Haem-containing enzymes (peroxidase and catalase) are widely distributed among prokaryotes and eukaryotes and play a vital role in H2O2 detoxification. But, to date, no haem-containing enzymatic defence against toxic H2O2 has been discovered in Leishmania species. We cloned, expressed and purified an unusual plant-like APX (ascorbate peroxidase) from Leishmania major (LmAPX) and characterized its catalytic parameters under steady-state conditions. Examination of its protein sequence indicated approx. 30–60% identity with other APXs. The N-terminal extension of LmAPX is characterized by a charged region followed by a stretch of 22 amino acids containing a transmembrane domain. To understand how the transmembrane domain influences the structure–function of LmAPX, we generated, purified and extensively characterized a variant that lacked the transmembrane domain. Eliminating the transmembrane domain had no impact on substrate-binding affinity but slowed down ascorbate oxidation and increased resistance to H2O2-dependent inactivation in the absence of electron donor by 480-fold. Spectral studies show that H2O2 can quickly oxidize the native enzyme to compound (II), which subsequently is reduced back to the native enzyme by an electron donor. In contrast, ascorbate-free transmembrane domain-containing enzyme did not react with H2O2, as revealed by the absence of compound (II) formation. Our findings suggest that the single copy LmAPX gene may play an important role in detoxification of H2O2 that is generated by endogenous processes and as a result of external influences such as the oxidative burst of infected host macrophages or during drug metabolism by Leishmania.
Under physiological conditions, APX (ascorbate peroxidase) catalyses the oxidation of ascorbate with H2O2 through the well-known peroxidative one-electron transfer mechanism . It has been established that APX catalyses these reactions through the following steps:
where H2A represents the reducing substrate (ascorbate) and A is dehydroascorbate. Compounds (I) and (II), being two and one-electron oxidation states above native ferriperoxidase respectively oxidize ascorbate (H2A) by two one-electron transfer reactions with the formation of monodehydroascorbate radical (HA•), a fairly reactive and unstable species, which is reduced back to ascorbate and dehydroascorbate (A). Presteady-state mechanistic information for oxidation of ascorbate by native and recombinant pea cytosolic APXs [2,3] and for tea APX  is available. EPR and UV–visible spectroscopic features of compound (I) are consistent with the formation of a porphyrin π-cation radical intermediate [as found in HRP (horseradish peroxidase)], and not a protein-based radical species as found in cytochrome c peroxidase [5,6]. In addition to the known activity of APXs towards ascorbate, it is well known that these enzymes are rather indiscriminate in their choice of substrate and are able to catalyse the oxidation of non-physiological (often aromatic) substrate, in some cases at rates comparable with that of ascorbate itself .
For several reasons, substrate recognition and binding in APX is more complex than it might first appear. Although NMR-derived distance constraints for binding of ascorbate to APX are consistent with the existence of two distinct binding sites, one close to the 6-propionate (γ-meso position) and the other near the δ-meso position of the haem , site-directed mutagenesis together with chemical modification experiment are indicative of a single ascorbate interaction at the haem edge in the region of Arg172, Cys32 and the haem propionates (close to the γ-meso position) . The refined atomic positions in the ascorbate-bound APX crystal structure show H-bonds between the 2′-OH and 3′-OH groups of the ascorbate and the protein (Arg172), and between the 2′-OH group of the ascorbate and the (deprotonated) haem 6-propionate . This structure also shows that the side chain of Lys30 swings in from the solvent to provide an additional H-bond to the 6-OH group of the ascorbate for stabilization of the substrate binding . Additionally, steady-state oxidation of ascorbate by some APXs does not obey the normal (hyperbolic) Michaelis–Menten kinetics, suggesting either allosteric effect, which seems unlikely , or more than one substrate-binding site , or the disproportion of monodehydroascorbate molecules to give back ascorbate and dehydroascorbate . Interestingly, oxidation of the aromatic substrate guaiacol, which is thought to bind close to the δ-meso position , shows normal Michaelis kinetics.
In plants, H2O2 is continuously produced as a by-product of photorespiration, fatty acid β-oxidation, photosynthesis and oxidative phosphorylation. The H2O2-induced oxidative damage is minimized by the concerted action of powerful antioxidant enzymes. Most important among them are catalases, which are localized to peroxisomes, glyoxysomes and mitochondria and APX, which is located in both the chloroplasts and the cytoplasm [14,15].
