The Conserved Serine-Tyrosine Dipeptide in Leishmania donovani Hypoxanthine-guanine Phosphoribosyltransferase Is Essential for Catalytic Activity*

(Received for publication, December 2, 1996, and in revised form, January 21, 1997)

Armando Jardim and Buddy Ullman Dagger

From the Department of Biochemistry and Molecular Biology, Oregon Health Sciences University, Portland, Oregon 97201-3098

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

Crystal structures of hypoxanthine-guanine phosphoribosyltransferase (HGPRT) proteins have implied that the translocation of a flexible loop containing a highly conserved Ser-Tyr dipeptide is necessary for the protection of the proposed oxocarbonium ion transition state of the enzyme (Eads, J. C., Scapin, G. T., Xu, Y., Grubmeyer. C., and Sacchettini, J. C. (1994) Cell 78, 325-334; Schumacher, M. A., Carter, D., Roos, D. S., Ullman, B., and Brennan, R. G. (1996) Nature Struct. Biol. 3, 881-887). An essential role for this Ser-Tyr dyad in HGPRT catalysis has now been verified biochemically and genetically for the Leishmania donovani HGPRT employing a combination of protein modifying reagents and site-directed mutagenesis. Incubation of HGPRT with either tetranitromethane or diethyl pyrocarbonate inactivated the enzyme completely, and peptide sequence analysis revealed that tetranitromethane treatment modified the Tyr residue within the Ser95-Tyr96 dipeptide. Analysis of site-directed mutants confirmed that both amino acids were vital for phosphoribosylation activity. Mutant HGPRTs, S95A, S95E, Y96F, and Y96V, exhibited dramatic reductions in their catalytic capabilities of 2-3 orders of magnitude, whereas HGPRTs containing conservative substitutions, S95C and S95T, displayed only a 2-3-fold decrease in kcat. Km values for the substrates of the forward and reverse reactions were largely unchanged for all HGPRT constructs, except for a 4-5-fold decrease in the Km value of the Y96F and Y96V mutants for phosphoribosylpyrophosphate. Expression of L. donovani hgprt constructs in Escherichia coli indicated that wild type and S95T HGPRTs complemented bacterial phosphoribosyltransferase deficiencies, whereas the S95A and S95C mutants complemented weakly, and the S95E, Y96F, and Y96V HGPRT did not support bacterial growth. These data authenticate that the Ser-Tyr dipeptide that is conserved among all members of the HGPRT family is essential for phosphoribosylation of purine nucleobases by HGPRT.


INTRODUCTION

Protozoan parasites constitute a devastating public health and socioeconomic burden in the tropical and subtropical regions of the world. In the absence of effective vaccines, empirically derived drugs are the only means available to treat parasitic infections. However, the current arsenal of antiparasitic drugs is far from ideal, as these agents are moderately to highly toxic, possibly mutagenic and/or carcinogenic, and require protracted treatment regimens. The control of parasitic diseases has also been complicated by the emergence of drug-resistant strains (1), further underscoring the need for novel antiparasitic agents. Rational development of new drugs that selectively target protozoan pathogens requires the exploitation of biochemical or metabolic differences between parasite and host. As protozoan parasites, unlike mammalian cells, are auxotrophic for purines and consequently rely on purine appropriation from the host for survival and proliferation (2), the purine salvage pathway has stimulated considerable interest as a paradigm for antiparasitic drug development. One enzyme that is central to purine acquisition in many parasites (2) is hypoxanthine-guanine phosphoribosyltransferase (HGPRT)1 which catalyzes the phosphoribosylpyrophosphate (PRPP)-dependent phosphoribosylation of hypoxanthine and guanine to the nucleotide level. In some parasites, HGPRT also recognizes xanthine as a substrate (3) and is, therefore, termed a hypoxanthine-guanine-xanthine phosphoribosyltransferase (HGXPRT). Genes and cDNAs encoding HG(X)PRT (HG(X)PRT) from a number of protozoan parasites have now been cloned, sequenced, and overexpressed in Escherichia coli, and the recombinant HG(X)PRT enzymes have been purified and characterized kinetically (3). Multisequence alignments demonstrate several short regions of amino acid homology among HG(X)PRT family members that are flanked by much longer regions without significant sequence similarity. The crystal structures for the product-bound form of the human HGPRT (4) and the Tritrichomonas foetus HGXPRT (5) have revealed, however, a common core framework within the tertiary structure that is conserved among other members of the phosphoribosyltransferase (PRT) family (6-8). These structures have established functional roles for all of the conserved regions in HG(X)PRT proteins, except one motif that is located on a flexible loop distal to the active site. This flexible loop encompasses a Ser-Tyr dipeptide that is found in all HG(X)PRTs for which sequence is available. The recently determined crystal structures of the HGXPRT from Toxoplasma gondii in the absence of ligand and in the presence of product have demonstrated that this flexible loop in the apoenzyme is shifted toward the active site (9). Schumacher et al. (9) further conjectured that this loop in the apoenzyme is poised to cover the catalytic pocket to protect the transition state intermediate from hydrolysis by the bulk solvent.

These structures, however, have not established which amino acid residues are critical to the catalytic mechanism of HG(X)PRT proteins. To identify amino acids requisite for the phosphoribosylation reaction, we have employed chemical modification and site-directed mutagenesis to definitively establish a critical catalytic role for the Ser95-Tyr96 dipeptide of the HGPRT from the protozoan parasite Leishmania donovani.


EXPERIMENTAL PROCEDURES

Chemicals and Reagents

Acetic anhydride, diethyl pyrocarbonate (DEPC), diisopropyl fluorophosphate, ethyl dimethylaminopropyl carbodiimide, iodoacetate, p-chloromercuric benzoate, phenyl glyoxal, tetranitromethane (TNM), guanosine monophosphate dialdehyde, inosine monophosphate (IMP), PRPP, hypoxanthine, and guanine were obtained from Sigma. Radiolabeled [14C]guanine (56 mCi/mmol) was purchased from Moravek (Brea, CA). All other reagents were of the highest quality commercially available.

