Cloning and characterization of a leucyl aminopeptidase from three pathogenic Leishmania species.

Aminopeptidases are emerging as exciting novel drug targets and vaccine candidates in parasitic infections. In this study, we describe for the first time an aminopeptidase from three highly pathogenic Leishmania species. Intronless genes encoding a leucyl aminopeptidase (lap) were cloned from Leishmania amazonensis, Leishmania donovani, and Leishmania major, which encoded 60-kDa proteins that displayed homology to leucyl aminopeptidases from Gram-negative bacteria, plants, and mammals. The lap genes were present as a single copy in each genome, and lap mRNA was detected by reverse transcription-PCR in all life-cycle stages of L. amazonensis. Lap assembled into catalytically competent 360-kDa hexamers and demonstrated potent amidolytic activity against synthetic aminopeptidase substrates containing leucine, methionine, and cysteine residues, representing the most restricted substrate specificity of any leucyl aminopeptidase described to date. Optimal activity was observed against L-leucyl-7-amido-4-methylcoumarin (k(cat)/K(m) approximately 63 s(-1) x mm(-1)) with a pH optimum of 8.5. Leishmania Lap activity was inhibited by metal ion chelators and enhanced by divalent manganese, cobalt, and nickel cations, although only zinc was detected in the purified Lap by inductively coupled plasma atomic emission spectroscopy, indicating that zinc is the natural Lap cofactor. Activity was potently inhibited by bestatin and apstatin in a slow binding competitive fashion, with K(i)* values of 3 and 44 nm, respectively. Actinonin was a tight binding competitive inhibitor (K(i) approximately 1 nm), whereas arphamenine A (K(i) approximately 70 microm) and L-leucinol (K(i) approximately 100 microm) were non-tight binding competitive inhibitors. Lap was not secreted by Leishmania in vitro and was localized to the parasite cytosol.

Protozoan parasites of the genus Leishmania cause visceral, cutaneous, and mucosal diseases in humans, collectively referred to as leishmaniasis. Leishmaniasis is prevalent in 88 countries, with 12 million people currently infected, a further 350 million at risk, and 2 million new cases reported per year.
No vaccine exists, and therapies are inadequate (1). There is a pressing need for the identification of novel leishmanial virulence factors, drug targets, and vaccines to improve our understanding, prevention, and treatment of leishmaniasis.
The peptidases of parasitic protozoans (for review, see Ref. 2) are becoming increasingly important as virulence factors, drug targets, and vaccine candidates in parasitic infections. However, only three peptidases from Leishmania, lysosomal cysteine peptidases (3), a cell-surface metallopeptidase (4), and a cytosolic serine oligopeptidase (5) have received attention.
Aminopeptidases, which catalyze the removal of N-terminal amino acid residues from peptides and proteins (6), are emerging as novel and exciting anti-parasite targets. Vaccination of sheep with Fasciola leucyl aminopeptidase (Lap) 1 elicited high anti-Lap titers that conferred protection against fascioliasis and fascioliasis-related liver damage (7). Synthetic broad-spectrum aminopeptidase inhibitors (8) and 1,2-aminoalcohol inhibitors of Lap (9) have shown promise as drugs against malaria, and the aminopeptidase inhibitor arphamenine A has activity against Trypanosoma brucei, a kinetoplastid protozoan that is a close relative of Leishmania (10).
