HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 5, 3219-3225, February 1, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/5/3219    most recent M105905200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kicska, G. A. Articles by Schramm, V. L. Search for Related Content PubMed PubMed Citation Articles by Kicska, G. A. Articles by Schramm, V. L. Social Bookmarking                      What's this?

Transition State Analogue Inhibitors of Purine Nucleoside Phosphorylase from Plasmodium falciparum^*

Gregory A. Kicska, Peter C. Tyler^§, Gary B. Evans^§, Richard H. Furneaux^§, Kami Kim^¶**, and Vern L. Schramm

From the Departments of  Biochemistry, ^¶ Medicine (Division of Infectious Diseases), and  Microbiology and Immunology, Albert Einstein College of Medicine, Bronx, New York 10461 and the ^§ Carbohydrate Chemistry Team, Industrial Research Limited, Lower Hutt, New Zealand

Received for publication, June 25, 2001, and in revised form, November 8, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Immucillins are logically designed transition-state analogue inhibitors of mammalian purine nucleoside phosphorylase (PNP) that induce purine-less death of Plasmodium falciparum in cultured erythrocytes (Kicska, G. A., Tyler, P. C., Evans, G. B., Furneaux, R. H., Schramm, V. L., and Kim, K. (2002) J. Biol. Chem. 277, 3226-3231). PNP is present at high levels in human erythrocytes and in P. falciparum, but the Plasmodium enzyme has not been characterized. A search of the P. falciparum genome data base yielded an open reading frame similar to the PNP from Escherichia coli. PNP from P. falciparum (P. falciparum PNP) was cloned, overexpressed in E. coli, purified, and characterized. The primary amino acid sequence has 26% identity with E. coli PNP, has 20% identity with human PNP, and is phylogenetically unique among known PNPs with equal genetic distance between PNPs and uridine phosphorylases. Recombinant P. falciparum PNP is catalytically active for inosine and guanosine but is less active for uridine. The immucillins are powerful inhibitors of P. falciparum PNP. Immucillin-H is a slow onset tight binding inhibitor with a K_i* value of 0.6 nM. Eight related immucillins are also powerful inhibitors with dissociation constants from 0.9 to 20 nM. The K_m/K_i* value for immucillin-H is 9000, making this inhibitor the most powerful yet reported for P. falciparum PNP. The PNP from P. falciparum differs from the human enzyme by a lower K_m for inosine, decreased preference for deoxyguanosine, and reduced affinity for the immucillins, with the exception of 5'-deoxy-immucillin-H. These properties of P. falciparum PNP are consistent with a metabolic role in purine salvage and provide an explanation for the antibiotic effect of the immucillins on P. falciparum cultured in human erythrocytes.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Plasmodium falciparum is responsible for the majority of deaths due to malaria (1). An essential step in the life cycle of the parasites is cellular replication of the trophozoite in the human erythrocyte. RNA and DNA synthesis creates a need for large quantities of purines, which must be salvaged from the mammalian host, because P. falciparum is a purine auxotroph (2-5). This purine demand makes growth of Plasmodium sensitive to disruption of pathways for purine salvage (6). Hypoxanthine has been reported to be the major purine precursor for purine salvage, and P. falciparum growth is reduced by culturing in the presence of xanthine oxidase, which depletes both the medium and erythrocyte of hypoxanthine (7, 8). The sole pathway of hypoxanthine production in P. falciparum and in human erythrocytes is through the phosphorolysis of inosine to hypoxanthine, a reaction catalyzed by PNP¹ (3). These features of purine metabolism in humans and P. falciparum make the PNPs from both organisms of interest as potential targets for antimalarials.

Purine nucleoside phosphorylases from bacterial and mammalian sources belong to distinct families. Family 1 contains most bacterial PNPs, uridine phosphorylases from both bacteria and mammals and methylthioadenosine phosphorylase from Sulfolobus solfataricus. Proteins in this family invariably contain the signature consensus sequence (G/S/T)XG(L/I/V/M)GX(P/A)SX(P/A)SX(G/S/T/A)IX₃EL. Bacterial PNPs in this family are hexamers and are termed "high molecular mass" PNPs because the average aggregate mass of the hexamer is near 150 kDa (9, 10). The bacterial PNPs accept 6-aminopurines (e.g. adenosine) and accept 6-oxopurines (e.g. inosine) as substrates when a single PNP is expressed in the organism.

Family 2 includes all mammalian PNPs, some of the bacterial PNPs (Bacillus subtilis (11) and Bacillus stearothermophilus (12)), methylthioadenosine phosphorylases from eukaryotes and xanthosine phosphorylase from Escherichia coli (13). Proteins in this family invariably contain the consensus sequence (L/I/V)X₃GX₂HX(L/I/V/M/F/Y)X₄(L/I/V/M/F)X₃(A/T/V)X_1-2 (L/I/V/M)X(A/T/V)X₄(G/N)X_3-4(L/I/V/F/M)₂X₂(S/T/N)(S/A)XG(G/S)(L/I/V/M). The PNPs in Family 2 are homotrimers and are termed "low molecular mass" PNPs with a mean aggregate weight of 90 kDa. Family 2 PNPs accept inosine, guanosine, and 2'-deoxyguanosine as substrates but not adenosine. A third group of PNPs do not fit into either family (e.g. Cellulomonas sp. (14)).

