Adenine Aminohydrolase from Leishmania donovani

Background: Purine salvage in Leishmania is an indispensable nutritional process. Results: Adenine aminohydrolase, a key purine enzyme in Leishmania, has been characterized biochemically and genetically. Conclusion: Adenine aminohydrolase is a unique enzyme in purine salvage that converts 6-aminopurines into 6-oxypurines. Significance: Functional characterization of key enzymes is crucial for understanding purine salvage and ultimately for targeting the pathway with drugs. Adenine aminohydrolase (AAH) is an enzyme that is not present in mammalian cells and is found exclusively in Leishmania among the protozoan parasites that infect humans. AAH plays a paramount role in purine metabolism in this genus by steering 6-aminopurines into 6-oxypurines. Leishmania donovani AAH is 38 and 23% identical to Saccharomyces cerevisiae AAH and human adenosine deaminase enzymes, respectively, catalyzes adenine deamination to hypoxanthine with an apparent Km of 15.4 μm, and does not recognize adenosine as a substrate. Western blot analysis established that AAH is expressed in both life cycle stages of L. donovani, whereas subcellular fractionation and immunofluorescence studies confirmed that AAH is localized to the parasite cytosol. Deletion of the AAH locus in intact parasites established that AAH is not an essential gene and that Δaah cells are capable of salvaging the same range of purine nucleobases and nucleosides as wild type L. donovani. The Δaah null mutant was able to infect murine macrophages in vitro and in mice, although the parasite loads in both model systems were modestly reduced compared with wild type infections. The Δaah lesion was also introduced into a conditionally lethal Δhgprt/Δxprt mutant in which viability was dependent on pharmacologic ablation of AAH by 2′-deoxycoformycin. The Δaah/Δhgprt/Δxprt triple knock-out no longer required 2′-deoxycoformycin for growth and was avirulent in mice with no persistence after a 4-week infection. These genetic studies underscore the paramount importance of AAH to purine salvage by L. donovani.


Adenine aminohydrolase (AAH) is an enzyme that is not present in mammalian cells and is found exclusively in
Leishmania among the protozoan parasites that infect humans. AAH plays a paramount role in purine metabolism in this genus by steering 6-aminopurines into 6-oxypurines. Leishmania donovani AAH is 38 and 23% identical to Saccharomyces cerevisiae AAH and human adenosine deaminase enzymes, respectively, catalyzes adenine deamination to hypoxanthine with an apparent K m of 15.4 M, and does not recognize adenosine as a substrate. Western blot analysis established that AAH is expressed in both life cycle stages of L. donovani, whereas subcellular fractionation and immunofluorescence studies confirmed that AAH is localized to the parasite cytosol. Deletion of the AAH locus in intact parasites established that AAH is not an essential gene and that ⌬aah cells are capable of salvaging the same range of purine nucleobases and nucleosides as wild type L. donovani. The ⌬aah null mutant was able to infect murine macrophages in vitro and in mice, although the parasite loads in both model systems were modestly reduced compared with wild type infections. The ⌬aah lesion was also introduced into a conditionally lethal ⌬hgprt/⌬xprt mutant in which viability was dependent on pharmacologic ablation of AAH by 2-deoxycoformycin. The ⌬aah/ ⌬hgprt/⌬xprt triple knock-out no longer required 2-deoxycoformycin for growth and was avirulent in mice with no persistence after a 4-week infection. These genetic studies underscore the paramount importance of AAH to purine salvage by L. donovani.
Leishmania donovani is the etiologic agent of visceral leishmaniasis, a devastating and invariably fatal disease if untreated. L. donovani, like all Leishmania species, is a digenetic parasite that exists as flagellated extracellular promastigotes in the phlebotomine sandfly vector and as immotile intracellular amastig-otes within phagolysosomes of macrophages and reticuloendothelial cells of the infected mammalian host (supplemental Fig.  1). There is no effective vaccine against leishmaniasis, and the armamentarium of antileishmanial drugs is far from ideal. These drugs are expensive, toxic, and cumbersome to administer, and the emergence of drug resistance renders these therapeutic protocols too often ineffective (1)(2)(3). Thus, the exigency for new drugs, as well as new drug targets, is acute. The development of rational, selective, and effective anti-parasitic drug therapies depends upon the exploitation of fundamental biochemical and/or metabolic disparities between parasite and host. Possibly the most striking metabolic discrepancy between Leishmania and their mammalian hosts is the pathway for the synthesis of purine nucleotides. Leishmania, like all protozoan parasites studied to date, cannot synthesize the purine ring de novo (4 -6). Consequently, each genus of parasite expresses a unique complement of nutritionally indispensable salvage and interconversion enzymes that enable the acquisition of host purines.
