Development of metal-chelating inhibitors for the Class II fructose 1,6-bisphosphate (FBP) aldolase

Article abstract of DOI:10.1016/j.jinorgbio.2012.02.032

It has long been suggested that the essential and ubiquitous enzyme fructose 1,6-bisphosphate (FBP) aldolase could be a good drug target against bacteria and fungi, since lower organisms possess a metal-dependant (Class II) FBP aldolase, as opposed to higher organisms which possess a Schiff-base forming (Class I) FBP aldolase. We have tested the capacity of derivatives of the metal-chelating compound dipicolinic acid (DPA), as well a thiol-containing compound, to inhibit purified recombinant Class II FBP aldolases from Mycobacterium tuberculosis, Pseudomonas aeruginosa, Bacillus cereus, Bacillus anthracis, and from the Rice Blast causative agent Magnaporthe grisea. The aldolase from M. tuberculosis was the most sensitive to the metal-chelating inhibitors, with an IC₅₀ of 5.2 μM with 2,3- dimercaptopropanesulfonate (DMPS) and 28 μM with DPA. DMPS and the synthesized inhibitor 6-(phosphonomethyl)picolinic acid inhibited the enzyme in a time-dependent, competitive fashion, with second order rate constants of 273 and 270 M^{- 1} s^{- 1} respectively for the binding of these compounds to the M. tuberculosis aldolase's active site in the presence of the substrate FBP (K_M 27.9 μM). The most potent first generation inhibitors were modeled into the active site of the M. tuberculosis aldolase structure, with results indicating that the metal chelators tested cannot bind the catalytic zinc in a bidentate fashion while it remains in its catalytic location, and that most enzyme-ligand interactions involve the phosphate binding pocket residues.

Full text of DOI:10.1016/j.jinorgbio.2012.02.032

Journal of Inorganic Biochemistry 112 (2012) 49–58
Contents lists available at SciVerse ScienceDirect
Journal of Inorganic Biochemistry
journal homepage: www.elsevier.com/locate/jinorgbio
Development of metal-chelating inhibitors for the Class II fructose 1,6-bisphosphate
☆
(FBP) aldolase
Geneviève Labbé, Anthony P. Krismanich, Sarah de Groot, Timothy Rasmusson, Muhong Shang,
⁎
Matthew D.R. Brown, Gary I. Dmitrienko, J. Guy Guillemette
Department of Chemistry, University of Waterloo, 200 University Ave. W, Waterloo, ON, Canada N2L 3G1
a r t i c l e i n f o
a b s t r a c t
Article history:
It has long been suggested that the essential and ubiquitous enzyme fructose 1,6-bisphosphate (FBP) aldolase
could be a good drug target against bacteria and fungi, since lower organisms possess a metal-dependant
(Class II) FBP aldolase, as opposed to higher organisms which possess a Schiff-base forming (Class I) FBP al-
dolase. We have tested the capacity of derivatives of the metal-chelating compound dipicolinic acid (DPA), as
well a thiol-containing compound, to inhibit purified recombinant Class II FBP aldolases from Mycobacterium
tuberculosis, Pseudomonas aeruginosa, Bacillus cereus, Bacillus anthracis, and from the Rice Blast causative
agent Magnaporthe grisea. The aldolase from M. tuberculosis was the most sensitive to the metal-chelating
inhibitors, with an IC50 of 5.2 μM with 2,3-dimercaptopropanesulfonate (DMPS) and 28 μM with DPA.
DMPS and the synthesized inhibitor 6-(phosphonomethyl)picolinic acid inhibited the enzyme in a time-
dependent, competitive fashion, with second order rate constants of 273 and 270 M−1 s−1 respectively for
the binding of these compounds to the M. tuberculosis aldolase's active site in the presence of the substrate
FBP (KM 27.9 μM). The most potent first generation inhibitors were modeled into the active site of the
M. tuberculosis aldolase structure, with results indicating that the metal chelators tested cannot bind the cat-
alytic zinc in a bidentate fashion while it remains in its catalytic location, and that most enzyme-ligand inter-
actions involve the phosphate binding pocket residues.
Received 27 October 2011
Received in revised form 26 February 2012
Accepted 26 February 2012
Available online 10 March 2012
Keywords:
FBP aldolase
Competitive slow-binding inhibitors
© 2012 Elsevier Inc. All rights reserved.
1. Introduction
for the viability of these organisms [2–13]. The essential role of the
Class II FBP aldolase for the growth of M. tuberculosis on glucose or
The metal-dependant Class II fructose 1,6-bisphosphate (FBP) aldol-
ase has been proposed to be a good antimicrobial target because it is
present in many plant and human pathogens, but not in animals or
plants [1]. FBP aldolases (E.C. 4.1.2.13) catalyze the reversible aldol con-
densation of dihydroxyacetonephosphate (DHAP) and glyceraldehyde
3-phosphate (GAP) in glycolysis, gluconeogenesis, and the Calvin
cycle. The Class I aldolase forms a Schiff base using an active site lysine
residue with the keto group of the substrate; whereas the Class II en-
zyme uses a divalent metal ion as an electron sink to stabilize the carb-
anion formed on the 3rd carbon of the substrate. Attempts to disrupt
the Class II FBP aldolase genes from Mycobacterium tuberculosis,
Escherichia coli, Streptomyces galbus, Bacillus subtilis, Pseudomonas
aeruginosa, Streptococcus pneumoniae, Saccharomyces cerevisiae, and
Candida albicans by insertion or deletion mutagenesis have been unsuc-
cessful, thereby suggesting that the Class II FBP aldolases are essential
gluconeogenic substrates in vitro has been confirmed recently by
knock-out experiments, with a clear demonstration that the chromo-
somal copy of the Class II FBP aldolase can be deleted only in the pres-
ence of a rescue copy of the gene in that organism [14]. The Class II
FBP aldolase from M. tuberculosis was shown to be strongly expressed
in the granulomatous lung tissues of infected mice and guinea pigs [14].
The glycolysis/gluconeogenesis pathway, at the core of the central
metabolism of all cells, may therefore be a drug target. In this path-
way, we are targeting an essential metabolic enzyme, Class II FBP al-
dolase, which is not found in eukaryotic cells. We propose that the
blockage of this central pathway will inhibit the growth of the cell
by shutting down its core metabolism and all synthetic pathways
due to the accumulation of phosphorylated metabolites [15–17].
This will result in the inhibition of ribosomal RNA transcription and
effectively prevent ribosomes from being synthesized.
In this investigation, a series of commercially available and syn-
thetic compounds were screened to determine starting points for
the rational design of ligands for this class of enzyme. A few known
inhibitors of Class II aldolases, all derivatives of the reaction interme-
diate analogue phosphoglycolohydroxamic acid (PGH), have been
reported [1,18–21]. Unfortunately, it has been found that there are
significant side-effects associated with drugs bearing hydroxamic
☆
Inhibitor synthesis was carried out by A. K., T. R., M. R., and M. B. under the direc-
tion of G. I. D.; protein preparations and kinetic studies were carried out by G. L. and
S. G. under the direction of J. G. G.; molecular modeling was performed by G. L. and
G. I. D. The manuscript was written by G. L., A. K., G. I. D. and J. G. G.
⁎
Corresponding author at. Tel.: +1 519 888 4567x35954; fax: +1 519 746 0435.
