Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Biochemistry. Author manuscript; available in PMC 2010 October 13.
Published in final edited form as:
PMCID: PMC2771402

Structural and metabolic specificity of methylthiocoformycin for malarial adenosine deaminases


Plasmodium falciparum is a purine auxotroph requiring hypoxanthine as a key metabolic precursor. Erythrocyte adenine nucleotides are the source of the purine precursors, making adenosine deaminase (ADA) a key enzyme in the pathway of hypoxanthine formation. Methylthioadenosine (MTA) is a substrate for most malarial ADAs, but not for human ADA. The catalytic site specificity of malarial ADAs permits methylthiocoformycin (MT-coformycin) to act as a Plasmodium-specific transition state analogue with low affinity for human ADA (Tyler, P. C., Taylor, E. A., Fröhlich, R. G. G. and Schramm, V. L. (2007) J. Am. Chem. Soc. 129, 6872–6879). The structural basis for MTA and MT-coformycin specificity in malarial ADAs is the subject of speculation (Larson, E. T. et al. (2008) J. Mol. Biol. 381, 975–988). Here, the crystal structure of ADA from Plasmodium vivax in complex with MT-coformycin reveals an unprecedented binding geometry for 5’-methylthioribosyl groups in the malarial ADAs. Compared to malarial ADA complexes with adenosine or deoxycoformycin, 5’-methylthioribosyl groups are rotated 130°. A hydrogen bonding network between Asp172 and the 3'-hydroxyl of MT-coformycin is essential for recognition of the 5’-methylthioribosyl group. Water occupies the 5'-hydroxyl binding site when MT-coformycin is bound. Mutagenesis of Asp172 destroys the substrate specificity for MTA and MT-coformycin. Kinetic, mutagenic and structural analyses of PvADA and kinetic analysis of five other plasmodial ADAs establishes the unique structural basis for its specificity for MTA and MT-coformycin. Plasmodium gallinaceum ADA does not use MTA as a substrate, is not inhibited by MT-coformycin and is missing Asp172. Treatment of P. falciparum cultures with coformycin or MT-coformycin in the presence of MTA is effective in inhibiting parasite growth.

Malaria is caused by protozoan parasites of the Plasmodium genus. Within the four species of malaria parasite that infect humans, Plasmodium vivax and Plasmodium falciparum are the most prevalent species, with P. falciparum being responsible for most of the fatal cases (1). P. vivax has the widest global distribution and is responsible for most of the malaria cases in Central and South America and Asia (2). Plasmodium knowlesi is a primate malaria that is an emerging infectious disease of humans (3, 4). Malaria treatment by chemotherapeutic and vector control strategies have not prevented its widespread occurrence. Recent increases in the resistance of malaria parasites to drug treatment and in mosquito vectors to insecticides have renewed the demand for new chemotherapeutic strategies (5, 6). The 48 hr intraerythrocytic parasite growth phase requires robust nucleic acid synthesis, thus, targeting of purine salvage pathways provide a promising route for novel drug development.

All Plasmodium species are purine auxotrophs, salvaging host cell purines for synthesis of cofactors and nucleic acids (7, 8). In Plasmodium, adenosine is converted to hypoxanthine using adenosine deaminase (ADA) and purine nucleoside phosphorylase (PNP). IMP is formed from hypoxanthine by hypoxanthine-guanine-xanthine phosphoribosyl-transferase (HGXPRT). Inhibition of the purine salvage pathway with transition state analogue inhibitors of both human and Plasmodium PNP, such as Immucillin-H and 4’-deaza-1’-aza-2’-deoxy-1’-(9-methylene)-Immucillin-G (DADMe-ImmG), are lethal for P. falciparum in vitro (9, 10).

Coformycin is a picomolar, transition state analogue inhibitor of both human and Plasmodium ADAs (11). Coformycin alone does not inhibit parasite growth in cultured erythrocytes (10) but 2’-deoxycoformycin (d-coformycin, Pentostatin) is reported to cause decreased parasitemia in P. knowlesi-infected primates (12). Plasmodium species lack adenosine kinase (10, 13) and cannot incorporate exogenous adenosine directly into the adenylate pool. Thus, adenosine (or MTA) can only be salvaged after action of ADA in the parasite. Here, we demonstrate that P. falciparum in vitro growth is inhibited by coformycin or MT-coformycin with MTA as the purine source.

P. falciparum ADA (PfADA) also deaminates 5’-methylthioadenosine (MTA; Figure 1) in addition to adenosine (14). Thus, P. falciparum ADA serves the dual functions of adenosine salvage and recycling MTA formed from the synthesis of polyamines (14). Mammalian ADAs do not deaminate MTA and instead express a specific MTA phosphorylase for recycling of MTA (14). Mammalian erythrocytes do not synthesize polyamines. Thus, an intact polyamine synthetic pathway is important for the viability of malaria parasites (14, 15). In P. falciparum, MTA is deaminated by PfADA to 5’-methylthioinosine (MTI), a metabolite that has not been reported in mammalian metabolism (16). The Plasmodium PNP also serves a dual purpose by converting both inosine and MTI to hypoxanthine for conversion to IMP and incorporation into nucleic acids (14).