Parasitic protozoa of the order Kinetoplastida are the causative agents of several medically important tropical diseases including visceral (Leishmania donovani) and cutaneous (L. major) leishmaniasis. During an infective cycle of Leishmania in the vertebrate host, the parasite must survive in the rigorous oxidizing environment of the macrophage. In order to survive under such oxidative burst conditions, they must evade the toxic effects of nitric oxide (NO•), peroxynitrite (ONOO−), hydroxyl radicals (OH•), H2O2 and superoxide radicals (O2•−). However, Leishmania species use intracellular thiols , lipophosphoglycan , iron superoxide dismutase , HSP70 (heat-shock protein 70) , ovothiol A, trypanothione  and peroxidoxins  to overcome a variety of reactive oxygen and nitrogen species  during their life cycle. Unlike most eukaryotes, Leishmania lacks catalase and selenium-containing glutathione peroxidases, enzymes capable of rapidly metabolizing high levels of H2O2. Hence, the mechanism by which it withstands the toxic effects of H2O2 is still unclear. To date, no haem containing enzymatic defence against H2O2 has been identified in Leishmania. Partial genome sequencing of L. major confirms the presence of a subset containing the open reading frame that putatively codes for a protein homologous to unusual plant-like ascorbate-dependent haemoperoxidase. To understand better the structure–function aspects of the peroxidase protein, we cloned, expressed and characterized an APX-like protein from L. major, LmAPX. Our study reveals the physical and catalytic features of LmAPX, which shows marked susceptibility to H2O2 in ascorbate-depleted medium. Evidence has been presented to show that a 22 amino acid hydrophobic region present at the N-terminus of LmAPX plays an important role in controlling H2O2 susceptibility and ascorbate oxidation.
Detection of peroxidase activity in
Leishmania cell lysate
Promastigotes of L. major and L. donovani were cultured in DMEL (Dulbecco's modified Eagle's liquid) media and blood/agar media respectively at 22 °C. Promastigotes, grown up to stationary phase, were harvested by centrifugation at 6000 g for 10 min, and the pellets were resuspended in 10 ml of 50 mM Tris/HCl buffer (pH 7.5) containing 0.1 mM ascorbate, 1 mM PMSF and 1 mM protease inhibitor I and II (Roche Molecular Biochemicals, Indianapolis, IN, U.S.A.). The resuspended cells were disrupted by sonication and the lysate was centrifuged at 15000 g for 30 min. The supernatant was designated as the crude extract.
Genomic DNA isolation from
Cloning of LmAPX from genomic DNA
The sense primer1: 5′-AGGTAATGGCTGCGTAGCG (610 nt upstream) and the antisense primer2: 5′-CGTGTCCGAGGAGATACTAACG (458 nt downstream of putative APX gene) were used to amplify the desired portion (1980 bp) from L. major genomic DNA by PCR. The amplified product (1980 bp) was gel-purified by using the Qiagen kit. The coding region of full-length LmAPX, Δ12 LmAPX (12 amino acids deleted from the N-terminal sequence of LmAPX) and Δ34 LmAPX (34 amino acids deleted from the N-terminus sequence of LmAPX) were amplified by using the amplified product (1980 bp) as the template. The following sense primers: 5′-AAAGGATCCGGCACCTCGCGGCGAGCGAAAGGC, 5′-AAAAGGATCCACCGGCATCGCTGTCGGCACC and 5′-AAAAGGATCCGAGGAGCCGCCGTTCGACATC were used for the amplification of LmAPX, Δ12 LmAPX and Δ34 LmAPX respectively. The antisense primer 5′-AAAGGTACCTTAGCTCTCCGAAGCGGGTGCT was used in each case. Each of the amplified products was cloned into the BamHI and KpnI sites of the prokaryotic expression vector pTrcHisA (Invitrogen) and DNA was sequenced by using an automated DNA sequencer.
