Structure and mechanism of the 6-oxopurine nucleosidase from Trypanosoma brucei brucei.
ABSTRACT Trypanosomes are purine-auxotrophic parasites that depend upon nucleoside hydrolase (NH) activity to salvage nitrogenous bases necessary for nucleic acid and cofactor synthesis. Nonspecific and purine-specific NHs have been widely studied, yet little is known about the 6-oxopurine-specific isozymes, although they are thought to play a primary role in the catabolism of exogenously derived nucleosides. Here, we report the first functional and structural characterization of the inosine-guanosine-specific NH from Trypanosoma brucei brucei. The enzyme shows near diffusion-limited efficiency coupled with a clear specificity for 6-oxopurine nucleosides achieved through a catalytic selection of these substrates. Pre-steady-state kinetic analysis reveals ordered product release, and a rate-limiting structural rearrangement that is associated with the release of the product, ribose. The crystal structure of this trypanosomal NH determined to 2.5 Å resolution reveals distinctive features compared to those of both purine- and pyrimidine-specific isozymes in the framework of the conserved and versatile NH fold. Nanomolar iminoribitol-based inhibitors identified in this study represent important lead compounds for the development of novel therapeutic strategies against trypanosomal diseases.
- [show abstract] [hide abstract]
ABSTRACT: Human African trypanosomiasis is a major health problem in large regions of Africa. Current chemotherapeutic options are limited and far from ideal. A diverse range of drug targets has been identified and validated in trypanosomes. These include several organelles (glycosomes, acidocalcisomes, kinetoplast) that are not represented in the mammalian host and biochemical pathways that differ significantly from host counterparts (carbohydrate metabolism, protein and lipid modification, response to oxidative stress, cell cycle). However, there has been little progress in developing novel drugs. Pharmaceutical companies are unwilling to invest in the development of drugs for a market that comprises some of the worlds poorest people. This review highlights some of the most attractive drug targets in trypanosomes.Expert Review of Anticancer Therapy 07/2003; 1(1):157-65. · 3.22 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: African trypanosomes cause human sleeping sickness and livestock trypanosomiasis in sub-Saharan Africa. We present the sequence and analysis of the 11 megabase-sized chromosomes of Trypanosoma brucei. The 26-megabase genome contains 9068 predicted genes, including ~900 pseudogenes and ~1700 T. brucei-specific genes. Large subtelomeric arrays contain an archive of 806 variant surface glycoprotein (VSG) genes used by the parasite to evade the mammalian immune system. Most VSG genes are pseudogenes, which may be used to generate expressed mosaic genes by ectopic recombination. Comparisons of the cytoskeleton and endocytic trafficking systems with those of humans and other eukaryotic organisms reveal major differences. A comparison of metabolic pathways encoded by the genomes of T. brucei, T. cruzi, and Leishmania major reveals the least overall metabolic capability in T. brucei and the greatest in L. major. Horizontal transfer of genes of bacterial origin has contributed to some of the metabolic differences in these parasites, and a number of novel potential drug targets have been identified.01/2005;
- [show abstract] [hide abstract]
ABSTRACT: The pathways leading to purine and pyrimidine nucleotide production in members of the family Trypanosomatidae are discussed with special emphasis on data relating to pathogenic species published from 1974 to 1983 inclusive. Trypanosomes and leishmania in general lack a de novo purine biosynthetic pathway, but have a multiplicity of possible routes for purine salvage. In contrast, pyrimidine nucleotides can be produced by either de novo or salvage pathways. The properties of these pathways in trypanosomatids are compared and contrasted with those of their hosts.Molecular and Biochemical Parasitology 12/1984; 13(3):243-61. · 2.73 Impact Factor
pubs.acs.org/BiochemistryPublished on Web 09/08/2010
r2010 American Chemical Society
Biochemistry 2010, 49, 8999–9010
Structure and Mechanism of the 6-Oxopurine Nucleosidase from
Trypanosoma brucei brucei†,‡
An Vandemeulebroucke,§,@Claudia Minici,^Ilaria Bruno,^Laura Muzzolini,^Paola Tornaghi,^David W. Parkin,)
Wim Vers? ees,§Jan Steyaert,§and Massimo Degano*,^
§Department of Molecular and Cellular Interactions (VIB) and Structural Biology Brussels, Vrije Universiteit Brussel,
1050 Brussel, Belgium,
Department of Chemistry, Adelphi University, Garden City, New York 11530-0701, and
^Division of Immunology, Transplantation and Infectious Diseases, Scientific Institute San Raffaele, Milan, Italy.
@Present address: Department of Biochemistry, Albert Einstein College of Medicine of Yeshiva University, Bronx, NY 10461.
Received May 5, 2010; Revised Manuscript Received September 8, 2010
ABSTRACT: Trypanosomes are purine-auxotrophic parasites that depend upon nucleoside hydrolase (NH)
activity to salvage nitrogenous bases necessary for nucleic acid and cofactor synthesis. Nonspecific and
purine-specific NHs have been widely studied, yet little is known about the 6-oxopurine-specific isozymes,
although they are thought to play a primary role in the catabolism of exogenously derived nucleosides. Here,
we report the first functional and structural characterization of the inosine-guanosine-specific NH from
analysis reveals ordered product release, and a rate-limiting structural rearrangement that is associated with
the release of the product, ribose. The crystal structure of this trypanosomal NH determined to 2.5 A˚
resolution reveals distinctive features compared to those of both purine- and pyrimidine-specific isozymes in
the framework ofthe conservedand versatile NH fold. Nanomolar iminoribitol-based inhibitors identifiedin
this study represent important lead compounds for the development of novel therapeutic strategies against
Protozoan parasites of the Trypanosoma genus cause a variety
of diseases, including lethal sleeping sickness (1, 2). This patho-
logy caused by Trypanosoma brucei gambiense or Trypanosoma
brucei rhodesiense in humans and by Trypanosoma brucei brucei
in livestock is endemic in Central African countries, where it has
a dramatic social and economic impact. Every year, 250000-
300000 people are estimated to diefrom lack of diagnosis and/or
treatment. Chemotherapeutic intervention is currently poorly
effective, because it takes advantage of the action of compounds
that exhibit aspecific toxicity, such as melarsoprol, eflornithine,
or suramin. It is thus clear that the development of novel, highly
efficient drugs against trypanosomes represents an urgent need
for a large portion of the world population (3).
The completionofthesequence of the T.bruceibruceigenome
reinforced hopes for the identification of new molecular targets
enzymatic activitiesthatarerequired tosynthesizethenucleotide
inosine monophosphate from simple metabolites confirmed the
observation that this protozoan parasite is purine-auxotrophic.
sis are thus absorbed from the host, via the concerted actions of
extracellular nucleases, phosphodiesterases, and nucleotidases (5).
Neutral nucleosides are imported by the parasite and further
processed to the purine bases via the action of N-glycosidases, in
particular nucleoside hydrolases (NHs).1Trypanosomes are
unique in using the NH-catalyzed reaction to this end, because
prokaryotes and higher eukaryotes rely on nucleoside phosphor-
ylases (NP) for the same process. Indeed, neither NH-like genes
nor NH activity has ever been found in mammals (6, 7). This
dichotomy, together with the observation that more than 85% of
the nucleosides taken up by trypanosomes are hydrolyzed to the
free base (8), underscores the purine recycling pathway as an
inhibited by specific compounds that do not affect the activity of
human NPs, hence selectively blocking this crucial parasitic
Enzymes with NH activity have been extensively characterized,
from protozoa (8, 12-15), bacteria (16-19), yeast (20), insects
(21),nematodes (19),and archaea (22,23).NHsareclassified into
groups according to their specificity, including the nonspecific
inosine-uridine-preferring NHs (IU-NH), the pyrimidine-specific
NHs (cytidine-uridine-preferring, CU-NHs), the purine-specific
NHs (inosine-adenosine-guanosine-preferring, IAG-NHs), and
†Supported by a grant from the Institute for the Promotion of Innova-
tion through Science and Technology in Flanders (IWT-Vlaanderen, to
A.V.), a grant from FWO-Vlaanderen (W.V.), and research grants from
Cariplo Foundation and FIRB (M.D.).
‡Structure factors and coordinates have been deposited with the
Protein Data Bank as entry 3FZ0.