HGPRT Expression in E. coli and Purification of Recombinant Protein

L. donovani HGPRT was ligated into the pBAce expression vector and transformed into SPhi 606 (Delta pro-gpt-lac, thi, hpt) E. coli as described (10, 11). This strain lacks both the bacterial xanthine-guanine and hypoxanthine phosphoribosyltransferase activities (11). HGPRT expression was induced in low phosphate induction (LPI) medium as reported (10). One-liter bacterial cultures were harvested by centrifugation, and the cell pellets were resuspended in 15 ml of 20 mM Tris-HCl, pH 8.0, 10 mM MgCl2, 2.0 mM dithiothreitol (TMD) buffer and lysed with two passes through a French Press. Lysates were clarified by centrifugation at 31,000 × g for 30 min and supernatants applied to a mixed bed column of DEAE-cellulose and AG 50-8X (2.7 × 10 cm) pre-equilibrated with TMD buffer. Void volume fractions containing HGPRT activity, as determined radiometrically (10), were pooled and solid (NH4)2SO4 added to 40% saturation. Precipitates were removed by centrifugation, and the supernatants were brought to 80% saturation with (NH4)2SO4 to recover HGPRT. The protein pellets were dissolved in 4.0 ml of TMD buffer and dialyzed overnight at 4 °C against 4 liters of 20 mM sodium carbonate, pH 9.8, 10 mM MgCl2, 2.0 mM dithiothreitol. Dialysates were clarified by centrifugation and applied to a Q-Sepharose column (1.5 × 30 cm) equilibrated with 50 mM sodium carbonate, pH 9.8, 10 mM MgCl2 and developed with a 500-ml linear gradient of 0-0.4 M NaCl in the equilibration buffer. Column eluates were monitored at 280 nm and protein fractions assayed for HGPRT activity (10) and analyzed for purity by SDS-polyacrylamide gel electrophoresis (12). Fractions containing a homogeneous 24-kDa polypeptide were pooled, adjusted to pH 8.0 with 1.0 M Tris-HCl, pH 8.0, and concentrated to 10 mg/ml using a Centriprep concentrator (Amicon, Beverly, MA). All enzyme preparations were frozen in dry ice/acetone and stored at -70 °C, except the S95C mutant, which was used immediately after purification. The S95C HGPRT was relatively unstable and lost activity upon storage. All other HGPRT mutants, as well as the wild type enzyme, were stable at -70 °C for up to 6 months.

Chemical Modification of L. donovani HGPRT

HGPRT preparations were diluted to 10 µg/ml in 100 mM Tris phosphate, 10 mM MgCl2 (either pH 5.0, 7.0, or 9.0), and protein modifying agents were added to a final concentration of 10 mM, with the exception of guanosine monophosphate dialdehyde which was used at a concentration of 100 µM. Samples were incubated at 20 or 0 °C for 60 min. HGPRT activity was measured by adding 2.0 µl of the enzyme solution to 18 µl of a reaction mixture containing 40 µM [14C]guanine (56 mCi/mmol), 1.0 mM PRPP, 100 mM Tris-HCl, 10 mM MgCl2, and 2.0 mM dithiothreitol, pH 8.0, and incubating at 37 °C for 4 min. The HGPRT assay was linear with time and protein under these conditions. Fifteen-µl aliquots of the reaction mixtures were applied to DEAE-Whatman filters (Maidstone, United Kingdom) and washed with dH2O on a Millipore manifold (Bedford, MA) to remove unreacted [14C]guanine. [14C]GMP reaction product was quantitated by liquid scintillation.

Quantitation and Isolation of TNM-labeled Peptides

HGPRT (2.3 nmol) was dissolved in 1.0 ml of 100 mM Tris phosphate, 10 mM MgCl2 in the presence or absence of 10 mM PRPP and preincubated at 20 °C for 5 min. TNM (230 nmol) dissolved in 5 µl of methanol was then added to each sample, and the mixture was incubated for an additional 20 min at 20 °C. The reaction was then terminated by the addition of trichloroacetic acid to a final concentration of 15%. The HGPRT precipitate was dissolved in 1.0 ml of 100 mM sodium acetate, pH 4.7, and dialyzed overnight against 2.0 liters of 20 mM acetic acid to remove residual nitroformate chromophore (13). The dialyzed protein was lyophilized and dissolved in 200 µl of 50 mM ammonium bicarbonate, pH 8.0, and the levels of nitrotyrosine were quantitated spectrophotometrically at 428 nm using an extinction coefficient of 4100 M-1 cm-1 (13). The nitrated protein was then digested with the endoproteinase Glu-C (Sigma) at a molar ratio of 20 µl (HGPRT:Glu-C) for 20 h at 37 °C. Nitrated peptides were purified by high performance liquid chromatography (HPLC) using a reversed phase C18 column (Bio-Rad) equilibrated with 20 mM ammonium bicarbonate, pH 8.0, and developed with a 1%/min linear gradient of 50:50 methanol, 1-propanol. Fractions exhibiting 428 nm absorbance were collected, lyophilized, and subjected to N-terminal sequence analysis on a Applied Biosystems 473 pulsed liquid phase sequenator equipped with a 120 PTH analyzer (Foster City, CA).

Mass Spectrometry

To determine molecular masses of nitrated peptides, the peptides were fractionated by HPLC on a Brownlee C8 column (2.1 × 100 mm) (Applied Biosystems) and developed with a 1%/min linear gradient of 1% acetic acid, 2-propanol at a flow rate of 0.2 ml/min. The HPLC column effluent was interfaced via an electrospray source to a Fisons VG Quattro mass spectrometer (Cheshire, United Kingdom) to determine parent ion masses.

Site-directed Mutagenesis of the L. donovani HGPRT

Mutations within L. donovani HGPRT were generated by subcloning the EcoRI-XbaI fragment from the HGPRT-pBAce construct described by Allen et al. (10) into the corresponding sites in the pSelect mutagenesis vector (Promega). This EcoRI-XbaI fragment encompasses the entire HGPRT protein coding region, as well as the phoA promoter, thus enabling the expression of the mutated genes in E. coli without further subcloning. Single amino acid substitutions were introduced into HGPRT with 25-base pair mutagenic oligonucleotides containing the appropriate base replacements centrally located in the primer following the protocols outlined in the brochure provided by Promega.