The purine pathway of Leishmania is particularly convoluted and capable of incorporating virtually any purine nucleobase or nucleoside from the culture medium or host environment into the parasite nucleotide pool (4 -7). The parasite accommodates the following four enzymes that can convert preformed host purines to the nucleotide level: 1) hypoxanthine-guanine phosphoribosyltransferase (HGPRT) 2 ; 2) xanthine phosphoribosyltransferase (XPRT); 3) adenine phosphoribosyltransferase (APRT); and 4) adenosine kinase (AK) (Fig. 1) (4, 8 -11). HGPRT and XPRT are sequestered within the glycosome (12,13), a peroxisome-like subcellular microbody that is unique to trypanosomatid parasites (14 -16), and APRT is located in the cytosol (13). Leishmania also express numerous purine interconversion enzymes, most of which have human counterparts (4,6,17,18). One leishmanial purine interconversion enzyme, however, that lacks a mammalian equivalent is adenine aminohydrolase (AAH) (EC 3.5.4.2) (19), an enzyme that catalyzes the effectively irreversible deamination of adenine to hypoxanthine (17). The central role of AAH in purine metabolism in L. donovani was underscored by the isolation and characterization of a conditionally lethal mutant deficient in both HGPRT and XPRT (20). The ⌬hgprt/⌬xprt null mutant, unlike wild type parasites, cannot salvage 6-oxypurines and can only survive and grow in vitro when AAH is pharmacologically blocked with 2Ј-deoxycoformycin (dCF) and either adenine or adenosine is provided as the purine source (20). Thus, AAH in the promastigote funnels adenine and adenosine into hypoxanthine, a dead-end substrate for purine salvage by the ⌬hgprt/⌬xprt knock-out. Moreover, the ⌬hgprt/⌬xprt strain is highly but incompletely compromised in its capacity to sustain a visceral infection in mice (20) implying a central role for AAH in purine salvage by amastigotes as well.
AAH is a member of the aminohydrolase superfamily and is commonly found among bacteria (21)(22)(23)(24), but it has thus far only been described in a few lower eukaryotes, including Leishmania (17), Crithidia (25), Saccharomyces cerevisiae (26), and Aspergillus (27). The genomes of Trypanosoma brucei and Trypanosoma cruzi, parasites that are phylogenetically related to Leishmania, lack an AAH homolog (28 -30), and AAH activity is not found in plants (31), mammals (19), or any other protozoan parasite that infects humans (32). Mammals and several other genera of protozoan pathogens express an adenosine deaminase (ADA) activity, and the sequences of ADA and AAH enzymes are homologous (33)(34)(35). ADA in humans is notably critical for immune function because ADA deficiency in humans results in a severe combined immunodeficiency disease in which both the T and B limbs of the immune system are severely debilitated (33, 36 -38).
Previous studies on AAH enzymes from prokaryotes and yeasts have shown that the enzyme is selective for adenine among the naturally occurring purines, although AAH can also metabolize and be competitively inhibited by a variety of purine analogs (26,33,39). However, investigations on AAH from protozoan parasites are virtually nonexistent. Nolan and co-workers (17,25) first identified AAH in extracts from Crithidia fasciculata and four Leishmania species and demonstrated that AAH is inhibited by two potent inhibitors of mammalian ADA, coformycin and dCF (17,36,37). AAH measurements in cellular extracts prepared from L. donovani promastigotes and amastigotes strongly intimated that AAH was a promastigotespecific activity (6). However, later experiments from our laboratory demonstrated robust adenine metabolism in ⌬aprt L. donovani axenic amastigotes that was strongly inhibited by dCF implying that AAH was active in amastigotes (40).
To characterize the L. donovani AAH enzyme and evaluate AAH function in intact parasites, we cloned and sequenced L. donovani AAH and generated ⌬aah knock-outs in both wild type and ⌬hgprt/⌬xprt backgrounds by homologous gene replacement. The ⌬aah lesion in the ⌬hgprt/⌬xprt genetic background alleviated the requirement for dCF, and the growth and virulence phenotypes of this null mutant indicated a pivotal role for AAH in purine metabolism in both promastigotes and amastigotes. AAH was also determined to be cytosolic, unlike HGPRT and XPRT, and a biochemical analysis of the purified recombinant enzyme revealed AAH to be specific for adenine and a target for dCF inhibition. Furthermore, although the ⌬aah mutation in the wild type background did not profoundly affect parasitemias in macrophages or mice, introduction of this lesion into the ⌬hgprt/⌬xprt line reduced the parasite level in macrophages to zero, and there were no persistent ⌬aah/ ⌬hgprt/⌬xprt parasites recovered from mice after a 4-week infection underscoring the potential utility of this triple knockout strain as a live attenuated vaccine candidate.

