Mode of action of recombinant hypoxanthine-guanine phosphoribosyltransferase from Mycobacterium tuberculosis

Article abstract of DOI:10.1039/c5ra14918e

Tuberculosis (TB) is the second most important cause of mortality worldwide due to a single infectious agent, Mycobacterium tuberculosis. A better understanding of the purine salvage pathway can unveil details of the biology of M. tuberculosis that might be used to develop new strategies to combat this pathogen. Hypoxanthine-guanine phosphoribosyltransferase (HGPRT) is an enzyme from the purine phosphoribosyltransferase (PRTase) family and catalyzes the conversion of hypoxanthine or guanine and 5-phospho-α-d-ribose 1-diphosphate (PRPP) to, respectively, inosine 5′-monophosphate (IMP) or guanosine 5′-monophosphate (GMP), and pyrophosphate (PPi). Gel filtration chromatography has shown that recombinant M. tuberculosis HGPRT (MtHGPRT) is homodimeric. A sequential compulsory ordered enzyme mechanism with PRPP as the substrate that binds to free MtHGPRT enzyme and PPi as the first product to dissociate is proposed based on kinetic data and thermodynamics of ligand binding from isothermal titration calorimetry (ITC) results. ITC data have also provided thermodynamic signatures of non-covalent interactions for PRPP, IMP and GMP binding to free MtHGPRT. Thermodynamic activation parameters (E_a, ΔG^#, ΔS^#, ΔH^#) for the MtHGPRT-catalyzed chemical reaction, pre-steady-state kinetics, solvent kinetic isotope effects, equilibrium constants and pH-rate profiles are also presented. Pre-steady-state analysis reveals that there is an initial rapid phase (burst) followed by a slower phase, suggesting that product release is rate limiting. The data here described provide a better understanding of the mode of action of MtHGPRT.

Full text of DOI:10.1039/c5ra14918e

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Mode of action of recombinant hypoxanthine–
guanine phosphoribosyltransferase from
Mycobacterium tuberculosis
Cite this: RSC Adv., 2015, 5, 74671
Paulo C. Patta,†ab Leonardo K. B. Martinelli,†a Mariane Rotta,†ac Bruno L. Abbadi,†ab
ab
ab
*
*
Diogenes S. Santos†
and Luiz A. Basso†
Tuberculosis (TB) is the second most important cause of mortality worldwide due to a single infectious
agent, Mycobacterium tuberculosis. A better understanding of the purine salvage pathway can unveil
details of the biology of M. tuberculosis that might be used to develop new strategies to combat this
pathogen. Hypoxanthine–guanine phosphoribosyltransferase (HGPRT) is an enzyme from the purine
phosphoribosyltransferase (PRTase) family and catalyzes the conversion of hypoxanthine or guanine and
5-phospho-a-D-ribose 1-diphosphate (PRPP) to, respectively, inosine 50-monophosphate (IMP) or
guanosine 50-monophosphate (GMP), and pyrophosphate (PPi). Gel filtration chromatography has shown
that recombinant M. tuberculosis HGPRT (MtHGPRT) is homodimeric. A sequential compulsory ordered
enzyme mechanism with PRPP as the substrate that binds to free MtHGPRT enzyme and PPi as the first
product to dissociate is proposed based on kinetic data and thermodynamics of ligand binding from
isothermal titration calorimetry (ITC) results. ITC data have also provided thermodynamic signatures of
non-covalent interactions for PRPP, IMP and GMP binding to free MtHGPRT. Thermodynamic activation
parameters (Ea, DG#, DS#, DH#) for the MtHGPRT-catalyzed chemical reaction, pre-steady-state kinetics,
solvent kinetic isotope effects, equilibrium constants and pH-rate profiles are also presented. Pre-
steady-state analysis reveals that there is an initial rapid phase (burst) followed by a slower phase,
suggesting that product release is rate limiting. The data here described provide a better understanding
of the mode of action of MtHGPRT.
Received 27th July 2015
Accepted 28th August 2015
DOI: 10.1039/c5ra14918e
www.rsc.org/advances
been associated with co-infection with human immunode-
ciency virus (HIV),1,2 and the emergence of multi, extensively3
and recently, totally drug resistant strains of M. tuberculosis.4
Introduction
Mycobacterium tuberculosis is the major etiological agent of
human tuberculosis (TB) and is believed to infect one-third of
the world's population. This bacteria was responsible for 8.6
million new TB cases in 2012, which resulted in 1.3 million
deaths worldwide.1 According to the World Health Organization
(WHO), TB is the second most important cause of mortality
worldwide as a single infectious agent.1 Despite effective short-
course chemotherapy, the increasing global burden of TB has
TB is an ancient human disease,5 but little is known about
the nutritional adaptability of M. tuberculosis in the progression
of TB infection.6,7 It is still not clear whether M. tuberculosis
recycles complex nutrient molecules from the human host
using salvage pathways or relies on synthesis of essential
molecules from passive diffusible precursors via de novo
synthesis pathways. Purine and pyrimidine salvage pathways in
M. tuberculosis remain an incompletely explored possibility for
drug development as purine and pyrimidine biosynthesis are
essential steps for the cell, supplying building blocks for DNA
and RNA synthesis, among other biological roles.8 Accordingly,
elucidation of biochemical properties of the enzymes involved
in purine and pyrimidine salvage pathways should contribute to
a better understanding of the biology of M. tuberculosis, and,
hopefully, to the design of analogs that may selectively inhibit
M. tuberculosis replication and survival.9
aCentro de Pesquisas em Biologia Molecular e Funcional (CPBMF), Instituto Nacional
ˆ
´
´
de Ciencia e Tecnologia em Tuberculose (INCT-TB), Pontifıcia Universidade Catolica
´
do Rio Grande do Sul (PUCRS), 6681/92-A Av. Ipiranga – Tecnopuc – Predio 92A,
90619-900, Porto Alegre, RS, Brazil. E-mail: luiz.basso@pucrs.br; diogenes@pucrs.
br; Fax: +55-51-33203629; Tel: +55-51-33203629
b
´
˜
´
Programa de Pos-Graduaçao em Biologia Celular e Molecular, Pontifıcia Universidade
´
Catolica do Rio Grande do Sul (PUCRS), 6681/12-Av. Ipiranga, 90619-900, Porto
Alegre, RS, Brazil. Fax: +55-51-33203912; Tel: +55-51-33203912
cPrograma de Pos-Graduaçao em Medicina
´
˜
ˆ ´ ´
e Ciencias da Saude, Pontifıcia
The purine phosphoribosyltransferases (PRTases) form a
family of enzymes that transfer 5-phosphoribosyl from 5-phos-
pho-a-D-ribose 1-diphosphate (PRPP; a-D-5-phosphoribosylpyr-
´
Universidade Catolica do Rio Grande do Sul (PUCRS), 6690/30-Av. Ipiranga, 90610-
000, Porto Alegre, RS, Brazil. Fax: +55-51-33203318; Tel: +55-51-33203318
† The manuscript was written through contributions of all authors. All authors
have given approval to the nal version of the manuscript.
ophosphate; a-D-ribosyl diphosphate 5-phosphate) to
a
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 74671–74683 | 74671

View Article Online
RSC Advances
Paper
nitrogen-containing nucleophile (such as the imidazole N-9 of a ltration data suggest a homodimeric quarternary structure for
purine base) to form the corresponding b-substituted ribose 5- MtHGPRT. Thermodynamic activation parameters (Ea, DG#,
phosphate (such as purine nucleotides).10 Despite the lack of DS#, DH#) for MtHGPRT-catalyzed chemical reaction, solvent
clear sequence homology among the PRTases, these enzymes kinetic isotope effects, equilibrium constant (Keq) determina-
show tertiary and quaternary structure conservation.11 A PPRP tion, solvent kinetics isotope effects, transient kinetics
binding motif of 12 amino acids, which is also found in PRPP measurements and pH-rate proles results are also presented.
synthetases, is conserved among the PRTases involved in The absolute requirement of divalent magnesium ion for
nucleotide synthesis or salvage pathways.12 This is the main catalysis, the sequential kinetic mechanism, the presence of
unifying characteristic of type I PRTases. The type I purine PRPP binding motif, the homodimeric assembly in solution
PRTase-catalyzed reactions have been found to follow a indicate that MtHGPRT belongs to type I PRTases family of
sequential ordered bi–bi mechanism, in which PRPP binds to enzymes. It is hoped that the results presented here contributes
the free enzyme followed by the purine base, and ordered PPi to a better understanding of MtHGPRT mode of action, and may
and nucleotide products release.13–16 Another characteristic also be useful to chemical biologists interested in designing
feature among the type I PRTase structures is a long exible loss-of-function (inhibitors) or gain-of-function (activators)
loop closely associated with the active site, known as “the chemical compounds to reveal the biological role of MtHGPRT
catalytic loop”. In most of the determined structures of type I in the context of whole M. tuberculosis cells.