Complementation Analysis

Wild type and mutant HGPRT-pSelect constructs were transformed into SPhi 609 (Delta pro-gpt-lac, hpt, thi, pup, purHJ) E. coli, a purine auxotroph that lacks the same PRT activities as the SN606 cells (11). Transformants were isolated on LB plates containing 100 µg/ml ampicillin and 50 µg/ml streptomycin. A single bacterial colony was then picked and resuspended in 200 µl of LPI medium supplemented with 100 µg/ml ampicillin and 50 µg/ml streptomycin (10). Twenty-µl aliquots were then dispensed into 3.0 ml of LPI medium containing both antibiotics and adenine, adenine/hypoxanthine, adenine/guanine, or adenine/guanosine. Each nucleobase or nucleoside was present at a concentration of 125 µM. Cultures were incubated at 37 °C with vigorous shaking (300 rpm) for 8 h, and cell densities were measured spectrophotometrically at 600 nm.

Steady-state Kinetic Measurements of the Forward and Reverse Reactions of HGPRT

Initial rate measurements for the forward reaction were determined using a 1.0-cm path length on a Beckman DU640 spectrophotometer equipped with a kinetic software package. Reactions were carried out in 100 mM Tris-HCl, pH 8.0, 10 mM MgCl2, 2.0 mM dithiothreitol at 37 °C either in 1.0 mM PRPP and 1-100 µM guanine or hypoxanthine to determine the Km for the base or at 40 µM guanine and 25-800 µM PRPP to ascertain the Km for PRPP. IMP and GMP formation were monitored at 243 and 257 nm using extinction coefficients of 2200 M-1 cm-1 and 4200 M-1 cm-1, respectively. The reverse reaction in which IMP in the presence of PPi is converted to hypoxanthine and PRPP was monitored using the xanthine oxidase-coupled enzyme assay previously reported (14).


RESULTS

Effects of Protein Modifying Reagents on HGPRT Activity

To identify amino acid residues within HGPRT that are essential for catalytic activity, the purified enzyme was treated with a battery of protein modifying reagents, including iodoacetate, p-chloromercuric benzoate, acetic anhydride, ethyl dimethylaminopropyl carbodiimide, diisopropyl fluorophosphate, phenyl glyoxal, diethyl pyrocarbonate, and guanosine monophosphate dialdehyde. The specificity of these protein modifying reagents is indicated in Table I. Surprisingly, the only reagent among these that effectively reduced L. donovani HGPRT activity at 20 °C was DEPC (Table I). Although DEPC reacts preferentially with the imidazole side chains of His residues at pH 5.0-7.0 (15), the reagent can also modify to a lesser extent the epsilon -amino group of Lys and the hydroxyl group of Tyr residues under alkaline conditions (16). Incubation of HGPRT with 10 mM DEPC at pH 9.0 resulted in >95% loss in activity after 60 min at 0 °C (Table I). No loss in HGPRT activity was observed with DEPC at either pH 5.0 or 7.0, implying that a His residue is not requisite for HGPRT activity (Table I). Since acetic anhydride, a Lys modifying agent, failed to abrogate enzyme activity, this suggested that DEPC probably modified an essential Tyr residue. That Tyr was the amino acid within HGPRT being modified by DEPC was substantiated by changes in the aromatic region of the UV spectrum after treatment with DEPC. These spectral changes included a decrease in the 278 nm absorbance with a concomitant increase in the absorbances at 260 and 290 nm, implying a modification of a phenolic hydroxyl group (17). As DEPC carboethoxylations are relatively unstable (18), identification of the specific Tyr residue within the L. donovani HGPRT that was inactivated by DEPC was not attempted.

Table I.

Effects of protein modifying reagents on the L. donovani HGPRT activity


Reagent Target residues Percent enzyme activity

Iodoacetate Cys 101  ± 28
p-Chloromercuric benzoate Cys 96  ± 12
Acetic anhydride Lys 114  ± 11
Ethyl dimethylaminopropyl carbodiimide/ammonium acetate Asp, Glu 98  ± 15
Diisopropyl fluorophosphate Ser 104  ± 3
Phenyl glyoxal Arg 94  ± 6
DEPC,a pH 5.0 His, Lys, Tyr 99  ± 4
DEPC,a pH 7.0 His, Lys, Tyr 97  ± 2
DEPC,a pH 9.0 His, Lys, Tyr 4  ± 3
TNM,a pH 5.0 Tyr 96  ± 9
TNM,a pH 7.0 Tyr 100  ± 5
TNM,a pH 9.0 Tyr 20  ± 15
Guanosine monophosphate dialdehyde Lys, Arg 92  ± 11

a Reactions were incubated for 60 min at 0 °C.

Modification of L. donovani HGPRT with TNM

To establish that a Tyr residue is essential for HGPRT catalysis, the L. donovani enzyme was reacted with the Tyr-specific protein modifying reagent TNM, which nitrates Tyr residues at the meta positions in a pH-dependent manner (19). Incubation of HGPRT with a 100-fold molar excess of TNM at 0 °C abolished >80% of the enzymatic activity when the reaction was performed at pH >9.0, while no significant loss of activity by TNM was observed at either pH 5.0 or 7.0 (Table I).

To ascertain if the Tyr residue(s) modified by TNM were located proximal to the PRPP or nucleobase-binding sites of HGPRT (4), protection experiments were conducted at 20 °C in the presence of either PRPP or IMP. HGPRT inactivation in the presence of PRPP exhibited a concentration-dependent biphasic profile not observed for the TNM-treated enzyme in the absence of PRPP and resulted in a 50% reduction in specific activity after 5 min, followed by a much slower rate of inactivation (Fig. 1). Conversely, IMP did not protect HGPRT from TNM inactivation, even at a concentration as high as 10 mM, although the rate of inactivation at the higher IMP concentrations was slightly less (Fig. 1).