Expression and purification of proteins
pTrcHisA vector alone, the recombinant pTrcHisA/LmAPX, pTrcHisA/Δ12 LmAPX and pTrcHisA/Δ34 LmAPX vectors were used to transform Escherichia coli BL21D3 cells. Transformed cells were grown overnight in 50 ml Luria–Bertani broth containing 100 μg ml−1 ampicillin at 37 °C in a shaker. The overnight grown cultures were then inoculated into 500 ml Terrific broth (12 g of tryptone, 24 g of yeast extract, 9.4 g of dibasic potassium phosphate, 2.2 g of monobasic potassium phosphate and 4 ml of glycerol/litre of medium). When the culture reached an absorbance of 0.8 at 600 nm, 0.5 mM isopropyl β-D-thiogalactoside and 0.4 mM δ-aminolevulinic acid were added, and the bacteria were further grown for 18 h at 22 °C. Cells were harvested by centrifugation at 6000 g for 10 min, and the pellets were resuspended in 10 ml of 50 mM phosphate buffer (pH 7.5) containing 0.1 mM ascorbate, 150 mM NaCl, 1 mM PMSF, 1 mM protease inhibitors I and II (Roche Molecular Biochemicals) and 1 mg/ml lysozyme. The resuspended solution was kept for 1 h at 4 °C and then the cells were broken by sonication. The lysate was centrifuged at 15000 g for 30 min. The supernatant, designated as the crude extract, was loaded on to an Ni2+-nitrilotriacetate column. After loading the crude extract, the column was washed with washing buffer (50 mM phosphate buffer, pH 7.5, containing 0.1 mM ascorbate and 1 mM PMSF; 10 column volumes) and then washed further by 50 mM phosphate buffer (pH 5.25; 10 column volumes). The pure enzyme was eluted with either 50 mM phosphate or acetate buffer (pH 4.0) and then dialysed three times against 0.1 mM ascorbate and 50 mM phosphate buffer (pH 7.5) to adjust neutral pH. It is worth mentioning here that the enzymes eluted with either phosphate or acetate behaved identically as far as the optical spectra and kinetic parameters are concerned. In the case of ascorbate-free LmAPX preparation, whole purification was carried out in the absence of ascorbate.
Protein concentration determination
The haem was identified and quantified by the pyridine–haemochrome method . The molar absorption coefficient of LmAPX at 408 nm was 101 mM−1·cm−1.
Native forms of the Δ12 LmAPX and Δ34 LmAPX were analysed by gel-filtration chromatography using a Protein Pak SW 300 column (Nihon Waters, Japan, Tokyo) in an HPLC system (Waters), preequilibrated with 50 mM phosphate buffer (pH 6.5) and 250 mM NaCl. The column was run at room temperature with a flow rate of 0.5 ml/min. The absorbance was monitored at 280 nm. The column was calibrated using thyroglobulin (660 kDa), dimer BSA (132 kDa), ovalbumin (43 kDa) and RNAase (12.5 kDa).
Binding and kinetic measurement of LmAPX
All spectral studies were performed on a Shimadzu UV-1601 spectrophotometer using quartz cells of 1 cm light path. The difference spectrum of enzyme–ligand versus enzyme was obtained as described previously [26,27].
Compound (II) spectrum of LmAPX is stable and could be detected by a conventional spectrophotometer [compound (I) in LmAPX is a short-lived species]. Pseudo-first-order rate constants for compound (II) reduction (kobs) of LmAPX were obtained at 424 nm by the mixing of enzyme (1.0 μM) in the presence of various concentrations of ascorbate with equimolar amounts of H2O2. Monophasic transient traces were fitted to a single exponential process to obtain pseudo-first-order rate constants.
Inactivation of LmAPX by H2O2
The rate of inactivation of LmAPX by H2O2 was measured by preincubation of LmAPX with different concentrations of H2O2 in 50 mM phosphate buffer (pH 7.5) in the absence of electron donors. After the addition of H2O2, at various time intervals the incubation mixture was transferred to a cuvette containing 1 ml of assay mixture of 144 nM enzyme, 50 mM phosphate buffer (pH 7.5), 20 mM guaiacol and 0.3 mM H2O2. During the substrate protection study against inactivation, a high concentration of electron donor was added to the preincubation mixture containing the enzyme before the addition of H2O2.