Figure 1
Structures and atomic numbering for MT-coformycin and methylthioadenosine (MTA).

We synthesized 5’-methylthiolcoformycin (MT-coformycin; Figure 1) as a specific transition state analogue inhibitor of plasmodial ADAs based on their unusual specificity for both adenosine and MTA (11). MT-Coformycin is a sub-nanomolar inhibitor of PfADA and demonstrates >20,000 fold selectivity for PfADA relative to human ADA. This selectivity is remarkable since coformycin and d-coformycin are powerful picomolar inhibitors of both human and P. falciparum ADAs (11).

To understand the structural basis of recognition of PfADA for MTA and MT-coformycin, we overexpressed and characterized five additional ADAs from parasites with different host preferences: P. vivax (human), P. knowlesi (P. falciparum-like; simian host), P. cynomolgi (P. vivax-like; simian host), P. berghei (rodent host) and P. gallinaceum (avian host). Avian erythrocytes are nucleated, distinguishing them from mammalian red cells. Within the six plasmodial ADAs tested, only P. gallinaceum ADA (PgADA) does not have significant activity for MTA and consequently, MT-coformycin is a poor inhibitor. Sequence alignment revealed that PgADA differs in its catalytic site with an Asp172Glu replacement.

Recent crystal structures of P. vivax ADA (PvADA) revealed catalytic site interactions with adenosine and d-coformycin (17). Molecular modeling experiments hypothesized that the Plasmodium species enzymes can accommodate the 5’-methylthio substituent with only minor conformational changes to the catalytic site amino acids and to the ligand (17).

Here, we present the crystal structure of MT-coformycin bound to PvADA at 2.1 Å resolution. MT-Coformycin binds tightly as a consequence of a large change in the glycosidic torsion angle to reposition the 5’-methylthioribosyl group in a geometry previously unseen in other adenosine deaminases structures (17, 18). The 1.9 Å resolution crystal structure and kinetic properties of a mutant lacking Asp172 (PvADA-ΔAsp172) established the mechanism of MT-coformycin binding.


Cloning and expression of adenosine deaminase enzymes from different Plasmodium species

Orthologs of PfADA were located using the tblastn function (default settings) with the P. knowelsi, P. vivax, P. reichenowi, P. gallinaceum, and P. berghei genome sequence databases from Sequencing Groups at the Sanger Institute (, The Institute for Genomic Research Parasite Database ( and PlasmoDB ( Appropriate primers were designed (Supporting information, Table S1). P. cynomolgi ADA was cloned using degenerate primers based on the P. reichenowi sequence. In each strain ADA was predicted to reside on a contiguous DNA sequence to permit cloning from genomic DNA. Genomic DNA from P. berghei (ANKA strain), P. vivax (Sal-1, gift of Jane Carlton, NYU Langone Medical Center), P. gallinaceum (gift of Joseph Vinetz, UCSD School of Medicine), P. cynomolgi and P. knowlesi (gift of Clemens Kocken and Alan Thomas, Biomedical Primate Research Centre) was used for PCR amplification of the ADA gene from each species. The coding region of each enzyme, without the stop codon, was amplified by PCR and cloned into the pTrcHis2-TOPO vector (Invitrogen) with a C-terminal His6 tag and ampicillin selection cassette. Each plasmid was transformed into E. coli strain TOP10 (Invitrogen) and multiple clones of each DNA encoding for ADA were sequenced and the data confirmed from the P. knowlesi, P. vivax, P. gallinaceum and P. berghei genome predictions. The DNA sequence for P. cynomolgi ADA was determined and is reported as new data since no genome sequence data was available. The respective amino acid sequences of the malarial ADAs are reported (Supporting information, Figure S1). The recombinant enzymes were expressed by induction of 100 ml bacterial culture with 1 mM isopropyl-1-thio-β-D-galactopyranoside (IPTG) at 37 °C for 18 h and purified using nickel-nitrilotriacetic acid affinity chromatography (Ni-NTA spin column, Qiagen) according to the manufacturer’s instructions. The purified proteins were used for enzymatic assays without further purification. Enzyme concentrations were determined from the extinction coefficients at 280 nm (Supporting information, Table S1).

P. falciparum in vitro cell growth and inhibition assay

Coformycin and MT-coformycin were dissolved in water. Inhibition tests were carried out in flat-bottomed microtiter plates (Costar). The method described by Desjardins and colleagues (19) was used to determine the IC50 value and parasite DNA content was determined by DNA dye-binding fluorescence as described by Quashie and colleagues (20). For each condition, three experiments were carried out in duplicate. Synchronized P. falciparum cultures were grown in purine-rich medium (370 µM hypoxanthine, standard medium). Prior to growth inhibition experiments, schizont stage parasite cultures were split and one half was washed in purine-free medium and cultured in purine-free medium for 24 h while the other half was maintained in standard medium. Ring stage parasite cultures (200 µl per well, with 1% hematocrit and 1% parasitemia) were grown for 72 h in the presence of increasing inhibitor concentration in the presence of 100 µM MTA as sole purine source. After incubation, cells were harvested and analyzed for DNA content. Uninfected erythrocytes were used as background controls.