*To whom correspondence should be addressed. Telephone: þ39-
0226437152. Fax: þ39-0226434153. E-mail: firstname.lastname@example.org.
1Abbreviations: NH, nucleoside hydrolase; IU-NH, inosine-uridine-
preferring NH; IAG-NH, inosine-adenosine-guanosine-preferring NH;
IG-NH, inosine-guanosine-preferringNH; NP, nucleosidephosphorylase;
Tbb, Trypanosoma brucei brucei; Tv IAG-NH, NH from Trypanosoma
tag; PEG-MME, polyethylene glycol monomethyl ether; pNPR, p-nitro-
phenyl β-D-ribofuranoside; 7mGuo, 7-methylguanosine; Bis-Tris, bis-
(2-hydroxyethyl)aminotris(hydroxymethyl)methane; ImmH, Immucillin-H;
ImmA, Immucillin-A; 8-aza-ImmH, 8-aza-Immucillin-H.
9000Biochemistry, Vol. 49, No. 41, 2010Vandemeulebroucke et al.
the 6-oxopurine-specific isozymes (inosine-guanosine-preferring,
IG-NHs). Three distinct NH activities have been reported in
(24, 25). The structure and mechanism of IU- and IAG-NHs are
ion, via specific hydrogen bonds and interactions between the
conserved histidine residue is clearly involved in the catalytic
hydrogen bond in the nucleoside substrate and an aromatic
stacking of the purine ring with an active site tryptophan to
increase the pKaof the base, thus allowing direct protonation
of the base from the solvent (27). The crystal structures of both
isozyme-specific inhibitors (11, 15, 28). Nanomolar inhibitors for
both isozymes are currently available, and their structures in
complex with the NH enzymes can be exploited to augment their
efficacy and selectivity (28-30).
The IG-NH enzymes have been isolated from Crithidia
fasciculata (24) and T. brucei brucei (13) and are less well
understood, from both functional and structural points of
view. The kinetic parameters of this isozyme purified from
C. fasciculata revealed that the majority of the inosine imported
by the parasite from mammalian plasma would be catabolized
through the IG-NH, thus making it the primary target for the
design of a specific drug. The enzyme displayed an apparent
trimeric quaternary structure, thus differing from both IU-NHs
(tetrameric) (31) and IAG-NHs (dimeric) (15). Here we report the
cloning, expression, and first in-depth enzymatic and structural
The IG-NH differs from previously characterized NHs in amino
acid composition and surprisingly shares the same quaternary
structure with IU-NHs. The active site cavity is conserved in the
ribosyl binding portion but differs from those of both IU- and
IAG-NH isozymes in the residues involved in base discrimination.
submicromolar KM, hence indicating a crucial role of the 6-oxo
group in catalysis. Pre-steady-state kinetic analysis of guanosine
hydrolysis suggests a rate-limiting step involving the release of a
ribose product. The herein reported structure-function character-
ization provides the missing piece in the family of trypanosomal
Tbb IG-NH offers further opportunities for the development of a
specific therapeutic approach targeting the nucleotide salvage
pathway in parasites.
MATERIALS AND METHODS
Gene Identification and Cloning. The Tbb IG-NH enzyme
was originally isolated from the parasite (13, 25). Partial peptide
sequences obtained from the purified enzyme were used for
homology searches against the T. brucei brucei translated ESTs
from the genome sequencing project database (not completed at
the start of this study). Two distinct, partially overlapping ESTs
correctly matched two IG-NH peptide sequences. These could
be assembled, together with one EST containing the NH finger-
print sequence (DXDXXXDD) at the N-terminus, to a single
open reading frame. Two oligonucleotides were designed on the
NdeI and XhoI recognition sequences at their 50extremities.
These oligonucleotides were used in polymerase chain reactions
(PCRs) on a T. brucei brucei genomic DNA template (strain
TREU927) to amplify the ignh gene using the proofreading Pfu
SmaI-linearized pBSK(-) vector and sequenced on both DNA
strands using the automated dideoxy method. The N-terminal
sequence translated from the amplified gene is 100% identical to
that previously reported and is distinct from that of Tbb IAG-
the deposited genomic sequence (4), leading to two amino acid
substitutions (Lys98Gln and Leu228Ser). These differences were
confirmed in three independent clones obtained from different
PCRs, thus suggesting either mutations in the genomic DNA
used compared to the reference strain or errors in the deposited
genomic sequence. Multiple-sequence alignments were created
using CLUSTALW (32).
Protein Expression and Purification. For recombinant
protein expression, the ignh gene was subcloned between the
NdeI and XhoI sites of a pET28 vector previously digested with
the same restriction enzymes. The plasmid was transformed into
Rosetta(DE3) Escherichia coli cells for protein expression. Bac-
teria were grown to an OD600of 0.6, and protein expression was
induced by addition of 1.0 mM isopropyl β-thioglucopyranoside
and incubated with shaking for either 3 h at 37 ?C or 18 h at
were harvested by centrifugation at 5000g, resuspended in a
buffer containing 20 mM Tris (pH 8.0), 50 mM NaCl, 1 mg/mL
DNase, 1 mg/mL lysozyme, and 1? Complete EDTA free pro-
tease inhibitor cocktail (Roche), incubated at 37 ?C for 20 min,
for 45 min at 17000 rpm to separate the cell debris from the
intracellular soluble fraction. The soluble phase was incubated
a polypropylene column, washed extensively with binding buffer
finally eluted with the same buffer containing 0.5 M imidazole.
Elution fractions were pooled and dialyzed against a buffer
composed of 20 mM Tris (pH 8.4), 150 mM NaCl, and 250 mM
CaCl2. The N-terminal hexahistidine tag was removed by site-
specific proteolysis using thrombin at a 1:200 weight ratio,
leaving a GSH tripeptide fused to the N-terminus of the
recombinant enzyme. The IG-NH protein was further purified
via a Superdex 200 size exclusion chromatography column with
Protein was concentrated using ultrafiltration devices and stored
at 4 ?C. Dynamic light scattering was used to assess the ideal
storage buffer that prevented nonspecific aggregation and main-
(expressed per subunit) was determined spectrophotometrically
using an ε280of 52940 M-1cm-1(33). Typical enzyme prep-
arations have a specific activity of 120 μmol min-1mg-1with
inosine as the substrate.
Steady-State Kinetics. Initial rate kinetic measurements
were taken at 35 ?C in a 50 mM potassium phosphate buffer
(pH 7.0). Measurements were restricted to a maximal product
formation of 10% to ensure the linearity of the curves. Product
formation was assessed spectrophotometrically using the differ-
ence in absorption between the nucleoside and the base. The
following Δε values were used: -4.0 mM-1cm-1for guanosine
at 260 nm, -1.3 mM-1cm-1for inosine at 250 nm, -1.4 mM-1
cm-1for adenosine at 276 nm, -4.4 mM-1cm-1for 7-methyl-
guanosine at 258 nm, 1.45 mM-1cm-1for purine riboside at
ArticleBiochemistry, Vol. 49, No. 41, 20109001
275 nm, -3.27 mM-1cm-1for ethenoadenosine at 271.5 nm,
1.5 mM-1cm-1for cytidine at 291 nm, -1.9 mM-1cm-1for
uridine at 280 nm, 12 mM-1cm-1for p-nitrophenyl β-D-ribofu-
at 250 nm, -1.47 mM-1cm-1for xanthosine at 293 nm, -1.33
mM-1cm-1for 50-deoxyadenosine at 275 nm, and -1.64 mM-1
cm-1for 3-deazaadenosine at 263 nm. The data were fitted to the
Michaelis-Menten equation using Origin (Microcal). All kinetic
parameters were calculated per active site, making them indepen-
dent of the multimerization state of the enzyme.
Inhibition Studies. Steady-state inhibition studies were per-
formed at 35 ?C in a 50 mM phosphate buffer under saturating
substrate concentrations. The apparent inhibition constants for
the deoxyadenosines were determined with p-nitrophenyl β-D-
ribofuranoside (pNPR) as a substrate at pH 7. Assays were
initiated via addition of the enzyme to a solution containing a
for the Immucillins, inosine was used as a substrate in a coupled
assay using xanthine oxidase at pH 7.5 (optimal pH of xanthine
oxidase). This coupled assay has a higher sensitivity than the
anthine, is oxidized by the xanthine oxidase to uric acid, which
can be monitored spectrophotometrically at 293 nm (ε293of 12.9
the enzyme to the reaction mixture containing 500 μM inosine,
120 milliunits of xanthine oxidase/mL, and variable concentra-
tions of inhibitor. The kinetic constants for inosine hydrolysis
assay yields kinetic parameters identical to those of the direct
method. The data from the deoxyadenosine and the Immucillin
inhibition experiments (8-10 data points) were fit to the Dixon
linearization of the equation describing competitive inhibition:
where [S] is the substrate concentration, [I] is the inhibitor
concentration, [E]0is the enzyme concentration, kcatand KM
are the steady-state parameters for the substrate, and KIis the
by the high degree of structural similarity between the inhibitors
and the substrate, and by the reported structures of nucleoside
hydrolases in complex with inhibitors (29).