Primary structure analysis
The 0.9 kb genomic DNA fragment, coding for a 303 amino acid long LmAPX possesses 62.73% identity and 86.79% similarity with Trypanosoma cruzi APX, 35% identity and 60.72% similarity with pea cytosolic APX, 34% identity and 61.39% similarity with soya-bean cytosolic APX, and 31.35% identity and 64.35% similarity with chloroplast APX [30–33]. Similarity of the primary sequences of LmAPX with that of chloroplast, T. cruzi, pea and soya-bean cytosolic APXs indicate that LmAPX is related to the class I group of haemoperoxidases. However, the sequence identities between LmAPX with other classes of superfamily are found to be less than 18% (results not shown). The charged residues in the dimer interface of the pea APX are not similar to LmAPX. Figure 1(A) also provides a sequence alignment of the proximal cation-binding loop in various APXs. LmAPX has the side chain Thr209 residue instead of aspartic acid, indicating that the proximal cation-binding loop is very similar to the K+-binding site of APX . The other feature that distinguishes LmAPX from the plant enzymes is the presence of a sequence insertion, of unknown function, near the C-terminus containing charged amino acids. A notable feature that differentiates LmAPX from the cytosolic APX is its N-terminal extended portion. TargetP V1.0 prediction  result indicates that in LmAPX, the extended region of the N-terminal sequence codes for a positive charged region (12 amino acids) which is followed by a stretch of 22 amino acids containing a hydrophobic region that has the potential to form a transmembrane domain (Figure 1B). Sequence analysis predicts that the overall structural elements of LmAPX are quite similar to the cytosolic APX. Furthermore, Swiss-Model protein modelling also predicts that the entire LmAPX sequence is highly compatible with structures of the distal as well as proximal site of haem in cytosolic APX protein (Figure 2). All of the key residues on the proximal site of the haem are conserved between LmAPX and APX: His192, Trp208 and Asp253 in LmAPX are superimposed with the corresponding His163, Trp179 and Asp208. The key distal residues of LmAPX (His68, Trp67 and Arg64) are also found to be in identical position with respect to the distal site residues of APX (His42, Trp41 and Arg38). The most significant difference is that of the Phe201 residue in LmAPX, which substitutes Arg172 of the APX, the crucial residue for the ascorbate binding as well as oxidation. It is worth mentioning that in contrast with plant APX, the parasite-specific T. cruzi APX, which is more close to LmAPX, also lacks this arginine residue at the ascorbate-binding site .
Sequence alignment of APXs from different sources
Ribbon structural model showing the position of Lys55 and Phe201 in LmAPX relative to the ascorbate binding Lys30 and Arg172 in APX
Physical and spectral characteristics of LmAPX
Recombinant N-terminal histidine-tagged LmAPX protein is overexpressed in E. coli cells by induction with isopropyl β-D-thiogalactoside and δ-aminolevulinic acid, but total protein goes into inclusion bodies and cannot be purified in soluble form (results not shown). However, Δ12 LmAPX and Δ34 LmAPX are both expressed as active and soluble forms. To investigate the native state molecular mass of both the variants of LmAPX, purified proteins were subjected to gel filtration using HPLC. Results shown in Figure 3(A) indicate that Δ34 LmAPX eluted at a position expected of monomeric enzyme (33 kDa), whereas the elution pattern of Δ12 LmAPX showed that 70% of the protein were eluted as monomeric enzyme (35.5 kDa), 15% eluted as a dimeric state (71 kDa) and the rest 15% of the protein eluted at a wide range of molecular mass (oligomerization of the protein). This result suggests that the tendency of oligomerization of the Δ12 LmAPX may be due to hydrophobicity of the transmembrane domain. The Δ34 LmAPX is a monomer instead of a dimer as observed with cytosolic APX probably because the charged residues in the dimer interface of the pea APX are absent from LmAPX. The Δ12 LmAPX and Δ34 LmAPX migrated on denaturing SDS/polyacrylamide gel at a molecular mass of 35.5 and 33 kDa respectively, identical with their calculated molecular mass (Figure 3A, inset). The UV–visible spectra of ascorbate-free Δ34 LmAPX shows the presence of a Soret peak at 408 nm with secondary peaks at approx. 500 and 640 nm (Figure 3B). Addition of a 5 molar excess of H2O2 to the resting state of Δ34 LmAPX produces oxyferryl compound (II) [oxyferryl compound (II) is produced via compound (I), a very short-lived ferryl heam iron with porphyrin π cation radical ] absorbing at 420 nm at the Soret region with visible peaks at 532 and 560 nm. The calculated purity number Rz (A408/A280) for Δ34 LmAPX and Δ12 LmAPX were 0.98 (Figure 3B) and 0.9 (results not shown) respectively.