Fig. 1. Protection of L. donovani HGPRT from TNM inactivation by ligands. HGPRT (10 µg/ml) was preincubated at 20 °C for 5 min with no ligand (bullet ), 1 mM PRPP (open circle ), 10 mM PRPP (black-square), 1 mM IMP (black-triangle), or 10 mM IMP (square ) then treated with a 100 molar excess of TNM. At the indicated times, HGPRT activity was assayed.
[View Larger Version of this Image (16K GIF file)]


As nitration of Tyr side chains generates a chromophore with a lambda max at 428 nm, it is possible to quantitate spectroscopically the number of Tyr residues modified within a protein after TNM treatment (13). Treatment of L. donovani HGPRT with a 100-fold molar excess of TNM for 20 min at 20 °C in the absence of ligand resulted in the nitration of 1.7 ± 0.4 mol of Tyr/mol of HGPRT. Inclusion of 10 mM PRPP in the reaction mixture decreased the amount of Tyr nitration to 0.84 ± 0.10 mol of Tyr/mol of HGPRT. To determine the number of Tyr residues required for HGPRT catalysis, the extent of HGPRT inactivation as a function of TNM incubation time was correlated with the level of Tyr nitration. As demonstrated in Fig. 2, HGPRT inactivation was directly proportional to the number of modified Tyr residues in the protein after TNM treatment, and the nitration of one Tyr residue resulted in >80% loss of HGPRT activity.


Fig. 2. Titration of essential Tyr residues in HGPRT with TNM. Wild type HGPRT was treated with a 100-fold molar excess of TNM for various durations. Aliquots were removed at the indicated times and assayed for HGPRT activity (bullet ) and quantitated for the presence of Tyr(NO2) residues (open circle ) as described under "Experimental Procedures."
[View Larger Version of this Image (21K GIF file)]


Characterization of Nitrotyrosine Containing Peptides

To determine which Tyr residues in the L. donovani HGPRT were modified by TNM, TNM-treated enzyme was digested with endoproteinase Glu-C, and the nitrated peptides were resolved by HPLC and sequenced. As illustrated in Fig. 3, TNM treatment of HGPRT in the absence of PRPP yielded five 428 nm absorbing peaks diagnostic of nitrotyrosine-containing peptides. Untreated HGPRT did not contain any 428 nm absorbing peptides. Peptide NP1 was a tetrapeptide, Ser-Tyr-Arg-Glu, corresponding to residues 189-192 of the L. donovani HGPRT (10) (Table II). The second Edman degradation cycle yielded two phenylthiohydantoin (PTH)-derivatives, one eluting with the PTH-Tyr standard. The second was a novel PTH-derivative, presumably PTH-nitrotyrosine (PTH-Tyr(NO2)). This PTH-Tyr(NO2) derivative was found in all five 428-nm absorbing peptides obtained from Glu-C digestion of TNM-treated HGPRT. Peptides NP2 and NP3 both exhibited the same N-terminal Tyr-Tyr-Glu sequence. NP3 was the tripeptide, while NP2 was an incomplete Glu-C digestion product coinciding with residues 203-208 of the L. donovani HGPRT (Table II). Peptides NP4 and NP5 revealed the N-terminal sequence, Phe-Ile-Xaa-Ala-Ser-Ser-(Tyr/Tyr(NO2))-Gly-Thr, that corresponded to residues 90-98 of the enzyme. Mass spectroscopy analysis of NP4 and NP5 yielded parent molecular ions (M + H+) of 1281 and 1327 mass units, respectively. These masses approximate the molecular mass of the putative Glu-C fragment of this region of the protein with the sequence Phe-Ile-Cys-Ala-Ser-Ser-Tyr-Gly-Thr-Gly-Val-Glu (1234 Da). The differences between the calculated and observed masses for peptides NP4 and NP5 were 46 and 92 Da, respectively, which can be ascribed to the mononitration and dinitration of Tyr96.


Fig. 3. Isolation of Tyr(NO2)-containing peptides. L. donovani HGPRT was treated with TNM and digested with endoproteinase Glu-C as described under "Experimental Procedures." Nitrated peptides were isolated by reversed phase HPLC on a C18 column equilibrated with 20 mM NH4HCO3, pH 8.0, buffer then developed with a 1%/min 0-60 linear gradient of methanol:propanol (1:1) at a flow rate of 0.75 ml/min. The column effluent was monitored at 428 nm to detect Tyr(NO2)-containing peptides.
[View Larger Version of this Image (16K GIF file)]


Table II.

Microsequence analysis of Tyr(NO2) containing peptides isolated by endoproteinase Glu-C digest of TNM-treated L. donovani HGPRT


Peptide designation Peptide sequence Molecular mass (M + H+) Corresponding residues in HGPRT sequence

NP1 SYRE NDa 189 -192
NP2 YYEKPE ND 203 -208
NP3 YYE ND 203 -205
NP4 FIXASSYGTb 1281 90 -98
NP5 FIXASSYGT 1327 90 -98

a ND, not determined.
b X, unidentified PTH residue.

Of the four Tyr residues that are modified by TNM treatment of the L. donovani HGPRT, Tyr96, Tyr190, Tyr203, and Tyr204, multisequence alignments reveal that only Tyr96 is conserved in all members of the HG(X)PRT family, as well as in the xanthine-guanine PRT E. coli (20) and the guanine PRT from Giardia lamblia (21) (Fig. 4). Tyr190, Tyr203, and Tyr204 are all located in the C-terminal portion of the L. donovani HGPRT (10) and are not conserved among all HG(X)PRT family members. This multisequence alignment implies that the Tyr96 that is nitrated in peptides NP4 and NP5 is important for HGPRT catalysis, making this Tyr residue a primary candidate for further study by site-directed mutagenesis.


Fig. 4. Alignment of HGPRT flexible loop sequence containing the invariant Ser-Tyr dipeptide. The predicted amino acid sequences for the L. donovani, Crithidia fasciculata, Trypanosoma brucei, Trypanosoma cruzi, S. mansoni, and human HGPRTs (3), the Plasmodium falciparum, T. gondii, and T. foetus HG(X)PRTs (3), the E. coli XGPRT (20), and the G. lamblia GPRT (21) were aligned using the CLUSTAL V multiple sequence alignment software (43).
[View Larger Version of this Image (29K GIF file)]


Complementation Analysis

To validate the hypothesis that the conserved Ser95-Tyr96 dipeptide plays an integral role in the catalytic mechanism of the enzyme, a series of site-directed mutations was introduced into the L. donovani HGPRT at these positions. Ser95 was mutated to Ala, Cys, Glu, or Thr, whereas Tyr96 was changed to either Phe or Val (Fig. 5). These mutated constructs were then assessed for their ability to complement the bacterial gpt and hpt lesions within the SPhi 609 E. coli genetic background. Allen et al. (10) have previously demonstrated that the wild type L. donovani HGPRT can complement these bacterial prt lesions.


