Synthesis and biological evaluation of phosphate isosters of fosmidomycin and analogs as inhibitors of Escherichia coli and Mycobacterium smegmatis 1-deoxyxylulose 5-phosphate reductoisomerases

Article abstract of DOI:10.1016/j.bmc.2016.11.040

Hydroxamate analogs of fosfoxacin, the phosphate homolog of fosmidomycin, have been synthesized and their activity tested on Escherichia coli and Mycobacterium smegmatis DXRs. Except for compound 4b, the IC₅₀values of phosphate derivatives are

Full text of DOI:10.1016/j.bmc.2016.11.040

Accepted Manuscript
Synthesis and biological evaluation of phosphate isosters of fosmidomycin and
analogs as inhibitors of Escherichia coli and Mycobacterium smegmatis 1-de-
oxyxylulose 5-phosphate reductoisomerases
Mathilde Munier, Denis Tritsch, Fanny Krebs, Jérémy Esque, Andréa
Hemmerlin, Michel Rohmer, Roland H. Stote, Catherine Grosdemange-Billiard
PII:
S0968-0896(16)31115-4
DOI:
http://dx.doi.org/10.1016/j.bmc.2016.11.040
BMC 13407
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
2 November 2016
18 November 2016
21 November 2016
Please cite this article as: Munier, M., Tritsch, D., Krebs, F., Esque, J., Hemmerlin, A., Rohmer, M., Stote, R.H.,
Grosdemange-Billiard, C., Synthesis and biological evaluation of phosphate isosters of fosmidomycin and analogs
as inhibitors of Escherichia coli and Mycobacterium smegmatis 1-deoxyxylulose 5-phosphate reductoisomerases,
Bioorganic & Medicinal Chemistry (2016), doi: http://dx.doi.org/10.1016/j.bmc.2016.11.040
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthesis and biological evaluation of phosphate isosters of
fosmidomycin and analogs as inhibitors of Escherichia coli and
Mycobacterium
smegmatis
1-deoxyxylulose
5-phosphate
reductoisomerases
Mathilde MUNIERa, Denis TRITSCHa, Fanny KREBSa,b, Jérémy ESQUEb, Andréa
HEMMERLINc, Michel ROHMER a, Roland H. STOTEb, Catherine GROSDEMANGE-
BILLIARD* a
a
Laboratoire Chimie et Biochimie de Molécules Bioactives
–
Université de
Strasbourg/CNRS, UMR 7177, Institut Le Bel, 4 rue Blaise Pascal, 67081 Strasbourg, France
b
Department of Integrative Structural Biology, Institut de Génétique et de Biologie
Moléculaire et Cellulaire (IGBMC), Institut National de la Santé et de la Recherche Médicale
(INSERM) U964, Centre National de la Recherche Scientifique (CNRS) UMR 7104,
Université de Strasbourg, 67404 Illkirch, France.
c
Université de Strasbourg, CNRS, IBMP UPR 2357, 12 rue Général Zimmer F-67084
Strasbourg, France
Corresponding author: Tel.: +33-368-851-349; fax: +33-368-851-345; e-mail:
grosdemange@unistra.fr
Abstract
Hydroxamate analogs of fosfoxacin, the phosphate homolog of fosmidomycin, have been
synthesized and their activity tested on Escherichia coli and Mycobacterium smegmatis
DXRs. Except for compound 4b, the IC50 values of phosphate derivatives are approximately
10 fold higher than those of the corresponding phosphonates. Although their inhibitory activity
on Escherichia coli DXR is less efficient than their phosphonate analogs, we report the ability of
1

phosphate compounds to inhibit the growth of Escherichia coli. This work points out that the
uptake of fosfoxacin and its analogues is taking place via the GlpT and UhpT transporters. As
expected, these compounds are inefficient to inhibit the growth of M. smegmatis growth
inhibition probably due to a lack of uptake.
Keywords: antimicrobials, fosfoxacin, fosmidomycin, DXR, GlpT and UhpT transporters
Abbreviations: DXP: 1-deoxy-D-xylulose 5-phosphate; DXR: 1-deoxy-D-xylulose 5-
phosphate reductoisomerase; MEP: 2-C-methyl-D-erythritol 4-phosphate; GlpT: glycerol-3-
phosphate transporter; UhpT: hexose-6-phosphate transporter; FosR: fosmidomycin resistant
strain of E. coli
2

1. Introduction
In many pathogenic bacteria, the enzymes of the 2-C-methyl-D-erythrirol 4-phosphate (MEP)
pathway, involved in the biosynthesis of isopentenyl diphosphate and dimethylallyl
diphosphate, the two universal and essential units for the biosynthesis of isoprenoids are
recognized as valuable targets for the development of new antimicrobial compounds.1 Of
particular interest is the enzyme 1-deoxyxylulose 5-phosphate reductoisomerase (DXR) that
catalyzes the transformation of 1-deoxy-D-xylulose 5-phosphate (DXP) to 2-C-methyl-D-
erythritol 4-phosphate (MEP) via an isomerization and a NADPH dependent reduction
reaction.2–4 Fosmidomycin 1a and its close relative FR900098 2a, two natural retro-
hydroxamate phosphonic acids are efficient inhibitors of DXR (Figure 1).5–8 Bacteria,
however, rapidly develop resistance to fosmidomycin.9,10 In an attempt to improve the
efficiency of such inhibitors, numerous analogs have been synthesized. Unfortunately, to date,
the results appear rather disappointing as none of these compounds are significantly more
efficient than fosmidomycin. In previous work, we described the synthesis and the DXR
inhibiting activity of hydroxamate phosphonic acid 3a and of its N-methylated derivative 4a
(Fig. 1). The latter compound, which is as effective as fosmidomycin on DXR, has significant
antibacterial activity against wild type Escherichia coli and even against an E. coli mutant
resistant to fosmidomycin.11
Figure 1. DXR inhibitors.
Crystallographic structures of the DXR-NADPH-fosmidomycin complex show that the
phosphonate group of the inhibitor is located in the phosphate binding cleft of the substrate
DXP and is bound in a similar fashion as the phosphate group. Moreover, the hydroxamate
chelates the Mg2+ dication present in the active site.12,13
In the early 1990’s, fosfoxacin 1b, the phosphate analog of fosmidomycin, was isolated from
the filtrate of a Pseudomonas fluorescens PK-52 culture and was shown to have potent
3