EXPERIMENTAL PROCEDURES
Materials, Chemicals, and Reagents- [8-14 C]Adenosine (53 mCi/mmol), [8-14 C]adenine (50 mCi/mmol), and [8-14 C]hypoxanthine (51 mCi/mmol) were purchased from Moravek Biochemicals (Brea, CA), and [␣-32 P]dCTP was bought from MP Biomedicals (Irvine, CA). Unlabeled purine bases, nucleosides, and nucleotides were procured from Sigma or Fisher, and dCF was obtained from Tocris Bioscience (Ellisville, MO). Restriction enzymes were purchased from New England Biolabs (Beverley, MA). The TOPO TA Cloning kit, pCR2.1-TOPO vector, Champion TM pET200/D-TOPO expression vector, BL21 Star TM (DE3) One Shot TM competent cells, and the goat anti-rabbit Oregon Green-conjugated and goat antiguinea pig Rhodamine Red-conjugated secondary fluorescent antibodies were bought from Invitrogen. Goat anti-rabbit HRPconjugated and goat anti-mouse HRP-conjugated secondary antibodies were purchased from Thermo Scientific (Waltham, MA). The Complete Mini EDTA-free protease inhibitor tablets were procured from Roche Diagnostics; the Ni 2ϩ -NTA-agarose beads were from Qiagen (Valencia, CA), and the Biosafe TM Coomassie and protein assay kits were from Bio-Rad. All other chemicals and reagents were of the highest quality commercially available.
Cloning of AAH from L. donovani-The AAH gene was isolated from an L. donovani cosmid library that had been generated in the Supercos 1 cosmid vector (Stratagene, La Jolla, CA) (46) using the following strategy. First, the S. cerevisiae Aah1p AAH protein (NP_014258) was used as a query sequence to search the L. major genome (47) for homologs, and a single putative LmjFAAH ORF (LmjF35.2160) was identified that matched the query sequence with a smallest sum probability (P(N)) of 3.7 ϫ e Ϫ51 . Oligonucleotide primers were designed against the LmjF35.2160 ORF and utilized to amplify a fragment of the gene from L. major genomic DNA via standard PCR methodology (48,49). The identity of the PCR fragment was confirmed by sequencing, and the DNA fragment was used to probe the L. donovani cosmid library under stringent conditions. Several colonies containing the L. donovani AAH gene were isolated and purified, and AAH and its flanking regions were sequenced in both directions. Multisequence and pairwise sequence alignments of primary structures were accomplished using algorithms within the Vector NTI Suite 7.1 software (Invitrogen).
Creation of Null Mutants-The ⌬aah knock-out was generated from wild type parasites by double targeted gene replacement using the transfection parameters and plating techniques reported previously (44,55). The linearized drug resistance cassettes enclosing the AAH flanks were excised from pX63-PAC-⌬aah and pXG-BSD-⌬aah with HindIII-BglII and SgrAI-BamHI, respectively, and used to sequentially create AAH/⌬aah::PAC (AAH/aah) heterozygotes and ⌬aah::PAC/⌬aah::BSD (⌬aah) null mutants. The AAH/aah heterozygotes were isolated on semi-solid growth plates containing 20 g/ml puromycin supplemented with 100 M xanthine, whereas the ⌬aah knock-outs were selected on plates containing 20 g/ml puromycin, 20 g/ml blasticidin, and 100 M xanthine.
Generation of Episomally Complemented Lines-To generate an episomal AAH plasmid, forward 5Ј-GGATCCATGGCT-GATGCGACCCTTC-3Ј and reverse 5Ј-GGATCCTCACT-CATACGTCCCCCCGCACACGT-3Ј primers harboring 5Ј and 3Ј BamHI restriction sites were used to PCR-amplify the AAH full-length coding sequence from genomic DNA. The AAH gene was cloned into the pCR2.1-TOPO vector (Invitrogen), sequenced to ensure fidelity, and inserted into the BamHI site of the pXG-PHLEO expression plasmid that was donated by Dr. Stephen Beverley. The relevant portion of the pXG-PHLEO-AAH episome was sequenced to verify proper orientation of the AAH gene. A ⌬aah[pAAH] "add-back" line was created from the ⌬aah null mutant by transfecting ⌬aah parasites with the pXG-PHLEO-AAH episome. The XPRT addback ⌬aah/⌬hgprt/⌬xprt[pXPRT] parasites were selected after transfection of the ⌬aah/⌬hgprt/⌬xprt mutants with the pXG-BSD-XPRT episome that was previously employed to complement the ⌬xprt lesion in L. donovani (13,20,40), and the episomally complemented parasites were selected on 100 M xanthine, an unusable purine source for the parental ⌬aah/ ⌬hgprt/⌬xprt parasites.
AAH Expression and Purification-The AAH ORF was amplified by PCR from the AAH cosmid, ligated into the pET200/D-TOPO bacterial expression vector (Invitrogen), which automatically attaches a His 6 tag to the NH 2 terminus of the inserted gene product, and sequenced to ensure fidelity of the construct. The pET200/D-TOPO-AAH construct was then transformed into BL21 Star TM (DE3) One Shot Escherichia coli (Invitrogen). After induction in 0.5 mM isopropyl 1-thio-␤-D-galactopyranoside, the recombinant L. donovani AAH protein was purified to virtual homogeneity over a Ni 2ϩ -NTA-agarose column from E. coli extracts that had been prepared using a French press as described previously (56). An additional salt wash containing 32.5 mM imidazole and 1 M sodium chloride was added to reduce nonspecific binding, and the concentration of imidazole in the final wash buffer was increased from 20 to 32.5 mM. Recombinant AAH was eluted from the Ni 2ϩ -NTA-agarose with 250 mM imidazole as detailed previously (56). Separation of the purified recombinant AAH fractions on a 10% SDS-polyacrylamide gel and subsequent staining with Bio-safe TM Coomassie (Bio-Rad) confirmed the purity of the recombinant protein. A Thermo Labsystems Multiskan Ascent plate reader (Thermo Scientific) was employed at 600 nm to determine the protein concentration and yield of the purified AAH after addition of Bio-Rad protein assay reagent (56). Yields of purified recombinant AAH protein were ϳ5-8 mg/liter of bacterial culture. To stabilize the purified AAH protein for kinetic analysis, ZnSO 4 was added to a final concentration of 100 M, and kinetic studies were performed within 8 h of purification.
