Antimicrobial N-(2-chlorobenzyl)-substituted hydroxamate is an inhibitor of 1-deoxy-d-xylulose 5-phosphate synthase

Article abstract of DOI:10.1039/c3cc40758f

N-(2-Chlorobenzyl)-substituted hydroxamate, readily produced by hydrolysis of ketoclomazone, was identified as an inhibitor of 1-deoxy-d-xylulose 5-phosphate synthase (DXS), with an IC₅₀ value of 1.0 μM. The compound inhibited the growth of Haemophilus influenzae. A convenient spectroscopic method for assaying DXS using NADPH-lactate dehydrogenase (LDH) is also reported.

Full text of DOI:10.1039/c3cc40758f

Chemical Communications
www.rsc.org/chemcomm
Volume 49 | Number 49 | 21 June 2013 | Pages 5519–5606
ISSN 1359-7345
COMMUNICATION
Junko Ohkanda et al.
Antimicrobial N-(2-chlorobenzyl)-substituted hydroxamate is an inhibitor of 1-deoxy-D-xylulose
5-phosphate synthase

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Antimicrobial N-(2-chlorobenzyl)-substituted
hydroxamate is an inhibitor of 1-deoxy-D-xylulose
5-phosphate synthase†
Daisuke Hayashi,a Nobuo Kato,a Tomohisa Kuzuyama,b Yasuo Satoc and
Junko Ohkanda*a
Cite this: Chem. Commun., 2013,
49, 5535
Received 29th January 2013,
Accepted 27th February 2013
DOI: 10.1039/c3cc40758f
www.rsc.org/chemcomm
N-(2-Chlorobenzyl)-substituted hydroxamate, readily produced by
hydrolysis of ketoclomazone, was identified as an inhibitor of
1-deoxy-D-xylulose 5-phosphate synthase (DXS), with an IC50 value
of 1.0 lM. The compound inhibited the growth of Haemophilus
influenzae. A convenient spectroscopic method for assaying DXS
using NADPH–lactate dehydrogenase (LDH) is also reported.
Isoprenoids, a diverse group of compounds derived from the five-
carbon building units of isopentenyl diphosphate (IPP) and its
isomer, dimethylallyl diphosphate (DMAPP), are essential for sur-
vival in all organisms. Animals synthesize isoprenoids from meva-
lonic acid, whereas most pathogenic bacteria and the malaria
parasites utilize a distinct pathway for IPP and DMAPP synthesis,
the nonmevalonate (methylerythritol phosphate; MEP) pathway
(Scheme 1).1–4 Specific inhibitors of the MEP pathway therefore
hold promise as a new class of antibiotic, antituberculosis, and
antimalarial drugs.1–4 The most widely studied enzyme in the MEP
pathway is DXP reductoisomerase, which is responsible for the
second step in the pathway. A number of potent inhibitors have
been reported, such as the antimalarial drug fosmidomycin and its
Scheme 1 Isoprenoid biosynthesis via the MEP pathway.
derivatives.5,6 The first enzyme in the MEP pathway is 1-deoxy-D- report N-(2-chlorobenzyl)-substituted hydroxamate (1) as a new
xylulose-5-phosphate synthase (DXS), which catalyses the conden- inhibitor of Haemophilus influenzae DXS (HiDXS), with an IC50
sation of pyruvate and D-glyceraldehyde-3-phosphate in the value of 0.28 Æ 0.02 mg mLÀ1 (1.0 Æ 0.07 mM). This compound
presence of a cofactor, thiamine pyrophosphate, to generate was also found to suppress the growth of H. influenzae, with a
1-deoxy-D-xylulose-5-phosphate (DXP). Recent reports1–4,7–10 have minimum inhibitory concentration (MIC) value of 32 mg mLÀ1
.
highlighted DXS as a potential target for antimicrobial agents, These results may provide useful insights for further develop-
and identified several inhibitors, such as fluoropyruvate7 and ment of a new class of antimicrobial agents.
pyrazolopyrimidinone.8 However, only a limited number of
Ketoclomazone, N-(2-chlorobenzyl)-4,4-dimethylisoxazolidine-
inhibitors have been discovered, and their inhibition activities 3,5-dione (kcz, Scheme 2), a herbicide, inhibits Chlamydomonas
are rather moderate (10–400 mM). In this communication, we DXS and has an IC50 value of 0.1 mM.9 A more recent study
reported by Kuzuyama et al.10 revealed that kcz suppresses the
growth of H. influenzae (MIC = 12.5 mg mLÀ1). In addition, kinetic
a The Institute of Scientific and Industrial Research (ISIR), Osaka University,
studies demonstrated that kcz inhibits H. influenzae DXS
(HiDXS) at low micromolar concentration in a non-competitive
manner to pyruvate. This critical study prompted us to perform
further structure–activity-relationship studies based on kcz to
search for potent HiDXS inhibitors.
8-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan.
E-mail: johkanda@sanken.osaka-u.ac.jp; Fax: +81 6 6879 8474; Tel: +81 6 6979 8472
b Biotechnology Research Center, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku,
Tokyo 113-8657, Japan
c R&D Administration, Pharmaceuticals, Meiji Seika Kaisha, 2-4-16, Kyobashi,
Chuo-ku, Tokyo 104-8002, Japan
During the course of the synthetic study,11 we noticed that
mass analysis of kcz in methanol exclusively provided a peak
† Electronic supplementary information (ESI) available: Details of synthesis,
biochemical experiments and cell-based assays. See DOI: 10.1039/c3cc40758f
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 5535--5537 5535