Despite these observations, no aminopeptidase has ever been studied in any kinetoplastid protozoan parasite, including Leishmania. To address this, we report here the cloning, genetic analysis, and biochemical characterization of a novel aminopeptidase from three highly pathogenic Leishmania species. This aminopeptidase, which is responsible for the bulk of soluble leucyl aminopeptidase activity in Leishmania extracts, appears to represent the most phylogenetically distant branch of the M17 family of leucyl aminopeptidases (EC 3.4.11.1). Leishmania Laps exhibit the most restricted substrate specificity of any Lap described to date, confining activity almost exclusively to three amino acid residues, leucine, cysteine, and methionine. In addition to providing important new data on the enzymology of this interesting family of peptidases, the data we present here considerably expand our knowledge of the peptidolytic capacity of Leishmania, which thus far has been limited to three peptidases, by introducing a newly identified peptidase from these parasites. These data will also serve as a solid foundation for subsequent studies on the potential of this aminopeptidase as a drug target and vaccine candidate in Leishmania infections. . The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Parasite and Bacterial Strains-Virulent
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AF424691, AF424692, and AF424693.
Gene copy number was determined by Southern blot. The lap ORFs were excised from pNA151, pNA159, and pNA161 and served as templates in random prime labeling reactions (Rediprime TM II system, Amersham Biosciences) with 50 Ci of [␣-32 P]dCTP to generate 32 Plabeled lap DNA probes. Endonuclease-digested Leishmania genomic DNA (5-7.5 g) was resolved on a 0.7% (m/v) agarose gel, transferred to a Hybond Nϩ nylon membrane (Amersham Biosciences), and probed with the 32 P-labeled DNA lap probes. Blots were washed to a final stringency of 0.1ϫ SSC (1ϫ SSC ϭ 0.15 M NaCl and 0.015 M sodium citrate), 0.1% (m/v) SDS (60°C, 20 min; (5)), followed by autoradiography at Ϫ70°C for 16 h. Stage-specific lap expression was determined by reverse transcription (RT)-PCR. Total RNA from 3 ϫ 10 8 L. amazonensis amastigotes and log-(density Ϸ 1 ϫ 10 6 cells/ml) and stationary-phase (density Ϸ 5 ϫ 10 7 cells/ml) promastigotes was isolated with TRIzol ® reagent (Invitrogen) per the manufacturer's instructions. Poly(A) mRNA was isolated from total RNA with an Oligotex™ mRNA mini kit (Qiagen). RT was undertaken with ThermoScript™ reverse transcriptase (Invitrogen) per the manufacturer's instructions, primed with 2.5 M oligo(dT) 20 containing 500 ng of mRNA template. PCR reactions, carried out as described above, contained forward (5Ј-ATG CTT CGT CGC GTG CTG TCT CGC GG-3Ј) and reverse (5Ј-TTA AGG CTT GTT GTG GCG CAG GAA GT-3Ј) primers and 2 l of cDNA generated in the RT reaction (or 2 l of an RT reaction to which no reverse transcriptase had been added as negative control).
Substrate specificity of recombinant Lap was determined using fluorogenic substrates by preincubation of Lap (1 nM, 37°C, 1 min) in 50 mM Tris-HCl, 0.5 mM MnCl 2 , pH 8, followed by the addition of substrate. Initial steady-state velocity was determined by continuous assay for a range of substrate concentrations in a Hitachi F-2000 spectrofluorimeter ( ex ϭ 370 nm, ex ϭ 460 nm for AMC and ex ϭ 340 nm, ex ϭ 410 nm for ␤-naphthylamine), and a Amersham Biosciences Ultrospec 2000 at 405 nm for para-nitroaniline. The K m and V max were determined by hyperbolic regression of the kinetic data using Hyper 1.01 (obtained from Dr. J. S. Easterby, University of Liverpool, UK). The k cat was determined from k cat ϭ V max /[E] 0 , where [E] 0 represents the concentration of Lap monomers (by mass, assuming 100% activity). The [E] 0 for recombinant Lap was determined using molecular masses of 62.424, 62.337, and 62.569 kDa for the L. amazonensis, L. donovani, and L. major enzymes, respectively, which takes into account the added mass of the affinity tag.