Site-directed mutagenesis of P. vivax ADA

Site-directed mutagenesis used the QuickChange® Site-directed Mutagenesis kit (Stratagene) according to the manufacturer’s instructions. Appropriate primers were designed (Supporting information, Table S2) and the mutagenesis reactions were performed using the pTcrHis2-TOPO vector containing the PvADA sequence as the template. The final reaction mixture was transformed into E. coli strain X10-Gold (Stratagene). Multiple clones of each ADA mutant were sequenced to confirm the presence of the desired mutation (Supporting information, Table S2). The plasmids carrying the desired mutations were transformed into E. coli strain BL21-codon plus (DE3)-RIPL (Stratagene). The recombinant enzymes were expressed, purified and quantified as described above.

Enzymatic assays and inhibition studies

Recombinant proteins were used for enzymatic assays directly following purification. Adenosine deaminase activity was determined by monitoring the change in absorbance at 265 nm upon conversion of adenosine to inosine or MTA to MTI in 100 mM Tris-HCl buffer (pH 8.0) and varied concentrations of adenosine or MTA (14). Enzyme inhibition assays to determine the Ki value for initial and slow-onset inhibition constants (Ki*) were performed using different concentration of coformycin or MT-coformycin (generous gifts of Peter C. Tyler, Industrial Research, Ltd., New Zealand) and 200 µM adenosine. The inhibition constants were determined as described previously (11).

Protein purification and crystallization

Recombinant PvADA and the mutant PvADA-ΔAsp172 were expressed by induction of the bacterial culture with 1 mM IPTG at 30 °C for 18 h. The cells were ruptured by passage through a French press, the cell debris was removed by centrifugation, and the remaining supernatant was purified over a 3 ml Ni-NTA affinity column with elution by a step gradient of 10, 50, 75, 100 and 500 mM imidazole. Purified recombinant proteins were dialyzed overnight against 50 mM HEPES pH 7.5, 50 mM NaCl and 1 mM DTT. The final concentration of wild type PvADA for crystallization was 10 mg/ml in the presence of 1 mM MT-coformycin. The crystallization condition of 25% PEG3350, 100 mM Hepes pH 7.5 and 0.2 M MgCl2 was identified using Hampton Research Index HT screening by sitting-drop vapor diffusion. A condensed cluster of rod-shaped crystals was obtained. PvADA-ΔAsp172 (10 mg/ml) failed to co-crystallize with MT-coformycin but co-crystallized with 3 mM MTA in 0.2 M MgCl2, 0.1 M BisTris pH 5.5 and 25% PEG 3350. MTA was not present in the co-crystallized crystal (data not shown). These crystals were soaked with 2 mM MT-coformycin for 1 hour to obtain MT-coformycin bound to PvADA-ΔAsp172. Crystals were transferred into a fresh drop of the crystallization solution containing 20% glycerol and rapidly frozen in liquid nitrogen.

Data collection and processing

X-ray diffraction data of MT-coformycin bound to PvADA was collected at Beamline 24-ID-E equipped with a MD-2 microdiffractometer at the Advanced Photo Source of Argonne National Laboratory. The microdiffractometer was used to search for a well separated diffraction pattern among the crystal cluster. X-ray diffraction of MT-coformycin bound PvADA-ΔAsp172 was collected at Beamline X29A at Brookhaven National Laboratory. All data were processed with HKL2000 program suite and the data processing statistics are provided in Table 1 (21).

Table 1
X-ray Data Collection and Refinement Statistics.

Structure determination and refinement

The crystal structure of PvADA bound to MT-coformycin was determined by molecular replacement in Molrep (22) using the published structure of PvADA bound to d-coformycin (PDB:2PGR) as the search model. The model without MT-coformycin and Zn2+ ion was first built in COOT (23) and refined in Refmac5. The Zn2+ ion was added and refined based on the crystal studies by Larson and colleagues (17). The MT-coformycin was added last using the Fo-Fc map and refined in Refmac5 (24). The crystal structure of PvADA-ΔAsp172 bound to MT-coformycin was determined in the same way as the crystal structure of PvADA bound to MT-coformycin. In both crystal structures, His 253 and Asp310 coordinate the catalytic zinc ion and are the only residues whose torsion angles are in the disallowed region of the Ramachandron plot. The disfavored torsion angles of His254 and Asp310 are also observed and described in the other three published structures of PvADA with bound ligands (17). The final models were validated by Procheck in the absence of the disordered residues, including the first six amino acid residues and C-terminal linker with the his-tag (25) and the refinement statistics were summarized in Table 1. The ligand-omit mFo-DFc difference map and ligand-omit 2mFo-DFc electron density map, presented in figure 4C and and5C5C were calculated using phases from the final refined protein models from which the ligands were removed (26). Figure 3A, ,4,4, ,5,5, ,66 and and7,7, where oxygen, nitrogen and sulfur atoms were colored in red, blue and magenta, respectively were made by Pymol (