E-mail address: jguillem@uwaterloo.ca (J.G. Guillemette).
0162-0134/$ – see front matter © 2012 Elsevier Inc. All rights reserved.
doi:10.1016/j.jinorgbio.2012.02.032

50
G. Labbé et al. / Journal of Inorganic Biochemistry 112 (2012) 49–58
acid functions [22]. As a result, we have chosen to approach the dis-
covery of novel drug candidates that specifically inhibit the Class II al-
dolase over the mammalian Class I aldolase and that incorporate
other metal-binding moieties such as thiols and carboxylates. Ulti-
mately the synthesized compounds should be specific inhibitors of
the Class II enzymes and form a stable ternary complex with the en-
zyme and the active site zinc, instead of promoting the release of
the catalytic zinc ion (non-complexing inhibition).
albumin (BSA), 100 mM potassium acetate, 5% v/v DMSO and
23.75 mM Hepes, pH 7.3. The molecules tested for inhibition (except
compound 1, shown in Scheme 1) were dissolved in 100% DMSO to
produce 100 mM stock solutions. The compound 1 was dissolved di-
rectly in 50 mM Hepes, pH 7.3 to produce a 50 mM stock solution.
EDTA was diluted in the Hepes buffer and tested in the same way
but without DMSO. Assays were performed at 28 °C in quadruplicate
in 96-well flat bottom polystyrene plates (Corning, NY). The reac-
tion was initiated by the addition of FBP and monitored at 340 nm
for 10 min on a 96-well plate reader (Spectramax 190, Molecular
Devices, Sunnyvale, CA). With each enzyme studied, the FBP con-
Several derivatives of the metal chelator dipicolinic acid (DPA)
were synthesized and evaluated as FBP aldolase inhibitors. This com-
pound was chosen as the starting point for the synthesis of Class II
FBP aldolase inhibitors in part because it was observed to have anti-
fungal capabilities in our initial greenhouse tests (unpublished re-
sults). The antifungal and antibacterial capabilities of DPA were
recently confirmed by an independent team [23]. DPA was also cho-
sen as a starting point because it was shown to inhibit enzymes via
a formation of a ternary complex (enzyme-zinc-DPA). For example,
the removal of the zinc cofactor from bovine carbonic anhydrase
(BCA), which is coordinated by three His residues and located in a
deep narrow groove on the enzyme surface, was shown to occur
through the formation of a ternary complex with DPA [24]; whereas
the rate of the zinc removal from the enzyme by EDTA was shown
to be governed by the spontaneous dissociation rate of the zinc-
enzyme, with no evidence of ternary complex formation [25]. The in-
hibition of BCA by DPA was also found to be competitive with respect
to the substrate bicarbonate [26]. Our goal is to produce DPA deriva-
tives displaying improved competitive binding to the active site of
FBP aldolase with respect to the substrate FBP.
In order to maximize our chances of finding an effective inhibitor
against the microbial, metal-containing FBP aldolase (also called Class
II FBP aldolase), we decided to study the enzymes from several mi-
crobes, including both human and plant pathogens. In this report,
the results of studies of several commercial and newly synthesized
metal-binding compounds will be described in terms of their capacity
to inhibit the Class II FBP aldolases from human bacterial pathogens
M. tuberculosis, P. aeruginosa, B. cereus, B. anthracis, as well as from
the ascomycete fungus M. grisea, causative agent of rice blast. The
Class I FBP aldolase from rabbit muscle was used as a negative control
to assess the specificity of the compounds for the Class II enzymes.
The inhibition kinetics of the most potent compounds with the aldol-
ase from M. tuberculosis and the stability of the enzyme-inhibitor
complexes will be discussed. The crystallographic structure of the
M. tuberculosis Class II FBP aldolase was recently determined
with its substrates DHAP and G3P bound to the active site [5].
The compounds tested here were modeled into the structure of the
M. tuberculosis FBP aldolase. Suggestions as to how such inhibitors
might be modified to lead to improved potency are provided.
centration employed was at least 10 times its KM
.
The
concentration of inhibitor was varied depending on the amount of
inhibition recorded, but inhibitor concentrations of 20, 50, 75, 100,
200, 500, and 1000 μM were typically used to determine the IC50s.
The enzymes were pre-incubated with the inhibitor for 15 min in
the coupled assay mixture prior to the addition of the substrate
FBP. The compound 1 was freshly dissolved prior to the kinetic as-
says as it became less potent during long-term storage in solution.
DPA and racemic 2,3-dimercaptopropane-1-sulfonate (DMPS)
were purchased from Sigma-Aldrich (Mississauga, ON) and were
freshly diluted in buffer prior to the assays.
2.2. Second order rate constants determination
The M. tuberculosis aldolase was used for these assays. Four repli-
cates were done for each combination of substrate and inhibitor con-
centrations, and the reaction was monitored by absorbance readings
every six seconds for 10 min using a coupled assay. The assays were
started by addition of the enzyme to the substrate and inhibitor mix-
ture, without pre-incubation of the enzyme with the inhibitor. The
FBP concentrations used were 25, 40, 60, 80, 120, 200, and 500 μM.
The inhibitor concentrations were 20, 50, 100 and 500 μM of DMPS
and DPA; 20 and 50 μM of compound 1; 50 and 100 μM of compound
7, and 100 and 500 μM of EDTA. Eqs. (1) and (2) [30,31] presented
below were fitted to each progress curve by non-linear regression
using automatic outlier elimination (Rout coefficient Q set to 1% to
exclude
outliers,
with
no
weighting)
to
obtain
the
apparent constants k or A using the GraphPad Prism software (Graph-
Pad Software Inc, La Jolla, CA). A secondary plot of the apparent con-
stant A multiplied by the inhibitor concentration versus the FBP
concentration was then performed to obtain the on-rate constant
kon for each inhibitor (Eq. (3)), again using non-linear regression
with GraphPad Prism as described above.
ꢀ
ꢁ
ðreversible inhibitionÞ ½Pꢀ ¼ vst þ ðv0–vsÞ 1−e−kt =k
ð1Þ
ð2Þ
t
2. Materials and methods
ꢀ
ꢁ
ðirreversible inhibitionÞ ½Pꢀ ¼ ðv0=½IꢀAÞ 1−e−½IꢀAt
t
2.1. Inhibition screens
All buffers and solutions were prepared with deionized Milli Q
water (Millipore, Bedford, MA). The rabbit muscle Class I aldolase
was purchased from Sigma-Aldrich (Mississauga, ON). The Class II
FBP aldolases were purified as described previously [27–29]. The syn-
thesis of the inhibitors is summarized in Scheme 1, and is described
in Appendix A (Supplementary material). The structure of each of
the synthetic inhibitors was established by 1H NMR (300 MHz), 13C
NMR (75 MHz), and mass spectrometry, and the purity of the com-
pounds was assessed to be greater than 95% by examination of the
1H NMR spectra (Appendix A).