Stopped-Flow Analysis. The experimental procedures for
stopped-flow analysis were described previously (35). Stopped-
flow experiments were performed on an Applied Photophysics
SX18.MV stopped-flow spectrofluorimeter. All experiments
wereperformed at 5 ?C under pseudo-first-orderconditionswith
a minimal 4-fold excess of substrate or ligand over Tbb IG-NH.
Multipleturnovers of substrate by Tbb IG-NH were followed by
stopped-flow absorbance using the difference in absorption
between the nucleoside and the base with the Δε values men-
tioned above. Changes in the Tbb IG-NH tryptophan fluores-
of emitted radiation above 305 nm, using a 305 nm cutoff filter.
Quench-Flow Analysis. Rapid-quench measurements were
performed as described previously (35). A KinTek RQF-3
wasmixed withenzyme (40 μM) forvarious reactiontimes and the
reaction quenched with acid (0.333 M HCl). Samples were filtered
at 4 ?C using wetted Microcon centrifugal device micro concen-
column (100 mm ? 4.6 mm, ODS HYPERSIL RP-C18, 3 μm)
trifluoroacetic acid in 10 mM ammonium acetate (pH 5.0) and
monitored spectrophotometrically at 260 nm.
Analysis of Pre-Steady-State Data. Linear and nonlinear
curve fitting to progress curves was performed using Microcal
where y(t) is the observed signal at time t, i is the number of
transients, Aiis the amplitude of the ith transient, kiis the
observed rate constant (kobs) for the ith transient, v is the steady-
to make plots of observed rate constants and amplitudes versus
substrate, or ribose concentrations. The concentration depen-
dencies of the different observed rate constants were analyzed as
described previously (35).
The progress curves of multiple turnovers of guanosine and
7-methylguanosine (7mGuo) followed by stopped-flow absor-
the base ([B]obs) normalized for the enzyme concentration used
([E]0) as a function of time and were fitted to eq 3. This equation
and slower product release (Scheme 1) providing the character-
istic parameters for burst kinetics: the maximal burst rate (kobs),
ðk2þ k-2þ k3Þ2
From the saturated values of these parameters and eq 3, the
apparent rate constants of chemistry (k2and k-2) and product
release (k3) can be determined.
Crystallization. The Tbb IG-NH was crystallized using the
solution at 11 mg/mL and the precipitant solution consisting of
100 mM bis(2-hydroxyethyl)aminotris(hydroxymethyl)methane
(Bis-Tris) (pH 6.5), 200 mM ammonium sulfate, and 22% poly-
coverslip, inverted over a reservoir containing 0.7-1.0 mL of
precipitant, and kept at a constant temperature of 18 ?C. Crystals
grown using IG-NH expressed at 37 ?C exhibited regular, bipyr-
amidal shapes but diffracted X-rays to barely 6.0 A˚resolution and
were characterized by extremely high mosaicity. Instead, the same
prism-shaped crystals that diffracted X-rays beyond 3.0 A˚. This
the protein produced at higher temperatures, because enzyme
9002Biochemistry, Vol. 49, No. 41, 2010Vandemeulebroucke et al.
preparations from inductions of expression at 37 ?C exhibited
consistently higher polydispersity (∼25% vs 15%) as determined
from dynamic light scattering experiments. The kinetic parameters
of the enzyme, however, were not affected by the temperature of
induction. Crystals were cryoprotected using a solution with the
PEG-MME concentration increased to 34%, mounted on fiber
loops, and plunged and maintained in liquid nitrogen for storage
prior to data collection. Crystals were orthorhombic in the C2221
space group, with a cell volume consistent with four IG-NH
Structure Determination. Data from a single crystal of
unliganded IG-NH were collected on beamline ID14-EH4 of the
European Synchrotron Radiation Facility (Grenoble, France) at
2.5 A˚. Intensities were converted to structure factor amplitudes
using TRUNCATE (37). Initial phases were obtained with the
molecular replacement method as implemented in MOLREP (38)
using the structure of the CU-NH YeiK monomer as a search
model [Protein Data Bank (PDB) entry 1Q8F] after all ligands,
residues truncated to alanine (or glycine). Only two rotation
function solutions were apparent, but all four molecules could
An initial Fo- Fc(φc) map calculated using data to 2.5 A˚showed
The map, however, was heavily biased by the model phases, and
substantially untraceable. The density was manually inspected,
with them were removed from the model. Phases calculated to
5.0 A˚resolution using the modified YeiK model were refined and
extended in 1000 cycles of solvent flattening, histogram matching,
and noncrystallographic symmetry averaging as implemented in
DM (39). The molecular envelope was calculated using a 5.0 A˚
radius at each atom position. The final maps were of very good
quality, and both the mask preparation and the number of cycles
of phase refinement were crucial for the success of the procedure.
IG-NH molecule. The same procedure was also successful when
using the Tv IAG-NH or the Cf IU-NH as a search model for
phasing, underscoring the robustness of the approach that was
followed. The model was refined in cycles of manual rebuilding in
electron density maps in O (40), and maximum likelihood energy
minimization using REFMAC5 (41). TLS refinement was im-
individual group (42). MOLPROBITY (43) was used to validate
the geometric quality of the model during the refinement. Water
molecules were added at the later stages of refinement in positive
peaks of Fo- Fc(φc) maps greater than 3.5σ using ARP/wARP
molecule at each active site (Figure 1 of the Supporting In-
formation). The final crystallographic validators Rcrysand Rfree
are 0.189 and 0.225, respectively. Residues for which no electron
density was visible (the three amino acids remaining from the
construct at the N-terminal end in all subunits, amino acids
227-235 in chains A, C, and D; residues 256-273 in all chains)
were omitted from the model. All residues fall within the allowed
regions of a Ramachandran plot.
Identification of the IG-NH from T. brucei brucei. The
limited amino acid sequence available from the Tbb IG-NH was
used to identify the coding gene from the partial genomic
sequence available at the time of the investigation. The Tbb ignh
genome) is composed of 1074 bp, encoding a protein of 357
amino acids and a calculated Mrof 39365 Da. Among the so far
characterized NHs, the Tbb IG-NH protein displays the most
similarity to the Crithidia IU-NH (25% identical) and E. coli
CU-NHs YbeK/RihA and YeiK/RihB (27 and 20% identical,
respectively). Unexpectedly, a significantly lower level of amino
Tv IAG-NH. The alignment of the Tbb IG-NH amino acid
sequence with a nonspecific and purine-specific trypanosomal
NH highlights the presence of two major insertions (Figure 1),
loop at the C-terminal end of helix R9. While the residues
mediating ribosyl binding are strictly conserved in the IG-NH,
the amino acids involved in the interactions with the aglycone
isozymes. Indeed, His82 and His239 of the Crithidia IU-NH in
the Tbb IG-NH are substituted with Tyr and Pro, respectively.