An early characterization of the catalytic properties of P. falciparum PNP was accomplished in extracts of parasites cultured in erythrocytes from children with a genetic deficiency in PNP (15). This approach was used to distinguished the large excess of human erythrocyte PNP from the P. falciparum PNP activity and established the presence of PNP catalytic activity in P. falciparum. The present work provides a genetic approach to pure preparations of large quantities of P. falciparum PNP. These preparations have been used to define substrate specificity and transition state inhibitors.

We report the molecular cloning, expression, and isolation of P. falciparum PNP as well as the identification of powerful transition state inhibitors. The primary sequence is related to the E. coli PNP (Family 1 PNP), and sequence alignment shows conservation of most active site residues, even though the consensus pattern is not strictly observed.

The availability of P. falciparum PNP permits the exploration of the recently discovered purine-less death of P. falciparum caused by immucillins. Immucillins are picomolar inhibitors of mammalian PNP that inhibit P. falciparum growth in cultured human erythrocytes (16). We hypothesized that inhibition of P. falciparum growth requires inhibition of both human and P. falciparum PNPs, because both host and parasite contain high PNP activities (17, 18). Immucillins inhibit P. falciparum PNP with equilibrium dissociation constants (K_i*) as low as 0.6 nM, and represent the tightest binding inhibitors yet reported for this enzyme. The substrate specificity of P. falciparum PNP is consistent with the role of purine salvage, and the inhibition of P. falciparum PNP by immucillins is appropriate to target this pathway for novel antibiotic action against malaria.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Reagents-- Imm-H ((1S)-1-(9-deazahypoxanthin-9-yl)-1,4-dideoxy-1,4-imino-D-ribitol) and other immucillins were synthesized from D-gulonolactone and chemically protected 9-deazahypoxanthine (33). Purity and structure were established by NMR and purity confirmed by high-performance liquid chromatography. Nucleosides and deoxynucleosides were purchased from Sigma Chemical Co. (St. Louis, MO). 6-Methylpurine deoxyriboside was purchased from Invitrogen (Rockville, MD). Human erythrocyte purine nucleoside phosphorylase and xanthine oxidase from bovine buttermilk were obtained from Sigma (St. Louis, MO). Calf intestinal alkaline phosphatase was obtained from Promega (Madison, WI).

Identification and Cloning of P. falciparum PNP-- Sequence data for P. falciparum was obtained from The Sanger Center website (available at www.sanger.ac.uk/Projects/P_falciparum). Protein sequences for mammalian and bacterial PNPs were used to search the complete sequence data base from the P. falciparum 3D7 Genome Sequencing Centers (Sanger Center, Stanford, and The Institute for Genomic Research or the Gene Sequencing Tag Project at the University of Florida) (19). Preliminary sequence data for P. falciparum chromosomes 2, 10, 11, and 14 were obtained from The Institute for Genomic Research website (www.tigr.org). Sequencing of chromosomes 2, 10, 11, and 14 was part of the International Malaria Genome Sequencing Project and was supported by awards from the Burroughs Wellcome Fund and the U.S. Department of Defense. The algorithm, TBLASTN, was used to find candidate genes with primary sequence homology to human, bovine, mouse, yeast, E. coli, Baccillus subtilis I, B. subtilis II, B. streptococcus, or Treponema pallidum PNPs. A candidate sequence that was embedded in an open reading frame of appropriate length was targeted for molecular cloning.

Polymerase chain reaction (PCR) was used to amplify the coding region from genomic DNA isolated from FCB strain of P. falciparum (provided by Dr. Tom McDonald, Albert Einstein College of Medicine). Primers were designed to flank the ends of the open reading frame, creating a PCR product for blunt-end insertion into an expression vector. The sense primer was 5'-GATAATCTTTTACGCCATTTAAAAATAAG-3'. The antisense primer contained 5'-TCAGGCATATTTGGTAGCTAATTTTG-3' complementary to the intended sequence, except that a mutation was included to avoid primer-dimer formation. The primer altered the N-terminal sequence from MDNL~ to MALDNL~. The PCR cycling parameters included denaturation at 95 °C (1 min), annealing at 45 °C (1 min), and extension at 72 °C (1 min). After 35 cycles, a single band of the appropriate size was visualized by agarose gel analysis. The PCR product was ligated into the protein expression vector pTrcHis2 TOPO (Invitrogen, Carlsbad CA). The antisense primer contained a stop codon to prevent the expression of the His tag located on the vector. A clone with a correctly oriented insert was identified by restriction analysis with NcoI, PvuI, and EcoRI (Roche Molecular Biochemicals, Petersburg, VA) and confirmed with automated DNA sequencing (Albert Einstein College of Medicine DNA Sequencing Facility).