When t approaches infinity; Equation 2 becomes : ½Pꢀ∞ ¼ ðv0=½IꢀAÞ
ðcompetitive irreversible inhibitionÞ A ¼ kon=ð1 þ ½Sꢀ=KM
Þ
ð3Þ
where [P]t is the concentration of product formed at time t, v0 is the
initial velocity, vs is the end velocity, [I] is inhibitor concentration,
[S] is the concentration of substrate, KM is the Michaelis constant,
and kon is the second order rate constant for the binding of the com-
petitive inhibitor (Fig. 1). The description of the apparent constants k
and A depends on the type of inhibition (Competitive, Noncompeti-
tive or Uncompetitive). Note that the constant k in Eq. (1) is equiva-
lent to [I]A in Eq. (2).
The standard assay mixture (final volume 100 μL) contained the
FBP aldolase (0.003 to 0.020 U/mL), 0.3 mM NADH, 0.2 U/mL of rab-
bit muscle α-glycerophosphate dehydrogenase, 2.25 U/mL of rabbit
muscle triose phosphate isomerase, 0.2 mg/mL bovine serum

G. Labbé et al. / Journal of Inorganic Biochemistry 112 (2012) 49–58
51
Scheme 1. The reaction sequences used in the synthesis of inhibitors 1–7.
2.3. Docking and molecular modeling
previously published (PDB ID: 3EKZ) [5] before the file was uploaded
into AutoDockTools 1.5.4 [33] to calculate Gasteiger charges, merge
the non-polar hydrogens and to assign AD4 atom types. The Zn param-
eters were modified to a radius of 0.87 Angstroms and a charge of
+0.95e [34]. The dockings were done setting the enzyme (receptor)
as being rigid, and the inhibitors flexible, as calculated in the AutoDock
Tools quick setup option for ligands. The search was centered on
the catalytic zinc coordinates with a box of 15×15×15 Å. The docking
was done using the default search parameters of AutoDock Vina.
Inhibitors containing phosphonate groups were also minimized in
the M. tuberculosis monomer using the MMFF94 force field in Sybyl
X1.3 (Tripos, St. Louis, MO). The compounds were positioned approx-
imately in the space occupied by the substrates in the M. tuberculosis
aldolase structure (PDB ID: 3EKZ), the files were merged, the
substrates were deleted, and the whole enzyme-inhibitor complex
was minimized employing the anneal function within Sybyl X1.3. Be-
fore minimization, it was necessary to modify the crystallographic
The inhibitors were docked into the active site of the M. tuberculosis
aldolase using AutoDock Vina [32]. The substrates, the Na+ ion, and the
water molecules were deleted from a M. tuberculosis aldolase structure
Fig. 1. Scheme of competitive reversible inhibition with associated rate constants.

52
G. Labbé et al. / Journal of Inorganic Biochemistry 112 (2012) 49–58
structure so that the MMFF94 force field could assign proper atom
types. The surface bound (second) zinc ion was deleted and all formal
bonds to the metal ions (that were created automatically by the soft-
ware when the pdb file was imported) were deleted. Atom types
were modified to ensure that the lone pair electrons of the sp2- hy-
bridized nitrogen atoms on the imidazole rings could make contact
with the active site zinc ion. All water molecules in the pdb file
were retained as was the sodium ion in the phosphate binding pock-
et. Since the crystal structure lacks structural information about the
coordinates of certain stretches of amino acid residues, it was neces-
sary to “cap” the C-termini of incompletely defined parts of the struc-
ture as primary amides to avoid computational problems caused by
atoms with unspecified bonding arrangements. All hydrogen atoms
were added and the ionization states of the inhibitors were modified
before geometry optimization. In the case of DPA, both carboxylate
groups were adjusted to be in an ionized state. For compound 1, the
carboxylate group was assumed to be ionized and the phosphonic
acid group was assumed to be singly ionized such the anionic oxygen
atom was oriented towards the sodium ion. In the case of DMPS,
both the sulfonate group and the C-3 thiol group were adjusted to
be singly ionized. Prior to initiating the energy minimization, the
zinc ion, the sodium ion, manually docked inhibitor molecule, and
all amino acid residues within a 10 Å radius were chosen as the
“hot” region that was subjected to full geometry optimization. By de-
fault, an “interesting” region consisting of all amino acid residues
within 12 Å of the “hot” region was defined and allowed to influence
the geometry optimization in the “hot” region. A gradient conver-
gence criterion=0.05 kcal/mol*Å was employed to define geometry
optimization. In each case, this criterion was achieved in approxi-
mately 20,000 iterations.
amino acid sequence is the same as that of the aldolase from
B. anthracis str. ‘Ames Ancestor’, whose structure was deposited in
the Protein Data Bank recently (PDB ID: 3Q94)). Several potential in-
hibitory compounds were tested, including DPA derivatives (com-
pounds 1–7) (Scheme 1), in order to facilitate rational ligand
design. The IC50 values of these inhibitors after 15 min of incubation
were determined with the recombinant Class II FBP aldolases from
the targeted microorganisms (Table 1). It is relevant to note that the
presence of metal in the assay was found to significantly decrease
the inhibitory capacity of the compounds, as less than 30% inhibition
was observed after 15 min of incubation with 1 mM DPA or com-
pound 1 for the P. aeruginosa aldolase in the presence of 0.7 mM
CoCl2, whereas the IC50 values in the absence of metal were 200 μM
or less. An experiment was also performed to determine if the inhibi-
tion was reversible upon simple dilution in the assay mixture. The
concentrated enzyme (100× compared to the assay concentration,
or ~0.8 μM of M. tuberculosis aldolase) was incubated with 1 mM of
the inhibitors EDTA and DPA, or 250 μM of compound 1 for 15 min,
in the presence of 5% v/v DMSO. The enzyme was only ~60% inhibited
by the compound 1 in these conditions. The aldolase was in contrast
100% inhibited by the pre-incubation with 1 mM EDTA and DPA
under these conditions. The enzyme and inhibitor mixture was then
diluted in the assay such that the inhibitor concentration was 10 μM
(or 2.5 μM for compound 1), and the reaction was immediately
started by addition of FBP at a final concentration of 200 μM. The M.
tuberculosis aldolase recovered its full activity in all cases but with
DPA, where only ~5% of the activity was recovered.
According to the data presented in Table 1, the compounds tested
appear to be generally no better than EDTA at inhibiting the Class II
aldolases. However, these results were obtained after 15 min of pre-
incubation in the absence of the substrate FBP. While performing
similar experiments without pre-incubation with the inhibitory com-
pounds, it was observed that the compounds were much less effective
inhibitors when the substrate was present, pointing to a competitive
mechanism. Some inhibitors were therefore assayed by varying
both inhibitor and substrate concentration with and without the
15 min of pre-incubation of the enzyme with the inhibitor. However,
the analysis showed that the competitive, uncompetitive, or mixed
inhibition models did not fit the resulting data (Figure B1). The
compounds in fact exhibited a time-dependent pattern of inhibition
of the FBP cleavage reaction (see progress curves in the presence of
various compounds in Fig. 2). In spite of being reversible upon simple
dilution as described above, the inhibition did not appear reversible
under the coupled assay conditions, as the reverse reaction was too
slow to be detected. However the enzyme could be reactivated
The images of the docked and minimized compounds were pro-
duced using PyMOL (DeLano Scientific LLC).
3. Results
3.1. Inhibition screens
Several Class II FBP aldolases previously characterized in our
laboratory (those from M. tuberculosis, M. grisea, P. aeruginosa and
B. cereus) were used in this study, in order to cover most of the evo-
lutionary branches of the Class II aldolase phylogenetic tree [29].