The size-exclusion chromatography, SDS-polyacrylamide gel electrophoresis and light absorbance spectra recorded after purification of LmAPX
To characterize further the effects of transmembrane domain on LmAPX catalysis, turnover of other common electron donors were investigated. Table 1 shows the steady-state data of ascorbate, guaiacol and iodide oxidation of both Δ12 LmAPX and Δ34 LmAPX. In the case of ascorbate oxidation, both the truncated forms of LmAPX displayed non-Michaelis–Menten kinetics under steady-state condition. A linear dependence on substrate concentration was observed and saturation was not detected at any accessible concentration suggesting that the binding of ascorbate is weak probably due to the absence of Arg172. The full-kinetic profile of both mutants could not be generated and the direct determination of the kinetic parameters (Km, kcat) was not possible. Hence specific activities are calculated at 0.5 mM ascorbate concentration (a similar condition was used to measure the specific activity of pea cytosolic APX ). Table 1 showed that the transmembrane domain containing LmAPX has 5-fold higher activity compared with Δ34 LmAPX. The rate of ascorbate oxidation of Δ12 LmAPX was approx. 15-fold lower as compared with the pea cytosolic APX [2,3,9] but similar to the T. cruzi APX . The fact that variant LmAPX and TcAPX proteins each exhibit a lower ascorbate oxidation compared with pea APX supports the idea that the absence of the Arg172 side chain in parasite-specific APX may disrupt their ability to utilize ascorbate as a source of electron. As opposed to ascorbate oxidation both proteins exhibited Michaelis–Menten-type kinetics in the presence of guaiacol and iodide. From the Lineweaver–Burk plot, the calculated Km values of both Δ12 LmAPX and Δ34 LmAPX for guaiacol, iodide and H2O2 are very similar to each other (Table 1) indicating that the substrate affinity of the recombinant LmAPX is unaltered after removing the transmembrane domain. When Δ12 LmAPX and Δ34 LmAPX were purified from the overexpression system in the absence of ascorbate, the Δ12 LmAPX enzyme showed low activity and high Km value for H2O2, whereas Δ34 LmAPX was found to be highly active (Table 1) indicating that ascorbate is essential during the purification of Δ12 LmAPX for its stabilization. This result is consistent with the previous report where it has been shown that chloroplast APX (containing a transmembrane domain) is a labile enzyme in an ascorbate-depleted medium when compared with cytosolic APX .
|.||Ascorbate* .||H2O2† .||Guaiacol† .||Iodide† .|
|Enzyme .||Specific activity (units/mg) .||Km (μM) .||Km (mM) .||kcat (min−1) .||Km (mM) .||kcat (min−1) .|
|.||Ascorbate* .||H2O2† .||Guaiacol† .||Iodide† .|
|Enzyme .||Specific activity (units/mg) .||Km (μM) .||Km (mM) .||kcat (min−1) .||Km (mM) .||kcat (min−1) .|
H2O2-dependent inactivation of LmAPX
Chloroplast APX is known to be distinct from cytosolic APX with respect to H2O2 susceptibility in the absence of an electron donor . To ascertain whether Δ12 LmAPX (like chloroplast APX) and Δ34 LmAPX (like cytosolic APX) displayed similar properties, both the enzymes were preincubated with different concentrations of H2O2 that resulted in concentration and time-dependent irreversible inactivation of the enzyme following pseudo-first-order kinetics (Figures 4A and 4B). When kobs values obtained from the slope of each line were plotted against H2O2 concentration, a straight line (Figures 4A and 4B inset) was obtained from which a second-order rate constant was calculated to be 6000 M−1·min−1 for Δ12 LmAPX and 16.7 M−1·min−1 for Δ34 LmAPX at 30 °C. The result shows that Δ12 LmAPX is 480-fold more susceptible to H2O2-dependent inactivation as compared with Δ34 LmAPX. H2O2-dependent inactivation of both enzymes were protected by a high concentration of ascorbate or aromatic donor (guaiacol) indicating that electron donors scavenged the preincubating H2O2 by reducing the compound (I) and (II) back to the native state.