Fig. 5. Complementation analysis for wild type HGPRT and mutant hgprts. Wild type and mutant constructs in pSelect expression vectors were transformed into SPhi 609 E. coli and inoculated into LPI media containing adenine, adenine/hypoxanthine, adenine/guanine, or adenine/guanosine. All purines were present at a concentration of 125 µM. Cells were incubated at 37 °C with vigorous shaking for 8 h, conditions under which exponential growth was continuously maintained. Culture densities were then measured spectrophotometrically at 600 nm. Each bar represents the mean and standard deviation of three separate experiments.
[View Larger Version of this Image (35K GIF file)]


As shown in Fig. 5, SPhi 609 cells transformed with wild type HGPRT can grow in minimal medium supplemented with either hypoxanthine or guanine. If either hypoxanthine or guanine is omitted from the medium, SPhi 609 cells cannot survive because of the genetic defect (purHJ) in the purine biosynthetic pathway that confers purine auxotrophy. The adenine present in this minimal medium (see "Experimental Procedures") cannot serve as a source of guanylate nucleotides in the presence of high histidine levels in the media (11). In contrast, the SPhi 609 cells transformants containing mutant hgprt genes exhibited three distinct growth phenotypes (Fig. 5). The S95T transformant exhibited essentially a wild type growth rate when hypoxanthine was provided as a purine and a slightly reduced growth rate when guanine was added to the medium. The second phenotype was associated with the S95A and S95C mutations. These transformants complemented the bacterial prt mutations weakly, i.e. the cells grew, albeit slowly when either hypoxanthine or guanine was the 6-oxypurine in the medium. The third growth phenotype was observed with the transformants in which the L. donovani hgprt contained S95E, Y96F, or a Y96V mutation. These transformants were incapable of growth when either hypoxanthine or guanine was provided as a purine source (Fig. 5). All strains grew equivalently in minimal medium supplemented with guanosine, as the nucleoside can be converted to the nucleotide level via a route independent of HGPRT metabolism. Thus, this growth condition served as a positive control for each transformant. The distinctive growth phenotypes observed among the wild type and mutant transformants could not be ascribed to differences in protein production as all transformants grown in LPI medium supplemented with adenine plus guanosine produced similar amounts of HGPRT/hgprt protein, as assessed by Coomassie staining of cell lysates fractionated by SDS-polyacrylamide gel electrophoresis (data not shown).

Kinetic Studies

The effects of these single site mutations on the kinetics of the forward and reverse reactions of the enzyme were also characterized. Wild type HGPRT exhibited Km values of 6.1 ± 1.0 and 5.1 ± 0.8 µM for hypoxanthine and guanine, respectively. These values agree with those reported previously for the native (22, 23) and recombinant L. donovani HGPRT (10) enzymes. Although this enzyme exhibited similar affinities for both nucleobase substrates, the phosphoribosylation reaction was ~4 times more efficient with guanine than with hypoxanthine (Table III). The Km values of all of the mutant hgprt enzymes for hypoxanthine and guanine were similar to those obtained with the wild type protein and ranged from 4.7 ± 1.0 to 10.8 ± 2.4 µM for hypoxanthine and from 5.1 ± 0.8 to 12.2 ± 3.4 µM for guanine (Table III).

Table III.

Kinetic parameters for the forward reaction of wild type and mutant HGPRTs


Mutation Hypoxanthinea
Guaninea
PRPPb
Km kcat Km kcat Km kcat

µM s-1 µM s-1 µM s-1
Wild type 6.1  ± 1.0 8.54  ± 0.82 5.1  ± 0.8 41.27  ± 1.8 134  ± 22 36.3  ± 12.6
S95A 7.0  ± 0.7 0.59  ± 0.13 6.9  ± 0.2 0.52  ± 0.02 117  ± 27 0.54  ± 0.17
S95C 4.7  ± 0.5 3.15  ± 0.27 4.3  ± 1.1 9.82  ± 0.62 82  ± 4.0 5.75  ± 0.08
S95E 5.4  ± 0.8 0.07  ± 0.03 8.9  ± 3.0 0.95  ± 0.23 112  ± 8.0 0.34  ± 0.02
S95T 5.3  ± 0.6 1.97  ± 0.03 6.3  ± 1.3 9.1  ± 0.55 89  ± 5.0 7.8  ± 0.04
Y96F 6.8  ± 1.6 0.09  ± 0.015 8.3  ± 2.6 0.12  ± 0.08 28  ± 3.0 0.02  ± 0.008
Y96V 10.8  ± 2.4 0.072  ± 0.01 12.2  ± 3.4 0.04  ± 0.01 30  ± 7.5 0.03  ± 0.005

a Determined using 1 mM PRPP as the second substrate.
b Determined using 40 µM guanine as the second substrate.

Km values for the PRPP substrate were also determined. A Km value for PRPP of 134 ± 22 µM was obtained for wild type HGPRT, a value similar to that obtained for the recombinant human HGPRT (24). All the mutations at Ser95 did not significantly alter the PRPP Km value (Table III). Surprisingly, however, mutations at Tyr96 dropped the Km value of the enzyme for PRPP ~4-5 fold. Km values of 28 ± 3.0 and 30 ± 7.5 µM were obtained for the Y96F and Y96V mutations, respectively.

The most striking aspect of the kinetics of the forward reaction, however, was the dramatic decrease in the kcat values observed with hgprt enzymes containing Ser95-Tyr96 mutations. The kcat for the wild type enzyme was 8.5 ± 0.8 s-1 for hypoxanthine and 41.3 ± 1.8 s-1 for guanine (Table III). The conservative S95C and S95T substitutions decreased the kcat of the enzyme for both hypoxanthine and guanine ~2- and ~4-fold, respectively. However, the hgprts encompassing the S95A or S95E mutations exhibited a marked diminution in the turnover rate of the enzyme of 1-2 orders of magnitude (Table III). Most remarkable were the kcat values determined for the Y96F and Y96V mutations. The kcat values determined for the Y96F enzymes were 0.09 ± 0.015 and 0.12 ± 0.08 s-1 for hypoxanthine and guanine, respectively, while those for Y96V were 0.072 ± 0.01 s-1 for hypoxanthine and 0.04 ± 0.01 s-1 for guanine.