Figure 3
A: The solvent accessible surface map of a cut-surface diagram of PvADA with bound MT-coformycin. The solvent accessible map (colored in gray) shows an enclosed cavity comprising the active site of PvADA. The 5’-methylthiol group of MT-coformycin ...
Figure 4
The comparative geometry of MT-coformycin and d-coformycin in the active site of PvADA. A: MT-coformycin in the catalytic site of PvADA is colored in yellow. The hydrogen bonds between the ribosyl group of MT-coformycin and adjacent catalytic site molecules ...
Figure 5
The Asp172-Ile180 region of ADA structures consisting of PvADA with MT-coformycin, bovine ADA with 6-hydroxy-1,6-dihydropurine riboside and PvADA-ΔAsp172 with MT-coformycin. A: Inhibitor-bound ADAs are depicted as ribbon diagrams and colored in ...
Figure 6
The stereo view of overlaid structures of MT-coformycin bound to PvADA or to PvADA-ΔAsp172. The MT-coformycin and PvADA are shown in yellow while MT-coformycin in the geometry bound to PvADA-ΔAsp172 is colored in green. Phe132 rotates ...
Figure 7
Overlaid catalytic site residues from PvADA with bound MT-coformycin are compared to bovine ADA with bound 6-hydroxy-1,6-dihydropurine riboside. Only the MT-coformycin ligand is shown (colored in yellow). Residues from bovine ADA (PDB ID: 1KRM) are shown ...


Identification and characterization of ADA from various Plasmodium species

Open reading frames for ADA from various Plasmodium species were identified, placed in expression vectors and the expressed proteins were purified to characterize the kinetic parameters, substrate and inhibitor specificity. P. falciparum was used as a control since it had been previously characterized (11, 14). The Plasmodium ADA amino acid sequences (Supporting information, Figure S1) exhibit identity values ranging from 62% to 72% as compared to PfADA. The Km values for adenosine ranged from 32 µM (P. gallinaceum) to 120 µM (P. knowlesi) while the Km values for MTA varied from 4.4 µM (P. berghei) to 115 µM (P. falciparum) and had no detectable catalytic activity with P. gallinaceum ADA (Table 2). All ADAs had kcat/Km values near 104 M−1s−1 with adenosine as substrate. Similar values were obtained with MTA as a substrate with the exception of PgADA which showed no detectable activity for MTA under conditions that would have detected 0.1% of that activity.

Table 2
Kinetic constants for Plasmodium ADAs with adenosine and 5’-methylthioadenosine (MTA) as substrates.

The Ki values for coformycin ranged from 2.3 nM (P. berghei) to 14 nM (P. falciparum) and the Ki* (slow-onset inhibition constant) varied from 0.25 nM (P. berghei) to 0.71 nM (P. vivax). Inhibition constants for MT-coformycin ranged from 3.2 nM (P. falciparum) to 48 nM (P. knowlesi) and showed no detectable inhibition with P. gallinaceum. Only PfADA and PbADA showed slow-onset inhibition for MT-coformycin, to give Ki* values of 0.25 and 5 nM, respectively (Table 3).

Table 3
Inhibition constants of transition state analogue inhibitors for plasmodial ADAs

Effect of coformycin and MT-coformycin on P. falciparum DNA biosynthesis

The effect of inhibiting parasite ADA or parasite plus host ADAs was measured in P. falciparum 3D7 strain cultured in human erythrocytes. DNA content was analyzed following treatment with coformycin (inhibitor of both human and parasite ADAs) or MT-coformycin (inhibitor of parasite ADA) for 72 hr in the presence of MTA. The inhibitors reduced parasite DNA synthesis with IC50 values of 2 nM for coformycin and 5 nM for MT-coformycin when 100 µM MTA was provided as the exogenous purine source (Figure 2). Adenosine and MTI were also tested as exogenous purine source but no inhibition was detected (data not shown).

Figure 2
Inhibition of DNA synthesis in P. falciparum cultures treated with coformycin and MT-coformycin. Infected erythrocytes were cultured in the presence of 100 µM MTA in the presence of the indicated inhibitor concentrations for 72 hr, followed by ...

Characterization of PvADA mutants

Based on the crystal structure of PvADA with bound MT-coformycin (see below) three PvADA mutants were designed (Supporting information, Table S2) to test the importance of Asp172 in methylthio-group recognition. Kinetic constants were determined under the same conditions as the PvADA wild-type (Table 4). All Asp172 mutants lost the ability to deaminate MTA and lost their high affinity for MT-coformycin. PvADA-Glu172 showed weak catalytic activity at elevated enzyme concentration (data not shown). In contrast to the loss of 5’-methylthio group specificity, deletion of Asp172 produced only a small effect on the catalytic efficiency for adenosine and the binding of coformycin.

Table 4
Kinetic and inhibition constants for PvADA-mutants.

Structure of PvADA with MT-coformycin

The structure of PvADA in complex with MT-coformycin was solved and refined to final resolution of 2.1 Å. PvADA is a typical TIM barrel, composed of eight α-helices and eight β-strands plus 16 α-helices of accessory structure. With MT-coformycin bound in the active site, the Asp172-Ile180 region (part of the β3 strand and β3/α12 loop) shifts approximately 15 Å compared to the open structure reported for apo P. yoelli ADA (17). This movement forms a closed foot-shaped cavity in the active site with the 2’- and 3’-hydroxyl groups of the MT-coformycin in the heel portion of the foot shaped cavity (Figure 3A).