The Lap pH profile was determined by incubating Lap (2 nM, 37°C, 5 min) in constant ionic strength acetate/Mes/Tris (AMT) buffers (50 mM acetic acid, 50 mM Mes, 100 mM Tris-HCl (12), pH 4 -11) before the addition of L-Leu-AMC. The pH stability was determined by incubating Lap (20 nM, 37°C) in these AMT buffers (1 h) before assaying residual activity of a 5-l aliquot in pH 8 AMT buffer. Ionic strength dependence was investigated by assaying Lap activity in pH 8 AMT buffer over the ionic strength range 0.01-0.5.
Inhibition of Lap by peptidase inhibitors was determined by prein-

FIG. 2. Determination of Leishmania lap copy number by Southern blot.
Southern blot of Leishmania genomic DNA digests probed with the fulllength lap gene labeled with [␣-32 P]dCTP. Blots were washed to a high final stringency followed by autoradiography. The ORF of all three lap genes contain a single restriction site for AvaI, PstI, XhoI, and AccI whereas Hind III, BglII, and SmaI do not cleave within the lap ORF. The second (lower) band generated by AccI is not clearly evident in L. amazonensis and L. donovani digests because the AccI site is located very close (203 bp) to the 5Ј end of the coding sequence, yielding a small stretch of sequence with which the probe could hybridize.

FIG. 3. Expression of Lap mRNA and protein in Leishmania.
A, RT-PCR from L. amazonensis amastigotes (Am) and log-phase (Pr (log.)) and stationary-phase (Pr (stat.)) promastigote mRNA. An RT reaction to which no reverse transcriptase had been added (ϪRT) and plasmid pNA159 (containing the full-length L. amazonensis lap gene) served as negative and positive control templates, respectively, in the PCR reaction. B, soluble Leishmania extracts were resolved by Tris-Tricine SDS-PAGE under reducing conditions, blotted, probed with polyclonal rabbit anti-porcine cytosol leucyl aminopeptidase serum, and developed using enhanced chemiluminescence as described under "Materials and Methods." Recombinant L. amazonensis Lap from which the polyhistidine affinity tag had been removed (75 ng) served as a positive control.
cubation of Lap (2 nM, 37°C) with inhibitors in 50 mM Tris-HCl, 0.5 mM MnCl 2 , pH 8 for 5 min before the addition of substrate. In the presence of metal ion chelators, MnCl 2 was omitted from the assay buffer. The K i for non-tight binding reversible competitive inhibitors was determined as described in Salvesen and Nagase (13). The enzyme-catalyzed hydrolysis of L-Leu-AMC was monitored continuously to establish an uninhibited rate of substrate hydrolysis (v 0 ), after which a 20-fold molar excess of inhibitor over enzyme was added (in less than 5% of the total assay volume), and the new steady-state velocity in the presence of the inhibitor (v i ) was determined. The apparent inhibition constant in the presence of substrate (K i(app) ) was given by (14)), the method of Williams and Morrison (15) was employed. Briefly, Lap (5 nM) was incubated in 50 mM Tris-HCl, 0.5 mM MnCl 2 , pH 8 at 37°C alone or with inhibitor (at a concentration of 0.5-2.5 ϫ [E] 0 ) before assaying residual activity against L-Leu-AMC (20 -50 M). A plot of velocity in the presence of inhibitor (v i ) at inhibitor concentration [I] was fitted to the general integrated equation [E] 0 )) using non-linear regression with SigmaPlot ® (Jandel Scientific) to obtain K i(app) , the apparent inhibition constant in the presence of substrate (15). In both tight and non-tight binding cases, the true K i was calculated from K i ϭ K i(app) /1 ϩ [S]/K m , where [S] denotes substrate concentration. Slow binding inhibition was evaluated as described in Bienvenue et al. (16). L. amazonensis Lap (125 ng of protein) prewarmed to 37°C in 50 mM Tris-HCl, 0.5 mM MnCl 2 , pH 8, was added to a solution of apstatin (0 -1.25 M) or bestatin (0 -50 nM) in 50 mM Tris-HCl, 0.5 mM