More surprisingly, both tryptophan residues that mediate the
differences clearly point to a different mode of substrate binding
and catalytic mechanism.
analogues are summarized in Table 1. High kcat/KMratios were
found for the majority of the common purine nucleosides, with
the highest activity toward inosine and guanosine, although the
value for adenosine could not be accurately determined because
of the low KM. The catalytic rate constant (kcat) for inosine and
guanosine hydrolysis is 2 orders of magnitude higher than that
for adenosine hydrolysis, but the enzyme has a higher substrate
affinity (lower KM) for the latter. Instead, the IG-NH displays
103-104-fold lower catalytic efficiencies toward cytidine and
uridine substrates, resulting from both lower turnover rates and
higher KMvalues. The Tbb IG-NH displays hallmark features of
NH enzymes with respect to ribosyl moiety discrimination.
by a factor of 5 ? 104, while removal of the 20- and 30-OH groups
abolishes turnover of deoxyadenosines. 20-Deoxy- and 30-deoxy-
170 μM, respectively.
based on an iminoribitol scaffold are well-known NH inhibi-
tors (9, 10, 45, 46). A selection of immucillins was screened to
identify tight-binding inhibitors of Tbb IG-NH. Three Immucil-
lins, Immucillin-H (ImmH), Immucillin-A (ImmA), and 8-aza-
Immucillin-H (8-aza-ImmH), exhibited nanomolar apparent
inhibition constants (Table 1). None of these three tight-binding
inhibitors displayed slow-onset inhibition, a frequent event in
binding of iminoribitol-based compounds to NHs (29). The
highest-affinity inhibitor identified is ImmA, bearing an adenine
nucleobase mimic. This finding parallels the low KMmeasured
for adenosine and thus underscores the fact that the 6-amino
group in purines and analogues greatly enhances the affinity for
the active site of Tbb IG-NH.
lysis analyzed by stopped-flow absorption spectroscopy show
a very fast burst of guanine production upon mixing Tbb
IG-NHwith anexcessof guanosine(Figure2),indicating that the
ArticleBiochemistry, Vol. 49, No. 41, 2010 9003
rate-limiting step occurs after hydrolysis of the N-glycosidic bond
(chemistry) and thus is associated with product release. The
dependence of the observed burst rate (kobsin eq 3) on guanosine
concentration shows saturation behavior, permitting an estima-
active site titration also reveals that the maximal burst amplitude
corresponds to only half of the enzyme concentration used.
Multiple turnovers of the substrate analogue 7-methylguanosine
(7mGuo) catalyzed by Tbb IG-NH also follow burst kinetics
(Figure 2), showing a clear saturation behavior with a maximal
7-mGuo/mol of enzyme subunit.
IG-NH. Guanine formation is linear as a function of time, but
extrapolation of this linear guanine production back to the
product axis gives a positive intercept (data not shown). This
observation is a trademark for fast burst kinetics, with the burst
the dead time of a quench-flow instrument is greater than for a
stopped-flow instrument). The magnitude of the intercept equals
FIGURE 1: Alignment ofthe amino acidsequences ofthe Tbb IG-NH, the E. coliCU-NH YeiK (20% identical), the C.fasciculata IU-NH (25%
of the alignment. T defines the polypeptide turn structure and η the 310helix conformation.
Table 1: Activity Profile of TbbNH at pH 7 and 35 ?C
0.52 ( 0.04
0.0038 (0.0008 1451 (600 2.6 ( 1
4.5( 0.2 (8.3 ( 0.4) ? 106
<1>2.4 ? 105
3.6(0.7(1.9 ( 0.4) ? 106
271(49(1.9 ( 0.4) ? 103
(1.5 ( 0.3) ? 107
0.6( 0.1 (3.8 ( 0.8) ? 105
1.1(0.3(2.7 ( 0.6) ? 104
0.8(0.4(7 ( 3) ? 105
280(40(2.2 ( 0.4) ? 104
2.2(0.3(4.1 ( 0.6) ? 104
240(40(3.3 ( 0.6) ? 102
aDetermined at pH 7.5.
9004Biochemistry, Vol. 49, No. 41, 2010Vandemeulebroucke et al.
the burst amplitude and corresponds to 0.5 mol of guanine/mol
of enzyme subunit, confirming the observed burst amplitude
the reduced burst amplitudes for guanosine (0.5) and 7mGuo
(0.78) is a reversible chemistry step that is substrate-dependent.
The progress curves of multiple turnovers of guanosine and
7mGuo catalyzed by Tbb IG-NH can thus be explained by a
minimal three-step mechanism in which substrate binding is
followed by a reversible chemistry step followed by an overall
rate-limiting product release (Scheme 1).
Analyzing the progress curves of multiple turnovers of gua-
nosine and 7mGuo according to the equation describing such a
three-step mechanism with a reversible chemistry step (eq 3, see
Materials and Methods) allows the determination of the appar-
ent rate constants of chemistry (k2and k-2, the forward and
reverse rate constants, respectively) and product release (k3) as
described in Materials and Methods. The apparent rates of the
forward chemistry step (k2) and product release (k3) are very
comparable for guanosine and 7mGuo hydrolysis. On the other
hand, the apparent reverse rate of chemistry is almost 1 order of
magnitude faster for guanosine hydrolysis than for 7mGuo
hydrolysis (Table 2).
Fast Substrate Binding and Base Release. We have pre-
viously shown that multiple turnovers of purine nucleosides by
Tv IAG-NH analyzed by stopped-flow fluorescence allow the
assignment of transients to substrate binding and base release
also exist out of two transients (Figure 3). The first transient is
characterized by a rapid and large decrease in fluorescence, while
observed rate constant of the first transient has a linear concen-
tration dependency, which indicates that it originates from sub-
strate binding. The rate constant of the second transient has a
transients observed during purine nucleoside hydrolysis by Tv
IAG-NH (47), we propose that the second transient observed
during hydrolysis catalyzed by Tbb IG-NH also involves base
release. If this interpretation applies, the substrate concentration
dependency of the observed rates of the transients allows us to
determine the rate constants involved in substrate binding and
base release (Table 3). This analysis shows that the rate of base
release is 1 order of magnitude faster than the steady-state turn-
over rate. A fast rate of base release implies that the overall rate-
limiting step must occur after the release of base.
Ordered Product Release. To determine the order of
product release and validate the interpretation concerning the
FIGURE 2: Burstkineticsofguanosine[(a)60μM]and7mGuo[(b)50μM]turnoverbyTbbIG-NH(5μM)atpH7.0and5?C.Theconcentration
of formed base [B] has been normalized for the concentration of enzyme [E]0used during the experiment; hence, the concentration of base per
enzyme concentration ([B]/[E]0) is plotted as a function of time.
Table 2: Parameters (kobs, A0, and kcat) and Rate Constants of Burst
Kinetics of Guanosine and 7-Methylguanosine Turnover by TbbNH at pH
7.0 and 5 ?C, Using the Nomenclature of eq 3
220 ( 20
0.51 ( 0.03
1.3 ( 0.3
111 ( 12
106 ( 23
2.6 ( 0.6
100 ( 8
0.78 ( 0.01
5.2 ( 0.2
75 ( 8
18 ( 5
6.9 ( 0.4
FIGURE 3: Stopped-flow fluorescence progress curves of guanosine
[(a) 60 μM] and inosine [(b) 50 μM] hydrolysis by Tbb IG-NH
(0.5 μM) at pH 7.0 and 5 ?C.
Scheme 1: Minimal Three-Step Mechanism for the Multiple
Turnovers of Guanosine and 7mGuo by Tbb IG-NH
ArticleBiochemistry, Vol. 49, No. 41, 20109005
fluorescence progress curves, we examined the pre-steady-state
kinetics of hypoxanthine release using stopped-flow absorbance.
Hypoxanthine produced during turnover of inosine by Tbb IG-
NH wasdetectedwithaxanthineoxidase-coupled assay.Athigh
concentrations of xanthine oxidase, a burst is apparent and the
steady-state rate becomes independent of the xanthine oxidase
concentration and equals the previously determined rate of
The strong absorption by xanthine oxidase prevented the ob-
servation of a saturation of the burst rate or amplitude via an
increase in the xanthine oxidase concentration. It is thus not
possible to determine both the rate of base release and the active
site concentration from these progress curves. Nevertheless, the
observation of a hypoxanthine burst ascertains that base release
is faster than the steady-state turnover rate. Furthermore, this
implies that the overall rate-determining step occurs after base
release and is most likely ribose release.