Antibodies, Immunoblotting and DNA Manipulations-Polyclonal AAH antiserum was generated in rabbits by Open Biosystems (Huntsville, AL) using purified recombinant AAH as an immunogen and standard injection protocols. Monospecific polyclonal antibodies raised against purified recombinant L. donovani APRT, HGPRT, and XPRT proteins in rabbits have been described (9 -11). Inosine monophosphate dehydrogenase (IMPDH) antiserum was produced in guinea pigs as reported previously (57), and mouse antiserum against amastigote-specific A2 protein (58) was generously provided by Dr. Greg Matlashewski from McGill University Faculty of Medicine. Mouse monoclonal anti-␣-tubulin antibody (DM1A) was obtained from EMD Chemicals (Gibbstown, NJ). Western blotting protocols were performed as detailed previously (59) using goat anti-rabbit HRP-conjugated and goat anti-mouse HRPconjugated secondary antibodies (Thermo Scientific). Immunoblotting, isolation of genomic DNA, and Southern blot analysis were accomplished using conventional protocols (59). The hybridization probes harboring the full-length L. donovani AAH coding sequence or 5Ј UTR were PCR-amplified from the AAH cosmid and gel-purified using a Wizard SV gel and PCR clean-up kit (Promega, Madison, WI).
Glycosomal Fractionation-3-4 ϫ 10 9 L. donovani promastigotes were harvested at 4°C, washed once in 10 ml of cold PBS, centrifuged at 5000 ϫ g for 5-10 min, washed in 8 -10 ml of hypotonic buffer containing 2 mM EGTA, 2 mM DTT, and protease inhibitor tablets (Roche Diagnostics), and centrifuged at 5000 ϫ g for 5-10 min. The cell pellet was washed and spun as above, resuspended in a final volume of 4 ml of the hypotonic buffer, and placed on ice for 3-5 min. Cells were lysed after 10 -15 passages through a 27-28 1 ⁄ 2-gauge needle, and the solution was made isotonic by adding 1 ml of 4ϫ lysis buffer containing 200 mM HEPES-NaOH, pH 7.4, 1 M sucrose, 4 mM ATP, 4 mM EGTA, 8 mM DTT, and protease inhibitors. The solution was centrifuged at 5000 ϫ g for 5-10 min at 4°C to remove cell debris and nuclei. The supernatant was transferred to a new tube and spun at 45,000 ϫ g for 35-45 min at 4°C to separate the organellar fraction containing glycosomes from the cytosolic components. A sucrose gradient ranging from 20 to 70% was prepared as described previously (80) and allowed to equilibrate overnight at 4°C. The glycosome-containing organellar fraction was resuspended in a minimal volume of 25 mM HEPES, added to the sucrose gradient, and centrifuged for 6 h at 36,000 rpm in a Beckman SW41 rotor (Brea, CA). Gradient fractions were analyzed by standard Western blotting procedures (59) using polyclonal AAH and IMPDH antisera.
Antibody Purification and Immunofluorescence Assay-Purified AAH protein was bound to AminoLink coupling resin (Thermo Scientific) as detailed previously in the package insert, and crude AAH antisera was purified against AAH protein using an AminoLink immobilization kit (Thermo Scientific) according to the manufacturer's instructions. The immunofluorescence assay was performed on L. donovani promastigotes as described previously (12, 13, 57, 60) using a 1:1000 dilution of recombinant anti-AAH antibody and a 1:10,000 dilution of secondary goat anti-rabbit Oregon Green-conjugated antibody (Invitrogen). Parasites were co-stained with guinea pig anti-IMPDH antibodies (1:500) and goat anti-guinea pig Rhodamine Red-conjugated (1:10,000) secondary antibodies to visualize glycosomal IMPDH. Cells were visualized on a Zeiss Axiovert 200 inverted microscope (Carl Zeiss Microimaging, Thornwood, NY) with a ϫ63 oil immersion lens. Photographs were taken with a Zeiss AxioCam MR camera using Axiovision 4.2 software and compiled using Adobe Photoshop Creative Suite 4.