MnCl 2 , 5 M L-Leu-AMC, pH 8, and the reaction was monitored at 37°C over 800 s. Progress curves that are obtained using this procedure can be described by the general integrated equation P ϭ v s t ϩ (v 0 Ϫ v s )(1 Ϫ e Ϫkt )/k, where P represents the amount of product at time t, v 0 and v s represent the initial and final steady-state velocities, and k represents the apparent first-order rate constant for the establishment of the equilibrium between the initial (E⅐I) and mature (E⅐I*) enzyme-inhibitor complexes (17). Progress-curve data were fitted to this equation with SigmaPlot ® (Jandel Scientific), permitting the determination of v 0 and v s . The overall inhibition constant, K i *, is described by Preparation of Leishmania Extracts and Culture Supernatants-Leishmania (1 ϫ 10 9 cells) were harvested by centrifugation (1 000 ϫ g, 10 min, 4°C), resuspended in 100 mM Tris-HCl, 1 mM AEBSF, 10 M E-64, and 10 M pepstatin A, pH 8 (2 ml, 4°C), sonicated on ice with a Branson Sonifer 250 sonicator (duty cycle ϭ 20%, output ϭ 2; 2 ϫ 30 s), and centrifuged (15 000 ϫ g, 30 min, 4°C) to yield a crude soluble extract. Leishmania were evaluated for Lap secretion as described in Morty et al. (18). Mid-log-phase promastigote L. amazonensis (5 ϫ 10 7 cells/ml) were maintained in M199 medium described above. At 30-min intervals, aliquots (1 ml) were removed and centrifuged (1 500 ϫ g, 2 min, 25°C), and the Triton X-100 solubilized pellets and cell-free supernatants were tested for Lap activity against L-Leu-AMC.
Antibodies, Western Blotting, and Immunofluorescence-Polyclonal rabbit anti-porcine Lap (19) (21), and sequence alignments with other members of the M17 Laps indicate conservation of predicted catalytically important residues (Fig. 1A). The encoded Laps from Leishmania shared significant (92-96%) identity to each other (Fig. 1B); however, a phylogenetic comparison of the three full-length Leishmania Lap sequences with eight other M17 Laps indicated that the Leishmania enzymes were the most evolutionarily divergent members of this family, including homologues from prokaryotic and eukaryotic organisms (Fig. 1B). The E. coli and human Laps exhibit 30% identity, whereas L. amazonensis Lap exhibits 26 and 24% identity to E. coli and human Laps, respectively.
Leishmania genomic DNA digested with endonucleases that do not cut within the lap ORF (Hind III, BglII, and SmaI)  yielded single bands on a Southern blot when probed with full-length lap, whereas endonucleases that cut once within the lap ORF (AvaI, Pst I, XhoI, AccI) yielded two bands (Fig. 2). Thus, lap genes are present as a single copy per Leishmania genome.
Messenger RNA for Leishmania lap was detected by RT-PCR in amastigote and log-and stationary-phase promastigotes (Fig. 3A). Similarly, Lap protein was detected by Western blot in amastigote and log-and stationary-phase promastigote L. amazonensis using antiserum raised against porcine Lap (Fig. 3B). These data indicate that lap genes are expressed at the translational and transcriptional level in all life-cycle stages of Leishmania.
Enzymatic Properties of Recombinant Leishmania Lap-The full-length Leishmania lap genes were expressed in E. coli as catalytically active (Table I) The Lap activity eluted in a single, well resolved peak from a Sephacryl S-300 HR column at a molecular mass corresponding to 376 kDa (Fig. 5A). No activity was eluted at an elution volume corresponding to 62 kDa (the molecular mass of the translation products of Leishmania lap genes), although Lap monomers were detected in this fraction by Western blot (data not shown). These data may indicate that the translation product associates into catalytically competent homohexamers and suggests that Lap monomers may not be catalytically active. A similar observation has been reported for plant Laps (22).