Expression and Purification of PNP-- E. coli strain BL21(DE3) was freshly transformed and the culture grown at 37 °C in LB media with 100 µg/ml ampicillin to an A₆₀₀ of 0.25. After 500-fold dilution in fresh medium, cultures were grown to A₆₀₀ = 0.85 and induced with 0.6 mM isopropyl-1-thio--D-galactopyranoside. After 3 h at 37 °C, cells were pelleted and frozen at 70 °C for subsequent purification.

Cells were resuspended in 50 mM KPO₄, pH 7.5, 2 mM EDTA, 0.1% Triton X-100, and 0.2 mg/ml lysozyme in 5 ml of buffer/g of cells. Protein was extracted from cells with mechanical disruption using a French press with 16,000 p.s.i. pressure. The extract was centrifuged at 50,000 × g, 60 min 4 °C. The ammonium sulfate fraction precipitating between 35 and 70% saturation was resuspended and dialyzed into 50 mM imidazole, 5 mM KPO₄, pH 6.5. The dialysate was applied to a Q-Sepharose Fast Flow ion-exchange column (Amersham Biosciences, Inc.) and eluted with a gradient from 0 to 1 M NaCl in the imidazole phosphate buffer (5 column volumes). The PNP activity and a 27-kDa band on SDS-PAGE eluted between 450 to 750 mM NaCl. Protein fractions were concentrated to 2.5 ml then separated by gel filtration on a Superdex 200 HR 10/30 column (Amersham Biosciences, Inc.). Protein fractions containing PNP activity were concentrated to 25 mg/ml using an Ultrafree-15 Centrifugal filter device (Millipore) and subjected to ammonium sulfate fractionation, from 20 to 26% saturation, where the P. falciparum PNP precipitated. Purity was estimated to be greater than 95% based on denaturing polyacrylamide gel electrophoresis.

Enzyme concentration was determined by absorption in the far-UV near the peptide band at 205 nm (E = 37.0 + 120 × A₂₈₀/A₂₀₅) (20, 21). Protein concentration was also estimated by the assay of Bradford. Protein was stored at 10 mg/ml at 4 °C in 50 mM imidazole, 10 mM KPO₄, pH 7.0. No significant activity loss was noted after 12 weeks (<10%), and kinetic and inhibition assays were generally performed within 30 days of enzyme purification.

Sequence Comparison of P. falciparum PNP with Purine and Other Phosphorylases-- The protein sequence data bases from other organisms identified sequences similar to P. falciparum PNP. All non-redundant protein directories were searched using two iterations of position-specific iterated BLAST (PSI-BLAST) (BLOSUM62, inclusion threshold = 0.002) (22). This method generates a position-specific scoring matrix from an initial round of Gapped BLAST searching using conventional substitution scores (23). The position-specific scoring matrix was used for subsequent rounds of BLAST searching. High similarity hits were used for multiple sequence alignment with ClustalX (24) using default parameters and the BLOSUM series of tables (25). This alignment was used to calculate the similarity between sequences, using the formula [no. of identical residues/(alignment length  no. of gap residues)]. Identity between sequences was determined by pairwise alignment using the Lipman-Pearson method with a gap penalty = 4 and gap penalty length = 12. Phylogenetic trees were generated with ClustalX using the Neighbor-Joining method and excluding gaps (26) and displayed with NJPLOT. Reliability of phylogenetic structure was determined using bootstrapping analysis with 1000 replications (27). Values above 950 were considered to give high support to phylogenetic tree structure and above 800 are considered significant. Orthologues to P. falciparum PNP in other Plasmodium species were identified by using the P. falciparum PNP sequence to search multiple data bases using the BLAST algorithm (P. berghei, P. vivax, P. chabaudi, and P. yoelii) as described above.

Catalytic Activity and Determination of Kinetic Constants-- Partially purified P. falciparum PNP was used for kinetic and inhibition studies (60-70% as determined by denaturing polyacrylamide gel electrophoresis), some of which were repeated with highly purified enzyme. The results with both preparations were the same. Catalytic activity was determined in 50 mM KPO₄, pH 7.4 with variable inosine concentration (40 µM to 1 mM). Hypoxanthine was converted to uric acid in a coupled assay containing 160 milliunits/ml xanthine oxidase. The reaction was followed by measuring the formation of uric acid with E₂₉₃ = 12.9 mM¹ cm¹ (28). Uridine phosphorylase activity was measured in 50 mM KPO₄, pH 7.5. Product formation was determined by direct UV monitoring of the conversion of uridine to uracil at A₂₇₂ (E₂₆₀ = 2.9 mM¹ cm¹). Adenosine phosphorylase and 5'-methylthioadenosine phosphorylase activities were determined using continuous assays. In a coupled assay containing alkaline phosphatase, the reaction was followed by measuring formation of adenine from adenosine or 5'-methylthioadenosine (E₂₅₆ = 1.9 mM¹ cm¹). Guanosine and deoxyguanosine phosphorylase activity were also determined using a continuous assay. In a coupled assay containing alkaline phosphatase, the reaction was followed by monitoring the formation of guanine from guanosine or deoxyguanosine (E₂₅₈ = 5.2 mM¹ cm¹ (29)). Using similar methods, phosphorolysis of 6-methyl-purine deoxyriboside was monitored at 260 nm. In substrate specificity assays, sufficient P. falciparum PNP was added to reliably detect catalytic activity at 0.2% of the activity with inosine.