PRTase, the loop is usually highly disordered.10 This PRTase
family is of signicant interest in both human genetic diseases
and parasitic pathologies, such as Lesch–Nyhan syndrome
and Chagas' disease.15–17 One of the enzymes that belong to
Experimental
Materials
this family is the hypoxanthine–guanine phosphoribosyl-
transferase (HGPRT; EC 2.4.2.8) of purine salvage pathway.
HGPRTs are found in most microorganisms and mammals
and their reaction involves the ribophosphorylation in one
step of purine nucleobases (hypoxanthine and guanine) and
their analogues to their respective nucleoside 50-mono-
All chemicals were purchased from Sigma-Aldrich (Saint Louis,
USA), unless otherwise specied. The low molecular weight
(LMW) and high molecular weight (HMW) Gel Filtration Cali-
bration Kits were purchased from GE Healthcare. All kinetic
data analyses were carried out using SigmaPlot 10.0 (Systac
Soware, Inc., Melbourne, USA). Data are presented as mean ꢀ
standard deviation unless stated otherwise. ITC data analysis
was evaluated utilizing the Origin 7 SR4 soware (Microcal,
Inc.). All experiments were performed at 25 ꢁC using 50 mM 2-
amino-2-hydroxymethyl-propane-1,3-diol (Tris)–HCl pH 7.4
containing 12 mM MgCl2 (buffer A), in duplicates, unless
otherwise specied.
phosphate and pyrophosphate.18 HGPRT catalyzes the Mg2+
-
dependent reversible transfer of the 5-phosphoribosyl group
from PRPP to the N9 of guanine (Gua) (GPRT reaction) or
hypoxanthine (Hx) (HPRT reaction) to form the corresponding
ribonucleotides guanosine 50-monophosphate (GMP) and
inosine 50-monophosphate (IMP), releasing PPi.18 Investiga-
tion of purine salvage pathway enzymes of M. tuberculosis,
including HGPRT (MtHGPRT), might reveal insightful data on
the complex balance that exists between the bacillus and the
host.18 Recently, the rst crystal structure of MtHGPRT has
been reported.19 These authors have also described acyclic
nucleoside phosphonate (ANP) compounds that inhibit
MtHGPRT enzyme activity, and phosphoroamidates of ANPs
Overexpression and purication
The recombinant MtHGPRT was overexpressed and puried to
homogeneity as previously described.18
Oligomeric state determination
that inhibit the growth of M. tuberculosis.19 Incidentally, Patel The oligomeric state of homogenous MtHGPRT in solution was
and colleagues have employed computational strategies to determined by size exclusion liquid chromatography on a
design a series of fullerene–quinazolinones conjugates based HighLoad 10/30 Superdex-200 column (GE Healthcare). The
on homology models of MtHGPRT.20 These compounds were column was pre equilibrated and the sample (100 mL) was iso-
synthesized and showed antibacterial activity against M. cratically eluted with 1 column volume (CV) of 50 mM Tris–HCl
tuberculosis.20 Accordingly, these efforts suggest that pH 7.5 containing 200 mM NaCl at a ow rate of 0.4 mL minꢂ1
.
MtHGPRT is a potential target for the development of Protein elution was monitored at 215, 254, and 280 nm. The
chemotherapeutic agents to treat TB. However, the rational- LMW and HMW Gel Filtration Calibration Kits were used as
based drug design must, preferably, rely on structural and described by the manufacturer to prepare the calibration curve.
functional data. Accordingly, efforts to elucidate the mode of The elution volumes (Ve) of protein standards were used to
action of MtHGPRT appear to be warranted to improve our calculate their corresponding partition coefficients (KAV
)
understanding of purine metabolism in this human path- according to eqn (1). Blue dextran 2000 (GE Healthcare) was
ogen, and allowing a function-based approach to also be used used to determine the void volume (Vo) and Vt is the total bead
for MtHGPRT inhibitor design efforts.
volume of the column. The KAV value for each protein was
The present work describes the mode of action of recombi- plotted versus the logarithm of their corresponding molecular
nant MtHGPRT. True steady-state kinetic parameters and masses, giving a linear relationship. A volume of 100 mL (100
isothermal titration calorimetry (ITC) data indicate that mM) of MtHGPRT was loaded on the gel ltration column to
MtHGPRT follows
a sequential ordered mechanism. Gel obtain Ve. The partition coefficient (KAV) of the recombinant
74672 | RSC Adv., 2015, 5, 74671–74683
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
MtHPRT was calculated by data tting to eqn (1) and its dissociation constant for enzyme-substrate A binary complex
molecular mass derived from the linear relationship.
formation.
Ve ꢂ Vo
KAV
¼
(1)
Isothermal titration calorimetry
Vt ꢂ Vo
ITC experiments were carried out using an iTC200 Microcalo-
rimeter (Microcal, Inc., Pittsburgh, USA). The instrument
reference cell (200 mL) was loaded with Milli-Q water in all
experiments and sample cell (200 mL) was lled with either
100 mM or 80 mM of recombinant MtHGPRT. The injection
syringe (39.7 mL) was lled with substrates or products at
different concentrations: Hx (7 mM), PRPP (10 mM), IMP
(5 mM), GMP (2.5 mM) and PPi (1 mM), and the ligand
isotherms were measured by direct titration (ligand into
macromolecule). The same buffer preparation was used to
Steady-state kinetics
Recombinant MtHGPRT enzyme activity was measured by a
continuous spectrophotometric assay measuring the linear
increase in absorbance as a function of IMP or GMP formation
in quartz cuvettes (1 cm). The experiments were performed in a
UV-visible Shimadzu spectrophotometer UV2550 equipped with
a temperature-controlled cuvette holder. The kinetic properties
of MtHGPRT for Hx, Gua, and PRPP were spectrophotometri-
cally determined using the difference in molar absorptivity
between the nucleotide monophosphate and the free base as
described for Homo sapiens HGPRT (D3 ¼ 1900 Mꢂ1 cmꢂ1 at 245
nM for IMP conversion from Hx; and D3 ¼ 5900 Mꢂ1 cmꢂ1 at
257.5 nm for GMP conversion from Gua).15 Initial steady-state
rates were calculated from the linear portion of the reaction
curve under experimental conditions in which less than 5% of
the substrate was consumed. True steady-state kinetics
parameters were determined from initial velocity measure-
ments for HPRT reaction varying concentrations of Hx (10–150
mM) at varied-xed PRPP concentrations (200–4000 mM). For
GPRT reaction, initial velocity measurements were determined
varying concentrations of Gua (10–120 mM) at varied-xed PRPP
concentrations (200–3600 mM). All reactions started with addi-
tion of recombinant MtHGPRT, and all measurements were
performed at least in duplicates. Hyperbolic saturation curves
of initial rate data at single concentration of the xed substrate
and varying concentrations of the other were tted to the
Michaelis–Menten equation21 (eqn (2)), in which v is the initial
velocity, V is the apparent maximum initial velocity, A is the
varying substrate concentration and Km represents the apparent
Michaelis–Menten constant.
ꢁ
dissolve all ligands. The stirring speed was 500 RPM at 25 C
with constant pressure for all ITC experiments. The binding
reaction started with one injection of 0.5 mL of ligand to
prevent artifacts, followed by 19 injections of 2.0 mL each at
300 s intervals. Control titrations (ligand into buffer) were
performed in order to subtract the heats of dilution and
mixing for each experiment prior to data analysis. The heat
variation was monitored inside the cell allowing determina-
tion of binding enthalpy of the process (DH) and the equi-
librium association constant (Ka). The Gibbs free energy (DG)
and the entropy (DS) of binding were calculated using the
relationship described in eqn (5), in which R is the gas
constant (8.324 J Kꢂ1 molꢂ1 or 1.987 cal Kꢂ1 molꢂ1), and T is
the temperature in Kelvin (T ¼ ꢁC + 273.15).