Optimal amidolytic activity was observed against L-Leu-AMC (k cat /K m Ϸ 60 s Ϫ1 ⅐mM Ϫ1 ) ( Table I). Substrates L-Cys-AMC and L-Met-AMC were also efficiently hydrolyzed, although at a lower rate, with k cat values generally 20% of those seen for L-Leu-AMC. Substrates L-Ala-AMC and L-Ile-AMC were hydrolyzed considerably less efficiently (k cat /K m Ϸ 3 s Ϫ1 ⅐mM Ϫ1 ), and very poor activity was detected against L-Trp-AMC (k cat /K m Ϸ 1 s Ϫ1 ⅐mM Ϫ1 ). No activity was observed against any other Lamino acid amide, Cbz-L-Leu-␤-naphthylamine, or Cbz-Gly-Gly-Leu-para-nitroaniline. Removal of the N-terminal polyhistidine tag from L. amazonensis recombinant Lap did not alter the kinetic profile (Table I), illustrating that the tag did not interfere with the catalytic capacity.
Amidolytic activity against L-Leu-AMC was optimal at pH 8.5, with strong activity still detectable up to pH 10 (Fig. 5B). Activity rapidly declined under mildly acidic conditions (pH 6), although Lap was stable over a broad pH range (pH 4 -11) (Fig.  5B). Amidolytic activity was also influenced by ionic strength, with increasing ionic strength accompanied by reduced catalytic capacity (Fig. 5C), possibly by interfering with ionic enzyme-substrate interactions or disturbing ionic interactions that hold the hexameric complex together.
The Laps exhibited enhanced activity in the presence of several metal ions with the order of preference manganese Ͼ Ͼ magnesium Ͼ cobalt Ͼ nickel (Table II). Calcium, cupric, and ferrous ions were toxic, abrogating activity at 1 mM, whereas the zinc ion had little effect at 1 mM and was potently inhibitory at 10 mM. Titration of Mn(II) illustrated that the manganese anion optimally enhanced Lap activity at 0.5 mM (hence, this concentration was routinely used in Lap assays) and became inhibitory above 10 mM (Fig. 5D). After incubation with the metal-ion chelator 1,10-phenanthroline and subsequent removal of the chelator, metal-depleted Lap (apo-Lap) retained negligible activity (Ͻ0.01%). Incubation with Mn(II) at 1 and 5 mM (30 min, 37°C) reactivated the apo-Lap by up to 55% (data not shown), with Mg(II) having less of an effect (6 -8% reactivation at 5 mM). Both metal ion chelators EDTA (at 1 mM) and 1,10-phenanthroline (at 0.1 mM, but not its non-chelating analogue, 1,7-phenanthroline) were potent (89 -98% inhibition) inactivators of Lap, whereas the calcium-selective chelator EGTA was less effective (42% inhibition at 1 mM) ( Table III). The natural metal co-factor of Leishmania Lap was determined with ICP-AES. Using ICP-AES, a zinc-Lap subunit ratio of 2.12:1 was obtained, whereas the manganese-Lap subunit ratio was 0.05:1. These data suggest that zinc is the natural metal co-factor of Leishmania Lap, with two zinc atoms binding each Lap subunit.
Biphasic reaction progress curves were observed for the hydrolysis of L-Leu-AMC by Lap in the presence of the inhibitors bestatin and apstatin (Fig. 6), suggesting a slow binding inhibition mechanism (23). Both bestatin and apstatin were very potent Lap inhibitors, with overall inhibition constant (K i *) values of 3 and 44 nM respectively. In contrast, reaction progress curves for L-leucinol, arphamenine A, and actinonin remained linear for 600 s, and initial (60 s) reaction velocities indicated an elevated K m without changing the V max , suggesting simple competitive inhibition (data not shown). Actinonin proved to be a tight binding inhibitor ([E] 0 /K i Ͼ 0.01) with a K i ϭ 0.8 -2.9 nM for the three Laps (Table III), with weaker inhibition observed for arphamenine A (K i Ϸ 70 M) and L-leucinol (K i Ϸ 100 M).
