (k_ET) depends on the reaction standard free energy (ΔG°), the reorganization energy (λ), and the electronic coupling matrix (T_DA), described in the following equation:

$$k_{ET}=\frac{2\pi }{\mathit{ħ}}\hspace{0.17em}|T_{DA}{|}^2\frac{e^{-\frac{{(\mathrm{\Delta }G°+\hspace{0.17em}\lambda )}^2}{4\hspace{0.17em}\lambda \hspace{0.17em}{k}_{B}T}}}{\sqrt{4\pi \lambda \hspace{0.17em}{k}_{B}T}},$$

where ħ is the Planck constant, k_B is the Boltzmann constant, and T is the system temperature. The pathways method estimates the T_DA for a given donor–acceptor couple, which is related to the probability of ET transfer, and thus directly involved in the resulting ET rate. Further details about parameters adjustments are provided by Ferreiro et al. (61). To compute the T_DA for each case, 400 representative snapshots from the last 80 ns of MD trajectories were used. All molecular visualization and drawings were performed with the Visual Molecular Dynamics program (64).

Peptide Mapping-MS Analysis of DMPO-Adducts After Reaction with H₂O₂.

Mutant W233F (16 μM) enzymes were treated with equimolar concentrations of H₂O₂ in potassium phosphate buffer (100 mM, pH 7.4) at 25 °C in the presence of DMPO (100 mM). Excessive DMPO was removed by buffer exchange with ammonium bicarbonate buffer (50 mM, pH 8) using HiTrap desalting columns and enzyme-digested (overnight at 37 °C) with trypsin (trypsin/protein ratio of 20:1, sequencing grade; Promega). Tryptic digests were analyzed using a hybrid triple-quadruple linear ion trap mass spectrometer (QTRAP 4500; ABSciex) coupled online with a liquid chromatography system (Infinity 1260; Agilent). For the liquid chromatography-MS approach, peptides were separated on a reversed-phase column (Vydac 218TP; C18, 150 × 2.1 mm, 5 μm) and eluted with a linear gradient of acetonitrile (0.1% formic acid) from 0–45% in 90 min at a flow rate of 0.25 mL⋅min⁻¹. The electrospray voltage and declustering potential were set to 5.5 kV and 50 V, respectively, and the source temperature was 500 °C. Initial scanning was performed in the Q1 positive ion mode with the unit resolution at 1,000 Da⋅s⁻¹ using a mass range of 300–2,000 Da⋅s⁻¹. Sequences from peptides of interest were performed by HPLC-tandem MS. Enhanced product ion analysis from selected peptide ions was performed using Q3 as a linear ion trap, with an optimized collision energy of 50 V and a mass range of 50–1,500 Da at 10,000 Da⋅s⁻¹. Data were acquired using Analyst 1.6.2 software (ABSciex) and analyzed using the BioTool Kit designed for PeakView 2.1 (ABSciex).

Parasites.

T. cruzi epimastigotes (CL-Brener, wt) were cultured at 28 °C in brain heart infusion (BHI) containing heat-inactivated bovine fetal serum (10% vol/vol) as described previously (65). T. cruzi epimastigotes overexpressing TcAPx-CcP were obtained as described previously (5, 15). Transformed TcAPx-CcP (CL-Brener-pTEX-APX-9E10) overexpressing epimastigotes were cultured in BHI medium as above containing G418 (250 μg/mL; Sigma).

Immuno-EM Studies.