DG ¼ ꢂRT ln Ka ¼ DH ꢂ TDS
(5)
Pre-steady state kinetics
Pre-steady-state kinetic measurements of the reaction catalyzed
by MtHGPRT were performed to assess whether or not product
release is part of the rate-limiting step. The measurements were
carried out using an Applied Photophysics SX.18MV-R stopped-
ow spectrouorimeter on absorbance mode. The increase in
absorbance was monitored at 245 nm for HPRT reaction and
257.5 nm for the GPRT reaction for a period of 2 seconds with
400 points (1 mm slit width ¼ 4.65 nm spectral band) and an
optical path of 2 mm. The experimental conditions were 10 mM
MtHGPRT, 12 mM MgCl2, 125 mM Hx and 7 mM PRPP or 50 mM
Gua and 3.25 mM PRPP (mixing chamber concentrations). The
control experiments were performed as the experimental
conditions above in the absence of the enzyme. The dead time
of the stopped-ow equipment is 1.37 ms. The pre-steady-state
time course of the reaction was tted to eqn (6) that accounts for
a burst in product formation, in which Aobs is the (increasing)
value of observed absorbance due to product formation at any
time t, v0 is the steady-state rate, p is the burst (rapid increase)
in product formation, and k is the rst-order rate constant for
the rapid phase.22
VA
v ¼
(2)
Km þ A
The kcat values were calculated from eqn (3), in which [E]t
corresponds to the total concentration of enzyme subunits.
V
kcat
¼
(3)
½Eꢃ
t
Data from initial velocity measurements showing a pattern
of lines intersecting to the le of the y-axis in the double-
reciprocal plots (or Lineweaver–Burk plot) were tted to eqn
(4), which describes a sequential substrate binding and ternary
complex formation.
VAB
v ¼
(4)
KiaKb þ KaB þ KbA þ AB
For eqn (4), v is the initial velocity, V is the true maximum
initial velocity, A and B are the concentrations of the substrates,
Ka and Kb are their respective Michaelis constants, and Kia is the
Aobs ¼ v0t + p(1 ꢂ eꢂkt
)
(6)
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 74671–74683 | 74673

View Article Online
RSC Advances
Paper
D2O (0–90%) in 50 mM Tris–HCl pH 7.4 and 50 mM Tris–HCl
pH 8.5. Data were tted to eqn (12), which assumes isotope
effects on both V/K and V. In this equation, EV/K and EV are the
isotope effects minus 1 on V/K and V, respectively, and Fi is the
fraction of isotopic label in substrate A.24
Energy of activation
In order to access the energy of activation (Ea) of MtHGPRT for
the HPRT reaction, initial velocities were measured in the
presence of saturating concentrations of Hx (150 mM) and PRPP
(4000 mM). For GPRT reaction the concentrations were 120 mM
for Gua and 3600 mM for PRPP. The measurements were carried
out at temperatures ranging from 15 ꢁC to 35 ꢁC (from 288.15 to
308.15 K). MtHGPRT was incubated for several minutes at all
temperatures tested and assayed under standard conditions
(buffer A) to ensure enzyme stability. The Ea was calculated from
the slope (Ea/R) of the Arrhenius plot tting the data to eqn (7),
in which R is the gas constant (8.314 J molꢂ1 Kꢂ1) and the
constant A represents the product of the collision frequency (Z)
and a steric factor (p) based on the collision theory of enzyme
kinetics.23 Here, it is assumed a simplistic view to explain a
VA
ꢂ
ꢃ
v ¼
(12)
K 1 þ FiEV=K þ Að1 þ FiEV Þ
Determination of equilibrium constant
To determine whether or not MtHGPRT-catalyzed chemical reac-
tions are favorable processes, the equilibrium constants (Keq) were
identied at the point of equilibrium between substrates (Hx or
Gua, and PRPP) and products (IMP or GMP, and PPi) for,
respectively, HPRT and GPRT reactions. The Keq was measured by
xing the ratio of [IMP/Hx] or [GMP/Gua] at 1 and varying the ratio
of [PPi/PRPP]. For HPRT reaction, the range of [PPi/PRPP] was
from 0.0033 to 0.02 (PRPP: 1000–6000 mM; PPi ¼ 20 mM). For
GPRT reaction, the range of [PPi/PRPP] was from 0.0066 to 0.025
(PRPP: 800–3000 mM; PPi ¼ 20 mM). Specic activities (y-axis) were
plotted against the ratios (x-axis) and tted to a linear equation.
The point at which the curve crosses the abscissa is equal to Keq
(no net enzyme-catalyzed chemical reaction). The values for the
standard Gibbs free energy (DGꢁ) for MtHGPRT-catalyzed chem-
ical reactions were calculated from eqn (13), using the experi-
mentally determined Keq values, the gas constant (R) and absolute
temperature in Kelvin (T).
complex phenomenon and that
temperature.
A
is independent of
ꢀ
ꢁ
Ea
1
ln kcat ¼ ln A ꢂ
(7)
R
T
The enthalpy (DH#), entropy (DS#), and Gibbs free energy
(DG#) of activation were estimated using eqn (8)–(10) derived
from the transition state theory of enzymatic reactions:23
DH# ¼ Ea ꢂ RT
kB
(8)
(9)
ꢀ
ꢁ
DG# ¼ RT ln
þ ln T ꢂ ln kcat
h
DGꢁ ¼ ꢂRT ln Keq
(13)
and
DH# ꢂ DG#
DS#
¼
(10)
T
pH-rate proles
Prior to performing the pH-rate proles, the recombinant
enzyme stability was assessed over a wide pH range (4.5–10.5) by
incubation for 2 min at 25 ꢁC in 100 mM 2-(N-morpholino)
ethanesulphonic acid (MES)/N-2-hydroxyethylpiperazine-N-2-
ethanesulphonic acid (HEPES)/N-2-(N-cyclohexylamino)ethane-
sulfonic acid (CHES) buffer mixture, and then monitoring its
activity in buffer A.24 The dependence of steady-state kinetic
parameters on pH was determined by measuring initial velocities
in the presence of varying concentrations of one substrate and
saturating level of the other, in 100 mM MES/HEPES/CHES
buffer, over the following pH range: for HPRT reaction, 5.5–10
Energy values are in kJ molꢂ1, with kcat in sꢂ1, to conform to
the units of the Boltzmann (1.3805 ꢄ 10ꢂ23 J Kꢂ1) and Planck
(6.6256 ꢄ 10ꢂ34 J sꢂ1) constants, and R is as for eqn (5). Errors
on DG# were calculated using eqn (11).23
RTðkcat
Þ
Err
ðDGÞ
¼
(11)
Err
kcat
Solvent kinetic isotope effects (SKIE) and proton inventory
Solvent kinetic isotope effects were determined by measuring (10–120 mM varying concentrations of Hx and xed concentration
initial velocities for HPRT reaction using a saturating level of of PRPP at 4 mM, and 0.4–7 mM varying PRPP and xed
one substrate (Hx ¼ 120 mM; PRPP ¼ 4 mM) and varying concentration of Hx at 120 mM); and for GPRT reaction, 5–5.5
concentrations of the other (Hx: 10–120 mM; PRPP: 0.32–4 mM) (5–70 mM varying concentrations of Gua and xed concentration
in either H2O or 90% D2O. For GPRT reaction initial velocities of PRPP at 4 mM, and 0.65–4 mM varying PRPP and xed
measurements were assayed at saturating concentration of one concentration of Gua at 70 mM) and 6–10 (5–35 mM varying
substrate (Gua ¼ 50 mM; PRPP ¼ 4 mM) and varying concen- concentrations of Gua and xed concentration of PRPP at 3 mM,
trations of the other (PRPP: 0.32–4 mM; Gua: 10–50 mM) in and 0.32–3 mM varying PRPP and xed concentration of Gua at
either H2O or 90% D2O. Furthermore, the reactions were per- 35 mM). All measurements were performed at least in duplicates.
formed in 50 mM Tris–HCl pH 7.4, and in 50 mM Tris–HCl pH
The pH-rate prole was generated by plotting logarithm
8.5. The proton inventory was determined using saturating value of kcat or kcat/Km of the substrates versus the pH values (5.5
concentrations of both substrates (4000 mM PRPP and 120 mM to 10) and data were tted to eqn (14), in which y is the apparent
Hx, 3000 mM PRPP and Gua 50 mM) at different mole fractions of kinetic parameter (log kcat or log kcat/Km),
C is the pH-
74674 | RSC Adv., 2015, 5, 74671–74683
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
independent plateau value of y, H is the hydrogen ion concen- and C-terminal amino acids of the mobile loop of an adjacent
tration, and Ka is the apparent acid dissociation constant for the subunit of MtHGPRT have to be broken for catalysis.19 Disrup-
ionizing group.
tion of these interactions would result in the mobile loop
closing over the MtHGPRT active site for catalysis as proposed
for other 6-oxopurine PRTases.19 Interestingly, this proposal
would imply that there would be two forms of MtHGPRT in
solution: an inactive tetrameric form and an active dimeric one.