The amino acid sequences of the FBP aldolases from B. anthracis
and B. cereus are identical, therefore the results obtained for the
B. cereus aldolase are applicable to the B. anthracis enzyme as well
(the B. cereus strain ATCC 10987 was used in this study [29], and its
Table 1
Summary of IC50 values for the inhibition of FBP aldolases with DPA derivatives.
Source
organism
KM (μM)
IC50 (μM) obtained for each compoundb
(95% confidence interval)
for FBPa
DPA
1
2
3
4
5
6
EDTA
M. grisea
51
1
79
78
960
1370
540
430
740
53
(74–84)
28
(24–36)
150
(68–92)
57
(54–62)
78
(880–1,050)
650
(590–710)
1050
(1320–1420)
560
(520–600)
1100
(480–630)
190
(160–230)
N/I
(400–470)
(710–770)
81
(61–120)
740
(45–66)
48
(44–53)
8
M. tuberculosis
B. cereus
27.9 0.9
450 10
110c
(91–140)
170
(130–170)
95
(81–120)
N/I
(70–89)
130
(100–180)
N/I
(950–1170)
>1000
(990–1250)
>1500
(150–180)
270
(230–340)
680
(690–790)
240
(200–300)
N/I
(7–11)
41
(37–45)
N/I
P. aeruginosa
Rabbit muscle
35
2
>1000
N/I
5.1
N/I
–
(620–740)
N/I: no inhibition.
–: not determined.
a
Values taken from [29] and [42].
b
c
A one phase decay equation was fitted to the data by non-linear regression, and the 95% confidence interval for each IC50 value is indicated in brackets.
The inhibition is affected by pH. This compound was also tested at a higher pH with the M. tuberculosis aldolase (assay at pH 8.0 instead of pH 7.3), and the IC50 was found to be
~500 μM in these conditions. The other compounds were not tested at different pH.

G. Labbé et al. / Journal of Inorganic Biochemistry 112 (2012) 49–58
53
Fig. 2. Progress curves of FBP cleavage by M. tuberculosis Class II aldolase in the pres-
ence of different metal-chelating inhibitors. 3 nM of enzyme was used to start the re-
action in the presence of 500 μM FBP and the following compounds: A) negative
control (BSA, 3 μM); B) EDTA (5 mM); C) DMPS (500 μM); D) DPA (500 μM); and E)
compound 7 (500 μM). After subtracting the background NADH oxidation from the
data, the enzyme was calculated to be completely inactivated by the metal-chelating
compounds after 20 min.
upon the addition of zinc to the assay mixture (Figure B2). The inhi-
bition experiments were subsequently performed in the presence of
substrate (no pre-incubation) and analyzed using a time-dependent
inhibition model.
3.2. Inhibition mechanism determination
In order to determine the inhibition mechanism, we chose to use
the enzyme from M. tuberculosis, as it generally was the most sensi-
tive to the inhibitors during the initial screens (Table 1). Two addi-
tional compounds were tested: 7 and commercially available, thiol-
containing compound DMPS. DMPS was investigated because it is
already used for medical treatment in humans and animals, and it
was also found to be an inhibitor of another zinc metalloenzyme in
vitro [35]. The M. tuberculosis and M. grisea aldolases were inhibited
strongly by DMPS, with IC50 values of 5.2 0.4 μM and 31 3 μM, re-
spectively, while the rabbit muscle aldolase was not inhibited in the
presence of 2 mM of this compound. Compound 7 was also tested
with these enzymes, and inhibited the M. tuberculosis aldolase
with an IC50 of 21 3 μM and the M. grisea aldolase with an IC50 of
60 4 μM. However, compound 7 also inhibited the rabbit muscle al-
dolase by 37% at a concentration of 2 mM. The most potent inhibitors
from these screens (DPA, 1, 7, and DMPS) were chosen for the inhibi-
tion mechanism studies, and EDTA was used for comparison.
The slow-binding inhibition progress curves can be fitted to Eq.
(1) [30] (see Materials and methods section) in the case of reversible
inhibition. The inhibition of the Class II aldolases was shown to be re-
versible upon the addition of metal, but the end velocity (vs) obtained
in our assay conditions is too low to be detected. This could be due to
the background NADH oxidation which is relatively high in our
coupled assay, and this masks any potential small residual aldolase
activity. The observed enzyme inhibition is therefore not distinguish-
able from irreversible inactivation, so we will consider the end veloc-
ity (vs) in Eq. (1) to be negligible and use instead Eq. (2) [31] (see
Eq. (2) fitted to progress curves obtained in the presence of DMPS
in Figure B3). The type of irreversible inhibition (Eq. (2)) can be dis-
tinguished by suitable plots of A and [S] [31]. In the case of competi-
tive inhibition (Fig. 1), A is defined by Eq. (3), and thus a plot of 1/A
versus [S] should give a straight line. Some of these plots obtained
with the metal-chelating inhibitors are shown in Fig. 3. The results
Fig. 3. Variation of the apparent rate of inhibition in function of the substrate concen-
tration: Tsou's test for competitive irreversible inhibition. The apparent constant A was
obtained by fitting Eq. (2) using non-linear regression to the progress curves obtained
with the M. tuberculosis aldolase, in the presence of various concentrations of FBP and
different inhibitors: panel A) 500 μM compound 1 (◊); 100 μM DMPS (□); 100 μM
compound 7 (Δ); panel B) 500 μM EDTA (○); panel C) 500 μM DPA (○). Each point
represents the average value of 4 replicate assays. Straight lines indicate that the inhi-
bition is competitive with respect to the substrate FBP.

54
G. Labbé et al. / Journal of Inorganic Biochemistry 112 (2012) 49–58
indicate that the (apparent) irreversible inactivation is competitive
with the substrate for the inhibitors DMPS, EDTA, and compounds 1
and 7. However, the high dispersion of the calculated constant A for
these inhibitors at high substrate concentration means that non-
linearity cannot be totally excluded, and these compounds may
therefore not display purely competitive inhibition. In the case of
DPA, the resulting plot is non-linear, indicating that this compound
is a mixed inhibitor of the M. tuberculosis aldolase (Fig. 3C).
Since the inhibition appears irreversible under the assay condi-
tions used, the apparent KI cannot be determined for a one-step inhi-
bition mechanism (Fig. 1), as it would represent the ratio of the
forward and reverse inhibitor binding rate constants (kon /koff). The
binding capacity of the most potent inhibitors will instead be com-
pared using their second order binding rate constant kon (Eq. (3),
Fig. 1). The binding rate constant kon was determined using the equa-
tion associated with a competitive inhibition model for all com-
pounds for comparison purposes. The resulting plots are presented
in Fig. 4. The kon rate constants obtained with these inhibitors are
presented in Table 2. The apparent inhibition constants obtained
with DPA were also analyzed using a mixed inhibition equation (see
Figure B4), yielding a kon =73 4 M−1 s−1, which is ~13% lower
than the value presented in Table 2.
minimized complexes exhibited such interaction with the zinc ion
(Fig. 5, panel C, and Appendix C, Figure C2, panels A–C). The pyridine
ring is simply too large to permit such bidentate interaction in the
Class II FBP aldolase active site.
The minimized models do suggest that the internuclear distance
between the carboxylate groups of DPA is appropriate to allow signif-
icant monodentate binding between these two groups simultaneous-
ly with the zinc ion and the sodium ion in the phosphate binding site.