View Article Online
Communication
ChemComm
Scheme 2 Chemical structure of ketoclomazone, hydroxamate (1), and related
derivatives (2–4) evaluated in this study.
corresponding to the 1 : 1 methanol adduct ([M + Na]+ = 308.07,
Fig. S1 in ESI†). 1H NMR was used to analyse kcz in methanol-d4.
We found that the spectrum demonstrated two sets of singlet
peaks corresponding to the benzylmethylene and dimethyl
groups (5.08 and 4.90, and 1.42 and 1.40 ppm, respectively),
and that the peaks eventually converged into one pair after 72 h
(Fig. S2 in ESI†). These results suggest that kcz gradually reacts
with methanol. Furthermore, the 13C NMR spectrum of kcz in
methanol-d4 at 72 h showed that the resulting product provides
two carbonyl peaks (175.8 and 174.9 ppm, Fig. S3 in ESI†).
Based on these results, we concluded that kcz undergoes a ring
opening reaction following methanol attack at C-5 of the
isoxazolidine-3,5-dione, yielding the corresponding hydroxamate
derivative (Fig. S1 and S3 in ESI†). This observation motivated us
to chemically prepare hydroxamate 1 and related derivatives 2–4
and evaluate their DXS inhibitory activity (for preparation, see
ESI†). As expected, simple treatment of kcz with base readily
afforded compound 1 quantitatively.
To evaluate these compounds, we wanted to perform in vitro
enzyme assays using recombinant HiDXS to conveniently
determine the pyruvate concentration. Several assay methods
for DXS have been reported using 14C-pyruvate,8,12 or coupled
systems using DXP reductoisomerase.7,10 However, these procedures
require a radioactive substrate or additional recombinant enzymes,
which necessitate further expression and purification procedures.
We therefore coupled the HiDXS assay with the NADH-driven
lactate dehydrogenase (LDH) reaction, which has been widely
used to quantify pyruvate.13 In the LDH–NADH reaction,
pyruvate is converted to lactate in an NADH-dependent
manner. Therefore, the pyruvate concentration is correlated
with the amount of NADH consumed, which can be determined
by monitoring the change in UV absorbance. We also used
inexpensive D-glyceraldehyde (GA) as the substrate, and
designed the protocol for a 96-well plate format (in ESI†).
Briefly, stock solutions of compounds and pyruvate were placed
in a well and diluted with an assay buffer containing GA and
thiamine pyrophosphate (TPP). Recombinant HiDXS (in ESI†)
was added and the plate was incubated at 37 1C for 2 h. The
reaction mixture was then transferred to a fresh well containing
NADH and LDH, and the absorbance at 340 nm was measured.
Prior to evaluation of the compound, we wanted to verify that
this system could be used to estimate pyruvate concentration
accurately. Using this protocol with various concentrations of
pyruvate in the absence of HiDXS provided a linear relationship
Fig. 1 The % inhibition of DXS in the presence of 12.8 mg mLÀ1 of compounds
1–4 and kcz. Inset: dose–response curve for the inhibition of HiDXS by compound
1. All experiments were performed in duplicate.
between the initial concentration of pyruvate and the consump-
tion of NADH (Fig. S4 in ESI†), thereby validating the use of the
assay system for evaluating DXS inhibitory activity.
We then tested compound 1 for HiDXS inhibition. As shown
in Fig. 1 (inset), compound 1 significantly suppressed enzyme
activity in a dose–response manner, with an IC50 value of 0.28 Æ
0.02 mg mLÀ1 (1.0 mM), thus clearly demonstrating that 1 is
potent against HiDXS. We then compared the inhibition ability
of 1 to those of kcz and related compounds 2–4 at 12.8 mg mLÀ1
(Fig. 1). Similar to 1, kcz was found to be very active, completely
inhibiting HiDXS activity (99.3 Æ 0.0 and 98.9 Æ 0.1% inhibi-
tion for 1 and kcz, respectively). In contrast, as seen in 2–4,
absence of either a hydroxyl, dimethyl, or a carboxyl group in 1
significantly diminishes activity (11.4 Æ 5.4, 2.8 Æ 0.24, 3.13 Æ
0.28%, respectively), demonstrating that these structural
moieties are requisite for HiDXS inhibition.
Next, we tested the compounds for antibacterial activity
against H. influenzae (Table 1, also see ESI†). kcz inhibited
bacterial growth with a MIC value of 8 mg mLÀ1, which is
consistent with a previously reported result10 (12.5 mg mLÀ1).
The hydroxamate 1 suppressed growth at 32 mg mLÀ1, whereas
compounds 2 and 3 were not active. Interestingly, compound 4,
in which the carboxyl group in 1 was reduced to a primary
alcohol, showed moderate activity (64 mg mLÀ1), suggesting that
reducing the negative charge might improve cell permeation.
Finally, we wanted to examine whether the hydroxamate 1
targets DXS and inhibits bacterial proliferation. The free
alcohol form of DXP, 1-deoxy-D-xylulose (DX), has been used
Table 1 MIC values of compounds against Haemophilus influenzae ATCC43095
Compound
MICa (mg mLÀ1
)
Ketoclomazone (kcz)
8 (12.5)b
1
2
3
4
32
256
128
64
a
b
Minimum inhibitory concentration. Ref. 10.
c
5536 Chem. Commun., 2013, 49, 5535--5537
This journal is The Royal Society of Chemistry 2013