Whether or not there is an equilibrium between two forms of
MtHGPRT in solution will have to await for further evidence.
Notwithstanding, the results here reported show that recombi-
nant MtHGPRT is an active dimer in solution.
0
1 þ
1
B
B
@
C
C
A
C
log y ¼ log
(14)
H
Ka
Result and discussion
Oligomeric state determination
In order to determine the oligomeric state of the enzyme,
MtHGPRT (100 mM) was loaded on a Superdex 200 HR 10/30 size
Steady-state kinetics
exclusion column. The protein elution prole and data tting to The initial velocity experiments allowed the calculation of the
eqn (1) yielded an apparent molecular mass value of 46 kDa. As true steady-state kinetics parameters and a proposal of the
the MtHGPRT subunit molecular mass value is 22.251 kDa, the enzyme mechanism. The double-reciprocal plots showed a
gel ltration result suggests that MtHGPRT is a homodimer in pattern of lines intersecting to the le of the y-axis (Fig. 1),
solution, which suggests that this enzyme belongs to type I which is consistent with a sequential mechanism for the
PRTases as enzymes of this family are homodimers.25 It has forward reaction. The data were thus tted to eqn (4), yielding
recently been reported that MtHGPRT enzyme in complex with the true macroscopic steady-state kinetic constants for the
GMP, PPi and Mg2+ crystallized as a tetramer.19 This result is in forward reaction of MtHGPRT (Table 1). Divalent metal ion
disagreement with both the dimeric state of MtHGPRT in activation of substrate PRPP has already been demonstrated as
solution described here and other type I PRTases, which are essential for PRTase catalyzed reactions26–28 in which the
homodimers in solution.25 Interestingly, the crystal structure of dimagnesium Mg2:PRPP complex (and not PRPP alone) was
MtHGPRT in complex with an aza-acyclic bisphosphonate proved to be the true substrate for HGPRT reaction.28 No activity
containing a guanine base crystallized as a dimer.19 A mecha- could be detected when Mg2+ was omitted from the reaction
nism has been put forward in which interactions between Tyr93 mixture for the MtHGPRT (data not shown). These results are in
in the mobile loop (residues 90–106) of one subunit and Arg141 agreement with MtHGPRT belonging to type I PRTases, for
Fig. 1 Intersecting initial velocity patterns for MtHGPRT using either Hx or PRPP (A and B), and Gua or PRPP (C and D) as the variable substrate.
Each curve represents varied-fixed levels of the co-substrate.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 74671–74683 | 74675

View Article Online
RSC Advances
Paper
Table 1 True steady-state kinetic parameters for MtHGPRT
Substrate
Km (mM)
Vmax (U mgꢂ1
)
kcat (sꢂ1
)
kcat/Km (Mꢂ1 sꢂ1
)
Hypoxanthine
PRPPa
26 ꢀ 2
2.40 ꢀ 0.09
0.89 ꢀ 0.04
3.4 (ꢀ0.3) ꢄ 104
6.3 (ꢀ0.5) ꢄ 102
1.9 (ꢀ0.2) ꢄ 104
2.9 (ꢀ0.3) ꢄ 102
14.1 (ꢀ0.4) ꢄ 102
10 ꢀ 1
—
—
Guanine
0.52 ꢀ 0.01
0.193 ꢀ 0.004
PRPPb
6.5 (ꢀ0.7) ꢄ 102
—
—
a
HPRT reaction. b GPRT reaction.
which there is an absolute requirement of divalent magnesium equilibrium (Kd) from reciprocal of Ka. The ITC data for binding
ion for catalysis.25
of ligands to MtHGPRT are summarized in Table 2. These
The steady-state kinetics constants for MtHGPRT (Table 1) binding assays showed that PRPP, IMP and GMP can bind to
showed an approximately 54-fold larger value for Km of PRPP as free MtHGPRT (Fig. 2A–C). ITC data for these compounds were
compared to Hx, and 65-fold as compared to Gua. Higher tted to one set of sites model. No signicant heat change upon
overall Km values for PRPP compared to Hx or Gua have also titration of Hx and PPi to MtHGPRT was detected under the
been reported for Homo sapiens (HsHGPRT) (68-fold for Hx; 18- experimental conditions for ITC measurements (data not
fold for Gua)15 and Plasmodium falciparum (PfHGPRT) (27-fold shown).
for Hx; 53-fold for Gua).29 MtHGPRT showed a Km 2.6-fold lower
Formation of MtHGPRT:IMP and MtHGPRT:GMP binary
for Gua as compared to Hx. However, the kcat/Km value was 1.8- complexes (Fig. 2B and C) were detected by ITC measurements
fold larger for the HPRT reaction as compared to the GPRT and generated exothermic proles (heat released to the system).
reaction. As kcat/Km determines the specicity for competing On the other hand, formation of MtHGPRT:PRPP binary
substrates, MtHGPRT appears to be more efficient at using Hx complex (Fig. 2A) was characterized by an endothermic prole
than Gua as co-substrate. Interestingly, HsHGPRT has a kcat/Km (heat taken up from the system). No binding of either Hx or PPi
value of 1.3 ꢄ 107 Mꢂ1 sꢂ1 for HPRT reaction and 3.7 ꢄ 106 Mꢂ1 to free MtHGPRT was detected (data not shown). Fitting the ITC
sꢂ1 for GPRT reaction,15 indicating a higher efficiency at using data for PRPP, IMP and GMP binding to free MtHGPRT to one
both substrates when compared to MtHGPRT. As for MtHGPRT, set of sites binding model yielded n values (number of active
the apparent second-order rate constant value for HsHGPRT- sites) of 1.81 sites per monomer for GMP, 12 for PRPP and xed
catalyzed phosphoribosyl transfer to Hx is larger than to Gua.
as 1 for IMP. This value indicates the number of molecules
The double reciprocal patterns of lines intersecting to the bound to each enzyme active site with equal affinity. The value
le of the y-axis (Fig. 1) suggest that the reaction catalyzed by of 12 to stoichiometry (n) of the PRPP binding to the free
MtHGPRT obeys a sequential (either random or ordered) kinetic MtHGPRT might be related to the instability and purity of the
mechanism, which leads to the formation of a ternary complex compound (75% according to the supplier). It should be
capable of undergoing catalysis.24 Sequential mechanisms have pointed out that binding of Gua could not be evaluated by ITC
been suggested as one of the features shared by type I PRTa- studies due to limited solubility in aqueous solutions. Inci-
ses.25,27,30–32 The pattern of lines intersecting to the le of the y- dentally, attempts were made to measure whether or not Gua
axis rules out ping-pong (parallel lines), steady-state random binds to free MtHGPRT enzyme using protein uorescence
(that gives non-linear reciprocal plots), and rapid-equilibrium spectroscopy; however, no reliable data could be obtained due
ordered (one of the family of lines should cross at a single to high inner lter effects (data not shown). Accordingly, here it
value on the y-axis) mechanisms. However, the double- is assumed that the lack of binding of Hx to free enzyme serves
reciprocal plots alone cannot distinguish between rapid- as a surrogate to Gua (no binding to free MtHGPRT), and thus
equilibrium random and steady-state compulsory ordered bi the same kinetic mechanism is likely followed by both HPRT
bi mechanisms. Accordingly, ITC studies were performed to and GPRT reactions. The ITC measurements provided dissoci-
distinguish between these enzyme mechanisms.
ation constants values (Kd) for IMP (130 mM) and GMP (2.1 mM)
Isothermal titration calorimetry
Table 2 Binding parameters for MtHGPRT from ITC titration assays
As double reciprocal plots suggested a sequential kinetic
mechanism, substrate(s) and product(s) binding processes were
assayed by ITC at 25 ꢁC to ascertain the order, if any, of chemical
compound addition. The measure of heat taken up or released
upon binding of a ligand provides the binding enthalpy (DH) of
the process, an estimate for the stoichiometry of the interaction
(n) and the equilibrium binding constant (Ka). These values
allow the Gibbs free energy (DG) and the entropy (DS) of the
process to be calculated, as well as the dissociation constant at
PRPP
GMP
IMP
DG (kcal molꢂ1
DS
)
ꢂ5.9 (ꢀ0.1)
20.5 (ꢀ0.4)
ꢂ7.7 (ꢀ0.1)
ꢂ5.3 (ꢀ0.2)
ꢂ38 (ꢀ2)
ꢂ3.94 (ꢀ0.03)
(cal molꢂ1 Kꢂ1
)
DH (kcal molꢂ1
)
0.205 (ꢀ0.004)
ꢂ8.9 (ꢀ0.1)
ꢂ16.7 (ꢀ0.9)
Ka (Mꢂ1
Kd (mM)
)
2.2 (ꢀ0.3) ꢄ 104 4.8 (ꢀ0.5) ꢄ 105 7 (ꢀ1) ꢄ 103
48 (ꢀ7)
2.1 (ꢀ0.2)
1.3 (ꢀ0.2) ꢄ 102
74676 | RSC Adv., 2015, 5, 74671–74683
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Fig. 2 Ligand binding assays for binary complex formation: PRPP (A), IMP (B) and GMP (C). ITC data were fitted to one set of sites binding model.