The geometry optimization process did lead to significant alteration
of the environment of the zinc ion compared to that observed in the
X-ray crystal structure (PDB ID: 3EKZ) [5]. For example, some distor-
tion of the orientation of the imidazole rings of the histidine residues
around the zinc ion, as well as a new interaction of the carboxylate
group of D95, was observed. Thus the zinc ion was predicted to be
in a distorted trigonal bipyramidal environment.
In the case of the DPA analogues, 4, 5, and 6, the additional steric
bulk of the C-4 substituents appears to diminish the affinity for the
FBP aldolase active site (Table 1, and Appendix C, Figure C2, panels A–C).
It proved possible to construct and geometry optimize a model of
the pyridine carboxylate–phosphonate with the active site. In this
model the phosphonate group is able to occupy the phosphate bind-
ing site with significant interaction with the sodium ion while the
pyridine carboxylate group binds with the zinc ion (Fig. 5, panel D).
The environment of the zinc ion was observed to be altered in a sim-
ilar fashion to that observed in the energy minimized complex of DPA
with the enzyme. In compounds 2 and 3 the pyridine substituents ap-
pear to lead to increased unfavorable interaction with the active site
(Table 1 and Appendix C, Figure C2, panels D and E).
In the case of the pyridine dicarboxylate derivative, 7, all attempts
to construct an energy minimized active site bound model using a
similar strategy were unsuccessful. The very large substituent at C-4
of the pyridine ring in this compound led to bad contacts with active
site residues that could not be relieved in the molecular mechanics
minimization without gross distortions of the inhibitor structure.
Thus the relatively fast on rate observed for this compound is not ex-
plicable based on easy access or strong affinity for the H.E.I. binding
site. One possibility is that this compound binds to the two metal
ions via the carboxylate groups of the pyridine ring but with the pyr-
idine ring and the large C-4 substituent external to the active site. In
such a complex, the very large C-4 substituent might be capable of
blocking access to the active site by the substrate leading to what
would be competitive inhibition kinetically. Efforts to explore this
possibility through X-ray crystallographic experiments are planned.
In the case of DMPS, which contains one chiral centre and which
was analyzed as the racemic mixture, both the R and S enantiomers
were docked manually in the active site as defined by the coordinates
in the X-ray crystal structure of the M. tuberculosis FBP aldolase. In
both cases a fully extended conformation was assumed as a starting
point for the geometry optimization since only such a conformation
allows interaction with both the zinc ion and the phosphate binding
site. After energy minimization with the MMFF94 force field, the R-
enantiomer was found to have a slightly closer contact with the zinc
ion (C-3 thiolate ion to Zn2+ distance=2.32 Å for the S-isomer versus
2.29 Å for the R-isomer) and with the sodium ion (sulfonate oxygen
to Na+ distance=2.48 Å for the S-isomer versus 2.29 Å for the R-iso-
mer) (Fig. 5, panels E and F; Table C1).
As can be seen from the results presented in Table 2, the inhibitors
DMPS and compound 1 can inactivate the Class II FBP aldolase from
M. tuberculosis ~20 times faster than EDTA, and ~3 times faster than
DPA in the presence of the substrate FBP, whereas the inhibitor 7
can inactivate the enzyme almost 2 times faster than DPA in the pres-
ence of FBP. The apparent second order rate constants obtained for
the irreversible reaction with these inhibitors are over 3 orders of
magnitude lower than the second order rate constant determined
for the FBP cleavage reaction for the M. tuberculosis aldolase (kcat
/
Km =1.08×106 M−1 s−1 [29]). It is therefore estimated that these in-
hibitors have an affinity for the active site ~3 orders of magnitude
lower than the substrate FBP.
3.3. Molecular modeling
Initial efforts at exploring possible ways in which the pyridine car-
boxylate inhibitors bind to the active site of Class II aldolases were
carried out by docking flexible inhibitor molecules to a rigid model
of the protein. In each case, the docking exercise produced a model
in which the inhibitor interacted with the active site region in a man-
ner that is very different from the interaction of PGH, DHAP or the
High Energy Intermediate (H.E.I.) with the enzyme (e.g. see a com-
parison of the H.E.I. binding mode and the docked structure of DPA
determined using the rigid docking strategy in Fig. 5, panels A and
B, and rigid docking results for other pyridine carboxylate inhibitors
in Appendix C, Figure C1). The distances between H donors and ac-
ceptors from the ligands and the enzyme active site elements in all
the modeled compounds are presented in Appendix C, Table C1.
It was concluded that the enzyme likely must undergo significant
conformational alteration to accommodate the binding of the pyri-
dine carboxylate inhibitors so that a rigid docking process might not
be appropriate.
Thus construction of models by superimposition of the pyridine
carboxylate inhibitors on the H.E.I. in the X-ray crystal structure of
the complex ([5]) was undertaken. It proved possible to align a
model of DPA such that one carboxylate group could interact with
the zinc ion whereas the other carboxylate was in the vicinity of the
phosphate binding site. Bad contacts between the inhibitor and the
protein were then removed by geometry optimization with the An-
neal functionality within Sybyl X1.3 using the MMFF94 molecular
mechanics force field (see Materials and methods for more detail).
Interestingly, even though such pyridine dicarboxylates are potential-
ly capable of bidentate interaction with metal ions involving both
the carboxylate anion and the pyridine nitrogen, none of the energy
In the fully extended conformation, the two thiol/thiolate groups
are in an anti-periplanar relationship (expected to be lower in ener-
gy) for the S-enantiomer but in a gauche relationship (expected to
be higher in energy) for the R-enantiomer. Thus it is plausible that
both enantiomers contribute significantly to the observed inhibition
by racemic DMPS.
Although a bis-mercaptan, such as DMPS, has the potential to act
as a bidentate metal ion ligand to the active site zinc ion, the short
distance between the sulfonate group, that likely binds to the phos-
phate binding site, and the C-3 thiolate group allows only that sulfur

G. Labbé et al. / Journal of Inorganic Biochemistry 112 (2012) 49–58
55
Fig. 4. Secondary plots of the apparent inhibition constants obtained with the M. tuberculosis aldolase in the presence of various inhibitors, as a function of the FBP concentration.
The apparent inhibition constants (A*[I], represented by k in Eq. (1)) were obtained from 47 to 100 progress curves for each inhibitor, and are shown with their error bars repre-
senting the standard deviation. Eq. (3) defining the apparent constant A for competitive inhibition was fitted to the data by non-linear regression using the GraphPad Prism soft-
ware. The equation was fitted independently to the data from each inhibitor concentration (shown as lines in each Panel), and the global parameters were also calculated. The
global parameters calculated for each inhibitory compound are presented in Table 2. Panel A) DMPS; B) DPA; C) compound 1; D) Compound 7; and E) EDTA. The concentrations
of inhibitors used are 500 μM (●), 100 μM (■), 50 μM (▲), and 20 μM (♦).

56
G. Labbé et al. / Journal of Inorganic Biochemistry 112 (2012) 49–58
Table 2
located in the adjacent dimer subunit [5]. The fact that compound 7
partially inhibits the Class I FBP aldolase reaction at high concentra-
tion (Section 3.2) also suggests that this compound may be closer to
an FBP analog than the other metal-chelating compounds.