Kinetics of the inactivation of LmAPX by H2O2
Spectral properties of Δ12 LmAPX and Δ34 LmAPX with H2O2
Figure 5(A) shows the change in resting state of ascorbate-free Δ12 LmAPX spectrum when 20 molar excess of H2O2 was added at 25 °C. In contrast with other peroxidases, the initial spectrum was unaltered in the region of Soret and visible peaks of enzyme at 408, approx. 500 and 640 nm. This result indicated that ascorbate-free Δ12 LmAPX could not react with H2O2. Thus our spectral observation is consistent with the kinetic result where ascorbate-free Δ12 LmAPX was found to be catalytically inactive (Table 1). Figure 5(B) shows the change in the resting state of ascorbate-bound Δ12 LmAPX spectrum when 5-fold molar equivalent of H2O2 was added. The initial spectrum showed a higher absorbance and red-shifted to 420 nm at the Soret region with simultaneous appearance of a double hump at 532 and 560 nm in the visible region [1,37,38]. This initial spectrum is reminiscent of compound (II) of the other peroxidase. This compound (II) species returns to the ferric state of the enzyme within 30 s. These results suggest that ascorbate-bound Δ12 LmAPX is in active form, which is able to react with H2O2 to form compound (II) [via compound (I)] which subsequently reacts with ascorbate (electron donor) to form the native enzyme. The rate of compound (II) reduction by ascorbate for Δ12 LmAPX is very slow compared with pea APX [2,3] and follows monophasic kinetics (Figure 5C). The compound (II) reduction is linearly dependent on ascorbate concentration (Figure 5C). The second-order rate constant derived from this linear dependence was 6.01±0.03×103 M−1·s−1. In contrast with ascorbate-free Δ12 LmAPX, the ascorbate-free Δ34 LmAPX enzyme was found to be in an active state and exhibited peaks at 408, approx. 500 and 640 nm. When five molar excess of H2O2 was added to the ascorbate-free Δ34, the native enzyme immediately formed α/β bands at 532 and 560 nm with shifting of Soret band from 408 to 420 nm, and simultaneous decrease of absorbance at 640 nm. The ascorbate-free Δ34 LmAPX-oxidized species [compound (II)] is found to be stable for several hours. The red shift in the Soret band and double hump at the visible region of this species is very similar to the well-known peroxidase compound (II). Upon addition of 10 μM ascorbate to the compound (II) of ascorbate-free Δ34 LmAPX, 34 amino acids deleted from N-terminus sequence of LmAPX; the enzyme intermediate returned back to a native state within 60 s (Figure 5D, broken line), indicating that the ascorbate was oxidized by the LmAPX–H2O2 intermediate. A similar monophasic behaviour was observed for Δ34 LmAPX with a second-order rate constant of 1.57±0.08×103 M−1·s−1 (Figure 5E). This rate constant is approx. 4-fold lower than that of Δ12 LmAPX, which is consistent with our ascorbate oxidation data. The ascorbate-free Δ34 LmAPX was further studied to find out why the enzyme was inactivated in the presence of high concentrations of H2O2. Figure 5(F) showed that the Soret haem absorbance of the compound (II) was gradually decreasing with increasing intervals of time for high concentrations of H2O2 with the visible peak unaltered during decreasing Soret absorbance at 420 nm. This modified enzyme could not be reduced back to the native state by adding electron donors (results not shown), indicating that compound (II) is converted into the inactive species in the presence of a high concentration of H2O2. Thus our spectral observation is in line with the kinetic result where Δ12 LmAPX and Δ34 LmAPX were found to be susceptible to H2O2-dependent irreversible inactivation.
Spectral studies of LmAPX
As binding of ascorbate is a prerequisite for oxidation [9,10], the interaction of ascorbate with both Δ12 and Δ34 LmAPX were studied by optical difference spectroscopy [26,27] in the presence or absence of guaiacol. The binding of ascorbate gave a characteristic difference spectrum of the LmAPX–ascorbate complex versus LmAPX, having a maximum at 433 nm and a minimum at 413 nm (Figure 6A for Δ12 LmAPX and Figure 6B for Δ34 LmAPX). The apparent equilibrium dissociation constant, Kd, for the LmAPX–ascorbate complex as calculated from the plot of 1/ΔA versus 1/[ascorbate] (Figure 6C for Δ12 LmAPX and Figure 6D for Δ34 LmAPX) was 33 μM. When binding was studied in the presence of guaiacol, ascorbate also interacted with the LmAPX–guaiacol complex, showing a similar characteristic difference spectra at the Soret region; however, the nature of the binding was found to be competitive. This is further substantiated by the finding (Table 2) that the binding of ascorbate to both forms of LmAPX (Kd=33 μM) is significantly increased (Kd=100 μM for Δ12 and Kd=400 μM for Δ34 LmAPX) in the presence of guaiacol. This indicates that ascorbate interacts at a site close to the guaiacol (aromatic donor) binding site.