Inhibition Studies-- Inhibition of P. falciparum PNP was measured by addition of enzyme to the complete assay mixture containing inhibitor. Assays used conditions of excess substrate, so that inhibitor concentration exceeded enzyme concentration by at least 10-fold (30). The inhibition constant for the initial rate period was determined from fits of the initial data to the equation: v = (k_cat × A)/(K_m (1 + I/K_i)), where v = the initial reaction rate, k_cat = the maximal catalytic rate, A = substrate concentration, K_m = Michaelis constant, I = inhibition concentration and K_i = dissociation constant for the E·I complex. Reaction rates were monitored continuously to establish both the initial reaction rate (v) and to determine if a second steady-state rate (v_s) occurred following a slow-onset inhibition phase. This approach permitted the determination of K_i* from the data set, where the value K_i* is for competitive inhibition following the completion of slow onset inhibition; v_s = (k_cat × A)/(K_m (1 + I/K_i*) + A), where A is substrate concentration, K_m is the Michaels constant, and K_i* is the equilibrium inhibition constant that results after the slow-onset phase of inhibition (30, 31). The observed rate v_s is the rate following slow-onset inhibition, and k_cat is the uninhibited rate at saturating substrate concentration.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Isolation and Cloning of P. falciparum PNP-- Searching the P. falciparum genome data base with the protein sequence for E. coli PNP returned a single sequence with significant homology represented as independent entries (Fig. 1). Pairwise alignment revealed 26% identity to E. coli PNP and 24% identity to B. subtilis type II PNP. The P. falciparum gene encodes a single open reading frame corresponding to 245 amino acids with a molecular mass of 26,857 Da and a calculated isoelectric point at pH 6.3. The reading frame for P. falciparum PNP was amplified using PCR and DNA sequencing of the product showed no difference from the data base sequence, except for the changes introduced at the N terminus. Sequences similar to P. falciparum PNP were also identified in other Plasmodium species. These include P. berghei, P. vivax, and P. yoelii with 79%, 28%, and 78% identity to the P. falciparum PNP in regions of 168, 111, and 244 amino acid residues in length, respectively.

View larger version (50K): [in this window] [in a new window]   Fig. 1.   Protein sequence comparison of type 1 family PNPs with P. falciparum PNP. Residues blocked in black are identical to E. coli sequence. "C " indicates the residue range that contains the consensus signature. P_FALCIPARUM, P. falciparum PNP; PNP_ECOLI, E. coli PNP (9); PNP_HELPY, H. pylori PNP (43); PNPI_STEAR, B. stearothermophilus PNP II (44); PNPII_BSUB, B. subtilis PNP II (Ref. 11 and Wambutt, R., Wedler, H., Lapidus, A., Sorokin, A., and Ehrlich, D. (1997) Direct submission to Swiss-prot O34925); PNP_TREPA, T. pallidum PNP (46). Alignment was performed using ClustalX (see "Experimental Procedures").

Protein Sequence Alignment-- Searching the protein data base with the P. falciparum PNP sequence yielded homologous sequences containing the "purine and other phosphorylases" Family 1 signature (PROSITE, PS01232). There was weaker but significant homology to sequences with the Family 2 signature of purine other phosphorylases (data not shown). In a multiple sequence alignment of the putative P. falciparum PNP with the PNP, UDP, and MTAP protein sequences from mammalian and microbial sources, P. falciparum PNP had ~40% similarity to Family 1 phosphorylases that are bacterial PNPs and UDPs, and significant but less similarity (~20%) to Family 2 phosphorylases that are mostly of mammalian origin (data not shown). In contrast, Family 1 PNPs have an average similarity to each other of 62.8% and Family 2 PNPs have an average similarity to each other of 87%. P. falciparum PNP does not contain sufficient homology to either family to make it a typical member of either group.

The phylogenetic tree shows that P. falciparum PNP is an outlier in terms of genetic distance from both families of purine phosphorylases (Fig. 2). Phylogeny of the Plasmodium genus makes it more closely related to mammalian species than to bacteria. However, P. falciparum PNP is genetically more distant from mammalian PNPs (p < 0.01) and closer to bacterial PNPs (p < 0.01) (Fig. 2). With respect to all Family 1 phosphorylase members, it is closer to bacterial uridine phosphorylases than it is to other PNPs or to the bacterial methylthioadenosine phosphorylase from B. sulfolobus.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Phylogenetic tree of P. falciparum PNP, 10 PNPs, and 3 UDPs from Type 1 and 2 families of phosphorylases. The tree is derived using the NJ method (see "Experimental Procedures"). The scale bar of 0.05 represents evolutionary distance and is in units of substitutions per site. Values at the branch points are bootstrap values based on 1000 replications, and the value indicates the number of cases in which the species to the right appear as a cluster. PNP_HUMAN, human PNP (47); PNP_BOVINE, bovine PNP (48); UDP_ECOLI, E. coli uridine phosphorylase (35); UDPI_HALO, Halobacterium sp. uridine phosphorylase; UDP_VCOLI, Vibrio cholerae uridine phosphorylase (49). XNP_ECOLI, E. coli xanthosine phosphorylase (13); PNPI_STEAR, B. stearothermophilus PNP I (12); PNPI_BSUB, B. subtilis PNP I (50); MTAP_SULF, S. solfataricus methylthioadenosine phosphorylase (51). Other abbreviations are the same as in Figs. 1 and 2.