Subcellular Localization of Lap-Polyclonal antiserum raised against porcine Lap specifically reacted with Leishmania Lap in soluble parasite extracts and with recombinant Lap on Western blots (Fig. 3B). Immunocytochemical studies with these antibodies on permeabilized, fixed L. amazonensis promastigotes yielded a diffuse, uniform fluorescence throughout the parasite cytosol, with no apparent localization of the Lap to the cell membrane or to intracellular structures (Fig. 7). No Lap activity was detected in Leishmania cell-free culture supernatants, with 100% of the activity remaining cell-associated (data not shown). Taken together with the very poor activity of Lap at even mildly acid pH (which makes its presence in the lysosome unlikely), these data suggest that Lap is a cytosolic enzyme that is not secreted by Leishmania. This localization is consistent with the cytosolic location of plant Laps (24).
Leucyl Aminopeptidase Activity of Leishmania Soluble Extracts-Anti-porcine Lap IgG potently inhibited recombinant Leishmania Lap (125 ng) activity against L-Leu-AMC, L-Cys-AMC, and L-Met-AMC. Up to 96% of Lap activity was inhibited in the presence of 100 g⅐ml Ϫ1 anti-Lap IgG relative to Lap activity determined in the presence of 100 g⅐ml Ϫ1 pre-immune IgG (data not shown), providing a tool with which Lap activity in cell extracts could be selectively neutralized.
b The slow binding inhibition by apstatin and bestatin are illustrated in detail in Fig. 6.
porcine Lap IgG was used to determine the extent to which these aminopeptidase activities were attributable to Lap. The L-Leu-AMC activity of Leishmania-soluble extracts was almost completely (90 -95%) inhibited by anti-Lap IgG (Table IV), indicating that Lap was responsible for the bulk, if not all, L-Leu-AMC-hydrolyzing activity in Leishmania-soluble extracts. In contrast, L-Cys-AMC and L-Met-AMC hydrolytic activity of soluble Leishmania extracts was inhibited by 14 -40 and 14 -35%, respectively (Table IV), suggesting that other, unidentified aminopeptidases in addition to Lap are responsible for L-Cys-AMC and L-Met-AMC hydrolytic activity in extracts.

DISCUSSION
Peptidases of parasitic protozoa are currently the subject of intense investigation in the hope of identifying novel virulence factors, drug targets, and vaccine candidates. Only two Leishmania peptidases have received considerable attention to date. Lysosomal cysteine peptidases (for review, see Ref. 3) are virulence factors (25) that have been validated as drug targets (26) and vaccine candidates (27). A membrane-bound zinc metallopeptidase (gp63 or leishmanolysin) is also a virulence factor that shows promise as a vaccine (for review, see Ref. 4). Oligopeptidase B (5) and the proteasome (28) have also received attention. Although not evaluated in leishmaniasis, homologues of gp63 (29) and oligopeptidase B (30,31) are promising drug targets in African trypanosomiasis caused by the closely related kinetoplastid protozoan T. brucei.
Aminopeptidases have not been studied in kinetoplastid protozoans including Leishmania. However, Laps have emerged recently as novel and exciting vaccines and drug targets in other parasites. Vaccination with worm-derived Laps protect against challenge infections (7). Synthetic Lap inhibitors show promise as drugs against malaria (8,9) through inhibition of malarial aminopeptidases involved in terminal stages of hemoglobin degradation (32), which takes place in the parasite cytosol (33). Furthermore, arphamenine A (which we show here inhibits Leishmania Lap) has activity against kinetoplastid protozoans closely related to Leishmania (10). These reports prompted us to identify and characterize aminopeptidases from Leishmania.