(Table 2). A larger value for IMP as compared to GMP has also adjacent to the bound ligand and loss of conformational
been observed for HsHGPRT.15
freedom of free ligand).34 The negative DH values for IMP and
These ITC results and double-reciprocal plots suggest a GMP binding (Table 2) indicate that these processes are
sequential compulsory ordered mechanism for MtHGPRT, in accompanied by favourable redistribution of H-bonds and/or
which PRPP binds to free enzyme followed by Hx or Gua van der Waals interactions. The negative DS values for these
binding; and PPi is the rst product to dissociate followed by binding processes are likely coupled to a decrease in confor-
the respective monophosphate nucleoside (IMP or GMP) mational states upon binary complex formation. Accordingly,
release, leading to the regeneration of free enzyme (Fig. 3). A IMP or GMP dissociation from the binary complex to yield free
sequential mechanism has been proposed for HsHGPRT,15 in MtHPRT enzyme is likely to be accompanied by an increase in
which PRPP binds rst and release of the nucleotide as the last conformational states including the exible loop involved in
step of the reaction. Sequential mechanisms have also been binding of nitrogenous base, which is a characteristic feature of
suggested for others type I PRTases.27,33
type I PRTases.25 The ITC data also show that GMP binds 68-fold
The ITC results showed signicant heat changes upon ligand more strongly than IMP, probably due to additional interactions
(PRPP, IMP, or GMP) binding to free MtHGPRT enzyme (Fig. 2), made by the exocyclic amino group of GMP and MtHGPRT.15
thereby providing thermodynamic signatures of non-covalent The thermodynamic analysis of binding of PRPP gives a Kd of 48
interactions to each binding process. Observed enthalpies mM, with favourable entropic contribution and unfavorable
arise largely as a result of changes in interatomic interactions binding enthalpy (Table 2), indicating that this binding event is
(e.g., hydrogen bonds and/or van der Waals interactions), in likely accompanied by release of water molecules to bulk
which the sign indicates whether there is a net favourable solvent and unfavorable redistribution of the hydrogen bond
(negative DH) or unfavourable (positive DH) redistribution of network and/or van der Waals interactions between the reacting
the network of interactions between the reacting species species.34
(including solvent).34 Hydrophobic interactions are related to
The Gibbs free energy values for PRPP, GMP and IMP
the relative degrees of disorder in the free and bound systems binding to MtHGPRT (Table 2) show that these processes are
and thus these interactions are reected in the entropy change. favourable (negative value for DG). As indicated in eqn (5), DG
The release of “bound” water molecules from a surface to the consists of enthalpic and entropic contributions, and the
bulk solvent is usually a source of favourable entropy (positive results given in Table 2 are yet additional examples of entropy–
DS). A reduction in conformational states in either ligand or enthalpy compensation oen observed in biomolecular inter-
protein upon binary complex formation is entropically unfav- actions.35 If the ligand displays a signicant enthalpic contri-
ourable (negative DS) because this molecular recognition bution, sometimes this contribution is offset by a large entropic
process limits the external rotational and translational freedom compensation, which happens when we compare the Gibbs free
of both partners (for instance, structuring regions of the protein energy of binding of substrate and products to MtHGPRT. For
GMP and IMP binding, favorable formation of hydrogen bond
and/or van der Waals interactions are accompanied by a likely
decrease in conformational states of enzyme and ligand species,
thus compensating all gain in the enthalpy. On the other hand,
for PRPP the inverse happens: the unfavorable redistribution of
the hydrogen bonds is compensated by favorable entropic
Fig. 3 Proposed kinetic enzyme mechanism for MtHGPRT. This order
contribution (e.g., release of “bound” water molecules to
of substrate binding and product release is suggested on basis of the
solvent), compensating the penalty in enthalpy.35
Lineweaver–Burk plots (kinetics) and thermodynamics.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 74671–74683 | 74677

View Article Online
RSC Advances
Paper
than predicted by the collision theory and that the entropy value
for the enzyme:substrate(s) activated complex is lower than the
one for enzyme:substrate(s) in the ground state (there is loss of
degrees of freedom on going from the ground state to activated
state).
Energy of activation
The energy of activation for the enzyme-catalyzed chemical
reaction was assessed by measuring the dependence of kcat on
temperature for Hx and Gua (Fig. 4). These data were tted to
eqn (7). The results presented in Table 3 are derived from data
tting to eqn (7)–(11). These results provide the energy of acti-
vation (Ea), as well as the transition state enthalpy (DH#),
entropy (DS#) and Gibbs free energy (DG#).
Pre-steady state kinetics
To determine whether product release contributes to the rate
limiting step, pre-steady-state analysis of the reaction catalyzed
by MtHGPRT was performed. Fitting the pre-steady-state data to
eqn (6), which describes an initial rapid phase followed by a
slower linear phase, yielded a value of 1249 (ꢀ40) sꢂ1 for the
rst-order rate constant and a value of 0.0027 (ꢀ1 ꢄ 10ꢂ5) sꢂ1
for the steady-state rate of HPRT reaction (Fig. 5A), and a value
of 1024 (ꢀ32) sꢂ1 for the rst-order rate constant and a value of
0.0030 (ꢀ7 ꢄ 10ꢂ5) sꢂ1 for the steady-state rate of GPRT reaction
(Fig. 5B). The rst-order constant values for the rapid phase for
both reactions are larger than the steady-state rate, suggesting
fast formation of products in the MtHGPRT active site followed
by a slow rate-limiting step. The rates for the initial rapid phase
are likely faster than measured by the stopped-ow experiment
as the absorbance of control experiments indicate that part of
the signal was not detected. Hence, the burst observed in the
initial phase represents a fraction of the stopped-ow signal as
the remaining was lost in the dead-time of the equipment. The
observable fraction of the curve depends on the relationship
between the reaction rate and the dead-time, in which the
observed change (xobs), total change (xtot), the dead-time (td) and
the rate constant (k) can be described by the following equation:
ktd ¼ ln(xtot/xobs).22 This analysis indicates that the rate for the
initial rapid phase is at least 1930 sꢂ1 for HPRT and 1890 sꢂ1 for
GPRT reactions. These results are in good agreement with
stopped-ow data tting to eqn (6) (1249 sꢂ1 and 1024 sꢂ1 for
HPRT and GPRT reactions respectively).
The linearity of the Arrhenius plot (Fig. 4) suggests that there
is no change in the rate-limiting step over the temperature
range utilized in the assay for HPRT and GPRT reactions.21 The
Ea values for Hx and Gua, representing the minimum amount of
energy necessary to initiate the MtHGPRT-catalyzed chemical
reaction, were similar (Table 3). The values of free activation
energy (DG#) represent the energy barrier required for reactions
to occur. The DG# values can also be regarded as the variation of
the Gibbs energy between the enzyme:substrate(s) activated
complex and enzyme:substrate(s) complex in the ground state.
No differences in DG# values were observed for the substrates
studied here, suggesting a similar overall free activation energy
for HPRT and GPRT reactions. The constant A of eqn (7)
corresponds to the product of collision frequency (Z) and the
probability or steric factor (p) from the collision theory of
reaction rates. From the absolute rate theory, A ¼ pZ ¼ (kBT/h)
#
e
DS /R. Accordingly, this equation enables interpretation of the
probability factor in terms of entropy of activation. The negative
values for the entropy of activation (DS#) for HPRT and GPRT
reactions (Table 3) suggests that these reactions proceed slower
The value of 0.0027 sꢂ1 for change in absorbance at 245 nm
for the linear phase in the pre-steady-state experiment for the
HGPRT reaction corresponds to approximately 0.71 sꢂ1 (using
D3 ¼ 1900 Mꢂ1 cmꢂ1, optical path of 2 mm, and MtHGPRT
concentration of 10 mM). The value of 0.0030 sꢂ1 for the GPRT
reaction corresponds to approximately 0.25 sꢂ1 (D3 ¼ 5900 Mꢂ1
cmꢂ1 at 257.5 nm, optical path of 2 mm, and MtHGPRT
concentration of 10 mM). These values are in reasonable good
agreement with the kcat values determined by initial velocity
measurements in steady-state kinetics (0.89 sꢂ1 and 0.19 sꢂ1 for,
Fig.