Sequence alignment with only bacterial PNPs indicates that P. falciparum PNP has 11 of the 17 catalytic site residues identified in the E. coli PNP crystal structure (Table I and Fig. 1) (9, 32). Four of the six non-conserved residues are conservatively substituted, and another is nearby in the alignment. The general acid/base contact at N7 of the purine ring in E. coli PNP is Asp²⁰⁴, which is also in H-bond contact to the 6-amino group for adenosine phosphorolysis (33, 34). This region is not highly conserved between E. coli PNP and P. falciparum PNP, but Asp²⁰⁵ in P. falciparum PNP may function in the same way (Fig. 1). The displacement of Asp²⁰⁵ relative to the E. coli PNP is particularly interesting, because, unlike other Type 1 phosphorylases, adenosine is not a substrate for P. falciparum PNP ((15); see below).

Although P. falciparum PNP is phylogenetically closer to UDPs than other PNPs, it is missing important features of the UDPs (Fig. 2). The E. coli UDP contains a flexible loop (Tyr¹⁶³-Phe¹⁸⁰), which extends to neighboring subunits. The loop is missing in P. falciparum PNP and bacterial PNPs. Also absent is the E. coli UDP Asp⁵, which is involved in catalytic activity (35). Current crystallographic studies of E. coli UDP have not identified residues with specific functions in catalysis but only residues near the active site (35). These residues are no more conserved in P. falciparum PNP than they are in the bacterial PNPs.

The family 1 phosphorylase consensus pattern (G/S/T)XG(L/I/V/M)GX(P/A)SX(G/S/T/A)IX₃EL is not complete in P. falciparum PNP (Fig. 1). The 8th and 11th consensus residues (Ser and Ile) are substituted by Gly and Val. P. falciparum PNP shows no consistent features of the consensus pattern of the Type 2 phosphorylases.

Expression and Purification of PNP-- Bacterial recombinant P. falciparum PNP protein was expressed as ~5% of soluble protein in induced cultures (Fig. 3). Purine nucleoside phosphorylase activity in cell extracts was increased substantially relative to non-induced cultures. The specific catalytic activity of P. falciparum PNP increased at each purification step, and the final protein purity was estimated to be >95% (Fig. 3). The calculated molecular mass of the expressed protein with N-terminal Met is 27,041 amu. Matrix-assisted laser desorption analysis gave a mass of 26,906 Da, consistent with a mass of 26,910 Da expected for N-terminal Met-processed enzyme.

View larger version (62K): [in this window] [in a new window]   Fig. 3.   Expression and purification P. falciparum PNP. The lane labeled "Standard" is molecular weight standards. The lanes labeled "Control" and "Induced" represent uninduced cultures and the same culture 3 h after induction. The lane labeled "Purified" is the protein after the last purification step. The SDS-PAGE gel was stained with Coomassie Blue.

Substrate Specificity of P. falciparum PNP-- The substrate activity with inosine and guanosine gave k_cat/K_m values of 7.4 × 10⁴ and 1.3 × 10⁴ M¹ s¹, respectively (Table II). The K_m of 5 µM for inosine is less than the value of 40 µM reported for human PNP and is consistent with this nucleoside serving as an in vivo substrate (36). In contrast, deoxynucleosides are poor substrates with catalytic efficiencies of 10% or less than that for inosine. Mammalian PNPs have efficient phosphorolysis of 2'-deoxy-guanosine, because accumulation of this metabolite as dGTP causes T-cell immunodeficiency. The 29% sequence identity between E. coli UDP and P. falciparum PNP suggested functional similarity to this enzyme. This suggestion was supported by weak uridine phosphorylase catalytic activity for P. falciparum PNP compared with inosine and guanosine. Adenosine is not a substrate for this P. falciparum PNP, a trait shared with mammalian PNPs, however, bacterial PNPs do have activity with adenosine. Other substrates for bacterial PNPs include 6-methylpurine-deoxyriboside and ara-inosine, but these are inactive as substrates of P. falciparum PNP (Table II). Thus, the substrate specificity is distinct from both mammalian and bacterial PNPs. The primary protein sequence of P. falciparum PNP is similar also to bacterial uridine phosphorylases (e.g. E. coli, Fig. 2) and 5'-methylthioadenosine phosphorylase from S. solfataricus, and uridine but not 5'-methylthioadenosine is a substrate.