View Article Online
ChemComm
Communication
of hydroxamate 1. Interestingly, a similar herbicide, clomazone,
is structurally identical to kcz except that it lacks a C-5 carbonyl
group yet has no effect on bacterial growth1 or DXS activity.9 It
has been hypothesised that plants can metabolize clomazone
into kcz but bacteria cannot.1,10 This may also support the
possibility that after clomazone is oxidized to provide a carbonyl
group at the C-5 position, the resulting kcz eventually produces
hydroxamate 1, which targets DXS activity. Further experiments,
such as elucidating the metabolic pathway for clomazone and
kcz, are required.
In conclusion, we have demonstrated that N-(2-chlorobenzyl)-
substituted hydroxamate 1, which is readily derived from kcz by
hydrolysis, significantly inhibits HiDXS activity, with an IC50 value of
1.0 mM. To the best of our knowledge, this is the first example of a
DXS inhibitor exhibiting single-digit mM activity. Compound 1 also
inhibits the growth of H. influenzae. This inhibitory effect is negated
by the addition of DX, indicating that the inhibition of DXS in cells is
the reason for its antibacterial activity. These results suggest that
hydroxamates provide a potential scaffold for developing new
inhibitors for DXS. Compound 1, however, only moderately inhibits
bacterial growth. Further structure–activity-relationship studies, as
well as efforts to improve cell permeability, are necessary for
improving hydroxamate-based antibacterial agents. Work towards
this goal is currently in progress in our laboratory.
This study was supported by a grant provided by the Eisai
Research Foundation (J. O.). We thank Mr Tsuyoshi Matsuzaki
(ISIR, Osaka University) for HRMS analysis.
Notes and references
Fig. 2 Growth inhibition activity of kcz, fosmidomycin and 1 (0–128 mg mLÀ1
)
1 M. Rodriguez-Concepcion, Curr. Pharm. Des., 2004, 10, 2391.
2 T. Kuzuyama and H. Seto, Proc. Jpn. Acad., Ser. B, 2012, 88, 41.
3 H. Eoh, P. J. Brennan and D. C. Crick, Tuberculosis, 2009, 89, 1.
4 I. Hale, P. M. O’Neill, N. G. Berry, A. Odom and R. Sharma,
MedChemComm, 2012, 3, 418.
against H. influenzae was evaluated in the absence (white) and the presence
(gray) of 1-deoxy-D-xyluose (DX, a free alcohol form of DXP, 1 mg mLÀ1). The
standard deviation values are given for n = 3 (for fosmidomycin and 1) and 4 (for
kcz).
5 T. Kuzuyama, T. Shimizu, S. Takahashi and H. Seto, Tetrahedron
Lett., 1998, 39, 7913.
6 H. Jomaa, J. Wiesner, S. Sanderbrand, B. Altincicek, C. Weidemeyer,