AAH Activity Measurements in Intact Parasites-The ability of wild type and ⌬aah promastigotes and axenic amastigotes to deaminate [8-14 C]adenine (50 mCi/mmol) into [8-14 C]hypoxanthine was compared using TLC. 7.0 ϫ 10 7 parasites were washed and resuspended in 30 l of PBS, and [8-14 C]adenine (50 mCi/mmol) was added to a final concentration of 67 M. At various time points over a 1-h time course, 5 l of the parasite mixture were added to 2 l of glacial acetic acid to lyse the cells and terminate the reaction (25), spotted onto a PEI-cellulose TLC plate (Scientific Adsorbents Inc., Atlanta, GA), and run in n-butanol/water/glacial acetic acid at a ratio of 20:6:4 (v/v) (25). The TLC plate was exposed to x-ray film at Ϫ80°C and developed in a standard x-ray film developer. The ability of 0.67 mM dCF to inhibit AAH in wild type L. donovani promastigotes and axenic amastigotes was also assessed by this method; however, parasites had to be incubated with dCF for 15 min prior to starting the TLC reaction for notable inhibition to occur.
Radiometric AAH Assay on Pure Recombinant AAH Protein-To validate that AAH did not utilize adenosine as a substrate, a radiometric method was implemented to test whether an excess amount of purified AAH could deaminate [8-14 C]adenosine (51 mCi/mmol) to inosine over a 2-h time course. 5-l samples were spotted at the 0-and 2-h time points, and the TLC plate was run, exposed, and developed as described above. [8][9][10][11][12][13][14] C]Adenine (50 mCi/mmol) was used as a control to verify that the AAH protein employed in the assay was enzymatically active.
Spectrophotometric AAH Assay and Enzyme Inhibition Kinetics-The ability of purified recombinant AAH to convert adenine to hypoxanthine was also assessed by means of a coupled assay in which the hypoxanthine produced from adenine was converted by 1 unit of xanthine oxidase (Sigma) into uric acid, the accumulation of which was determined by measuring the increase in absorbance at 292 nm (61,62) in a Beckman DU 640 spectrophotometer (Beckman Coulter, Inc., Brea, CA). All kinetic studies with wild type AAH were carried out under experimental conditions that were linear with time and enzyme concentration. Briefly, Ϸ0.35 g of recombinant AAH was added to a quartz cuvette containing 1 ml of PBS, 1 unit of xanthine oxidase, and adenine varying in concentration from 500 nM to 50 M. The reaction was carried out at 25°C at pH 7.0 over a 90-s time course, and absorbance at 292 nm was measured every 15 s. The amount of uric acid produced was calculated from the A 292 reading at each time point. The molar extinction coefficient () of uric acid was experimentally calculated to be 11,590 M Ϫ1 ⅐ cm Ϫ1 under these reaction conditions using the Beer-Lambert equation A ϭ ⅐c⅐l, where A ϭ 292 nm, c ϭ 20 M, and l ϭ 1 cm. Michaelis-Menten analysis (Graph-PAD Prism 4.0) (63) was used to determine the K m value of AAH for adenine, and this value was verified by replotting the data using the Hanes-Woolf method (64).
The xanthine oxidase-coupled assay was also employed to assess the inhibitory effect of dCF on the purified recombinant L. donovani AAH enzyme. Ϸ0.30 g of AAH was incubated with the appropriate concentration of dCF for 15 min prior to the commencement of each assay, and uric acid production was quantified by measuring the absorbance at 292 nm in the absence of dCF and in the presence of either 10, 20, or 30 M dCF with adenine concentrations ranging from 2 to 50 M. The background absorbance of each concentration of dCF was sub-tracted from the total absorbance at 292 nm, and the AAH activity was plotted as a function of adenine concentration at each concentration of dCF. To determine the K i value for dCF, the K m value in the absence of dCF and the apparent K m value at each dCF concentration were calculated using Michaelis-Menten analysis (GraphPAD Prism 4.0) (63) and replotted as a function of dCF concentration (65,66).
Mouse Infections-Groups of five 7-week-old female BALB/c mice (Charles River Laboratories, Wilmington, MA) were inoculated by tail vein injection with 5 ϫ 10 6 late log-stationary phase wild type, ⌬aah, or ⌬aah/⌬hgprt/⌬xprt promastigotes (60,67). Prior to injection, wild type and ⌬aah parasites were cycled back and forth several times between promastigote and axenic amastigote forms (43) to revitalize ancillary virulence determinants, as well as passaged through mice for 12 days to eliminate parasites that had become attenuated due to prolonged in vitro culture. Four weeks post-infection, mice were sacrificed, and their livers and spleens were harvested as reported previously (60,67). Single cell suspensions from mouse organs were prepared by passage through a 70-m cell strainer (Falcon), and parasitemias were then determined in 96-well microtiter plates using a standardized limiting dilution assay (69). The organ-derived wild type and ⌬aah parasites were titered in modified DME-L (40) supplemented with 5% FBS and 100 M xanthine, and the ⌬aah/⌬hgprt/⌬xprt mutants were grown in medium containing 100 M adenine.