The possibility that the decreases in the turnover rates observed for the mutated L. donovani hgprts could be attributed to an increased affinity for one of the reaction products was assessed by studying the kinetics of the reverse reaction, i.e. the formation of hypoxanthine and PRPP from IMP and PPi. Using the xanthine oxidase-coupled enzyme assay described originally by Giacomello and Salerno (14), Km values for IMP and pyrophosphate of 90 ± 24 and 103 ± 16 µM were obtained for wild type HGPRT (Table IV). Similar Km values for both substrates were obtained for the S95E, S95T, Y96F, and Y96V hgprts (Table IV). As with the forward reaction, the most significant consequence of the site-directed mutations within the Ser-Tyr dipeptide was the ~2-450 fold decrease in the kcat for pyrophosphorolysis of IMP. The extent of the reduction in catalytic capability of an individual mutant in the reverse direction was roughly proportional to the decrease in kcat for the forward reaction (Tables III and IV).

Table IV.

Kinetic parameters for the reverse reaction of the wild type and mutant HGPRTs


Mutation Inosine monophosphate
Pyrophosphate
Km kcat Km kcat

µM s-1 × 10-3 µM s-1 × 10-3
Wild type 90  ± 24 77  ± 13 103  ± 16 90  ± 10
S95A 103  ± 8 5.6  ± 1.1 63  ± 10 6.2  ± 0.4
S95E 130  ± 25 2.0  ± 0.4 83  ± 8.0 2.0  ± 0.4
S95T 105  ± 17 39  ± 9.0 105  ± 5.0 38  ± 2.0
Y96F 115  ± 9.0 8.0  ± 0.7 107  ± 7.0 8.0  ± 0.3
Y96V 115  ± 15 0.22  ± 0.04 115  ± 15 0.2  ± 0.02


DISCUSSION

The impact of protein modifying reagents and site-directed mutagenesis on the activity of L. donovani HGPRT has established an essential function for the Ser-Tyr dipeptide that is conserved among all members of the HG(X)PRT family. Both TNM and DEPC introduce bulky moieties onto Tyr side chains, presumably increasing steric hindrance at the active site cleft and thereby precipitating a loss of enzyme function. Moreover, addition of a NO2 group to the Tyr ring profoundly decreases the pKa of the phenolic hydroxyl from ~10.5 to ~7.2 (25-27). This pKa reduction at the Tyr96 of the L. donovani HGPRT would markedly disrupt the hydrogen bonding potential of this residue (26, 27) and also introduces a negative charge on the flexible loop at physiologic pH that could either impede PRPP entry into the catalytic pocket or prevent the loop from covering the active site during catalysis.

The requirement of Tyr96 for HGPRT activity was confirmed by mutational analysis of this residue. Mutant hgprts in which Tyr96 was changed to either Phe or Val exhibited substantial diminutions in kcat values for both the forward and reverse reactions and failed to complement bacterial hpt and gpt lesions. A similar study of Ser95 mutants also revealed a prominent role of Ser95 in HGPRT catalysis, although the effects of mutations at this position on the kcat value were not as dramatic as those observed for the Tyr96 mutations. Moreover, several of the Ser95 mutants could complement bacterial hpt and gpt lesions. That Ser95 is important for HGPRT activity is not particularly surprising in view of the fact that a S103R mutation at the analogous site in the human enzyme results in a marked diminution of enzyme activity and leads to the clinical manifestation of gout (28). Although no changes in the Km values of any of the Leishmania hgprt mutants for nucleobases were observed, it is interesting to note that the S103R mutation in the human enzyme increased the Km value for hypoxanthine 74-fold (28).

The kinetic data on the site-directed mutants revealed that the genetic alterations created within the L. donovani HGPRT did not affect the overall substrate specificity of the enzyme nor was the affinity of the enzyme for nucleobase substrates appreciably affected, although a consistent but small decrease in the PRPP Km value of the Tyr96 mutants was obtained. Clearly, the kinetic parameter most significantly affected by the mutations was the turnover rate, implying that this highly conserved motif is integral to the catalytic mechanism. Delta Delta Gapp values (29) calculated from the specificity constants (kcat/Km obtained for guanine) revealed differences between wild type HGPRT and mutant hgprts in the stabilization energy for enzyme-substrate complex. The destabilizing effects of the mutations in the L. donovani HGPRT can be separated into two groups. The first includes the S95C and S95T mutants that exhibited Delta Delta Gapp values of 3.2 and 4.3 kJ/mol, respectively. These small energetic changes can be attributed to non-ideal geometries for hydrogen bond formation or to diminished hydrogen bond dissociation energies (30). The second group of destabilizing mutations includes the S95A, S95E, Y96F, and Y96V mutants for which Delta Delta Gapp values of 11.5, 10.5, 15.8, and 20.7 kJ/mol, respectively, were calculated. The magnitudes of these free energy changes are commensurate with loss of a hydrogen bond to a charged acceptor group (31), intimating that the hydroxyl groups of both Ser95 and Tyr96 play important roles in transition state stabilization.

The findings obtained with the L. donovani HGPRT incubated with protein modifying reagents contrast with previous studies on the human and Schistosoma mansoni HGPRT enzymes. The human and S. mansoni HGPRTs were inactivated by sulfydryl reagents (32, 33), whereas the L. donovani HGPRT was refractory to inactivation by either iodoacetate or p-chloromercuric benzoate. Covalent labeling experiments with [14C]iodoacetate identified the Cys22 and Cys25 residues of the human and schistosomal HGPRTs, respectively, as crucial for enzyme activity (33, 34). The insensitivity of the L. donovani enzyme to sulfydryl inactivation does not, however, support a general role for Cys residues in HGPRT catalysis.