4 Arrhenius plot for Hx and Gua substrates (temperature respectively, HPRT and GPRT reactions).
dependence of log kcat).
The observation of burst in product formation in the time
course of the transient phase suggests rapid formation and
build-up of products in the MtHGPRT active sites.22 If a burst is
observed during the transient phase, and the concentration of
IMP or GMP produced is approximately equal to free initial
MtHGPRT subunit concentration, it would indicate that the
chemical step of the reaction is much faster than the release of
product(s) (PPi, IMP or GMP). The size of the burst phases for
HPRT and GPRT reactions are 40 mM and 10 mM of product
formed before being released from the active sites of MtHGPRT
to solution (Fig. 5). As the concentration of enzyme used in the
Table 3 Thermodynamic activation parameters for MtHGPRTa
Parameter
HPRT
GPRT
Ea (kcal molꢂ1
)
6.20 ꢀ 0.09
5.10 ꢀ 0.03
DH# (kcal molꢂ1
)
5.60 ꢀ 0.08
ꢂ39.6 ꢀ 0.5
17 ꢀ 0.2
4.50 ꢀ 0.02
ꢂ46.6 ꢀ 0.2
18 ꢀ 0.1
DS# (cal molꢂ1 Kꢂ1
)
DG# (kcal molꢂ1
)
a
All values were determined at 25 ꢁC (298.15 K).
74678 | RSC Adv., 2015, 5, 74671–74683
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Fig. 5 Stopped-flow trace for product formation by measuring the increase in absorbance at 245 nm for HPRT (A) and 257.5 nm for GPRT (B)
reactions. Data were fitted to eqn (6) for an initial rapid phase followed by a linear phase. The inset represents the biphasic profile of the
experiment in a smaller scale of time (500 ms).
experiment was 10 mM, it can be concluded that release of parameters are independent of pH (as will be shown below in
product(s) and/or any step linked to it (e.g., enzyme conforma- the pH-rate proles).
tional changes) contribute to the rate-limiting step in the mode
For the experiments performed at pH 7.4 (Table 4), the SKIE
D2O
of action of MtHGPRT. A burst in IMP and GMP product parameters for HPRT reaction (D OVPRPP
,
D OV/KPRPP
,
V ,
Hx
2
2
formation has also been observed for human HGPRT in rapid D OV/KHx) suggest modest, if any, participation of the proton
quench experiments,15 which prompted the authors to propose solvent in catalysis and binding. At pH 8.5 (Table 4), the SKIE
that phosphoribosyl transfer is fast and that release of the results show the similar effects for
nucleotide product (IMP or GMP) limits the overall rate for the
forward reaction.15
2
D2O
2
V
,
D OV/KPRPP, and
PRPP
D2O
V
, whereas D OV/KHx is inverse. The SKIE for GPRT reaction
2
Hx
D2O
2
2
(
D OV/KPRPP
,
V
,
D OV/KGua) also suggest modest, if any,
Gua
participation of proton transfer from the solvent in catalysis and
binding at pH 7.4 (Table 4). At pH 8.5, the SKIE for GPRT
reaction show similar effects (as compared to pH 7.4) for D OV/
2
Solvent kinetics isotope effects (SKIE) and proton inventory
D2O
, and D OV/KGua (Table 4). However, there is an
2
KPRPP
inverse effect for
,
V
Gua
To evaluate the contribution of proton transfer from the solvent
to the rate of phosphoribosyl transfer of MtHGPRT-catalyzed
reaction, SKIE were determined by data tting to eqn (12),
yielding the results given in Table 4. Isotope effects on V report
on events following the ternary complex formation capable of
undergoing catalysis (fully loaded enzyme), which include the
chemical steps, possible enzyme conformational changes, and
product release (leading to regeneration of free enzyme).
Solvent isotope effects on V/K report on the contribution of the
proton transfer in steps in the reaction mechanism from
binding of the isotopically labeled chemical compound
(solvent) to the rst irreversible step, usually considered to be
the release of the rst product (that is, all rate constants from
reactant binding until the rst irreversible step).36 Accordingly,
to evaluate the contribution, if any, of proton transfer from
solvent to a rate-limiting step, measurements of solvent isotope
effects on V and V/K were carried out. As rule of thumb, deute-
rium accumulates where binding is tighter (that is, fraction-
ation factor is larger than one). Transition state proton
contributes the reciprocal of its respective fractionation factor
to the solvent isotope effect, whereas the contribution of a
reactant state proton to the solvent isotope effect is equal to its
fractionation factor.37 Although isotope effects measurements
are to be carried out at pH values in which V and V/K values are
not dependent on pH, we deemed appropriate to determine
SKIE and proton inventory at pH values of 7.4 and 8.5. The
former would more closely mimic the pH value found in the
context of whole cell, and the latter to provide data for solvent
isotope effects at a pH value in which the steady-state kinetic
D2O
V
at pH 7.4 (0.71 ꢀ 0.04) that is negli-
PRPP
gible at pH 8.5 (1.04 ꢀ 0.04) for the GPRT reaction. Before trying
to interpret these results, it should be pointed out that there
have been reported inverse isotope effects or V or V/K of
unknown origin.38,39 The expression of deuterium kinetic
isotope effect on V includes the intrinsic isotope effect,
commitment factors (forward and reverse) and equilibrium
isotope effect.37 The inverse effect on D2OVPRPP at pH 7.4 suggests
that deuterium accumulates (tighter binding) in a transition
state from ternary complex to the rst irreversible step as
transition state protons contribute the reciprocal of its frac-
tionation factor to the kinetic solvent isotope effect (providing
equilibrium isotope effect is normal). Yet at pH 8.5, the inverse
D2O
SKIE
V
at pH 7.4 has become a modest, if any, normal
PRPP
contribution of the proton solvent in catalysis and binding. It
has been pointed out that the rate limitation of the chemical
Table 4 Solvent kinetic isotope effect for MtHGPRT
Parameter
SKIE pH 7.4
SKIE pH 8.5
Comments
D2O
V
1.2 ꢀ 0.1
1.3 ꢀ 0.2
1.24 ꢀ 0.07
0.95 ꢀ 0.13
1.3 ꢀ 0.1
HPRT
HPRT
HPRT
HPRT
GPRT
GPRT
GPRT
GPRT
PRPP
D OV/KPRPP
2
D2O
V
Hx
1.09 ꢀ 0.08
1.3 ꢀ 0.3
D OV/KHx
0.7 ꢀ 0.2
2
D2O
V
PRPP
0.71 ꢀ 0.04
0.9 ꢀ 0.2
1.04 ꢀ 0.04
0.74 ꢀ 0.09
0.98 ꢀ 0.04
0.78 ꢀ 0.01
D OV/KPRPP
2
D2O
V
Gua
0.92 ꢀ 0.07
0.8 ꢀ 0.3
D OV/KGua
2
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 74671–74683 | 74679

View Article Online
RSC Advances
Paper
step can vary as the pH changes, and the degree of the isotope intracellular concentrations of substrates and products need to
effects on V and V/K may change depending on the contribution be determined. There are other M. tuberculosis enzymes that
of the isotope-sensitive step to overall rate limitation at satu- catalyze reactions that could provide free bases and PRPP. For
rating and limiting reactant concentrations, respectively.24
instance, purine nucleoside phosphorylase (PNP), involved in
Proton inventory studies were performed at pH 7.4 and pH the metabolism of both purine and pyrimidine.40 PNP catalyzes
8.5 at saturating concentrations of substrates (Fig. 6). To ruled the reversible phosphorolysis of the N-glycosidic bond of a-
out any effect that elevate mole fraction of D2O might have on purine (deoxy)ribonucleosides to generate b-(deoxy)ribose 1-
the ionization constants of assay buffer, the pH of the reaction phosphate and the corresponding purine bases.41,42
mixtures containing the highest concentration of D2O (90%)
was measured. No pH change in the presence of D2O was
observed, indicating that the use of equivalent buffers
(concentrations of all solutes are the same) guaranteed that
rates were being measured at equivalent positions on the pH
and pD-rate proles.37 The proton inventory data show that the
modest SKIE on V (Table 4) arises from a single protonic site
(Fig. 6). In agreement with the SKIE on V (Table 4), a modest
normal effect with a single protonic site involvement was
observed for the HPRT reaction at pH 7.4 (Fig. 6A) and at pH 8.5
(Fig. 6A-inset). For the GPRT reaction, no protonic site
involvement was observed at pH 8.5 (Fig. 6B-inset), whereas at
pH 7.4 a modest normal effect suggests limited participation, if
any, of a single protonic site on V (Fig. 6B). Solvent isotope
effects lead to isotope exchanges at hundreds of protic positions
of the enzyme, which precludes any assignment to a particular
chemical group.