Geometry of MT-Coformycin in PvADA

The 8-R-hydroxy-diazepine ring of MT-coformycin mimics the 6-R-hydroxyl tetrahedral Meisenheimer intermediate, similar to the transition state formed by water attack at C6 of adenosine. The 8-R-hydroxy-group of MT-coformycin replaces the attacking water nucleophile and is chemically stable since the position normally occupied by the leaving group amine in the adenosine transition-state is replaced by a hydrogen. This hydrogen is facing the solvent, consistent with solvent water acting as the proton donor to NH2 to form the NH3 leaving group. The diazepine ring is positioned near the toe of a foot-shaped cavity (Figure 3), where in addition to the zinc ion interactions, N1H of the ring donates a hydrogen bond to Glu229. This is a specific transition state interaction since in the normal reaction, N1 of adenosine is a H-bond acceptor while at the transition state, it is rehybridized to N1H to become a H-bond donor, as in the bound MT-coformycin. His253, Asp310 and Asp311 and the backbone of Gly201 interact with MT-coformycin similar to the interactions with adenosine, the normal substrate (Figure 4A and 4C).

The 5'-methylthio group is displaced from the 5'-hydroxyl group binding site and is replaced by a structurally defined water molecule held in place by hydrogen bonds with His44, Asp46 and the 2'- and 3’-hydroxyl groups of MT-coformycin (Figure 3A and Figure 4A). The 3'-hydroxyl group of MT-coformycin also hydrogen bonds with Asp172. This geometry of hydrogen bonds requires a 130° rotation of the ribosyl group around the glycosyl torsion angle (Table 5) with respect to bound adenosine (17). The position of the 5’-methylthio group (Figures 4A and 4B) causes it to be positioned almost 180° relative to the 5’-hydroxyl group found with bound adenosine, guanosine and d-coformycin (17). Despite the dramatic alteration of the MT-ribose group geometry with respect to bound ribosyl groups, only a slight shift of the diazepine ring (approximately 0.4 Å) occurs relative to the position of d-coformycin in the active site (Figure 4A and 4B). These changes place the 5'-methylthio group near the ankle region of the foot-shaped cavity, directed toward to protein surface (Figure 3A). The 5'-methylthio group fits closely into a hydrophobic cavity without room for crystallographically ordered water molecules. The geometry of bound MT-coformycin provides a sharp contrast to adenosine, guanosine and d-coformycin where the 5'-hydroxyl group hydrogen bonds with His44 and Asp46 and points toward the protein core (Figures 4A and 4B).

Table 5
Glycosyl torsion angles

MT-coformycin in the active site of PvADA-ΔAsp172

Based on the structural comparison between bovine ADA with bound 6-hydroxy-1,6-dihydropurine riboside, (Figure 5A, structure in gray) and PvADA with MT-coformycin bound, (Figure 5B, structure in yellow) and the primary sequence alignment between mammalian and plasmodial ADAs (Supplementary Figure S1), we hypothesized that an extra amino acid insertion into the PvADA is responsible for the observed conformational differences between mammalian and plasmodial ADAs. We deleted Asp172 of PvADA to shift Thr173 to the Thr172 position (Figures 5B and 5C). In mammalian ADAs, a Met is present at the equivalent position. The crystal structure of PvADA-ΔAsp172 in complex with MT-coformycin shows an open Asp172-Ile180 region (Figure 5A, structure in green), similar to bovine ADA in complex with the inhibitor (Figure 5A, structure in gray). In PvADA-ΔAsp172, with Thr172 now replacing the Asp group, the hydrogen bond with the 3’-hydroxyl group of MT-coformycin is lost (Figure 5B and 5C). In PvADA-ΔAsp172, the ribosyl group of MT-coformycin rotates to the orientation found for d-coformycin in PvADA (Figure 6). Phe132, located near His44 and Asp46, swings 30° away from the active site to accommodate the 5’-methylthio group (Figure 6). The unfavorable nature of this geometry for 5’-methylthioribose group binding is apparent in a 200-fold decrease in affinity for MT-coformycin (Table 4). The relative energetic contributions from methylthio and ribosyl group interactions are not available from these structures, but cause an approximate 3 kcal/mol energy loss.