RESULTS
Multiple Sequence Alignment of L. donovani AAH-The L. donovani AAH (AAY98748) and its flanking sequences were isolated from a cosmid clone using the putative L. major AAH (LmjF. 35.2160) as a hybridization probe. The L. donovani AAH encodes a protein of 362 amino acids with a molecular mass of ϳ40.8 kDa. The L. donovani AAH encompasses a potential peroxisomal targeting signal 2 (PTS2) (underlined in Fig. 2), a somewhat divergent NH 2 -terminal topogenic sequence that directs proteins to the glycosome (16,70). This PTS2-type sequence is also found in the primary structures of the Leishmania mexicana and Leishmania infantum AAHs but not in the Leishmania major or Leishmania braziliensis homologs ( Fig. 2 and data not shown). A multiple sequence alignment of L. donovani AAH with other eukaryotic AAH and ADA proteins revealed many residues that are conserved throughout (Fig. 2). Notably, the L. donovani AAH contained all four histidine residues, His 20 , His 22 , His 213 , and His 237 , responsible for coordinating the Zn 2ϩ ion that is present in the active site of the crystal structures of the Plasmodium vivax and mouse ADAs (35,(71)(72)(73)(74). Indeed, these four His residues are conserved among all of the aligned AAH and ADA proteins (Fig. 2). Pairwise alignments of the L. donovani AAH primary structure with each of the Leishmania and S. cerevisiae counterparts revealed 94 and 38% amino acid identities with the other leishmanial and the S. cerevisiae Aah1p AAH sequences, respectively. Comparison of the L. donovani AAH with functionally characterized ADA proteins from Plasmodium falciparum, P. vivax (75), E. coli (76), Mus musculus (38,77), and Homo sapiens (78,79) demonstrated that the AAH from L. donovani was 23-25% identical to the two Plasmodium ADAs, 27% identical to the E. coli ADA, and 23-24% identical to the two mammalian ADAs.
Confirmation of ⌬aah and ⌬aah/⌬hgprt/⌬xprt Genotypes and Western Blot Analysis-⌬aah and ⌬aah/⌬hgprt/⌬xprt cell lines were created by targeted gene replacement strategies as described under "Experimental Procedures." Southern blot analysis of genomic DNA from wild type, ⌬aah, ⌬aah[pAAH], ⌬hgprt/⌬xprt, ⌬aah/⌬hgprt/⌬xprt, and ⌬aah/⌬hgprt/ ⌬xprt[pXPRT] parasites was then carried out to verify the homologous gene replacements of both AAH alleles and to confirm the genotype of the ⌬aah/⌬hgprt/⌬xprt triple mutant (Fig.  3, A-D). The full-length coding sequences of AAH, XPRT, and HGPRT and the 5Ј UTR of AAH were used as hybridization probes (Fig. 3, A-D). The hybridization signals in genomic DNA samples prepared from wild type and genetically manipulated parasites corresponded to the sizes of the restriction fragments predicted from the sequences of the AAH, XPRT, and HGPRT loci in the L. donovani cosmids (9, 10), as well as from the annotated L. infantum sequences of the same three loci, and the absence of hybridization bands in the knock-out cell lines reflected the specific gene rearrangements at the pertinent loci (Fig. 3, A-D).
Localization of AAH-AAH was localized to the cytosol by both subcellular fractionation and immunofluorescence assay. Separation of L. donovani lysates into organellar and cytosolic components by differential centrifugation and subsequent immunoblotting revealed that AAH was localized virtually exclusively to the cytosolic compartment (Fig. 4A). The differential centrifugation protocol that was employed has been shown previously to yield intact glycosomes (80,81). The AAH co-sedimented with APRT, a cytosolic marker (13), whereas IMPDH, a known glycosomal enzyme (57), was associated exclusively with the organellar (45,000 ϫ g) pellet (Fig. 4A). Further fractionation of the organellar pellet on a sucrose gra-dient revealed that IMPDH sedimented with the glycosomes (lower portion of the gradient, fractions 15-27), whereas AAH, which marginally contaminated the 45,000 ϫ g pellet, could be detected at the top of the sucrose gradient (fractions 1-14) with other cytosolic components ( Fig. 4B and data not shown). Immunofluorescence assay validated the cytosolic context for AAH in wild type and ⌬aah[pAAH] promastigotes and confirmed the absence of AAH in ⌬aah parasites (Fig. 5). The cytosolic milieu for AAH was supported by the lack of colocalization with IMPDH, a known glycosomal protein (Fig. 5) (57).