pH-rate prole
To probe for acid–base catalysis and likely residues involved in
catalysis and/or substrate binding, pH-rate proles were deter-
mined for the steady-state kinetic parameters of MtHGPRT. In
this experiment, initial velocities measurements were assayed
in a broad range of pH values. The pH-rate proles are shown in
Fig. 8. Data from the pH-rate prole of HPRT reaction for kcat
and kcat/KPRPP were tted to eqn (14), yielding values of,
respectively, 6.3 (ꢀ1.2) and 7.0 (ꢀ0.9) (Fig. 8A). For the GPRT
reaction, data tting to eqn (14) yielded a value of 6.1 (ꢀ2.8) for
kcat/KPRPP and 6.8 (ꢀ2.0) for kcat/KGua (Fig. 8B). There appears to
be no obvious ionizing group to be predicted from the kcat/KHx
data for HPRT reaction, and, interestingly, kcat data for GPRT
reaction (Fig. 8A and B). The enzyme was stable over the pH
range (HPRT reaction: 5.5–10; GPRT reaction: 5–10) used in pH-
rate proles (data not shown).
The pH dependence of kcat is concerned with the chemical
step and its value follows the pKa of groups that play critical
Determination of equilibrium constant
Plotting the MtHGPRT enzyme activity as a function of [PPi]/ roles in catalysis. The pH-rate data of kcat for HPRT reaction
[PRPP] ratio gives a straight line for HPRT and GPRT reac- showed a prole of a curve with slope of +1 that goes to zero as a
tions (Fig. 7). The analysis of the equilibrium experiments for function of increasing pH values (Fig. 8A). In HsHGPRT, Asp137
HPRT and GPRT reactions yielded values of 0.0271 for Keq of Hx (Asp126 in MtHGPRT) has been proposed to act as a general
(Fig. 7A) and 0.0357 for Keq of Gua (Fig. 5B). The standard free acid/base for catalysis.43 Although attempts of tting the pH-
energy (DGꢁ) can thus be calculated by eqn (13). This analysis rate data of kcat for GPRT reaction (Fig. 8B) to an equation
ꢂ1
gives values for DGꢁ at 25 C (298.15 K) of 8.95 kJ mol (2.13 that describes “hollows” were made,44 no convergence to any
kcal molꢂ1) for Hx and 8.27 kJ molꢂ1 (1.97 kcal molꢂ1) for Gua. parameters could be achieved. A similar prole was shown for
These results suggest that HPRT and GPRT reactions are not the HPRT reaction catalyzed by HsHGPRT enzyme.43 Xu and
favorable processes at equilibrium. The PRTase reactions Grubmeyer concluded that this “hollow” possibly arises from
display a wide range of Keq, from 0.1 for OPRTase to a value of slow protonic equilibria for this reaction, in which there are
300 for APRTase.15 However, to show whether or not phos- slow proton transfers between enzymatic residues and solvent.43
phoribosyl transfer is a favorable process (DG < 0) in vivo, the The possible role of Asp126 in the chemical mechanism of
ꢁ
Fig. 6 Proton inventory at different mole fractions of D2O (0–90%). Enzyme activity measurements were carried out at saturating concen-
trations of substrates. (A) HPRT reaction: 4000 mM PRPP and 120 mM Hx at pH 7.4. (B) GPRT reaction: 3000 mM PRPP and Gua 50 mM at pH 7.4.
The insets show the measurements of enzyme activity at pH 8.5.
74680 | RSC Adv., 2015, 5, 74671–74683
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Fig. 7 Plot of enzyme activity against different [PPi]/[PRPP] ratios to determine the equilibrium constant for HPRT (A) and GPRT (B) reactions. The
ratios for [IMP]/[Hx] and [GMP]/[Gua] were fixed at 1.
Fig. 8 Dependence of steady-state kinetic parameters on different pH values for HPRT (A) and GPRT (B) reactions.
MtHGPRT is presented in Fig. 9. The transition state shown in Asp137 (Asp126 in MtHGPRT) and Lys165 could play a role in
Fig. 9 is as proposed by Eads et al.12 based on the HsHGPRT transition-state stabilization via protonation or hydrogen
crystal structure. These authors propose a transition state with bonding at N7.12 The pH-rate data of kcat for HPRT reaction
oxocarbonium character at the ribose C10–O40, with a weak (Fig. 8A) appears to support the proposal that Asp126 plays a
bond to the pyrophosphate group and a weak glycosidic bond role in transition-state stabilization in MtHGPRT.
between the C10 of ribose and N9 of the purine ring. It is
The pH dependence of kcat/Km relates to the required (or
interestingly to point out that Eads et al. have proposed that the preferred) protonation states for binding and/or subsequent
N7 protonated tautomer of Gua (or Hx) is the species that catalysis of groups in either the substrate or the enzyme form
undergoes catalysis in the forward direction, and that the it combines with.44 The data for kcat/KPRPP suggest that an
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 74671–74683 | 74681

View Article Online
RSC Advances
Paper
leading to the proposed kinetic mechanism (Fig. 3). Determi-
nation of the thermodynamic parameters of MtHGPRT-
catalyzed chemical reactions suggests a similar overall free
activation energy for HPRT and GPRT reactions, and that these
reactions proceed at slower rate than predicted by the collision
theory. The stopped-ow results suggest that product release
participates in the rate-limiting step. SKIE data suggest proton
transfer from solvent is not likely involved in the rate-limiting
step. Determination of equilibrium constant values for HPRT
and GPRT reactions suggest that these processes are not
favorable at equilibrium. The analyses of pH-rate proles indi-
cated that Glu122 and Asp123 residues are likely to play roles in
catalysis and/or PRPP binding. These data appear to be borne
out by the crystallographic structure of MtHGPRT and sequence
alignment,19 showing a conserved PRPP binding motif.12 Site-
directed mutagenesis efforts should thus be pursued to
provide a solid basis for the role, if any, of Glu122 and Asp123 in
binding and/or catalysis in the mode of action of MtHGPRT. In
addition, M. tuberculosis hpt (Rv3624c) gene replacement efforts
should be carried out to evaluate whether or not MtHGPRT is
essential for growth and/or plays any role in latency. The
elucidation of the crystal structure of MtHGPRT in complex with
products GMP and PPi and Mg2+ by Eng et al.19 should aid in the
design of structure-based enzyme inhibitors. The rational
design of enzyme inhibitors should, preferentially, be based on
structure and functional data. Accordingly, the results pre-
sented here can be useful to these efforts as it provides data on
the mode of action of MtHGPRT. MtHGPRT inhibitors may be
both tested as anti-TB agents and used as tools for chemical
biologists to carry out loss-of-function experiments to reveal the
biological role of MtHGPRT in the context of whole M. tuber-
culosis cells.45
Fig. 9 Proposed chemical mechanism for MtHGPRT.
ionizable group with acid dissociation constant of 7.0 for HPRT
reaction and 6.1 for GPRT reaction needs to be deprotonated for
PRPP binding and/or catalysis (Fig. 8A and B). A highly
conserved PRPP binding loop is found in MtHGPRT, starting at
Val118 and extends to Leu131,19 and the presence of two acidic
amino acid is believed to be critical for PRPP binding to type I
PRTases.25 Interestingly, Xu and Grubmeyer found a bell-
shaped prole for kcat/KPRPP data of HsHGPRT, suggesting
Asp137 as the residue with pKa of 7.1 that needs to be depro-
tonated and Lys165 as the residue with pKa of 8.8 that needs to
be protonated for productive PRPP binding.43 The authors
concluded that Lys165 in human HGPRT (Lys154 in MtHGPRT)
is involved in ground-state interactions with substrates for
nucleotide formation reaction.43 The crystal structure of
MtHGPRT in complex with GMP and PPi and Mg2+ shows that
the side chain of Lys154 forms a hydrogen bond with the 6-oxo
group of the purine ring, and the authors proposed that this
residue plays a role in substrate specicity.19 However, contrary
to the results for human HGPRT,43 no ionizable group that
needs to be protonated was observed in the MtHPRT pH-rate
prole of kcat/KPRPP (Fig. 8). As the slopes of kcat/KPRPP for
HPRT and GPRT reactions were +1, it thus tempting to suggest
that the carboxyl side chain of Glu122 or Asp123, which are
located in the PRPP binding loop,19 is the likely residue that
needs to be deprotonated for productive PRPP binding for
MtHGPRT catalysis to occur.