Immucillins Inhibit P. falciparum PNP-- Immucillin-H and immucillin-G were designed as transition state analogues of mammalian PNP and were modified to explore the inhibitor specificity of P. falciparum PNP. Deletions of sugar hydroxyls help quantitate energy of protein-ribose contacts. Substitution of C8 with F or a methyl group alters the N7 pK_a, the proposed proton acceptor in the transition-state. Nine of eleven immucillins bind considerably tighter to P. falciparum PNP than substrates (Table II, Table III, Fig. 4). Several immucillins exhibit slow-onset binding kinetics characteristic of transition-state analogues (Fig. 5). Imm-H and Imm-G are both 9-deazapurine iminoribitol analogues of inosine and guanosine. Immucillin-H caused inhibition during initial rate assays, followed by slow-onset of a more powerful inhibition phase. Immucillin-G demonstrated only a single inhibitory phase. The equilibrium dissociation constants (K_i*) of 0.6 and 0.9 nM were obtained for Imm-H and Imm-G, respectively (Table III). These constants are in agreement with the relative K_m values for these substrates (15 and Table II). Immucillins with altered sugar properties include the 2'-deoxy-analogues of Imm-H and Imm-G, which exhibit K_i* = 2.2 nM and 8.5 nM, respectively. The 3'-deoxy-Imm-H is a stronger inhibitor with a K_i* = 1.7 nM. Immucillins with an altered pK_a value for the N7 proton include 8-F-Imm-H and 8-Me-Imm-H. These substitutions alter the strength of the interaction between the purine ring and the enzymatic groups that activate the purine leaving groups to achieve the transition state. They exhibit K_i* values of 19.6 nM and >10 µM, respectively. The 0.6 nM K_i* exhibited by Imm-H is the result of slow-onset inhibition, in which the initial interaction (K_i = 29 nM) strengthens by a factor of 48 to form the inhibited enzyme (Table III, Fig. 5). This is the most powerful inhibitor yet described for P. falciparum PNP, and exhibits a K_m/K_i* = 9000. 2-Delta-C2-Imm-H PZ has a deletion at C2 in the purine ring, and a 5-membered pyrrazolo- ring. With a K_i value of 2.7 nM, it binds with affinity similar to the 2'-deoxy-immucillin analogues. Ara-Imm-H is a good inhibitor (K_i* = 1.2 nM) even though ara-inosine is not a substrate. In contrast, Imm-A is an analogue of adenosine, and binds poorly to the enzyme with K_i > 10 µM.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Structures of the immucillin inhibitors.

View larger version (35K): [in this window] [in a new window]   Fig. 5.   Slow-onset inhibition of P. falciparum PNP by Imm-H (inset). Reactions were initiated by the addition of enzyme to substrate and the indicated concentrations of inhibitor (see "Experimental Procedures"). The ordinate scale indicates uric acid formation from the coupled assay with xanthine oxidase. Rate of product formation at the 0- to 5-min interval can be used to determine the initial dissociation constant for the inhibitor (Ki). The rate at 25-30 min is post-slow-onset inhibition and can be used to calculate the equilibrium dissociation constant, including the slow-onset phase (Ki*, outer panel).

Immucillins are known to inhibit human hypoxanthine-guanine phosphoribosyltransferase and Plasmodium hypoxanthine-guanine-xanthine phosphoribosyl transferase when phosphorylated to the 5'-monophosphate (37). Metabolic phosphorylation can be prevented by elimination of the 5'-hydroxy, as in 5'-deoxy-Imm-H, which is incapable of being 5'-phosphorylated. Despite the loss of the 5'-hydroxy group, this analogue still inhibits P. falciparum PNP with a K_i* of 7.4 nM, to give a K_m/K_i* of ~10³.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Plasmodium falciparum is a purine auxotroph and depends on an adequate supply of hypoxanthine for growth. In both the red cell and in the parasite, hypoxanthine is produced from the nucleotide pool by the PNP-catalyzed phosphorolysis of inosine. The immucillin inhibitors of mammalian PNP have been shown to kill P. falciparum cultured in human erythrocytes, and this effect can be reversed by hypoxanthine addition (16). The catalytic activity of PNP is high in erythrocytes and is even higher in P. falciparum (17, 18). A goal of this work was to determine if the P. falciparum PNP was inhibited by the immucillins and if this inhibitor specificity differed for the human and P. falciparum enzymes. Cloning and isolation of P. falciparum PNP enabled the kinetic characterization of powerful inhibitors with demonstrated potential as antimalarials.

The Plasmodium genome data base contains a single sequence with significant homology to bacterial PNPs and UDPs. The expressed P. falciparum PNP protein is similar in molecular weight, kinetic properties, and isoelectric value to the PNP described from extracts of P. falciparum (15). Purified P. falciparum PNP was characterized by high affinity for inosine and a low catalytic turnover number; kinetic properties associated with salvage pathways (38). The kinetic constants and substrate specificity were markedly different from the properties of E. coli and human PNPs (15, 39). Mass spectrometry of the purified protein gave the expected mass for the P. falciparum PNP construct expressed in E. coli.