Function of plasmodial and mammalian ADAs

Malarial parasites express relatively large quantities of ADA protein, suggesting its metabolic importance in the essential purine salvage pathway. Coformycin (10) and d-coformycin (27) are weak inhibitors of parasite growth in cultured erythrocytes but d-coformycin has been reported to decrease the parasitemia in primates infected with P. knowlesi (12). The action of ADA inhibitors in vivo suggests ADA as a potential target for antimalarials. ADA from P. falciparum also functions to deaminate MTA, a by-product of polyamine synthesis. In Plasmodium, ADA converts MTA to MTI and purine nucleoside phosphorylases converts MTI to hypoxanthine and 5-methylthio-α-D-ribose-1-phosphate. These enzymes form the only known pathway for MTA catabolism in Plasmodium, a necessary step for MTA recycling to methionine and S-adenosylmethionine. Human ADA has not evolved for MTA deamination activity since a methylthioadenosine phosphorylase (MTAP) and adenine phosphoribosyltransferase recycle the purine base in humans (28). Neither of these enzymes is encoded in P. falciparum genome (13). We tested the effect of coformycin and MT-coformycin in P. falciparum cultures in the presence of 100 µM MTA as an exogenous purine source (Figure 2). Under these conditions, coformycin and MT-coformycin reduce DNA synthesis of P. falciparum in vitro by approximately 80%, supporting plasmodial ADA as the pathway for MTA metabolism. Equivalent inhibition of DNA synthesis by coformycin and MT-coformycin establishes the parasite ADA as the target since coformycin is a pM inhibitor of host and parasite ADAs while MT-coformycin inhibits only the plasmodial ADA. Use of MT-coformycin as a potential antimalarial avoids the neurotoxicity of ADA inhibitors (such as Pentostatin, d-coformycin) in humans (29).

Species specificity for MT-coformycin action

Other Plasmodium species were examined for the substrate specificity of their ADAs. Plasmodium parasites that infect mammals (P. vivax, P. berghei, P. knowlesi and P. cynomolgi) showed robust catalytic efficiency for both adenosine and MTA. In contrast, P. gallinaceum that infects bird erythrocytes had no significant catalytic ability with MTA and was not inhibited by MT-coformycin (Table 2 and Table 3). Avian erythrocytes differ from those in mammals in that they have an average life of 35 days, are nucleated, larger in size and oblate ellipsoid in shape. The presence of nucleus, ribosomes, Golgi and mitochondria creates a different metabolic environment for the malaria parasite, as purines and polyamines can be salvaged or synthesized de novo in avian erythrocytes (30). The ADA specificity of P. gallinaceum suggests that activity for MTA deamination is unnecessary in this parasite as no MTA would be formed within the parasites if polyamines are salvaged from the host. This species difference provides a convenient tool to explore the catalytic site elements involved in 5’-methylthio group recognition. As shown below, the replacement of Asp172 with Glu172 in P. gallinaceum is important in its restricted activity for 5’-methylthioribosyl groups.

ADA catalytic site determinants for MTA recognition

Based on the crystal structures, sequence alignment and mutagenesis (see below), Asp172 is essential for the methylthio-specificity (Table 4). The crystal structure of MT-coformycin bound to PvADA established a remarkable spatial re-arrangement of methylthio-derivates. Although most plasmodial ADAs bind both coformycin and MT-coformycin with pM affinity, the orientation of the 5’-methylthioribosyl group of MT-coformycin is altered relative to d-coformycin. In nucleosides and d-coformycin, the 5’-hydroxyl group is H-bonded to His44 and Asp46. The 5’-methylthio group is not accommodated in the 5’-hydroxyl binding site, and the ribose is rotated by 130° to permit the 3'-hydroxyl group to hydrogen bond to Asp172. The 5’-hydroxyl group is replaced by a water molecule and the 5’-methylthio group relocates to a more hydrophobic region of the catalytic site, near Phe132 (Fig. 6). This rotation and the Asp172 interaction with the 3’-hydroxyl group is critical to permit MT-coformycin and MTA binding. In plasmodial ADAs, an Asp172 signature at the 5’-binding site indicates the ability of ADAs to accept MTA as a substrate and to be inhibited by tight binding of MT-coformycin. In mammalian ADAs, Met152 occupies the residue equivalent to PvADA-ΔAsp172. Met152 prevents binding of 5’-methylthio-derivatives due to a spatial clash, and this region of the sequence is completely conserved in mouse, bovine and human ADAs (Figure 7).

Substrate and inhibitor interactions in plasmodial ADAs

Of the six species of plasmodial ADAs examined here, only P. gallinaceum, in which Asp172 is replaced by Glu, is catalytically inactive with MTA. Mutated PvADAs (Asp172Ala, Asp172Glu and ΔAsp172) show catalytic characteristics similar to P. gallinaceum ADA. The hydrogen bond between the 3’-hydroxyl group of methylthio-derivates and Asp172 is required to permit methylthio-derivative binding and thereby assist in anchoring the purine or diazepine rings. Critically, this geometric change in the methylthioribose occurs while still permitting the purine group to achieve alignment with the catalytic site Zn2+ as needed for activation of the water nucleophile at the reaction center. Without the hydrogen bond between the 3’-hydroxyl group and Asp172, the methylthioribosyl group adopts the ribosyl conformation found in adenosine and d-coformycin binding. In that case, the methylthio-group is unfavorably positioned in the hydrophilic pocket near Asp46, His44 and a crystallographic water site. Kinetic evidence for this ribosyl conformational shift comes from the 200-fold weaker inhibition of MT-coformycin for PvADA-ΔAsp172 than for PvADA. Likewise, the P. gallinaceum ADA, containing Asp172Glu, binds MT-coformycin 14,500-fold weaker than PvADA.