In view of the multiplicity of conserved motifs among HG(X)PRT proteins, the failure of other protein modifying reagents to inactivate the L. donovani HGPRT is somewhat surprising (3). Several of these motifs are embedded within the core of the protein and may not be accessible to chemical modifying reagents (4). The Glu-Asp dipeptide that is situated within the PRPP binding motif that is found among all PRT family members (35) is, however, surface-exposed (4, 5, 9). Yet, the carboxyl-modifying reagent ethyl dimethylaminopropyl carbodiimide did not inactivate the protein. From the three-dimensional structures of HG(X)PRT proteins, it has been suggested that this Glu-Asp dipeptide functions in the binding of the ribose moiety of PRPP via a Mg2+ bridge and in the stabilization of the transition state complex (4, 9). It is possible that the free carboxyl moieties in the Glu-Asp dipeptide are shielded from chemical modification by either coordination to Mg2+ ions or by formation of internal salt bridges. The recently determined crystal structure of the Salmonella typhimurium OPRT (36) and the T. gondii HGXPRT (9) have revealed a Mg2+ ion complexed to this dyad.

The data with protein modifying reagents and mutants clearly confirm an essential role of the Ser-Tyr dipeptide in HGPRT catalysis. The human HGPRT (4) and T. gondii (9) HGXPRT crystal structures reveal that this Ser-Tyr is located on a solvent-exposed loop that exhibits weak electron density, intimating that this motif possesses a high degree of flexibility. This loop consists of residues 103-117 of the human enzyme and residues 96-106 of the T. gondii HGXPRT. It has been conjectured that this loop functions in a structural role to protect the oxocarbonium ion transition state of the enzyme from nucleophilic attack by water (4). A similar protective function has been proposed for an analogous loop (residues 98-119) in the S. typhimurium OPRT. Precedence for such a flexible loop functioning in a gating capacity to occlude solvent molecules from catalytic pockets has been observed for a variety of other proteins, including triose-phosphate isomerase (39), human immunodeficiency virus type I protease (38), and tyrosyl-tRNA synthetase (31). A second and perhaps more significant role for the Ser-Tyr dipeptide may be in the formation of the charge transfer complex in the transition state intermediate. Several lines of evidence support this idea. First, the structure of OPRT co-crystallized with orotate and PRPP reveals that Lys100 and Lys103 within the flexible loop of OPRT are translocated to within 6 Å of the active site of the enzyme-substrate complex (36) and may form contacts with the PPi moiety of PRPP (39). The crystal structures of the T. gondii HGXPRT apoenzyme and product-bound form also reveal a shift of this loop toward the active site (9). Moreover, the Arg100 and Arg102 residues within the human and S. mansoni HGPRT proteins, respectively, can be specifically cross-linked by the product analog 2',3'-GMP dialdehyde indicating that these residues are close to the active site (33). These observations coupled with the present data on the L. donovani enzyme imply that the hydroxyl groups of the Ser and Tyr residues may function as hydrogen bond donors. Alternatively, the Tyr96 may serve as a general acid that can protonate the PPi moiety of PRPP, thereby facilitating the scission of the C1-O bond of PRPP with the concomitant formation of the C1-N9 amino glycoside linkage of the nucleotide product. This latter mechanism may be similar to that observed for the acid-catalyzed hydrolysis of glucose 1-phosphate (40). Precedence for Tyr hydroxyls participating as general acids in reaction mechanisms has been obtained for the ketosteroid isomerase of Pseudomonas testosteroni (41) and the tyrosine phenol-lyase of Citrobacter freundii (42). The role of a general acid in PRT catalysis has also been proposed for the highly conserved Lys103 of OPRTs, since mutation of this residue in the S. typhimurium enzyme caused a 1000-fold decrease in the kcat value of the enzyme without affecting the Km values for either the forward or reverse reaction (39).


FOOTNOTES

*   This work was supported by in part Grant AI-23682 from the National Institute of Allergy and Infectious Diseases (to B. U.) and Grant HFSPO-LT-283/94 from the Human Frontiers Science Program Organization (to A. J.).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.
Dagger    Burroughs Wellcome Fund Scholar in Molecular Parasitology and supported in part by a grant from the Burroughs Wellcome Fund. To whom all correspondence should be addressed. Tel.: 503-494-8437; Fax: 503-494-8393; E-mail: ullmanb{at}ohsu.edu.
1   The abbreviations used are: HGPRT, hypoxanthine-guanine phosphoribosyltransferase; PRPP, 5-phosphoribosyl-1-pyrophosphate; HGXPRT, hypoxanthine-guanine-xanthine phosphoribosyltransferase; PRT, phosphoribosyltransferase; LPI medium, low phosphate induction medium; IMP, inosine monophosphate; GMP, guanosine monophosphate; HPLC, high performance liquid chromatography; PTH, phenylthiohydantoin; TNM, tetranitromethane; DEPC, diethyl pyrocarbonate.

ACKNOWLEDGEMENTS

We thank Dustin Lippert and Sandy Kielland of the University of Victoria Microchemistry Center for their technical expertise and assistance in performing the protein microsequencing and mass spectroscopy analysis.