Difference spectra of LmAPX–ascorbate complexes
|.||Spectrum of complex (nm) .||.||.|
|Enzyme .||Minimum .||Maximum .||Kd (μM) .||Δϵpeak-trough (mM−1·cm−1) .|
|.||Spectrum of complex (nm) .||.||.|
|Enzyme .||Minimum .||Maximum .||Kd (μM) .||Δϵpeak-trough (mM−1·cm−1) .|
Intracellular pathogen Leishmania possesses a strong antioxidant defence against the oxidants released by the macrophage under oxidative burst condition, for its survival and replication. In most pathogenic organisms, haemoproteins, e.g. catalase or peroxidase, play a major role in detoxification of H2O2, an oxidant. But to date, this type of enzymatic machinery has not been reported in Leishmania. This study for the first time describes the properties of an unusual plant-like haem-containing APX (LmAPX) from L. major. The sequence homology of LmAPX suggests that it belongs to a broad family of peroxidase evolutionarily related to class I peroxidase . The sequence alignment studies detail the marked similarities among LmAPX with other APXs in the proximal/distal sides of the haem. In fact, His68, Trp67 and Arg64 on the distal haem side are found to be absolutely conserved. These distal histidine and arginine residues have been implied to work in concert in an acid/base catalysed cleavage of the peroxide O–O bond [40,41]. Swiss-Model protein modelling also predicts that the amino acid residues at the proximal haem side of LmAPX, Asp253, proximal ligand His192 and Trp208 are identical with APX. Analysis of the crystal structure of a ascorbate and aromatic donor-bound APX, mutational and chemical modification studies have identified two distinct binding sites, the first of which contains a negative charged ascorbate-binding domain near the exposed γ-site of haem, whereas the second is thought to be a neutral aromatic donor-binding domain near the exposed δ-site of haem [2,9,42]. When the ascorbate-binding site of LmAPX is compared with that of APX, Lys55 in LmAPX was found to be identical with Lys30 of APX but Phe201 appears to be substituted for Arg172 that reportedly interacts with the 2′-OH and 3′-OH groups of ascorbate. Indeed, Arg172 is absolutely conserved among all other APXs examined so far except in TcAPX, which is the only known APX in the trypanosomatid family. Similar to LmAPX, TcAPX also lacks this residue where an asparagine residue (Asn216) occupies the homologous position. The specific activity of ascorbate and guaiacol oxidation in LmAPX is lower compared with the plant APX, probably due to slower rate of electron transfer from electron donor to compound (II). Compound (I) in ascorbate-bound LmAPX is short-lived and its spectrum decays too rapidly to capture by conventional spectrophotometer. Hence the reduction of compound (II) to ferric state is a rate-limiting step in the LmAPX catalysis. The rate of compound (II) reduction by ascorbate for both variants of LmAPX are monophasic and are slower than that obtained from the kobs-fast of pea APX compound (II) reduction . These results correlate well with what is seen in the steady-state kinetics data. Difference spectroscopic studies show that despite the absence of the crucial arginine residue at a position homologous to Arg172 of APX, the affinity (Kd) of LmAPX for ascorbate is still in the micromolar range, suggesting that other residues might be involved in ascorbate binding. It is possible that APXs from lower eukaryotes (TcAPX, LmAPX) probably utilizes some distinct ascorbate-binding mechanism.
The primary sequence of this peroxidase (LmAPX) was found to be comprised of a hydrophobic transmembrane motif at the N-terminus. It is well established that the general function of N-terminal hydrophobic transmembrane motif is to anchor the proteins to the membrane of different organelles including Golgi, ER, vacuoles, synaptic vesicles, mitochondria and peroxisomes . Apart from this, the novel aspects that set apart Δ12 LmAPX from its transmembrane domain deleted counterpart (Δ34 LmAPX) are (i) Δ12 LmAPX has a higher tendency for oligomerization than Δ34 LmAPX at the native state; (ii) the enzyme Δ12 LmAPX in an ascorbate-free system is catalytically inactive, whereas ascorbate-free Δ34 LmAPX is catalytically active; (iii) ascorbate oxidation rates of Δ12 LmAPX are greater than Δ34 LmAPX at physiological pH in an ascorbate-supplemented system; and (iv) the second-order rate constant of Δ12 LmAPX inactivation by H2O2 is 480-fold higher compared with Δ34 LmAPX in the absence of electron donors. These properties of the N-terminal transmembrane domain of this enzyme provide a unique perspective on LmAPX structure–function.