The specificity of the antibodies used for the immunolocalization studies was evaluated by Western blot of protein extracts (10 μg) from rat-derived cardiomyocytes (H9c2; American Type Culture Collection) and from the different parasite stages (amastigotes, trypomastigotes, and epimastigotes; Fig. S5). Cardiomyocytes were seeded in culture flasks (25 cm²) and infected as above. After 48 h of infection, an equal volume of a freshly prepared mixture of 4% paraformaldehyde (PFA; % vol/vol) and 0.4% glutaraldehyde (GA; % vol/vol) in sodium phosphate buffer (0.1 M, pH 7.4) was added to the cultures to reach a final concentration of 2% PFA and 0.2% GA (% vol/vol). The cells were fixed for 30 min at room temperature, after which the fixative was refreshed and kept at 4 °C. Samples were rinsed in sodium phosphate buffer (0.1 M, pH 7.4), and after overnight infiltration in sucrose (2.3 M), the cells were frozen in liquid nitrogen. Ultrathin cryosectioning and immunogold labeling were performed as described previously (66). Sections were picked up with a drop of a 1:1 mixture of sucrose (2.3 M) and 1.8% (vol/vol) methyl-cellulose (25 centipoise; Fluka AG), thawed, and transferred to formvar carbon-coated copper grids. The sections were subsequently immunolabeled at room temperature by floating on drops containing the diluted antibody [anti–TcAPx-CcP, anti-MPX, and antimonoclonal c-Myc9E10 antibody (Santa Cruz Biotechnology) and 10 nm of protein A gold]. After incubation, the sections were washed in distilled water, stained for 5 min with uranyl-oxalate (pH 7.0), washed again, and embedded in a mixture of 1.8% (vol/vol) methyl cellulose and 0.3% uranyl acetate at 4 °C. The sections were evaluated using a JEOL 1200CX electron microscope.

Cellular Infections.

Parasites were differentiated to the infective metacyclic stage, and metacyclic forms were purified by overnight incubation with fresh human serum as previously described (5). Metacyclic trypomastigotes were used to infect confluent Vero cells (American Type Culture Collection) at 37 °C in a 5% CO₂ atmosphere. Culture-derived trypomastigotes were used to infect macrophages (J774A.1, ATCC-TIB-67; American Type Culture Collection) and cardiomyocytes (H9c2) cultured in DMEM (Sigma) supplemented with l-glutamine (2 mM), penicillin (100 units/mL), streptomycin (100 mg/mL), and heat-inactivated FBS (10% vol/vol) at 37 °C in a 5% CO₂ atmosphere. Macrophages and/or cardiomyocytes were seeded in Lab-Tek chamber slides and infected with culture-derived trypomastigotes from wt and TcAPx-CcP overexpressers (parasite/cell ratio of 5:1) (7). Nonengulfed parasites were removed by washing twice in Dulbecco’s PBS at pH 7.4 (Sigma), and cells were further incubated for 24–96 h in DMEM at 37 °C. Infected cells were fixed in a 4% (vol/vol) fresh PFA solution in PBS for 10 min at room temperature, washed with PBS containing glycine (100 mM), and permeabilized for 5 min with 0.1% (vol/vol) Triton X-100 in PBS. The number of parasites per 100 macrophages and/or cardiomyocytes was determined by DAPI staining (5 μg/mL). Preparations were analyzed using a microscope (Nikon Eclipse TE-200) at a magnification of 1,000×, and digital photographs of infected cells were recorded. At least 2,500 cells from three independent experiments were counted. Results are expressed as the number of amastigotes per 100 cells and represent the mean of three independent experiments.

Animal Infections.

Male BALB/c or female C57BL/6 mice (7–10 wk old, respectively) were inoculated i.p. (five mice per group) with culture-derived trypomastigotes (1.5 × 10⁷ trypomastigotes) from wt (CL-Brener strain) or TcAPx-CcP overexpressers. The different susceptibility of these murine models to acute T. cruzi infection was previously characterized (67). Blood trypomastigote count (parasitemia) was assayed on blood (3 μL) drawn from the tail tips of mice, and the number of trypomastigotes per 32 fields was recorded from fresh blood in Neubauer chambers under a microscope (400× magnification).

Infection studies were approved by the Animal Ethics Committee (registration no. 070153-000119-5), registered with the Ethics Commission for the Use of Animals for Experimentation [Comisión de Ética en Uso de Animales (CEUA), Veterinary School, Uruguay], and mice were handled according to their guidelines.