We report here the cloning, genetic analysis, and biochemical characterization of a novel aminopeptidase from three highly pathogenic Leishmania species. In contrast to the gp63 (4) and cysteine peptidase (25) genes of Leishmania, the intronless lap genes were present as a single copy per genome and were expressed in all Leishmania life-cycle stages. Their encoded proteins exhibited 20 -30% identity with members of the M17 cytosolic Lap family of metallopeptidases. This represents the first identification of a M17 Lap from any protozoan parasite. To date, the only other aminopeptidase conclusively identified in protozoans is an aminopeptidase N homologue, a member of the M1 family of zinc metallopeptidases (EC subclass 3.4.11), from Plasmodium falciparum (34). The Leishmania Laps described here lack the canonical HEXXHX 18 E and GAMEN motifs that define the M1 family of zinc-dependent aminopepti-  dases (21), indicating that they are not related to the P. falciparum aminopeptidase.
Intriguingly, Leishmania Laps represent the most evolutionarily distant group of the M17 Laps, clustering into their own branch of the M17 Lap phylogenetic tree that diverges even before the mammalian and bacterial Laps diverge from each other. Polyclonal rabbit antiserum raised against V. cholerae Lap did not recognize Leishmania Lap on a Western blot (data not shown), whereas antiserum to porcine Lap did, indicating that Leishmania Lap is antigenically more similar to other eukaryotic Laps than it is to bacterial Laps. We see no evidence of post-translational processing of Leishmania Lap, since Laps were detected by Western blot at a molecular mass that corresponds to the size of the translational product of the full coding sequence. These data contrast with reports that plant (24) and bacterial (20) Laps, which are processed from a 55-to a 34-kDa form that retain catalytic activity.
Optimal amidolytic activity of Leishmania Lap was observed against L-Leu-AMC, and it is likely that these Laps are responsible for the aminopeptidase activity seen against the Leu-Ile-Ala-Tyr peptide in Leishmania extracts (35). We show here that Leishmania Lap exhibits the most restricted substrate specificity of any Lap described to date, confining aminopeptidase activity almost exclusively to three amino acid residues, Leu, Cys, and Met, with very poor activity observed against Ala, Ile, and Trp. Blocking of the N terminus with a Cbz group (Cbz-L-Leu-␤-naphthylamine) or extending the N terminus of the peptide chain (Cbz-L-Gly-L-Gly-L-Leu-para-nitroaniline) prevented hydrolysis by Leishmania Lap, indicating that Lap was a strict aminopeptidase devoid of endopeptidase activity. In contrast, other plant, animal, and bacterial M17 Laps exhibit a very broad substrate specificity that includes Leu, Met, Arg, Ala, Ile, Phe, Val, Thr, and Tyr, with activity, albeit reduced, also observed against Gly, Asp, Pro, and Trp (22). Leishmania Laps possessed potent activity against L-Cys-AMC. The cysteinyl aminopeptidase activity we report here for Leishmania Lap indicates a novel and unique property of Leishmania Laps, since it has never been observed in any M17 Lap previously. This activity is usually a property of cysteinyl aminopeptidase/oxytocinase (EC 3.4.11.3), which belongs to the M1 family of zinc metallopeptidases (36).