Conflict of interest
The data for kcat/KGua of GPRT reaction were tted to eqn
(14), yielding a pKa value of 6.8 (ꢀ2.0). These results suggest that
there is a residue that needs to be deprotonated for productive
binding of Gua (Fig. 8B). The side chain of conserved Asp126
has been shown to be rotated away from the N7 of the purine
ring of GMP and to form a hydrogen bond with the main chain
of nitrogen of Gly128 in the 50-phosphate binding site.19 For the
HPRT reaction, no equation could be tted to the kcat/KHx data
(Fig. 8A) and it appears that there is no ionizable group involved
in Hx binding in the 5.5 to 10 pH range.
The authors declare no competing nancial interest.
Acknowledgements
This work was supported by funds awarded by Decit/SCTIE/
MSMCT-CNPq-FNDCT-CAPES to National Institute of Science
and Technology on Tuberculosis (INCT-TB) to D. S. S. and L. A.
B. L. A. B. and D. S. S. also acknowledge nancial support
awarded by FAPERGS-CNPq-PRONEX-2009. L. A. B. (CNPq,
520182/99-5) and D. S. S. (CNPq, 304051/1975-06) are Research
Career Awardee of the National Research Council of Brazil
(CNPq). P. C. P., B. L. A., and M. R. acknowledge scholarships
awarded by CAPES and CNPq. L. K. B. M. is a Post-Doctoral
Fellow of FAPERGS/CAPES.
Conclusions
In this work we present experimental efforts to determine the
mode of action of MtHGPRT, showing that this enzyme is a type
I PRTase. The sequential ordered bi–bi mechanism and the
homodimeric quaternary conformation are characteristics of
this family of enzymes.25 MtHGPRT is less efficient using either
Hx or Gua, when compared to HsHGPRT. ITC experiments
demonstrated that PRPP, GMP and IMP binds to the free
MtHGPRT and no binding of Hx or PPi could be observed,
References
1 World Health Organization, Global Tuberculosis Report 2013,
WHO Press, Geneva, 2014.
2 R. G. Ducati, A. Ruffino-Netto, L. A. Basso and D. S. Santos,
Mem. Inst. Oswaldo Cruz, 2006, 101(7), 697–714.
74682 | RSC Adv., 2015, 5, 74671–74683
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
3 A. Jain and R. Mondal, FEMS Immunol. Med. Microbiol., 2008, 24 P. F. Cook and W. W. Cleland, Enzyme Kinetics and
53, 145–150.
Mechanism, Garland Science Publishing, New York, 2007,
ch. 9, pp. 253–323.
25 S. C. Sinha and J. L. Smith, Curr. Opin. Struct. Biol., 2001, 11,
733–739.
26 J. Victor, A. Leo-Mensah and D. L. Sloan, Biochemistry, 1979,
18, 3597–3604.
27 A. Breda, L. A. Rosado, D. M. Lorenzini, L. A. Basso and
D. S. Santos, Mol. BioSyst., 2012, 8, 572–586.
4 A. A. Velayati, P. Farnia, M. R. Masjedi, T. A. Ibrahim and
P. Tabarsi, et al., Eur. Respir. J., 2009, 34, 1202–1203.
5 T. R. Rustad, A. M. Sherrid, K. J. Minch and D. R. Sherman,
Cell. Microbiol., 2009, 11, 1151–1159.
6 W. Eisenreich, T. Dandekar, J. Heesemann and W. Goebel,
Nat. Rev. Microbiol., 2010, 8, 401–412.
7 M. Niederweis, Microbiology, 2008, 154, 679–692.
8 K. A. Kantardjieff, C. Vasquez, P. Castro, N. M. Warfel, 28 C. Salerno and A. Giacomello, J. Biol. Chem., 1981, 256, 3671–
B. S. Rho, T. Lekin, C. Y. Kim, B. W. Segelke, 3673.
T. C. Terwilliger and B. Rupp, Acta Crystallogr., Sect. D: 29 I. N. S. Subbayya and H. Balaran, FEBS Lett., 2002, 521, 72–
Biol. Crystallogr., 2004, 61, 355–364. 76.
9 W. B. Parker and M. C. Long, Curr. Pharm. Des., 2007, 13, 30 S. R. Krungkrai, S. Aoki, N. M. Q. Palacpac, D. Sato,
599–608.
T. Mitamura, J. Krungkrai and T. Horii, Mol. Biochem.
Parasitol., 2004, 134, 245–255.
31 G. P. Wang, C. Lundegaard, K. F. Jensen and C. Grubmeyer,
Biochemistry, 1999, 38, 275–283.
10 N. Munagala, V. J. Basus and C. C. Wang, Biochemistry, 2001,
40, 4303–4311.
11 D. L. Musick, Crit. Rev. Biochem., 1981, 11, 1–34.
12 J. C. Eads, G. Scapin, Y. Xu, C. Grubmeyer and 32 A. D. Villela, R. G. Ducati, L. A. Rosado, C. J. Bloch,
J. C. Sacchettini, Cell, 1994, 78, 325–334.
13 N. R. Munagala, M. S. Chin and C. C. Wang, Biochemistry,
1998, 37, 4045–4051.
14 J. P. Page, N. R. Munagala and C. C. Wang, Eur. J. Biochem.,
2001, 259, 565–571.
15 Y. Xu, J. Eads, J. C. Sacchettini and C. Grubmeyer,
Biochemistry, 1997, 36, 3700–3712.
16 L. Yuan, S. P. Craig III, J. H. McKerrow and C. C. Wang,
Biochemistry, 1992, 31, 806–810.
M. V. Prates, D. C. Gonçalves, C. H. I. Ramos, L. A. Basso
and D. S. Santos, PLoS One, 2013, 8, e56445.
33 P. Kwong, M. L. Doyle, D. J. Casper, C. Cicala, S. A. Leavitt
and S. Majeed, Nature, 2002, 420, 678–682.
34 J. E. Ladbury and M. L. Doyle, Biocalorimetry II, Wiley,
London, 2004.
35 D. J. Chodera and L. D. Mobley, Annu. Rev. Biophys., 2013, 42,
121–142.
36 D. B. Northrop, Biochemistry, 1975, 14, 2644–2651.
17 M. A. Wenck, F. J. Medrano, A. E. Eakin and S. P. Craig III, 37 P. F. Cook, Enzyme Mechanism from Isotope Effects, CRC
Biochim. Biophys. Acta, 2004, 1700, 11–18. Press, Boca Raton, 1991, pp. 203–228.
18 G. Biazus, C. Z. Schneider, M. S. Palma, L. A. Basso and 38 M. P. Patel, W. S. Liu, J. West, D. Tew, T. D. Meek and
D. S. Santos, Protein Expression Purif., 2009, 66, 185–190. S. H. Thrall, Biochemistry, 2005, 44(50), 16753–16765.
19 W. S. Eng, D. Hockova, P. Spacek, Z. Janeba, N. P. West and 39 R. G. Silva, L. P. S. Carvalho, J. S. Blanchard, D. S. Santos and
K. Woods, et al., J. Med. Chem., 2015, 58, 4822–4838. L. A. Basso, Biochemistry, 2005, 45(43), 13064–13073.
20 M. B. Patel, S. P. Kumar, N. N. Valand, Y. T. Jasrai and 40 L. Mascia, M. Cappiello, S. Cherri and P. L. Ipata, Biochim.
S. K. Menon, J. Mol. Model., 2013, 19, 3201–3217. Biophys. Acta, 2000, 1474, 70–74.
´
21 I. H. Segel, in Enzyme kinetics, behavior and analysis of rapid 41 H. M. Kalckar, J. Biol. Chem., 1947, 167, 429–443.
equilibrium and steady-state enzyme systems, John Wiley and 42 D. J. Porter, J. Biol. Chem., 1992, 267, 7342–7351.
Sons, Inc., New York, 1975, ch. 2, pp. 18–89. 43 Y. Xu and C. Grubmeyer, Biochemistry, 1998, 37, 4114–4124.
22 K. Hiromi, Kinetics of Fast Enzyme Reactions: Theory and 44 P. F. Cook and W. W. Cleland, Enzyme Kinetics and
Practice, Kodansha Ltd., Tokyo, 1979, ch. 4, pp. 188–253.
23 T. Lonhienne, E. Baise, G. Feller, V. Bouriotis and C. Gerday,
Biochim. Biophys. Acta, 2001, 1545, 349–356.
Mechanism, Garland Science Publishing, New York, 2007,
ch. 10, pp. 325–366.
45 A. C. Bishop and V. L. Chen, J. Chem. Biol., 2009, 2, 1–9.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 74671–74683 | 74683