Expression of AAH Protein in Intact Parasites-A previous study using crude cell extracts reported that AAH activity is expressed exclusively in L. donovani promastigotes (6), although adenine metabolism experiments in L. donovani axenic amastigotes suggested otherwise (40). To definitively determine whether AAH protein is stage-specific or expressed in both L. donovani life cycle stages, Western blot analysis was performed on lysates of wild type, ⌬aah, and ⌬aah[pAAH] promastigotes and axenically derived amastigotes. Immunoblotting with monoclonal A2 antibody, an amastigote-specific marker (58), revealed the existence of A2 proteins in the axenically derived amastigote forms of all three strains, but no A2 expression was observed in wild type, ⌬aah, and ⌬aah[pAAH]   , ⌬hgprt/⌬xprt, ⌬aah/ ⌬hgprt/⌬xprt, and ⌬aah/⌬hgprt/⌬xprt[pXPRT] parasites to grow in different purine sources was compared. Parasites were seeded at a density of 5 ϫ 10 4 parasites/ml, incubated for 7-10 days in 100 M purine, and enumerated via hemocytometer. Parasite proliferation in adenine (Ade) or adenosine was also assessed in the presence of 20 M dCF. The data depicted are plotted on a log scale and are the averages and standard errors of three independent determinations. A Student's t test was performed for the data comparing wild type and ⌬hgprt/⌬xprt growth in adenine and adenosine in the absence of dCF, and p values of Ͻ0.0001 were calculated. Gua, guanine; Hyp, hypoxanthine; Xan, xanthine. promastigotes (Fig. 7). AAH, however, is robustly expressed in promastigotes and axenic amastigotes in all cell lines with an intact AAH locus but not, as expected, in cell lines harboring the ⌬aah lesion (Fig. 7).
AAH Activity in Intact Parasites-To examine whether AAH activity is present in both life cycle stages of L. donovani and to validate AAH as a cellular target of dCF, AAH activity was measured in intact wild type promastigotes and axenic amastigotes in the absence and presence of dCF (Fig. 8). ⌬aah promastigotes and axenic amastigotes were employed as negative controls. Both life cycle forms exhibited the ability to produce [ 14 C]hypoxanthine from extracellular [ 14 C]adenine (Fig. 8) in wild type cells, and adenine deamination to hypoxanthine was not detected in ⌬aah parasites (Fig. 8). Addition of 0.67 mM dCF to wild type parasites dramatically reduced adenine conversion to hypoxanthine in both promastigotes and axenically derived amastigotes (Fig. 8).
Purification and Substrate Specificity of AAH-Recombinant His 6 -tagged L. donovani AAH was purified to essential homogeneity over a Ni 2ϩ -NTA column (data not shown), and the purified AAH was tested for its ability to deaminate [ 14 C]adenine and [ 14 C]adenosine to [ 14 C]hypoxanthine and [ 14 C]inosine, respectively. These experiments were conducted with an excess of protein to be able to detect negligible recognition of either purine as a substrate. Separation of the reactants and products by TLC revealed that an overload of purified L. donovani AAH protein robustly deaminated [ 14 C]adenine to [ 14 C]hypoxanthine virtually instantaneously, although no [ 14 C]adenosine to [ 14 C]inosine conversion was detected even after a 2-h incubation with AAH (Fig. 9).
Kinetic Analysis of AAH Activity-Freshly purified recombinant His 6 -AAH was enzymatically active and deaminated adenine robustly. However, upon storage the enzymatic activity diminished and was completely lost after 48 h at 4°C. Addition of 100 M ZnSO 4 to the purified recombinant AAH preparations significantly stabilized the activity of the enzyme and was therefore routinely added prior to all kinetic assessments (data not shown). Michaelis-Menten and Hanes-Woolf plots revealed that the apparent K m of AAH for adenine was ϳ15.4 M with a V max of ϳ0.41 mol⅐s Ϫ1 ⅐mg protein Ϫ1 . A k cat value of 16.75 s Ϫ1 was then calculated from the kinetic data, and the catalytic efficiency (k cat /K m ) was computed to be 1.09 s Ϫ1 ⅐M Ϫ1 (Fig. 10A).
Inhibition of AAH by dCF was quantified at fixed amounts of dCF as a function of adenine concentration, and the rate of production of hypoxanthine was calculated from the change in absorbance at 292 nm. The K m value in the absence and presence of either 10, 20, or 30 M dCF was determined by Michaelis-Menten analysis (Fig. 10B) (63). The K m and apparent K m values were plotted as a function of dCF concentration (65,66), and the K i of dCF was determined to be ϳ23.1 M (Fig. 10C).
Macrophage and Mouse Infectivity-To ascertain whether a genetic deficiency in AAH compromised the ability of L. donovani to infect host cells, infectivity assays with stationary phase promastigotes were performed in both primary macrophages and mice. Wild type and ⌬aah[pAAH] parasites were capable of infecting and sustaining robust infections in murine peritoneal macrophages, whereas the parasitemias of the ⌬aah, ⌬hgprt/ ⌬xprt, and ⌬aah/⌬hgprt/⌬xprt mutants were reduced to ϳ25, 5, and 0%, respectively, relative to that for wild type parasites (Fig. 11A). ⌬aah/⌬hgprt/⌬xprt[pXPRT] add-back parasites exhibited parasitemia levels that were equivalent to wild type (Fig. 11A). The percentages of macrophages infected with wild  1 and 2), ⌬aah (lanes 3 and 4), and ⌬aah[pAAH] (lanes 5 and 6), promastigotes (P) (lanes 1, 3, and 5), and axenic amastigotes (A) (lanes 2, 4, and 6) were analyzed by Western blotting using mouse monoclonal antibodies raised against the amastigote-specific A2 protein family and polyclonal rabbit anti-AAH antibodies. The blots were also probed with polyclonal rabbit anti-HGPRT antibodies as loading controls.   Because ⌬aah parasites exhibited an ϳ75% reduction in parasite loads in macrophages, the ability of the null mutant to infect BALB/c mice, a well characterized rodent model for L. donovani infections (67,(82)(83)(84), was evaluated. Parasite burdens in the livers of mice infected with the ⌬aah knock-out were also found to be reduced by ϳ75% when compared with mice infected with the wild type strain, whereas the parasitemias in spleens of these mice were virtually identical (Fig.  11B). The parasite load recovered from mice inoculated with the ⌬aah parasites was robust, however, compared with that observed after infection with the ⌬aah/⌬hgprt/⌬xprt triple mutant (Fig. 11B). No parasites were recovered from the livers or spleens of any of the five mice injected with this strain. All parasites emerging from the mouse infections were analyzed by Western blotting to confirm that the original phenotype of each of the strains was maintained during the course of the animal experiments (data not shown).