Second order rate constants for the binding of chelating inhibitors to the recombinant
Class II aldolase from M. tuberculosis.
Each value is calculated from a secondary plot (Eq. (3)) using the apparent constant A
(Eq. (2)) obtained for multiple assays done simultaneously in a microtiter plate in the
presence of various concentrations of FBP and inhibitor.
Upon further kinetic analyses, it was noticed that the inhibition by
these compounds was weaker in the presence of the FBP substrate,
implying a competitive mechanism. A time-dependent analysis of re-
action progress curves of the cleavage of FBP by the M. tuberculosis al-
dolase in the presence of the most potent inhibitors showed that
compounds 1, 7, and DMPS were better competitive inhibitors than
DPA, according to their second order binding rate constants. EDTA
was comparatively a weak inhibitor of the enzyme in the presence
of substrate, being 6 times less potent than DPA. The first-order ap-
parent rate constant for inhibition of FBP cleavage (k, or A[I] in Eqs.
(1) and (2)) were dependent on the concentration of the inhibitors
EDTA, DPA, compound 1, Compound 7 and DMPS (over the concen-
tration range 25 μM–500 μM inhibitor), confirming that the inhibition
mechanism involves the formation of a ternary complex [36] (Fig. 4).
In our study, the enzymes were reactivated by the addition of
small amounts of zinc (i.e. ~10 times less zinc than the inhibitor con-
centration), which indicates that the inhibitors are forming an unsta-
ble ternary complex with the active site and the zinc ion. The DPA
derivatives tested therefore do not have a sufficient affinity for the ac-
tive site to form a stable complex in the presence of extraneous zinc.
In contrast, the inhibitor PGH was shown to form a stable complex
with the cobalt-dependant Bacillus stearothermophilus Class II FBP al-
dolase, as the inhibited enzyme was not reactivated in the presence of
excess cobalt [37]. Our data however shows that some chelating in-
hibitors, such as DMPS and compound 1, are better able to compete
with the substrate FBP for the access to the active site zinc in the
M. tuberculosis aldolase in comparison with DPA or EDTA, which
means that some of the studied compounds have a higher affinity
for the active site than others.
The in silico docking studies indicate that the DPA derivatives can-
not form bidentate bonds with the zinc ion while it remains in its cat-
alytic location, due to both the restricted size and geometry of the
DHAP binding pocket, where the zinc ion is located; and the limited
solvent-exposed surface of the catalytic metal. This is in agreement
with the conclusions of a recent report on the inhibition of the
G. lamblia Class II FBP aldolase by metal-chelating aromatic com-
pounds [38]. The authors found that the orientation of a 3-hydroxy-
2-pyridone ligand in the binding pocket was not consistent with
strong bonds with the active site zinc in the crystal structure. They
concluded the zinc stabilizes the reaction intermediate (enediolate
form of DHAP), but does not provide significant binding energy in
the enzyme-substrate complex [38].
It appears that the presence of a substrate analog in the active site
is necessary to induce conformational changes in Class II FBP aldol-
ases that cause the zinc ion to move from a buried location to a
solvent-exposed position, which then allows strong interactions be-
tween the zinc ion and the enzyme substrates/inhibitors [38–41].
The slow inhibition observed with the metal-chelating compounds
may be related to the buried location of the catalytic zinc in the
absence of substrate. In our study, molecular modeling results con-
firmed that most interactions between the metal-chelating inhibitors
and the enzyme active site involve the phosphate binding pocket res-
idues. However, these results also show that some inhibitors can
alter the catalytic zinc environment. It is relevant to point out that
in some minimized models of M. tuberculosis aldolase with the che-
lating inhibitors, the interaction of the catalytic zinc with its ligand
His212 is considerably weakened compared to the interaction seen
in the original crystal structure (Fig. 5A). This weakened interac-
tion with His212 is concomitant with the new interaction of the
carboxylate group of Asp95 with the catalytic zinc mentioned in
Section 3.3. This is the case for the minimized models with DPA,
Compounds 4, 5 and 6 (Figs. 5C and C2A-C2C), as well as R-DMPS,
Inhibitor
kon M−1 s−1
R2 value of secondary plot
DPA
1
7
DMPS
EDTA
84
270
153
273
14
1
6
3
2
1
0.9879
0.9476
0.9575
0.9941
0.8683
centre and not the C-2 sulfur centre to interact with the zinc ion. Cur-
rent efforts in this group are aimed at creating synthetic analogues
DMPS in which the spatial relationships among the two thiol groups
and the sulfonate ion are such as to allow favorable binding to the
phosphate binding site as well as more potent binding to the zinc
ion in a bidentate fashion. Preliminary molecular modeling studies
suggest that this may be feasible.
4. Discussion
Although the Class II FBP aldolases are potential targets for antimi-
crobial therapy and for the development of new fungicides, there
have been few potent Class II FBP aldolase inhibitors reported in
the literature. In order to create specific inhibitors for the metal-
dependent Class II aldolases, compounds that have metal-chelating
groups were tested. The results were compared with the general
metal ion chelators, EDTA and DMPS. EDTA has an affinity for zinc
which is >104 higher than that of DPA, based on the equilibrium con-
stants for Zn2+ binding to these chelators [36]. The rate of enzyme in-
hibition by DPA and its derivatives is therefore expected to be faster
than the rate of inhibition by EDTA only when the active site zinc is
much more accessible to DPA compared to EDTA. In this study, inhibi-
tion by EDTA was used as a reference to evaluate the capability of DPA
derivatives to compete with the aldolase substrate for access to active
site zinc. This is considered to be in direct correlation with the DPA
derivatives' affinity for the enzyme's active site.
The IC50 results show that in the absence of substrate, the active
site zinc is accessible to both DPA and EDTA, based on the similar
IC50 values obtained for these compounds for most Class II aldolases
tested. The molecule 2,6-pyridinedicarboxylic acid (or dipicolinate,
DPA) was the most potent inhibitor according to our initial com-
pound screen. Derivatives of DPA were then synthesized and IC50
values were determined with a wider range of FBP aldolases. IC50
values are generally similar with DPA and compound 1, which differs
from DPA by the substitution of the carboxylate with a phospho-
methyl group at the 6-position of the pyridine ring. Addition of
other substituents (in compounds 2 to 6) resulted in higher IC50
values for most enzymes. Two exceptions are compounds 5 and 6, re-
spectively with hydroxymethyl and formyl substituents in position 4
of the pyridine ring, which had IC50s comparable to DPA and com-
pound 1 with the M. tuberculosis aldolase. The generally poorer inhi-
bition observed with compounds 2 to 4 in M. tuberculosis aldolase
initially did not appear to be due to steric hindrance in the active
site of the enzyme since compound 7, with a 3,5-dicarboxyphenyl
substituent in position 4 of the pyridine ring was found to be better
than DPA in inhibiting this enzyme (Table 2). However, molecular
modeling revealed that the substituents in position 3 and 4 of the pyr-
idine ring of compounds 2 and 3 make unfavorable contacts with key
active site residues, unlike compound 1. The improved inhibition
obtained with compound 7 could be the result of favorable interac-
tions of its dicarboxyphenyl substituent with GAP binding residues

G. Labbé et al. / Journal of Inorganic Biochemistry 112 (2012) 49–58
57
Fig. 5. M. tuberculosis FBP aldolase (PDB ID 3EKZ) with substrate DHAP (High Energy Intermediate) and modeled inhibitors. The DHAP enediolate form (High Energy Intermediare)
and the docked compounds are represented as sticks with carbons in white, oxygens in red, nitrogen in blue, sulfur in yellow, and phosphorus in orange. The active site residues are
shown as sticks with colored carbon aroms: D95 in yellow, N27 in cyan, S255 in blue, D276 in black, T277 in red, G253 in pink. The zinc ligands H96, H212 and H252 are shown as
gray sticks. The catalytic zinc (Zn1) is represented as a gray sphere (bottom left) and the Na+ is represented as a purple sphere (right). The electrostatic interactions between 2 and
3.6 Å are shown as black dashes. A) DHAP High Energy Intermediate (H.E.I.); B) docked DPA; C) to F) DPA, compound 1 R-DMPS, and S-DMPS minimized using the anneal function
of Sybyl (Tripos, St. Louis, MO). The figures were produced using PyMOL (DeLano Scientific LLC).