Two categories have been described to classify the PNP primary sequences, however, the P. falciparum PNP does not fit easily into either. The primary sequence is closest to the bacterial PNPs (Family 1 phosphorylases) but has greater genetic distance from bacterial PNPs than from UDPs and retains some UDP catalytic activity (Family 2 phosphorylases). Despite these similarities, P. falciparum PNP does not contain a complete signature sequence of either and has the unique property of low K_m, low k_cat to facilitate capture of substrate (see below). Phylogenetically, the enzyme does not cluster with any of the known PNPs. Other examples of atypical PNPs include the Cellulomonas and Mycobacterium species, but they cluster together near the Family 2 trimeric phosphorylases and share little sequence homology to P. falciparum PNP. The atypical PNP from P. falciparum is the only one associated with a purine auxotroph, and its structure reflects the unique selective pressures associated with its metabolic role in the purine salvage pathway. As additional genomic data accumulates from other purine auxotrophs, a more complete comparison of PNPs from other purine-requiring organisms, may reveal a third family of phosphorylases.

The inhibitor specificity of P. falciparum PNP reveals Imm-H as a 0.6 nM slow-onset inhibitor; followed in potency by Imm-G and ara-Imm-H at 0.9 and 1.2 nM, respectively. In the human enzyme, dissociation constants for these inhibitors follow a different order with Imm-G the best inhibitor at 29 pM, with similar K_i* values for 2'-deoxy-Imm-H, 2'-deoxy-Imm-G, and Imm-H (Table III). This specificity reflects the different metabolic roles of the enzymes, with deoxyguanosine the primary substrate for the human enzyme and inosine as that for P. falciparum PNP. Contacts with the 2'-hydroxyl group differ considerably between human and P. falciparum PNPs, because 2'-deoxy-Imm-H binds more favorably to the human enzyme than does Imm-H. In contrast, Imm-H binds more favorably than 2'-deoxy-Imm-H to P. falciparum PNP. However, this specificity is reversed for the 5'-deoxy-Imm-H, because for P. falciparum PNP it binds only 11-fold more weakly than Imm-H, whereas in the human enzyme 5'-deoxy-Imm-H binds 130-fold less than Imm-H. The structure of bovine PNP in complex with Imm-H and phosphate reveals a 2.9-Å hydrogen bond between His²⁵⁷ and the 5'-hydroxyl; however, this group does not appear to be conserved in the E. coli or P. falciparum enzymes (9, 34). Differences in the 2'-hydroxyl contacts for human and P. falciparum PNP are also evident from the 2'-deoxy-immucillins mentioned above, although the protein contact at this position is a conserved Met in both human and P. falciparum PNP (Table I).

A comparison of catalytic residues from P. falciparum PNP and E. coli PNP shows that most residues are conserved but the two enzymes have different substrate specificity. Of the six non-conserved residues, five are associated with base binding, thus explaining the difference in base specificity. The isozyme-specific phosphorolysis of 6-methylpurine-deoxyriboside by E. coli PNP has been used to selectively kill cells transfected with the enzyme (40, 41). Although the substrate range for P. falciparum PNP resembles the mammalian enzyme, the active site differs. It may therefore be possible to use prodrugs specific for P. falciparum PNP to achieve selective toxicity against the parasite. Differences in the purine-binding portion of the catalytic sites are also apparent in inhibitor binding. For example, fluoro and methyl substitutions at the 8 position of the purine show 4- and 200-fold increases in the K_i* for the human enzyme but resulted in a 30- and >1000-fold increase in the constant for P. falciparum PNP.

The primary metabolic role for human PNP is phosphorolysis of deoxyguanosine, because the genetic deficiency of PNP causes a T-cell deficiency due to accumulation of dGTP (42). The increase in serum deoxyguanosine is responsible for this pathology; consequently, human PNP has high catalytic activity with deoxyguanosine, to give a k_cat/K_m of 1.3 × 10⁶ M¹ s¹ (15). In P. falciparum, purine salvage is essential, and inosine serves as a major purine precursor in the erythrocyte. Thus, the catalytic efficiency of P. falciparum PNP is the greatest for this substrate and less for deoxynucleosides. Low k_cat values are observed in P. falciparum PNP and are accompanied by low K_m values. These kinetic features are hallmarks of metabolite-salvage pathways, permitting efficient capture of substrate. Malarial hypoxanthine-guanine-xanthine phosphoribosyl transferase exhibits similar kinetic properties and is responsible for converting the hypoxanthine formed by P. falciparum PNP to inosine 5'-phosphate (37). However, the low k_cat of P. falciparum PNP could also be an artifact of overexpression.

The immucillins inhibit P. falciparum PNP with dissociation constants as low as 0.6 nM, making them the most powerful inhibitors known for this target of the purine salvage pathway in P. falciparum. Dissociation constants for several immucillin analogues indicate differences in catalytic features that may be used for development of isozyme-specific inhibitors and substrates. The immucillins described here provide a first generation approach to the design of antipurine antibiotics against purine salvage pathways in P. falciparum. The companion report (16) demonstrates that immucillins induce purineless death in P. falciparum cultured in human erythrocytes. These findings provide the information necessary to design a second generation of related inhibitors, which might provide species-specific inhibition of P. falciparum PNP and be of use in the treatment and prevention of malaria.

	FOOTNOTES

^* This work was supported in part by Research Grant GM41916 from the National Institutes of Health and by Medical Scientist Training Program Training Grant GM07288.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EBI Data Bank with accession number(s) AF426159.