DISCUSSION
The purine acquisition pathway of Leishmania is indispensable, particularly elaborate, and includes a unique assortment of purine salvage and interconversion enzymes that have been FIGURE 10. AAH stability and kinetics. A, purified recombinant L. donovani AAH was incubated with 1 unit of xanthine oxidase and varying concentrations of adenine over a 90-s time course, and the absorbance was monitored at 292 nm. Kinetic constants were computed using Michaelis-Menten and Hanes-Woolf (inset) algorithms from GraphPad Prism. B, inhibition of AAH activity by dCF was followed over a 90-s time course, with the formation of adenine (Ade) monitored as detailed previously above at 0 (f), 10 (OE), 20 (), or 30 (ࡗ) M dCF. Kinetic constants were computed via the Michaelis-Menten algorithm within GraphPad Prism. C, apparent K m values calculated in B were replotted as a function of dCF concentration, and the K i value for dCF was determined from the negative value of the x-intercept. FIGURE 11. Macrophage parasitemia and mouse infectivity. A, mouse peritoneal macrophages were infected with wild type, ⌬aah, ⌬aah[pAAH], ⌬hgprt/⌬xprt, ⌬aah/⌬hgprt/⌬xprt, or ⌬aah/⌬hgprt/⌬xprt[pXPRT] stationary phase promastigotes at a multiplicity of infection of 10 parasites/macrophage. Cells were stained after 72 h and amastigotes enumerated visually. The results are the averages and standard errors of three independent determinations (n ϭ 3) plotted on a log scale. B, three separate groups of five BALB/c mice were infected with wild type (1), ⌬aah (2), or ⌬aah/⌬hgprt/⌬xprt (3) late log phase promastigotes. Mice were sacrificed 4 weeks post-infection, and parasite loads in livers and spleens quantified using limiting dilution.
identified through biochemical investigations (4, 6, 9 -11, 85, ⌬xprt double mutant is markedly compromised in its capability to infect BALB/c mice (20), an outcome that attests that effectively the entire salvageable pool of purines in the mammalian host to which L. donovani amastigotes are exposed consists of nucleobase substrates of HGPRT and XPRT, parasite loads were still measurable, ϳ10,000/g liver and Ͻ100/g spleen. The persistence of ⌬hgprt/⌬xprt parasites in mice could be ascribed either to residual salvage of host-supplied adenine or adenosine through APRT or AK, to a limited capacity of APRT to recognize hypoxanthine that is derived either directly from the host, or as a result of adenine deamination by AAH and/or APRT amplification (20,60). In contrast, no surviving parasites were detected after a 4-week infection of mice (Fig. 11B) with the ⌬aah/⌬hgprt/⌬xprt strain, and no ⌬aah/⌬hgprt/⌬xprt parasites were observed after a 72-h infection of peritoneal macrophages (Fig. 11A). Complementation of the ⌬aah/⌬hgprt/ ⌬xprt triple knock-out with XPRT restored parasite loads in macrophages to wild type levels, proving that the majority of the intracellular purine nutrients that Leishmania have access to are 6-oxypurines. Furthermore, the remarkable absence of persistence of ⌬aah/⌬hgprt/⌬xprt triple knock-out parasites in mice (Fig. 11B), coupled with the finite but low parasitemia caused by the ⌬hgprt/⌬xprt null line (20), provides strong genetic support that AAH is key to the lingering amastigote survival of the ⌬hgprt/⌬xprt double knock-outs. The dearth of persistent ⌬aah/⌬hgprt/⌬xprt parasites observed in livers and spleens of susceptible BALB/c mice inoculated with the ⌬aah/ ⌬hgprt/⌬xprt knock-out 4 weeks post-inoculation touts the triple knock-out as a potential live attenuated vaccine candidate for visceral leishmaniasis. Although low level persistent infections are thought to be essential for generating T-cell protective immunity to leishmanial infections (68,90,91), the presence of persistent parasites obviously precludes their utility as a live vaccine. Whether ⌬aah/⌬hgprt/⌬xprt parasites persist for a long enough interval to engender enduring protective immunity remains to be investigated.