Plasmodial ADAs are capable of using adenosine and MT-adenosine do so by accommodating the 5’-ribosyl and 5’-methylthioribosyl groups in different geometries. The 5’-ribosyl groups of substrates and inhibitors form hydrogen bonds with His44 and Asn46 in a 5’-hydroxyl group site. The 5’-methylthioribosyl substrates and inhibitors bind with the ribosyl groups in a different geometry with a hydrogen bond between the 3’-hydroxyl and Asp172. Mutation of Asp172 eliminates efficient deamination of MTA and MT-coformycin binding. A critical feature of this unusual geometrically-linked specificity in ADA is the ability to rotate the ribose groups of methylthioribosyl-derivatives in the active site of plasmodial ADAs while maintaining the register of the catalytic site elements with the site of hydrolytic deamination. Humans use MTAP to convert MTA to adenine and 5-methylthioribose 1-phosphate by phosphorolysis and human ADA does not deaminate MTA. Plasmodium has no MTAP and most species deal with MTA by deamination using the double-specificity ADAs featuring Asp172 as an essential catalytic site specificity element. A replacement of Asp172 on ADAs by intentional mutation causes loss of MTA deaminase activity and MT-coformycin binding. The absence of Asp172 in the case of P. gallinaceum has the same effect. Catalytic site flexibility in the malarial ADAs permits efficient purine metabolism with fewer expressed proteins.

Supplementary Material



adenosine deaminase
purine nucleoside phosphorylase
hypoxanthine-guanine-xanthine phosphorobosyl transferase
inosine 5’-monophosphate
aspartic acid
glutamic acid


This work was supported by NIH Research Grants AI049512, Training Grants F31 AI05665 and CM007288, U.S. Army Research Grant W81XWH-05-2-0025 and in part by a contract from the Medicines in Malaria (MMV) Consortium, Geneva, Switzerland. Parts of this manuscript were published in a thesis submitted by DCM in partial fulfillment of the requirements for the Degree Doctor of Philosophy in the Sue Golding Graduate Division of Medical Sciences, Albert Einstein College of Medicine, Yeshiva University. Data of X-ray diffraction for this study were measured at Beamline X29A of the National Synchrotron Light Source and Beamline 24-ID-E at the Northeastern Collaborative Access Team Beamlines of the Advanced Photon Source. Financial support of Beamline X29A of the National Synchrotron Light Source comes principally from the Offices of Biological and Environmental Research and of Basic Energy Sciences of the US Department of Energy, and from the National Center for Research Resources the National Institutes of Health. The Northeastern Collaborative Access Team Beamlines of the Advanced Photon Source is supported by award RR-15301 from the National Center for Research Resources at the National Institute of Health. Use of the Advanced Photon Source is supported by the U.S. Department of Energy, Office of Basic Energy Sciences, under Contract No. DE-AC02-06CH11357.


Primer sequences used for cloning of adenosine deaminase from different Plasmodium species and primer sequences used for site-directed mutagenesis of P. vivax ADA. Sequence alignment of plasmodial and mammalian ADAs. This material is available free of charge via the Internet at