Two-Step Ribose Binding. We studied the kinetics of the
binding of ribose to the Tbb IG-NH using stopped-flow fluor-
with a fixed amount of enzyme, yielding fluorescence transients
that could be fit to a single exponential. The observed rate
constants of these transients have a hyperbolic concentration
dependency, consistent with a two-step binding mechanism. In
the forward direction of the hydrolytic reaction, this slow
unimolecular step converts the tightly bound enzyme3ribose
complex into a loosely bound form from which ribose can easily
dissociate. The rate of this unimolecular step is very comparable
of guanosine and 7mGuo (Table 3). On the basis of these data, we
step on the reaction coordinate [k4=2.7 s-1(see Scheme 2)],
supporting the interpretation that base release precedes ribose
ordered, which was shown for the purine-specific Tv IAG-NH
(35, 48), and unlike what was observed for the C. fasciculata
Crystal Structure of the Tbb IG-NH. We determined the
structure of the unliganded Tbb IG-NH using single-crystal
X-ray diffraction to 2.5 A˚(Table 4). The IG-NH subunit
structure resembles the overall NH fold (31) with a parallel core
β-sheet flanked by R-helices, and the catalytic site located in a
of Tbb IG-NH and the E. coli RihB pyrimidine nucleosidase (6)
superimpose with a root-mean-square deviation (rmsd) of
1.50 A˚using 275 homologous CR positions. Conversely, the Tv
IAG-NH (15) displays a rmsd of 1.43 A˚over only 243 amino
acids. Hence, the overall structure of the IG-NH subunit is more
similar to that of CU- and IU-NHs than that of the functionally
homologous IAG-NHs. The major structural differences
between IG-NH and both IU- and IAG-NHs are the presence
of two 310helices, η1 (spanning residues 146-150) and η2
the subunit to which the element belongs but is positioned in
front of the opening of the active site of the neighboring subunit.
Another insertion between helices R7 andR8 adopts an extended
conformation and folds into a hairpin structure that resembles
two short antiparallel β-strands. This region is, however, highly
flexible and could only be traced in one IG-NH subunit. The
crossover region of the core β-sheet, linking strands β3 and β4, is
IU-NHs, with two random coil regions flanking a single R-helix
(Figure 2 of the Supporting Information).
Tetrameric Quaternary Structure of IG-NH. The recom-
binant Tbb IG-NH displays an apparent tetrameric quaternary
structure in size exclusion chromatography and dynamic light
scattering experiments (not shown), differing from the apparent
homotrimer reported for the Crithidia IG-NH enzyme (24).
Table 3: Overview of the Kinetic Constants for TbbNH at pH 7 and 5 ?C,
Using the Nomenclature of Scheme 2
3.09 ( 0.01
9.0 ( 0.6
111 ( 12
106 ( 23
29 ( 2
2.7 ( 2.5a
2.6 ( 0.6b
270 ( 15a
65 ( 13a
2.76 ( 0.09
4 ( 2
6.8 ( 0.2
2.7 ( 2.5a
75 ( 8
18 ( 6
2.7 ( 2.5a
6.9 ( 0.4b
270 ( 15a
65 ( 13a
270 ( 15a
65 ( 13a
aRate constant determined via ribose binding stopped-flow fluorescence
flow absorbance experiments.
bRate constant determined via multiple-turnover stopped-
Scheme 2: Minimal Kinetic Scheme for the Hydrolysis of
6-Oxopurine Nucleosides by the Tbb IG-NH
Table 4: Data Collection and Refinement Statistics
cell parametersa = 115.89 A˚
b = 124.48 A˚
c = 203.88 A˚
R = β = γ = 90?
resolution range (A˚)
total no. of reflections (I > -3σ)
no. of unique reflections
Wilson B factor (A˚2)
no. of reflections (F > 0)
rmsd for bonds (A˚)
rmsd for angles (deg)
residues in favored, allowed
Ramachandran regions (%)
atomic displacement parameter (A˚2)
protein (chains A, B, C, D)
65.5, 65.9, 65.4, 65.3
i||Ii(hkl)| - |ÆI(hkl)æ||/P
Fobs(hkl)| - |Fcalc(hkl)||/P
cRfreeis the same as Rcrysbut was
9006 Biochemistry, Vol. 49, No. 41, 2010Vandemeulebroucke et al.
The Tbb IG-NH crystallized also as a homotetramer in the crystal
asymmetric unit, with an internal 222 symmetry that closely
resembles the tetrameric assembly of IU- and CU-NHs, being
mediated by interactions between the same structural elements.
Instead, the quaternary structure of Tbb IG-NH shares no resem-
dimer of dimers, with two distinct protein-protein interaction
surfaces. The total protein surfaces buried at the interfaces are
for the Crithidia IU-NH, respectively. The role of quaternary
structure assembly in modulating NH function is still elusive. The
active sites are entirely composed of residues from a single subunit,
consistent with the lack of cooperativity observed in the kinetic
A Highly Hydrophobic Base-Binding Subsite. The active
of the core β-sheet. The site can be subdivided into a hydrophilic
outer portion that interacts with the 6-oxopurine and stabilizes
the negative charge that builds up in the base at the transition
state of the hydrolysis reaction. The bottom of the cavity is lined
carbonyl oxygen of Leu131 involved in metal binding interac-
tions. The active site Ca2þis bound to the enzyme in a highly
conserved octacoordination geometry, with three water mole-
cules completing the ligation of the ion. All amino acid residues
that are involved in ribosyl binding in other NHs are strictly
conserved in the Tbb IG-NH, including Asn40, Glu177, and
The outer portion of the catalytic cavity is remarkably hydro-
phobic in character, lined by the side chain atoms of Trp80,
Phe83, Phe178, Trp205, and Pro279. The hydrophilic residues
that have a potential for interactions with the purine base are
geometry. This strong interaction suggests a fair amount of rigi-
dity of these amino acids, and possibly a weak propensity for
structural rearrangement upon binding of substrate. Although
we were unsuccessful in crystallizing the Tbb IG-NH in complex
site with other NHs make it fairly straightforward to model the
substrate-enzyme interactions. A superposition of the different
structures of ligand-bound IAG-NH, CU-NH, and IU-NH
enzymes onto the IG-NH positions the ribosyl moiety of the
substrates at the expected position, coordinating the Ca2þion.
The orientation of the base with respect to the ribosyl moiety
is instead forced by the characteristics of the catalytic cavity
(Figure 6). The hypoxanthine base is forced by the walls of the
FIGURE 4: Structure of the Tbb IG-NH. (a) Overall structure of
the Tbb IG-NH subunit, with R-helices depicted as pink ribbons,
IG-NH maintains the overall NH fold and shares a higher degree of
(b) Structure of the Tbb IG-NH tetramer. The quaternary structure
of Tbb IG-NH resembles the IU-NH tetramer, a dimer of dimers
assembled using conserved structural elements.
FIGURE 5: StereoviewoftheactivesiteofTbbIG-NH.TheactivesiteresiduesinvolvedinCa2þchelationandinteractionswiththeribosemoiety
in NHs are depicted with green carbon atoms. The residues potentially involved in interactions with the nitrogenous base are depicted with pink
carbon atoms.A molecule ofinosine was modeled bysuperposition ofthe structure ofthe E. coli CU-NH YeiK incomplex withinosine without
further intervention. Only residues within 4.5 A˚of the inosine molecule are shown.
ArticleBiochemistry, Vol. 49, No. 41, 20109007
active site ina conformation thatis intermediatebetween the syn
and anti configurations. Interestingly, the conformation closely
NH (27). In this conformation, inosine can interact with Trp80
via stacking interactions and is also positioned in close van der
Pro279. This orientation also allows the face-to-face interaction
with the side chain amide of Asn171, and possible hydrogen
provided by amino acids 256-273 from the flexible loop linking
helix R11 to strand β10, disordered in the structure presented
here, that includes four Asp residues, one His, and one Arg.
Moreover, helix R13 also points the hydrophilic side chains of
Thr321, Gln323, Ser324, and Lys327 toward the active site and
could mediate interactions with the substrate via a swinging
Lacking the enzymatic activities required for purine base
biosynthesis, trypanosomes have evolved efficient scavenging
systems (4). The NH-dependent scavenging pathway is charac-
terized by three distinct NH isozymes with different substrate
specificities and kinetic properties. Here we report the first full
molecular, functional, and structural characterization of the yet
uncharacterized isozyme, the 6-oxopurine-specific IG-NH from
T. brucei brucei.
The steady-state kinetic analysis of Tbb IG-NH (Table 1)
shows that the highest activity is found for inosine and
guanosine exemplary for an IG-specific NH, although the
IG specificity is less pronounced compared to that of the
enzyme from C. fasciculata (24). The enzyme is very efficient
with specificity constants (kcat/KM)for inosine and guanosine
almost equal to the rate expected for a diffusion-limited
reaction. Both the IG-NHs and the IAG-NHs are thus very
efficient enzymes for purine hydrolysis (25, 26), underscoring
their crucial role in purine catabolism in trypanosomes.