Metalloaminopeptidases exhibit a broad range of metal-ion dependence. Mammalian M13 Laps typically utilize Zn(II) (37), whereas other aminopeptidases are reported to utilize Mn(II) (38), Fe(II) (39), and Zn(II) (40). We show in this study that Leishmania Laps were inactivated by incubation with metalion chelators and that activity is enhanced by Mn(II), Mg(II), and Co(II) at micromolar and Ni(II) at millimolar concentrations, whereas only Mn(II) and Mg(II) could re-activate the metal-depleted Lap. Although magnesium (67 nmol⅐mg Ϫ 1 ) and trace amount of zinc have been detected in the L. major cytoplasm (41), data on the cytoplasmic levels of Co(II), Mn(II), and Ni(II) are not available. 2 Analysis of the metal content of L. amazonensis Lap by ICP-AES yielded a 2.12:1 zinc-Lap subunit ratio, whereas manganese was detected at a manganese-Lap subunit ratio of 0.05:1. These data indicate that zinc is most likely the relevant metal cofactor for Leishmania Lap, binding at a ratio of two zinc atoms/Lap subunit, which is consistent with the binding ratio observed for bovine lens Lap (37). The two zinc-binding sites of bovine lens (42) and plant (43) Lap have been identified and consist of site 1, which readily exchanges Zn(II) for other divalent metal cations including Mn(II), Mg(II), and Co(II), and site 2, which binds Zn(II) much more strongly and retains its Zn(II) under condi-tions that allow exchange of the Zn(II) in site 1. It is likely that the activation of Leishmania Lap we see in this study with Mn(II), Mg(II), Co(II), and Ni(II) results from substitution of the site 1 Zn(II) with these metal ions. Indeed, substitution of the site 1 Zn(II) of porcine kidney Lap with Mn(II), Mg(II), and Co(II) has been shown to activate that enzyme by elevating the k cat (44). Similarly, the phenanthroline-generated apo-Lap we generated in this study probably retained the site 2 zinc and was reactivated by treatment with Mn(II) and Mg(II) that bound at the site 1 and restored catalytic activity to the enzyme.
Metallopeptidase inhibitors phosphoramidon, thiorphan, or puromycin did not inhibit Leishmania Laps, setting them apart from thermolysin, enkephalinase, and cytosolic alanyl aminopeptidases. Neither was activity influenced by reducing agents. The leucine derivative L-leucinol inhibited Leishmania Lap (K i Ϸ 100 M), although 5-fold less efficiently than the inhibition reported against bacterial Laps (16), with arphamenine A exhibiting similar potency (K i Ϸ 70 M). The broadspectrum aminopeptidase inhibitor bestatin and aminopeptidase P inhibitor apstatin were potent, slow competitive inhibitors of Leishmania Laps. This mechanism of inhibition by bestatin has been reported for bacterial and bovine Laps (45,46), and the K i * reported here for Leishmania Lap (3 nM) closely approximated that observed for the inhibition of bovine lens Lap by bestatin (1.3 nM) (46). However, although the inhibition of porcine Lap by apstatin has been observed (47), the mechanism was not investigated, and we report here for the first time a slow, competitive binding mode of action by apstatin. The peptide deformylase inhibitor actinonin was a tight ([E] 0 /K i Ϸ 3) and fast competitive inhibitor (K i Ϸ 1 nM) of Leishmania Laps, approximating the potency of actinonin inhibition of peptide deformylase (K i ϭ 10 nM (48)).
It thus appears that Leishmania Laps have evolved into their own subgroup of M17 Laps, demonstrating a unique, restricted substrate specificity that includes cysteine residues. The presence of Lap in all life-cycle stages of Leishmania together with its apparent potential as a drug target in other parasitic infections strongly suggests that it represents a drug target in Leishmania. Because the Laps described here appear to be entirely responsible for leucyl aminopeptidase activity in Leishmania extracts, their selective targeting by inhibitors may interfere with parasite viability. This contrasts with the methionyl and cysteinyl aminopeptidase activities that seem to be undertaken by at least one other aminopeptidase in addition to Lap. Furthermore, the identification in this study of potent Leishmania Lap inhibitors, which are well tolerated in experimental animals (in mice, the oral LD 50 of bestatin is Ͼ4 g⅐kg Ϫ1 (49), and the intraperitoneal LD 50 of arphamenine A is Ͼ500 mg⅐kg Ϫ1 (50)), will provide a good starting point to explore the anti-Leishmania effects of Lap inhibitors.