M. Hintz, I. Tu¨bachova, M. Eberl, J. Zeidler, H. K. Lichtenthaler,
D. Soldati and E. Beck, Science, 1999, 285, 1573.
7 B. Altincicek, M. Hintz, S. Sanderbrand, J. Wiesner, E. Beck and
H. Jomaa, FEMS Microbiol. Lett., 2000, 190, 329.
8 J. Mao, H. Eoh, R. He, Y. Wang, B. Wan, S. G. Franzblau, D. C. Crick
and A. P. Kozikowski, Bioorg. Med. Chem. Lett., 2008, 18, 5320.
9 C. Mueller, J. Schwender, J. Zeidler and H. K. Lichtenthaler,
Biochem. Soc. Trans., 2000, 28, 792.
10 Y. Matsue, H. Mizuno, T. Tomita, T. Asami, M. Nishiyama and
T. Kuzuyama, J. Antibiot., 2010, 63, 583.
11 Ketoclomazone was synthesized by a similar procedure to that
reported previously in the literature: H. Sakakibara, H. Kasahara,
N. Ueda, M. Kojima, K. Takei and S. Hishiyama, Proc. Natl. Acad. Sci.
U. S. A., 2005, 102, 9972.
12 A. M. Bailey, S. Mahapatra, P. J. Brennan and D. C. Crick, Glycobiology,
2002, 12, 813.
13 S. de Marcos, J. Galban, C. Alonso and J. R. Castillo, Analyst, 1997,
122, 355.
as an exogenous nutrient for MEP pathway-dependent bacteria,10
as DX is likely to be converted by bacterial D-xylulokinase14 to DXP
and therefore rescues the second step of the MEP pathway. Thus,
we synthesized DX,15 added it to a bacterial culture in the
presence of compound 1, and observed whether bacterial growth
was restored. Addition of 0.1% DX to the medium apparently
negated the antimicrobial activity of kcz (Fig. 2a), as reported
previously,10 whereas in the case of fosmidomycin, a selective
inhibitor of DXP reductoisomerase, addition of DX had no effect
on bacterial growth (Fig. 2b). In contrast, there was a clear pattern
showing that compound 1 suppresses bacterial growth only in the
absence of DX (Fig. 2c), showing that 1 inhibits DXS in H. influenzae
cells and results in bacterial growth inhibition.
It is important to note that kcz exhibits stronger cell-based
activity than 1, despite the fact that these compounds were
found to be equally potent towards HiDXS. This may suggest
that the hydrophobic kcz may permeate cells more readily than
the anionic hydroxamate. Taken together with the susceptibility
of kcz to ring opening, kcz could be a pro-drug for the generation
14 J. Wungsintaweekul, S. Herz, S. Hecht, W. Eisenreich, R. Feicht,
F. Rohdich, A. Bacher and M. H. Zenk, Eur. J. Biochem., 2001,
268, 310.
15 DX was synthesized by the procedure reported in the literature:
D. Shabat, B. List, R. A. Lerner and C. F. Barbas III, Tetrahedron Lett.,
1999, 40, 1347.
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 5535--5537 5537