Indeed, their catalytic efficiencies are at least 1 order of
magnitude greater than those of the nonspecific isozymes
from other flagellates (14, 15), or even pyrimidine-specific
NHs from bacteria and yeast (49). Moreover, the KMvalues
for purine nucleosides are in the low micromolar range for
IG- and IAG-NHs, while in IU- and CU-NHs, they approach
the millimolar values. The low KMvalues exhibited by both
IG- and IAG-NHs are likely to facilitate the capture of
substrates in the trypanosome. Because the concentration
of inosine in human plasma is estimated to be 1 μM (24), it is
obvious that IG- and IAG-NHs play the central role in
rescuing nucleosides from the host for the synthesis of
RNA, DNA, and cofactors. In fact, a recent study showed
that a nanomolar inhibitor of the Tbb IAG-NH is effective in
killing trypanosomes in a murine model of infection (11). The
IAG-NH from Tbb has lower kcatvalues and higher KM
latter is the main NH in Tbb for inosine and guanosine
processing, while the IAG-NH probably focuses on adeno-
sine hydrolysis (25).
The Tbb IG-NH’s activity profile clearly shows that the
exocyclic group at position 6 of the purine base ring plays a
crucial role in catalysis. A functional group with H-acceptor
substrates, demonstrated by the kinetic parameters of inosine,
turnover rate and a higher affinity are observed when the purine
ring contains a functional group with H-donor capacities linked
to the C6 atom, as in purine riboside, 2-aminopurine ribose, and
The IG-NH is like the IAG-NHs, truly purine-specific (see
Table 1), with kcat/KMratios 1000-1000000 times higher for the
naturally occurring purine nucleosides then for the pyrimidine
nucleosides (Table 1), yet the only requirement for efficient
catalysis seems to be the presence of the purine ring, because
on the purine ring, is as effectively catalyzed as that of the other
purine nucleosides. Adenosine displays a submicromolar KM
coupled to a kcat2 orders of magnitude lower than the kcatfor
6-oxopurinenucleosides(0.24 s-1). Thisfindingsuggeststhatthe
substrate specificity of Tbb IG-NH is determined by catalytic
turnover rather than specific binding interactions and takes
advantage of the presence of the carbonyl group at C6 of the
purine ring to lower the transition-state barrier. This could be
effectively achieved by stabilizing the partial negative charge
developing inthe purine ringthrough resonance formsdelocaliz-
ing the negative charge to the C6 carbonyl. The same delocaliza-
tion is not as favorable for the adenine base, because the
events have been proposed to play a role in the catalytic mecha-
nism of bacterial purine nucleoside phosphorylases (50).
Tbb IG-NH is the only so far characterized NH that is able to
catalyze the hydrolysis of 3-deazaadenosine, indicating that N3
of the purine ring is less important for catalysis than for NHs
from other specificity classes. The low KMfor 3-deazaadenosine
moreover shows that the purine N3 atom is not involved in
any binding interactions. The substrate analogue p-nitrophenyl
β-D-ribofuranoside (pNPR) does not require protonation at the
leaving group but is susceptible to O-glycosidic bond hydrolysis
when the ribosyl moiety is converted to the oxocarbenium ion.
pNPR is a relatively poor substrate for the Tbb IG-NH, hence
indicating that the enzyme attains a considerable fraction of
its catalytic power via leaving group activation rather than via
ribosyl distortion. Nevertheless, the kinetic constants of the
deoxynucleosides (Table 1) highlight the critical importance in
FIGURE 6: Molecular surface at the Tbb IG-NH active site. The
molecular surface of the active site shows a tight fit of the hypox-
anthine base (based on the model generated by superposition of the
unliganded Tbb IG-NH structure with the YeiK-inosine complex
structure) and a parallel π-π interaction with residue Trp80.
9008 Biochemistry, Vol. 49, No. 41, 2010Vandemeulebroucke et al.
catalysis of all three hydroxyl groups of the ribose moiety of the
On the basis of the pre-steady-state analysis, we can suggest a
minimal kinetic scheme (Scheme 2) for the hydrolysis of 6-oxo-
a reversible chemistry step followed by an ordered product release
and Table 3) is followed by a slow isomerization (k4and k-4in
Scheme 2 and Table 3) from a tightly bound (E3R0) to a loosely
bound (E3R) enzyme3ribose complex prior to ribose dissociation.
This isomerization is the overall rate-determining step of the
The finding that this slow step prior to ribose dissociation
(2.7 s-1at 5 ?C, k4in Scheme 2 and Table 1), which should be
independent of the substrate used, is faster than the catalytic
turnover of adenosine (0.24 s-1at 35 ?C, kcatin Table 1) could
suggest that the rate-limiting step for adenosine hydrolysis
catalyzed by Tbb IG-NH is shifted from product release to a
very slow inefficient chemistry step. Therefore, the overall
catalytic turnover rate will probably be determined by chemistry
for adenosine hydrolysis, while the catalytic turnover rate
of 6-oxopurine nucleoside hydrolysis is determined by ribose
sides and other purine nucleosides points to an even greater
degree of discrimination between adenosine and 6-oxopurine
kcatand further corroborates our statement above that the
substrate specificity is determined by the catalytic (chemistry)
Here we demonstrate that the chemistry step of the hydrolysis
of guanosine and 7mGuo catalyzed by Tbb IG-NH is reversible.
No reversibility of the chemistry step was discovered for purine
hydrolysis by the Tv IAG-NH (35). Moreover, a higher reversi-
bility of the chemistry step is observed for guanosine than for
a guanosine analogue with an activated (positively charged)
purine ring that needs no further leaving group activation by
enzymatic protonation (51). Thus, the observed difference in the
on-enzyme reversibility could be explained on the basis of the
lower nucleophilicity of the 7-methylguanine base toward ribose
compared to guanine. The reason for the difference between the
on-enzyme reversibility of guanosine hydrolysis catalyzed by Tv
IAG-NH and Tbb IG-NH remains to be clarified.
Overall, this pre steady-state kinetic analysis of the Tbb IG-
NH demonstrates that the IAG- and IG-specific nucleoside
hydrolases have very comparable kinetic mechanisms (35). In
both mechanisms, the isomerization necessary to convert a
tightly bound enzyme3ribose complex to a loosely bound com-
plex prior to ribose dissociation is the overall rate-determining
step. Hence, it is very tempting to suggest that this mechanism is
generally applicable for the nucleoside hydrolase family, or at
least for the purine-specific NHs. The two flexible loops named
For all specificity classes, two distinct conformations of these
loops have been observed: an open ligand free form and a closed
ligand-bound state (29, 30). For the Tv IAG-NH, it was shown
that flexible loop II (corresponding to flexible residues 256-273
in Tbb IG-NH) is involved in the overall rate-determining
is observed. Therefore, it is likely that this loop restructuring is
involved in ribose release and determines the steady-state turn-
over rates of this Tbb IG-NH, perhaps even of all NHs. This
hypothesis further implies that the flexible loop II could be a
conserved functional element in the catalytic mechanism of the
nucleoside hydrolase family.
The Tbb IG-NH structure represents yet another variation
from the standard NH fold (26, 31). Despite its functional and
mechanistic similarities to IAG-NHs, the IG-NH displays
more tertiary and quaternary structure similarity to IU- and
CU-NHs (Figure 2 of the Supporting Information). However,
while the Tbb IG-NH maintains the trademark Ca2þ-contain-
ing active site for ribosyl binding and discrimination, the
region devoted to base interaction displays unique features
compared to the other specificity classes. The walls of the
active site cavity of NHs are devoted to binding interactions
with the substrate nitrogenous base,and its stabilization along
the reaction coordinate (27, 28, 49). A conserved histidine
residue, whose mutation to alanine reduces the catalytic
efficiencyoftheCf IU-NHby afactorof3800,makingitlikely
to act as a general acid (7), is a trademark of IU-NHs. The Tbb
IG-NH lacks this histidine residue, thus reflecting the closer
functional similarity with IAG-NH. However, this is achieved
NH active site clearly suggests an important role of aromatic
stacking mediated by residue Trp80 in the catalytic mechanism,
similar to the pKa-increasing stacking proposed for the Tv IAG-
NH (27). In contrast to IAG-NH, the conformation of the
docked inosine (intermediate between syn and anti) does not
charge developing in the purine ring at the transition state
(Figure 6 and Figure 3 of the Supporting Information). Hence,
hydrophilic or charged residues, the ones present in the flexible
loop II linking helix R11 to strand β10 in the Tbb IG-NH, are
likely to participate in the leaving group activation and/or
in the unliganded IG-NH structure is more than 7 A˚from the
active site, could perhaps relocate upon substrate binding. The
residues in this helix are conserved in all closely related, try-
panosomal IG-NH homologues, supporting a possible role in
The molecular characterization of protozoal NHs has long
among the first nanomolar NH inhibitors identified and are
effective against IU-, IAG-, and IG-NHs. These compounds
have proven to be effective in killing Plasmodiumfalciparum in
human erythrocytes (53), a pathogenic protozoan that causes
malaria and is also purine-auxotrophic, though it relies for
base salvage on purine nucleoside phosphorylase and not
NHs. Moreover, Immucillins are approved for use against T
cell lymphomas as PNP inhibitors (54). The availability of
detailed kinetic and structural information for all specificity
classes of the trypanosomal NHs now provides a formidable
template for specific targeting of the protozoal nucleobase
salvage pathway. Indeed, N-arylmethyl-substituted iminori-
bitols are selective, potent inhibitors of purine-specific NHs
(55) and have proven to be effective in trypansomal infection
models (11). Further development of these compounds will
take advantage of the fine understanding of the NH structure
and enzymatic properties.