1. Snow RW, Guerra CA, Noor AM, Myint HY, Hay SI. The global distribution of clinical episodes of Plasmodium falciparum malaria. Nature. 2005;434:214–217. [PMC free article] [PubMed]
2. Mendis K, Sina BJ, Marchesini P, Carter R. The neglected burden of Plasmodium vivax malaria. Am. J. Trop. Med. Hyg. 2001;64:97–106. [PubMed]
3. Jongwutiwes S, Putaporntip C, Iwasaki T, Sata T, Kanbara H. Naturally acquired Plasmodium knowlesi malaria in human, Thailand. Emerg. Infect. Dis. 2004;10:2211–2213. [PMC free article] [PubMed]
4. Cox-Singh J, Davis TM, Lee KS, Shamsul SS, Matusop A, Ratnam S, Rahman HA, Conway DJ, Singh B. Plasmodium knowlesi malaria in humans is widely distributed and potentially life threatening. Clin. Infect. Dis. 2008;46:165–171. [PMC free article] [PubMed]
5. Hyde JE. Drug-resistant malaria. Trends Parasitol. 2005;21:494–498. [PMC free article] [PubMed]
6. Hyde JE. Drug-resistant malaria - an insight. Febs J. 2007;274:4688–4698. [PMC free article] [PubMed]
7. de Koning HP, Bridges DJ, Burchmore RJ. Purine and pyrimidine transport in pathogenic protozoa: from biology to therapy. FEMS Microbiol. Rev. 2005;29:987–1020. [PubMed]
8. Hyde JE. Targeting purine and pyrimidine metabolism in human apicomplexan parasites. Curr. Drug Targets. 2007;8:31–47. [PMC free article] [PubMed]
9. Kicska GA, Tyler PC, Evans GB, Furneaux RH, Schramm VL, Kim K. Purine-less death in Plasmodium falciparum induced by immucillin-H, a transition state analogue of purine nucleoside phosphorylase. J. Biol. Chem. 2002;277:3226–3231. [PubMed]
10. Cassera MB, Hazleton KZ, Riegelhaupt PM, Merino EF, Luo M, Akabas MH, Schramm VL. Erythrocytic adenosine monophosphate as an alternative purine source in Plasmodium falciparum. J. Biol. Chem. 2008;283:32889–32899. [PubMed]
11. Tyler PC, Taylor EA, Frohlich RF, Schramm VL. Synthesis of 5'-methylthio coformycins: specific inhibitors for malarial adenosine deaminase. J. Am. Chem. Soc. 2007;129:6872–6879. [PMC free article] [PubMed]
12. Webster HK, Wiesmann WP, Pavia CS. Adenosine deaminase in malaria infection: effect of 2'-deoxycoformycin in vivo. Adv. Exp. Med. Biol. 1984;165(Pt A):225–229. [PubMed]
13. Gardner MJ, Hall N, Fung E, White O, Berriman M, Hyman RW, Carlton JM, Pain A, Nelson KE, Bowman S, Paulsen IT, James K, Eisen JA, Rutherford K, Salzberg SL, Craig A, Kyes S, Chan MS, Nene V, Shallom SJ, Suh B, Peterson J, Angiuoli S, Pertea M, Allen J, Selengut J, Haft D, Mather MW, Vaidya AB, Martin DM, Fairlamb AH, Fraunholz MJ, Roos DS, Ralph SA, McFadden GI, Cummings LM, Subramanian GM, Mungall C, Venter JC, Carucci DJ, Hoffman SL, Newbold C, Davis RW, Fraser CM, Barrell B. Genome sequence of the human malaria parasite Plasmodium falciparum. Nature. 2002;419:498–511. [PubMed]
14. Ting LM, Shi W, Lewandowicz A, Singh V, Mwakingwe A, Birck MR, Ringia EA, Bench G, Madrid DC, Tyler PC, Evans GB, Furneaux RH, Schramm VL, Kim K. Targeting a novel Plasmodium falciparum purine recycling pathway with specific immucillins. J. Biol. Chem. 2005;280:9547–9554. [PubMed]
15. Trackman PC, Abeles RH. Methionine synthesis from 5'-SMethylthioadenosine. Resolution of enzyme activities and identification of 1-phospho-5-S methylthioribulose. J. Biol. Chem. 1983;258:6717–6720. [PubMed]
16. Olszewski KL, Morrisey JM, Wilinski D, Burns JM, Vaidya AB, Rabinowitz JD, Llinas M. Host-parasite interactions revealed by Plasmodium falciparum metabolomics. Cell Host Microbe. 2009;5:191–199. [PMC free article] [PubMed]
17. Larson ET, Deng W, Krumm BE, Napuli A, Mueller N, Van Voorhis WC, Buckner FS, Fan E, Lauricella A, DeTitta G, Luft J, Zucker F, Hol WG, Verlinde CL, Merritt EA. Structures of substrate- and inhibitor-bound adenosine deaminase from a human malaria parasite show a dramatic conformational change and shed light on drug selectivity. J. Mol. Biol. 2008;381:975–988. [PMC free article] [PubMed]
18. Wang Z, Quiocho FA. Complexes of adenosine deaminase with two potent inhibitors: X-ray structures in four independent molecules at pH of maximum activity. Biochemistry. 1998;37:8314–8324. [PubMed]
19. Desjardins RE, Canfield CJ, Haynes JD, Chulay JD. Quantitative assessment of antimalarial activity in vitro by a semiautomated microdilution technique, Antimicrob. Agents Chemother. 1979;16:710–718. [PMC free article] [PubMed]
20. Quashie NB, de Koning HP, Ranford-Cartwright LC. An improved and highly sensitive microfluorimetric method for assessing susceptibility of Plasmodium falciparum to antimalarial drugs in vitro. Malar. J. 2006;5:95. [PMC free article] [PubMed]
21. Otwinowski Z, Minor W. Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods in Enzymology. 1997;276:307–326.
22. Vagin AA, Teplyakov A. MOLREP: an Automated Program for Molecular Replacement. J. Appl. Cryst. 1997;30:1022–1025.
23. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 2004;60:2126–2132. [PubMed]
24. Murshudov GN, Vagin AA, Dodson EJ. Refinement of macromolecular structures by the maximum-likelihood method, Acta Crystallogr. D Biol. Crystallogr. 1997;53:240–255. [PubMed]
25. Laskowski RA, McArthur MW, Moss DS, Thornton JM. PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Cryst. 1993;265:283–291.
26. Read RJ. Improved Fourier coefficients for maps using phases from partial structures with errors. Acta Crystal. 1986;A42:140–149.
27. Roth E, Jr, Ogasawara N, Schulman S. The deamination of adenosine and adenosine monophosphate in Plasmodium falciparum-infected human erythrocytes: in vitro use of 2'deoxycoformycin and AMP deaminase-deficient red cells. Blood. 1989;74:1121–1125. [PubMed]
28. Singh V, Schramm VL. Transition-state structure of human 5'-methylthioadenosine phosphorylase. J. Am. Chem. Soc. 2006;128:14691–14696. [PMC free article] [PubMed]
29. Sauter C, Lamanna N, Weiss MA. Pentostatin in chronic lymphocytic leukemia. Expert Opin. Drug Metab. Toxicol. 2008;4:1217–1222. [PubMed]
30. Stevens L. Avian Biochemistry and molecular biology. New York: Cambridge University Press; 1996.