and S-DMPS (Fig. 5E and F). It is possible that this zinc ligand ex-
change in the presence of the inhibitors constitutes the first step of
a slow ligand exchange process, which could lead to a bidentate in-
teraction of the metal ion with the chelating inhibitors as the zinc
leaves its catalytic location. This could possibly lead to the removal
of the catalytic zinc from the enzyme when the metal-inhibitor
complex leaves the active site. These modeling results are also con-
sistent with previous reports of the considerable mobility of the
catalytic zinc in Class II FBP aldolases [35–38].
5. Conclusions
The strongest inhibitors in our study according to the second order
binding rate constants (kon) were DMPS and compound 1, which re-
spectively possess a sulfonate and a phosphonate group in addition
to zinc-chelating functions. Molecular modeling results suggest that
the strongest interaction between these compounds and the enzyme
occurs via the phosphate binding pocket ligands. Molecular modeling
results also suggest that the compounds cannot bind the catalytic zinc

58
G. Labbé et al. / Journal of Inorganic Biochemistry 112 (2012) 49–58
LaBonte, D.C. Lamb, N. Lan, H. Liang, H. Liao, L. Liu, C. Luo, M. Lussier, R. Mao, P.
Menard, S.L. Ooi, J.L. Revuelta, C.J. Roberts, M. Rose, P. Ross-Macdonald, B.
Scherens, G. Schimmack, B. Shafer, D.D. Shoemaker, S. Sookhai-Mahadeo, R.K.
Storms, J.N. Strathern, G. Valle, M. Voet, G. Volckaert, C.Y. Wang, T.R. Ward, J.
Wilhelmy, E.A. Winzeler, Y. Yang, G. Yen, E. Youngman, K. Yu, H. Bussey, J.D.
Boeke, M. Snyder, P. Philippsen, R.W. Davis, M. Johnston, Nature 418 (2002)
387–391.
ion in a bidentate fashion while the metal is chelated by the active site
zinc ligands, but indicate that several inhibitors can alter the zinc en-
vironment. In order to eliminate the possibility of zinc removal from
the enzyme, the first generation inhibitory compounds produced in
this study could be modified by the addition of R-groups extending
out of the FBP binding pocket to promote stronger interactions and
improved specificity for the Class II FBP aldolase from different path-
ogens. Longer-chain FBP analogues containing metal-chelating func-
tions have been shown recently to be very potent inhibitors of the
S. cerevisiae, C. albicans, H. pylori, M. tuberculosis, M. bovis and Y. pestis
Class II aldolases [18,21]. The use of metal-chelating substrate ana-
logues, like the ones reported in the present study, is therefore very
promising for the development of new drugs targeting the Class II
FBP aldolase.
[10] M.A. Jacobs, A. Alwood, I. Thaipisuttikul, D. Spencer, E. Haugen, S. Ernst, O. Will, R.
Kaul, C. Raymond, R. Levy, L. Chun-Rong, D. Guenthner, D. Bovee, M.V. Olson, C.
Manoil, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 14339–14344.
[11] K. Kobayashi, S.D. Ehrlich, A. Albertini, G. Amati, K.K. Andersen, M. Arnaud, K. Asai,
S. Ashikaga, S. Aymerich, P. Bessieres, F. Boland, S.C. Brignell, S. Bron, K. Bunai, J.
Chapuis, L.C. Christiansen, A. Danchin, M. Debarbouille, E. Dervyn, E. Deuerling,
K. Devine, S.K. Devine, O. Dreesen, J. Errington, S. Fillinger, S.J. Foster, Y. Fujita,
A. Galizzi, R. Gardan, C. Eschevins, T. Fukushima, K. Haga, C.R. Harwood, M.
Hecker, D. Hosoya, M.F. Hullo, H. Kakeshita, D. Karamata, Y. Kasahara, F.
Kawamura, K. Koga, P. Koski, R. Kuwana, D. Imamura, M. Ishimaru, S. Ishikawa,
I. Ishio, D. Le Coq, A. Masson, C. Mauel, R. Meima, R.P. Mellado, A. Moir, S.
Moriya, E. Nagakawa, H. Nanamiya, S. Nakai, P. Nygaard, M. Ogura, T. Ohanan,
M. O'Reilly, M. O'Rourke, Z. Pragai, H.M. Pooley, G. Rapoport, J.P. Rawlins, L.A.
Rivas, C. Rivolta, A. Sadaie, Y. Sadaie, M. Sarvas, T. Sato, H.H. Saxild, E. Scanlan,
W. Schumann, J.F. Seegers, J. Sekiguchi, A. Sekowska, S.J. Seror, M. Simon, P.
Stragier, R. Studer, H. Takamatsu, T. Tanaka, M. Takeuchi, H.B. Thomaides, V.
Vagner, J.M. van Dijl, K. Watabe, A. Wipat, H. Yamamoto, M. Yamamoto, Y.
Yamamoto, K. Yamane, K. Yata, K. Yoshida, H. Yoshikawa, U. Zuber, N.
Ogasawara, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 4678–4683.
[12] C.M. Sassetti, D.H. Boyd, E.J. Rubin, Mol. Microbiol. 48 (2003) 77–84.
[13] U.F. Wehmeier, FEMS Microbiol. Lett. 197 (2001) 53–58.
6. Abbreviations
BSA
bovine serum albumin
DHAP
DMPS
DPA
FBP
dihydroxyacetone phosphate
2,3-dimercaptopropane-1-sulfonate
dipicolinic acid
Fructose 1,6-bisphosphate
[14] M. de la Paz Santangelo, P.M. Gest, M.E. Guerin, M. Coincon, H. Pham, G. Ryan, S.E.
Puckett, J.S. Spencer, M. Gonzalez-Juarrero, R. Daher, A.J. Lenaerts, D.
Schnappinger, M. Therisod, S. Ehrt, J. Sygusch, M. Jackson, J. Biol. Chem. 286
(2011) 40219–40231.
GAP
H.E.I.
PGH
glyceraldehyde-3-phosphate
High Energy Intermediate (enediolate form of DHAP)
phosphoglycolohydroxamic acid
[15] A. Bock, F.C. Neidhardt, J. Bacteriol. 92 (1966) 470–476.