^** A Burroughs Wellcome New Investigator in Molecular Parasitology.

To whom correspondence should be addressed: Dept. of Biochemistry, Albert Einstein College of Medicine, 1300 Morris Park Ave., Bronx, NY 10461. Tel.: 718-430-2813; Fax: 718-430-8565; E-mail: vern@aecom.yu.edu.

Published, JBC Papers in Press, November 13, 2001, DOI 10.1074/jbc.M105905200

	ABBREVIATIONS

The abbreviations used are: PNP, purine nucleoside phosphorylase; P. falciparum PNP, Plasmodium falciparum PNP; MTAP, methylthioadenosine phosphorylase; K_i, the rapidly reversible inhibitory dissociation constant; K_i*, the equilibrium inhibition dissociation constant following slow-onset inhibition; Imm, immucillin.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				D. C. Madrid, L.-M. Ting, K. L. Waller, V. L. Schramm, and K. Kim Plasmodium falciparum Purine Nucleoside Phosphorylase Is Critical for Viability of Malaria Parasites J. Biol. Chem., December 19, 2008; 283(51): 35899 - 35907. [Abstract] [Full Text] [PDF]

				M. J. Downie, K. Kirk, and C. B. Mamoun Purine Salvage Pathways in the Intraerythrocytic Malaria Parasite Plasmodium falciparum Eukaryot. Cell, August 1, 2008; 7(8): 1231 - 1237. [Full Text] [PDF]

				K. Chaudhary, L. M. Ting, K. Kim, and D. S. Roos Toxoplasma gondii Purine Nucleoside Phosphorylase Biochemical Characterization, Inhibitor Profiles, and Comparison with the Plasmodium falciparum Ortholog J. Biol. Chem., September 1, 2006; 281(35): 25652 - 25658. [Abstract] [Full Text] [PDF]

				A. Lewandowicz, E. A. T. Ringia, L.-M. Ting, K. Kim, P. C. Tyler, G. B. Evans, O. V. Zubkova, S. Mee, G. F. Painter, D. H. Lenz, et al. Energetic Mapping of Transition State Analogue Interactions with Human and Plasmodium falciparum Purine Nucleoside Phosphorylases J. Biol. Chem., August 26, 2005; 280(34): 30320 - 30328. [Abstract] [Full Text] [PDF]

				J. E. Lee, V. Singh, G. B. Evans, P. C. Tyler, R. H. Furneaux, K. A. Cornell, M. K. Riscoe, V. L. Schramm, and P. L. Howell Structural Rationale for the Affinity of Pico- and Femtomolar Transition State Analogues of Escherichia coli 5'-Methylthioadenosine/S-Adenosylhomocysteine Nucleosidase J. Biol. Chem., May 6, 2005; 280(18): 18274 - 18282. [Abstract] [Full Text] [PDF]

				L.-M. Ting, W. Shi, A. Lewandowicz, V. Singh, A. Mwakingwe, M. R. Birck, E. A. T. Ringia, G. Bench, D. C. Madrid, P. C. Tyler, et al. Targeting a Novel Plasmodium falciparum Purine Recycling Pathway with Specific Immucillins J. Biol. Chem., March 11, 2005; 280(10): 9547 - 9554. [Abstract] [Full Text] [PDF]

				J.-W. Ting, M.-F. Wu, C.-T. Tsai, C.-C. Lin, I.-C. Guo, and C.-Y. Chang Identification and characterization of a novel gene of grouper iridovirus encoding a purine nucleoside phosphorylase J. Gen. Virol., October 1, 2004; 85(10): 2883 - 2892. [Abstract] [Full Text] [PDF]

				W. Shi, L.-M. Ting, G. A. Kicska, A. Lewandowicz, P. C. Tyler, G. B. Evans, R. H. Furneaux, K. Kim, S. C. Almo, and V. L. Schramm Plasmodium falciparum Purine Nucleoside Phosphorylase: CRYSTAL STRUCTURES, IMMUCILLIN INHIBITORS, AND DUAL CATALYTIC FUNCTION J. Biol. Chem., April 30, 2004; 279(18): 18103 - 18106. [Abstract] [Full Text] [PDF]

				A. Lewandowicz, P. C. Tyler, G. B. Evans, R. H. Furneaux, and V. L. Schramm Achieving the Ultimate Physiological Goal in Transition State Analogue Inhibitors for Purine Nucleoside Phosphorylase J. Biol. Chem., August 22, 2003; 278(34): 31465 - 31468. [Abstract] [Full Text] [PDF]

				G. A. Kicska, P. C. Tyler, G. B. Evans, R. H. Furneaux, V. L. Schramm, and K. Kim Purine-less Death in Plasmodium falciparum Induced by Immucillin-H, a Transition State Analogue of Purine Nucleoside Phosphorylase J. Biol. Chem., January 25, 2002; 277(5): 3226 - 3231. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/5/3219    most recent M105905200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kicska, G. A. Articles by Schramm, V. L. Search for Related Content PubMed PubMed Citation Articles by Kicska, G. A. Articles by Schramm, V. L. Social Bookmarking                      What's this?