ArticleBiochemistry, Vol. 49, No. 41, 2010 9009
We acknowledge the use of beamline ID14-EH4 at the
European Synchrotron Radiation Facility.
SUPPORTING INFORMATION AVAILABLE
is available free of charge via the Internet at http://pubs.acs.org.
1. Barrett, M. P., Burchmore, R. J., Stich, A., Lazzari, J. O., Frasch,
A. C., Cazzulo, J. J., and Krishna, S. (2003) The trypanosomiases.
Lancet 362, 1469–1480.
2. Markell, E. K., John, D. T., and Krotoski, W. A. (1999) Markell and
Voge’s Medical Prasitology, 8th ed., W. B. Saunders Co., Philadelphia.
3. Naula, C., and Burchmore, R. (2003) A plethora of targets, a paucity
of drugs: Progress towards the development of novel chemotherapies
for human African trypanosomiasis. Expert Rev. Anti-Infect. Ther. 1,
4. Berriman, M., Ghedin, E., Hertz-Fowler, C., Blandin, G., Renauld,
H., Bartholomeu, D. C., Lennard, N. J., Caler, E., Hamlin, N. E.,
Haas, B., Bohme, U., Hannick, L., Aslett, M. A., Shallom, J.,
Marcello, L., Hou, L., Wickstead, B., Alsmark, U. C., Arrowsmith,
C., Atkin, R. J., Barron, A. J., Bringaud, F., Brooks, K., Carrington,
M., Cherevach, I., Chillingworth, T. J., Churcher, C., Clark, L. N.,
Corton, C. H., Cronin, A., Davies, R. M., Doggett, J., Djikeng, A.,
Feldblyum, T., Field, M. C., Fraser, A., Goodhead, I., Hance, Z.,
Harper, D., Harris, B. R., Hauser, H., Hostetler, J., Ivens, A., Jagels,
K., Johnson, D., Johnson, J., Jones, K., Kerhornou, A. X., Koo, H.,
Larke, N., Landfear, S., Larkin, C., Leech, V., Line, A., Lord, A.,
Macleod, A., Mooney, P. J., Moule, S., Martin, D. M., Morgan,
G. W., Mungall, K., Norbertczak, H., Ormond, D., Pai, G., Peacock,
C. S., Peterson, J., Quail, M. A., Rabbinowitsch, E., Rajandream,
M. A., Reitter, C., Salzberg, S. L., Sanders, M., Schobel, S., Sharp,
S., Simmonds, M., Simpson, A. J., Tallon, L., Turner, C. M., Tait,
A., Tivey, A. R., Van Aken, S., Walker, D., Wanless, D., Wang, S.,
White, B., White, O., Whitehead, S., Woodward, J., Wortman, J.,
Adams, M. D., Embley, T. M., Gull, K., Ullu, E., Barry, J. D.,
Fairlamb, A. H., Opperdoes, F., Barrell, B. G., Donelson, J. E., Hall,
N., Fraser, C. M., Melville, S. E., and El-Sayed, N. M. (2005) The
metabolism in the Trypanosomatidae. Mol. Biochem. Parasitol. 13,
6. Giabbai, B., and Degano, M. (2004) Crystal structure to 1.7 A˚of the
Escherichia coli pyrimidine nucleoside hydrolase YeiK, a novel
candidate for cancer gene therapy. Structure 12, 739–749.
7. Gopaul, D. N., Meyer, S. L., Degano, M., Sacchettini, J. C., and
Schramm, V. L. (1996) Inosine-uridine nucleoside hydrolase from
Crithidia fasciculata. Genetic characterization, crystallization, and
identification of histidine 241 as a catalytic site residue. Biochemistry
8. Parkin, D. W., Horenstein, B.A., Abdulah, D. R., Estupinan, B.,and
Schramm, V. L. (1991) Nucleoside hydrolase from Crithidia fascicu-
lata. Metabolic role, purification, specificity, and kinetic mechanism.
J. Biol. Chem. 266, 20658–20665.
9. Berg, M., Bal, G., Goeminne, A., Van der Veken, P., Versees, W.,
Steyaert, J., Haemers, A., and Augustyns, K. (2009) Synthesis of
bicyclic N-arylmethyl-substituted iminoribitol derivatives as selective
nucleoside hydrolase inhibitors. ChemMedChem 4, 249–260.
10. Miles, R. W., Tyler, P. C., Evans, G. B., Furneaux, R. H., Parkin,
D. W., and Schramm, V. L. (1999) Iminoribitol transition state
11. Berg, M., Kohl, L., Van der Veken, P., Joossens, J., Al-Salabi, M. I.,
P., Haemers, A., Degano, M., Maes, L., de Koning, H. P., and
Augustyns, K. (2010) Evaluation of nucleoside hydrolase inhibi-
tors for treatment of African trypanosomiasis. Antimicrob. Agents
Chemother. 54, 1900–1908.
12. Miller, R. L., Sabourin, C. L., Krenitsky, T. A., Berens, R. L., and
Marr, J. J. (1984) Nucleoside hydrolases from Trypanosoma cruzi.
J. Biol. Chem. 259, 5073–5077.
13. Pelle, R., Schramm, V. L., and Parkin, D. W. (1998) Molecular
cloning and expression of a purine-specific N-ribohydrolase from
Trypanosoma brucei brucei. Sequence, expression, and molecular
analysis. J. Biol. Chem. 273, 2118–2126.
14. Shi, W., Schramm, V. L., and Almo, S. C. (1999) Nucleoside hydro-
lase fromLeishmaniamajor.Cloning, expression, catalytic properties,
transition state inhibitors, and the 2.5-A˚ crystal structure. J. Biol.
Chem. 274, 21114–21120.
15. Versees, W., Decanniere, K., Pelle, R., Depoorter, J., Brosens, E.,
Parkin, D. W., and Steyaert, J. (2001) Structure and function of a
novel purine specific nucleoside hydrolase from Trypanosoma vivax.
J. Mol. Biol. 307, 1363–1379.
16. Petersen, C., and Moller, L. B. (2001) The RihA, RihB, and RihC
ribonucleoside hydrolases of Escherichia coli. Substrate specificity,
gene expression, and regulation. J. Biol. Chem. 276, 884–894.
17. Hunt, C., Gillani, N., Farone, A., Rezaei, M., and Kline, P. C. (2005)
Kinetic isotope effects of nucleoside hydrolase from Escherichia coli.
Biochim. Biophys. Acta 1751, 140–149.
18. Ogawa, J., Takeda, S., Xie, S. X., Hatanaka, H., Ashikari, T.,
Amachi, T., and Shimizu, S. (2001) Purification, characterization,
and gene cloningof purine nucleosidase from Ochrobactrum anthropi.
Appl. Environ. Microbiol. 67, 1783–1787.
19. Versees,W.,VanHolsbeke,E.,DeVos,S.,Decanniere,K.,Zegers, I.,
and Steyaert, J. (2003) Cloning, preliminary characterization and
crystallization of nucleoside hydrolases from Caenorhabditis elegans
and Campylobacter jejuni. Acta Crystallogr. D59, 1087–1089.