[16] D.A. Schneider, R.L. Gourse, J. Bacteriol. 185 (2003) 6192–6194.
[17] C.H. Su, J.P. Merlie, H. Goldfine, J. Bacteriol. 122 (1975) 565–569.
[18] M. Fonvielle, M. Coincon, R. Daher, N. Desbenoit, K. Kosieradzka, N. Barilone, B.
Gicquel, J. Sygusch, M. Jackson, M. Therisod, Chemistry 14 (2008) 8521–8529.
[19] M. Fonvielle, P. Weber, K. Dabkowska, M. Therisod, Bioorg. Med. Chem. Lett. 14
(2004) 2923–2926.
[20] S. Gavalda, R. Braga, C. Dax, A. Vigroux, C. Blonski, Bioorg. Med. Chem. Lett. 15
(2005) 5375–5377.
[21] R. Daher, M. Coincon, M. Fonvielle, P.M. Gest, M.E. Guerin, M. Jackson, J. Sygusch,
M. Therisod, J. Med. Chem. 53 (2010) 7836–7842.
Acknowledgements
G.L. is the grateful recipient of an Ontario Graduate Scholarship, a
UW President's Graduate Scholarship, and a UW Provost's/Faculty of
Science Graduate Women's Incentive Fund. This research was sup-
ported by the Natural Science and Engineering Research Council of
Canada Grants 18351, 927-2008 and 235003.
[22] G. Rosenblum, S.O. Meroueh, O. Kleifeld, S. Brown, S.P. Singson, R. Fridman, S.
Mobashery, I. Sagi, J. Biol. Chem. 278 (2003) 27009–27015.
[23] A.T. Colak, F. Colak, O.Z. Yesilel, O. Buyukgungor, J. Mol. Struct. 936 (2009) 67–74.
[24] Y. Kidani, J. Hirose, H. Koike, J. Biochem. 79 (1976) 43–51.
[25] Y. Kidani, J. Hirose, J. Biochem. 81 (1977) 1383–1391.
Appendix A. Supplementary data
[26] Y. Pocker, C.T. Fong, Biochemistry 19 (1980) 2045–2050.
[27] G. Labbe, J. Bezaire, S.d. Groot, C. How, T. Rasmusson, J. Yaeck, E. Jervis, G.I.
Dmitrienko, J. Guy Guillemette, Protein Expr. Purif. 51 (2007) 110–119.
[28] P.C. Ramsaywak, G. Labbe, S. Siemann, G.I. Dmitrienko, J.G. Guillemette, Protein
Expr. Purif. 37 (2004) 220–228.
Supplementary data to this article can be found online at doi:10.
1016/j.jinorgbio.2012.02.032.
References
[29] G. Labbe, S. de Groot, T. Rasmusson, G. Milojevic, G.I. Dmitrienko, J.G. Guillemette,
Protein Expr. Purif. 80 (2011) 224–233.
[1] D.J. Lewis, G. Lowe, J. Chem. Soc. Chem. Commun. (1973) 713–715.
[2] T. Baba, T. Ara, M. Hasegawa, Y. Takai, Y. Okumura, M. Baba, K.A. Datsenko, M.
Tomita, B.L. Wanner, H. Mori, Mol. Syst. Biol. 2 (2006) 2006.0008.
[30] J.F. Morrison, C.T. Walsh, Adv. Enzymol. Relat. Areas Mol. Biol. 61 (1988) 201–301.
[31] C.L. Tsou, Adv. Enzymol. Relat. Areas Mol. Biol. 61 (1988) 381–436.
[32] O. Trott, A.J. Olson, J. Comput. Chem. 31 (2010) 455–461.
[3] G. Lamichhane, M. Zignol, N.J. Blades, D.E. Geiman, A. Dougherty, J. Grosset, K.W.
Broman, W.R. Bishai, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 7213–7218.
[4] N.T. Liberati, J.M. Urbach, S. Miyata, D.G. Lee, E. Drenkard, G. Wu, J. Villanueva, T.
Wei, F.M. Ausubel, Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 2833–2838.
[5] S.D. Pegan, K. Rukseree, S.G. Franzblau, A.D. Mesecar, J. Mol. Biol. 386 (2009)
1038–1053.
[33] G.M. Morris, R. Huey, W. Lindstrom, M.F. Sanner, R.K. Belew, D.S. Goodsell, A.J.
Olson, J. Comput. Chem. 30 (2009) 2785–2791.
[34] X. Hu, W.H. Shelver, J. Mol. Graph. Model. 22 (2003) 115–126.
[35] S. Siemann, A.J. Clarke, T. Viswanatha, G.I. Dmitrienko, Biochemistry 42 (2003)
1673–1683.
[36] D.S. Auld, Methods Enzymol. 248 (1995) 228–242.
[6] J.H. Song, K.S. Ko, J.Y. Lee, J.Y. Baek, W.S. Oh, H.S. Yoon, J.Y. Jeong, J. Chun, Mol.
Cells 19 (2005) 365–374.
[37] D.J. Lewis, G. Lowe, Eur. J. Biochem. 80 (1977) 119–133.
[38] Z. Li, Z. Liu, D.W. Cho, J. Zou, M. Gong, R.M. Breece, A. Galkin, L. Li, H. Zhao, G.D.
Maestas, D.L. Tierney, O. Herzberg, D. Dunaway-Mariano, P.S. Mariano, J. Inorg.
Biochem. 105 (2010) 509–517.
[39] N.S. Blom, S. Tetreault, R. Coulombe, J. Sygusch, Nat. Struct. Biol.
856–862.
[40] T. Izard, J. Sygusch, J. Biol. Chem. 279 (2004) 11825–11833.
[41] D.R. Hall, L.E. Kemp, G.A. Leonard, K. Marshall, A. Berry, W.N. Hunter, Acta Crystal-
logr. D: Biol. Crystallogr. 59 (2003) 611–614.
[7] A. Rodaki, T. Young, A.J. Brown, Eukaryot. Cell 5 (2006) 1371–1377.
[8] S.Y. Gerdes, M.D. Scholle, J.W. Campbell, G. Balazsi, E. Ravasz, M.D. Daugherty, A.L.
Somera, N.C. Kyrpides, I. Anderson, M.S. Gelfand, A. Bhattacharya, V. Kapatral, M.
D'Souza, M.V. Baev, Y. Grechkin, F. Mseeh, M.Y. Fonstein, R. Overbeek, A.L.
Barabasi, Z.N. Oltvai, A.L. Osterman, J. Bacteriol. 185 (2003) 5673–5684.
[9] G. Giaever, A.M. Chu, L. Ni, C. Connelly, L. Riles, S. Veronneau, S. Dow, A. Lucau-Danila,
K. Anderson, B. Andre, A.P. Arkin, A. Astromoff, M. El Bakkoury, R. Bangham, R.
Benito, S. Brachat, S. Campanaro, M. Curtiss, K. Davis, A. Deutschbauer, K.D. Entian,
P. Flaherty, F. Foury, D.J. Garfinkel, M. Gerstein, D. Gotte, U. Guldener, J.H.
Hegemann, S. Hempel, Z. Herman, D.F. Jaramillo, D.E. Kelly, S.L. Kelly, P. Kotter, D.
3 (1996)
[42] M. St-Jean, C. Blonski, J. Sygusch, Biochemistry 48 (2009) 4528–4537.