REFERENCES

  1. Grogl, M., Thomason, T. N., and Franke, E. D. (1992) Am. J. Trop. Med. Hyg. 47, 117-126 [Medline] [Order article via Infotrieve]
  2. Berens, R. L., Krug, E. C., and Marr, J. J. (1995) Biochemistry and Molecular Biology of Parasites, pp. 89-117, Academic Press Ltd., New York
  3. Ullman, B., and Carter, D. (1995) Infect. Agents Dis. 4, 29-40 [Medline] [Order article via Infotrieve]
  4. Eads, J. C., Scapin, G., Xu, Y., Grubmeyer, C., and Sacchettini, J. C. (1994) Cell 78, 325-334 [Medline] [Order article via Infotrieve]
  5. Somoza, J. R., Chin, M. S., Focia, P. J., Wang, C. C., and Fletterick, R. J. (1996) Biochemistry 35, 7032-7040 [CrossRef][Medline] [Order article via Infotrieve]
  6. Scapin, G., Grubmeyer, C., and Sacchettini, J. C. (1994) Biochemistry 33, 1287-1294 [Medline] [Order article via Infotrieve]
  7. Henriksen, A., Aghajari, N., Jensen, K. F., and Gajhede, M. (1996) Biochemistry 35, 3803-3809 [CrossRef][Medline] [Order article via Infotrieve]
  8. Smith, J. C., Zaluzec, E. J., Wery, J., Niu, L., Switzer, R. L., Zalkni, H., and Satow, Y. (1994) Science 264, 1327-1333
  9. Schumacher, M. A., Carter, D., Roos, D. S., Ullman, B., and Brennan, R. G. (1996) Nature Struct. Biol. 3, 881-887 [Medline] [Order article via Infotrieve]
  10. Allen, T. E., Hwang, H. Y., Jardim, A., Olafson, R. W., and Ullman, B. (1995) Mol. Biochem. Parasitol. 73, 133-143 [CrossRef][Medline] [Order article via Infotrieve]
  11. Jochimsen, B., Nygaard, P., and Vestergaard, T. (1975) Mol. Gen. Genet. 143, 85-91 [Medline] [Order article via Infotrieve]
  12. Laemmli, U. K. (1970) Nature 227, 680-685 [Medline] [Order article via Infotrieve]
  13. Means, G. E., and Feeney, R. E. (1971) Chemical Modification of Proteins, pp. 105-32, Holden-Day, Inc., San Francisco, CA
  14. Giacomello, A., and Salerno, C. (1978) J. Biol. Chem. 253, 6038-6044 [Abstract]
  15. Holbrook, J. J., and Ingram, V. A. (1973) Biochem. J. 131, 729-738 [Medline] [Order article via Infotrieve]
  16. Melchior, W. B., Jr., and Fahreny, D. (1970) Biochemistry 9, 251-258 [Medline] [Order article via Infotrieve]
  17. Glazer, A. N. (1976) in The Proteins (Neurath, H., and Hill, R. L., eds), 3rd Ed., Vol. 2, pp. 1-103, Academic Press, New York
  18. Bergmann, M., and Zervas, L. (1928) Hoppe-Seylers Z. Phys. Chem. 175, 145-153
  19. Sokolovsky, M., Riordan, J. F., and Valle, B. L. (1967) Biochem. Biophys. Res. Commun. 27, 20-25 [Medline] [Order article via Infotrieve]
  20. Pratt, D., and Subramani, S. (1983) Nucleic Acids Res. 11, 8817-8823 [Abstract]
  21. Sommer, J. M., Ma, H., and Wang, C. C. (1996) Mol. Biochem. Parasitol. 78, 185-193 [CrossRef][Medline] [Order article via Infotrieve]
  22. Tuttle, J. V., and Krenitsky, T. A. (1980) J. Biol. Chem. 255, 909-916 [Free Full Text]
  23. Allen, T. E., Henschel, E. V., Coons, T., Cross, L., Conley, J., and Ullman, B. (1989) Mol. Biochem. Parasitol. 33, 273-281 [Medline] [Order article via Infotrieve]
  24. Free, M. L., Gordon, R. B., Keough, D. T., Beachman, I. R., Emmerson, B. T., and de Jersey, J. (1990) Biochim. Biophys. Acta 1087, 205-211 [Medline] [Order article via Infotrieve]
  25. Riordan, J. F., and Valle, B. L. (1972) Methods Enzymol. 25, 515-521
  26. Richards, P. G., Walton, D. J., and Heptinstall, J. (1996) Biochem. J. 315, 473-479 [Medline] [Order article via Infotrieve]
  27. Bray, M. R., and Clarke, A. J. (1995) Biochemistry 34, 2006-2014 [Medline] [Order article via Infotrieve]
  28. Wilson, J. M., Stout, T., Palella, T. D., Davidson, L. B., Kelley, W. N., and Caskey, C. T. (1986) J. Clin. Invest. 77, 188-195 [Medline] [Order article via Infotrieve]
  29. Fersht, A. R. (1988) Biochemistry 27, 1577-1580 [Medline] [Order article via Infotrieve]
  30. Weiner, S. J., Kollman, P. A., Case, D. A., Singh, U. C., Ghio, C., Alagona, G., Profeta, S., and Weiner, P. (1984) J. Am. Chem. Soc. 108, 765-784
  31. Fersht, A. R., Knill-Jones, J. W., Bedouelle, H., and Winter, H. (1988) Biochemistry 27, 1581-1587 [Medline] [Order article via Infotrieve]
  32. Krenitsky, T. A., and Papaioannou, R. (1969) J. Biol. Chem. 244, 1271-1277 [Abstract/Free Full Text]
  33. Kanaani, J., Maltby, D., Focia, P., and Wang, C. C. (1995) Biochemistry 34, 14987-14996 [Medline] [Order article via Infotrieve]
  34. Keough, D. T., Emmerson, B. T., and de Jersey, J. (1991) Biochim. Biophys. Acta 1096, 95-100 [Medline] [Order article via Infotrieve]
  35. Mushegian, A. R., and Koonin, E. L. (1994) Protein Sci. 3, 1081-1088 [Abstract/Free Full Text]
  36. Scapin, G., Ozturk, D. H., Grubmeyer, C., and Sacchettini, J. C. (1995) Biochemistry 34, 10744-10754 [Medline] [Order article via Infotrieve]
  37. Lolis, E., and Petsko, G. (1990) Biochemistry 29, 6619-6625 [Medline] [Order article via Infotrieve]
  38. Wlodawer, A., and Erickson, J. W. (1993) Annu. Rev. Biochem. 62, 543-585 [CrossRef][Medline] [Order article via Infotrieve]
  39. Ozturk, D. H., Dorfman, R. H., Scapin, J., Sacchettini, J. C., and Grubmeyer, C. (1995) Biochemistry 34, 10764-10770 [Medline] [Order article via Infotrieve]
  40. Bunton, C. A., Llewellyn, D. R., Oldham, K. G., and Vernon, C. A. (1958) J. Chem. Soc. 3588-3594
  41. Zhao, Q., Mildvan, A. S., and Talalay, P. (1995) Biochemistry 34, 426-434 [Medline] [Order article via Infotrieve]
  42. Chen, H. Y., Demidkina, T. V., and Phillips, R. S. (1995) Biochemistry 34, 12276-12283 [Medline] [Order article via Infotrieve]
  43. Higgins, D., Bleasby, A. J., and Fuchs, R. (1992) Comput. Appl. Biosci. 8, 189-191 [Abstract]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.