20. Kurtz, J. E., Exinger, F., Erbs, P., and Jund, R. (2002) The URH1
uridine ribohydrolase of Saccharomyces cerevisiae. Curr. Genet. 41,
21. Ribeiro, J. M., and Valenzuela, J. G. (2003) The salivary purine
nucleosidase of the mosquito Aedes aegypti. Insect Biochem. Mol.
Biol. 33, 13–22.
22. Porcelli, M., Concilio, L., Peluso, I., Marabotti, A., Facchiano, A.,
lase from the archaeon Sulfolobus solfataricus: Biochemical charac-
terization and homology modeling. FEBS J. 275, 1900–1914.
23. Porcelli, M., Peluso,I., Marabotti, A., Facchiano, A., andCacciapuoti,
G. (2009) Biochemical characterization and homology modeling of a
purine-specific ribonucleoside hydrolase from the archaeon Sulfolobus
solfataricus: Insights into mechanisms of protein stabilization. Arch.
Biochem. Biophys. 483, 55–65.
24. Estupinan, B., and Schramm, V. L. (1994) Guanosine-inosine-preferring
nucleoside N-glycohydrolase from Crithidia fasciculata. J. Biol. Chem.
25. Parkin, D. W. (1996) Purine-specific nucleoside N-ribohydrolase
from Trypanosoma brucei brucei. Purification, specificity, and kinetic
mechanism. J. Biol. Chem. 271, 21713–21719.
26. Versees, W., and Steyaert, J. (2003) Catalysis by nucleoside hydro-
lases. Curr. Opin. Struct. Biol. 13, 731–738.
27. Versees, W., Decanniere, K., Van Holsbeke, E., Devroede, N., and
Steyaert, J. (2002) Enzyme-substrate interactions in the purine-
specific nucleoside hydrolase from Trypanosoma vivax. J. Biol. Chem.
28. Degano, M., Almo, S. C., Sacchettini, J. C., andSchramm, V. L. (1998)
Trypanosomal nucleoside hydrolase. A novel mechanism from the
structure with a transition-state inhibitor. Biochemistry 37, 6277–6285.
29. Versees, W., Barlow, J., and Steyaert, J. (2006) Transition-state
complex of the purine-specific nucleoside hydrolase of T. vivax:
Enzyme conformational changes and implications for catalysis.
J. Mol. Biol. 359, 331–346.
30. Versees, W., Goeminne, A., Berg, M., Vandemeulebroucke, A.,
of T. vivax nucleoside hydrolase in complex with new potent and
specific inhibitors. Biochim. Biophys. Acta 1794, 953–960.
31. Degano, M., Gopaul, D. N., Scapin, G., Schramm, V. L., and
Sacchettini, J. C. (1996) Three-dimensional structure of the inosine-
uridine nucleoside N-ribohydrolase from Crithidia fasciculata. Bio-
chemistry 35, 5971–5981.
W: Improving the sensitivity of progressive multiple sequence align-
ment through sequence weighting, position-specific gap penalties and
weight matrix choice. Nucleic Acids Res. 22, 4673–4680.
33. Pace, C. N., Vajdos, F., Fee, L., Grimsley, G., and Gray, T. (1995)
How to measure and predict the molar absorption coefficient of a
protein. Protein Sci. 4, 2411–2423.
34. Cook, P. F., and Cleland, W. W. (2007) Initial velocity studies:
Presence of added inhibitors. In Enzyme Kinetics and Mechanism
(Shank, D., Rogers, R. L., and Scholl, S., Eds.) p 121, Taylor and Francis
Group, New York.
9010 Biochemistry, Vol. 49, No. 41, 2010Vandemeulebroucke et al.
35. Vandemeulebroucke, A., Versees, W., Steyaert, J., and Barlow, J. N.
lase reveal complex substrate binding, product base release, and two
apparent rates of chemistry. Biochemistry 45, 9307–9318.
36. Kabsch, W. (1988) Automatic indexing of rotation diffraction pat-
terns. J. Appl. Crystallogr. 21, 67–72.
37. French, G. S., and Wilson, K. S. (1978) On the treatment of negative
intensity observations. Acta Crystallogr. A34, 517–525.
38. Vagin, A. A., and Isupov, M. N. (2001) Spherically averaged phased
translation function and its application to the search for molecules
and fragments in electron-densitymaps.ActaCrystallogr. D57,1451–
39. Cowtan, K. (1994) Joint CCP4 and ESF-EACBM Newsletter on
Protein Crystallography 31, 34–38.
40. Jones, T. A., Zou, J., Cowan, S. W., and Kjeldgaard, M. (1991)
Improved methods for building protein models in electron density
maps and the location of errors in these models. Acta Crystallogr.
41. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Refine-
ment of macromolecular structures by the maximum-likelihood
method. Acta Crystallogr. D53, 240–255.
42. Winn, M. D., Isupov, M. N., and Murshudov, G. N. (2001) Use of
TLS parameters to model anisotropic displacements in macromole-
cular refinement. Acta Crystallogr. D57, 122–133.
(2004) MOLPROBITY: Structure validation and all-atom contact
analysis for nucleic acids and their complexes. Nucleic Acids Res. 32,
44. Morris,R. J., Perrakis, A.,and Lamzin,V.S. (2003)ARP/wARPand
automatic interpretation of protein electron density maps. Methods
Enzymol. 374, 229–244.
P., De Prol, S., Versees, W., Steyaert, J., Apers, S., Haemers, A., and
Augustyns, K. (2007) 1,2,3-Triazolylalkylribitol derivatives as nucleoside
hydrolase inhibitors. Bioorg. Med. Chem. Lett. 17, 2523–2526.
46. Goeminne, A., McNaughton, M., Bal, G., Surpateanu, G., Van
Der Veken, P., De Prol, S., Versees, W., Steyaert, J., Haemers, A.,
and Augustyns, K. (2008) Synthesis and biochemical evaluation of
guanidino-alkyl-ribitol derivatives as nucleoside hydrolase inhi-
bitors. Eur. J. Med. Chem. 43, 315–326.
47. Vandemeulebroucke, A., Versees, W., De Vos, S., Van Holsbeke, E.,
and Steyaert, J. (2003) Pre-steady-state analysis of the nucleoside
hydrolase of Trypanosoma vivax. Evidence for half-of-the-sites reac-
tivity and rate-limiting product release. Biochemistry 42, 12902–
48. Barlow, J. N., and Steyaert, J. (2007) Examination of the mechanism
and energetic contribution of leaving group activation in the purine-
specific nucleoside hydrolase from Trypanosoma vivax. Biochim.
Biophys. Acta 1774, 1451–1461.
49. Iovane, E., Giabbai, B., Muzzolini, L., Matafora, V., Fornili, A.,
Minici, C., Giannese, F., and Degano, M. (2008) Structural basis for
substrate specificity in group I nucleoside hydrolases. Biochemistry
50. Deng, H., Cahill, S. M., Abad, J. L., Lewandowicz, A., Callender,
R. H., Schramm, V. L., and Jones, R. A. (2004) Active site contacts in
the purine nucleoside phosphorylase-hypoxanthine complex by
NMR and ab initio calculations. Biochemistry 43, 15966–15974.
51. Versees, W., Loverix, S., Vandemeulebroucke, A., Geerlings, P., and
Steyaert, J.(2004)Leaving groupactivationby aromatic stacking: An
alternative to general acid catalysis. J. Mol. Biol. 338, 1–6.
52. Vandemeulebroucke, A., De Vos, S., Van Holsbeke, E., Steyaert, J.,
and Versees, W. (2008) A flexible loop as a functional element in the
catalytic mechanism of nucleoside hydrolase from Trypanosoma
vivax. J. Biol. Chem. 283, 22272–22282.
53. Kicska, G. A., Tyler, P. C., Evans, G. B., Furneaux, R. H., Schramm,
induced by immucillin-H, a transition state analogue of purine
nucleoside phosphorylase. J. Biol. Chem. 277, 3226–3231.
54. Schramm, V. L. (2004) Immucillins as antibiotics for T-cell prolifera-
tion and malaria. Nucleosides, Nucleotides Nucleic Acids 23, 1305–
55. Goeminne, A., Berg, M., McNaughton, M., Bal, G., Surpateanu, G.,
and Augustyns, K. (2008) N-Arylmethyl substituted iminoribitol
derivatives as inhibitors of a purine specific nucleoside hydrolase.
Bioorg. Med. Chem. 16, 6752–6763.