Further insight into this unexpected difference between Δ12 and Δ34 LmAPX was gained from the formation of the enzyme intermediates when the peroxidase reaction cycles were analysed spectrometrically. The reason for the lack of measurable peroxidase activity in ascorbate-free Δ12 LmAPX could be that the recombinant protein, despite its solubility, is incorrectly folded. However, a similar lack of peroxidase activity was observed in the native ascorbate-free chloroplast APX, which is purified from plants . In general, several active enzymes are very unstable in the absence of substrate because the active site of substrate binding was found to be improperly folded. But the ascorbate-free Δ12 LmAPX enzyme can bind ascorbate with similar affinity as compared with Δ34 LmAPX, which rules out the possibility of incorrect folding at the ascorbate-binding site in ascorbate-free Δ12 LmAPX during purification. The spectral and kinetic evidence strongly support the view that the ascorbate-free Δ12 LmAPX enzyme is catalytically inactive due to its incapability of reaction with H2O2. This is probably because the H2O2 entry channel of ascorbate-free Δ12 LmAPX was hindered by the transmembrane domain.
Using previously established methods [43,44], we made an effort to examine the reactions of LmAPX with H2O2 in the absence of electron donors by kinetic analysis of their inactivation reactions, compound (III) and P670 species formation. It was found that H2O2 acted as a mechanism-based (suicide) inactivator in APX ; although, important differences were noticed between these enzymes. For HRP, it was established that a large stoichiometric excess of H2O2 was required for inactivation , whereas APX was extremely sensitive to inactivation. Although the Δ12 LmAPX is more sensitive to H2O2-dependent inactivation compared with Δ34 LmAPX, since both of them share identical active-site residues in catalysis, we suggest a similar overall mechanism involved in inactivation for both the enzymes. This inactivation process may be connected to the spontaneous reduction of compound (II) to an inactive species [compound (II)-like] followed by a decrease in haem Soret spectra. The compound P670-like species, which is formed in HRP-C under a high concentration of H2O2 , is not detected in the inactive state of both enzymes. No distinct spectral shift was observed for inactive species. A similar phenomenon occurs in both APX  and HRP-A2, where a P670-like species is not detected . This was probably because the P670 species of LmAPX is inherently unstable and difficult to detect. The decreasing haem Soret spectra indicate that H2O2-dependent haem degradation is occurring in ascorbate-free-LmAPX.
In view of the fact that ascorbate leads to Soret spectral changes in LmAPX, it is logical to conclude that ascorbate binds near the exposed haem edge. However, the non-saturation kinetics of ascorbate peroxidation does not correlate with the calculated Kd of ascorbate binding, which is consistent with a previously reported observation . Therefore the spectrally derived ascorbate Kd value of native LmAPX is different from that for kinetically active enzyme–substrate complexes under steady-state conditions. It has been predicted from ascorbate-dependent compound (II) reduction studies (biphasic rate constant at high ascorbate concentration) that APX has two competent ascorbate-binding sites for electron transfer including high-affinity (near the γ-haem edge) and low-affinity (δ-meso edge) binding sites . Since we did not observe biphasic reduction of compound (II) with ascorbate, the possibility of multiple ascorbate-binding sites may be ruled out at least in our case. Interestingly, our binding studies indicate that ascorbate interacts at the haem edge as an electron donor since it competes with guaiacol. An alternative possibility might be that the binding of guaiacol to the native LmAPX perturbs the conformation of its ascorbate-binding site leading to lowered affinity of the site for the ascorbate (Kd) observed as apparent competition between ascorbate and guaiacol. The unambiguous identification of the actual ascorbate-binding site should, however, wait until the X-ray crystal structure of LmAPX is solved.
In various stages of its life cycle, the Leishmania species may come in contact with H2O2 as a result of direct stimulation of the macrophage respiratory burst [47,48]. This could occur in vivo during initial infection of promastigotes or in passage from one macrophage to another for amastigotes [47,48]. In the parasite, the reactive oxygen species are generated by endogenous processes and as a result of external influences such as host immune responses and drug metabolism [47,49]. In 1985, it was reported that the Leishmania amastigotes can scavenge a large amount of H2O2 . The removal of H2O2 by amastigotes was markedly inhibited by aminotriazole or sodium azide, which is an inhibitor of haem-containing enzymes, e.g. catalase or peroxidase . Our preliminary results suggested that the Leishmania cell lysate has peroxidase activity (S. Adak, unpublished work). In the absence of catalase, the single copy APX gene may play a vital role in protecting this parasite against oxidative damage. This unusual LmAPX could thus be the fundamentals of a rational approach to the design and discovery of drugs against Leishmania infections.
We thank Dr R.K. Banerjee for critically reviewing this paper before submission, A. Chakraborty for assisting in genomic DNA purification, R. Datta and B. Sen for helpful discussions, S. Ghosh for excellent technical assistance and also all the members of L. major genome project for co-operation. This work was supported by CSIR Network Project SMM003, Government of India.