antibiotic activity against Gram-negative and Gram-positive bacteria.14 Latter on, fosfoxacin
and its N-acetylated analog 2b were synthetized and evaluated as inhibitors of the DXR from
Synechocystis sp. PCC6803. They were found to be more potent than fosmidomycin.15
Despite these promising results, no further investigations of fosfoxacin on other microbial
DXRs have been reported.
In this work, we synthesized the phosphate derivative of fosmidomycin, fosfoxacin, and
several other related hydroxamate analogs (2-4, Fig. 1) in order to evaluate their effect on the
activity of DXRs from E. coli and Mycobacterium smegmatis. This study also allowed us to
investigate the influence of the replacement of a phosphonate with a phosphate group. As
fosfoxacin was previously shown to inhibit the growth of several bacteria, it is obvious that
this phosphate ester can cross bacterial envelopes, including that of E. coli. In this study, we
also compared the efficiency of the different phosphate DXR inhibitors to inhibit the growth
of E. coli. However, as for fosmidomycin and its phosphonate analogs, the phosphate
derivatives should not be capable of crossing the bacterial wall and of inhibiting the growth of
M. smegmatis.
2. Results
2.1. Synthesis of phosphate derivatives 1b-4b
Fosfoxacin 1b and its acetylated homolog 2b were synthesized according to previously
described methods.15 The spectral data were in accordance with those published in the
literature.
The syntheses of compounds 3b and 4b are outlined in Scheme 1. The O-benzyl hydroxamate
derivative 6a was obtained in one step from commercially available β-propiolactone 5 using
O-benzyl hydroxylamine hydrochloride in presence of an excess of LiHMDS.16 Compound 6a
was obtained as a mixture of two Z and E conformers in a 7:3 ratio, respectively, due to the
1
restricted rotation around the C-N bond. The H NMR spectrum of 6a shows two signals for
the CH2CON(H)OBn protons: one major signal at 2.33 ppm corresponding to the protons of
the Z conformer and a minor signal at 2.50 ppm corresponding to the protons of the E
conformer.17,18 The selective N-methylation versus O-methylation of 6a was accomplished
under basic conditions. The methyl group was introduced by reaction of 6a with K2CO3 in
1
acetone under reflux followed by addition of methyl iodide. The H NMR spectrum of 2b
shows one signal for CH2CON(CH3)OBn protons at 2.62 ppm corresponding to the protons of
4

the E conformer. Introduction of dibenzyl phosphate was made using standard
phosphoramidite chemistry. Alcohols 6a and 6b were coupled with dibenzyl N,N-
diisopropylphosphoramidite in the presence of 1H-tetrazole to yield dibenzylphosphite
derivatives, which were oxidized to phosphate (7a and 7b) using m-chloroperbenzoic acid.19
Scheme 1. Reagents and conditions. (a) (i). H2NOBn.HCl, LiHMDS, THF, -78 °C (ii). TBAF.3(H2O), THF, rt, 80 % ; (b)
MeI, K2CO3, acetone, 60 °C, 81 % ; (c) (i). 1H-tetrazole, (BnO)2PN(iPr)2, DCM, rt ; (ii). m-CPBA, DCM, 0 °C, 7a (60 %),
7b (85 %) ; (d) H2, Pd/C, EtOH, rt 3b (99 %), 4b (56 %).
A mixture of two Z and E conformers in a 8:2 ratio was obtained for the phosphotriester 7a,
whereas for the N-methylated homolog 7b only the E conformer was detected in deuterated
chloroform solution. Removal of the protective benzyl groups of 7a and 7b was performed by
catalytic hydrogenolysis with palladium over charcoal at atmospheric pressure and room
temperature in ethanol. The phosphate esters 3b and 4b required no further purification and
were obtained with overall yields of 48 % and 29 %, respectively. After deprotection, as
1
determined by H NMR in deuterated methanol, the target compounds 3b and 4b were
obtained as a unique conformer. For molecule 3b, the signal at 2.45 ppm corresponding to
CH2CON(H)OH was less deshielded than the signal at 2.87 ppm corresponding to
CH2CON(CH3)OH of the molecule 4b, suggesting that compound 3b was in a Z conformation
whereas molecule 4b was present as an E conformer.17,18 To confirm the assignments of 4b, a
NOESY experiment was realized. As a result, the spectrum displayed correlations
between the singlet at ca. 3.21 ppm assigned to the N-Me protons and the triplet at ca.
2.87 ppm corresponding of the -CH2CONROH protons, indicating that these protons are
within 5 Angstroms through space and corroborated the presence of a E conformer. All
5

phosphate compounds were tested against recombinant His-tagged DXR of E. coli and M.
smegmatis for inhibition.
2.2. Inhibition of E. coli and M. smegmatis H-DXR with compounds 1-4
The inhibition potency of the phosphate derivatives, characterized by their IC50 value was
determined as described previously.11 Assuming that the phosphate analogs act as slow
binding inhibitors like fosmidomycin, they were pre-incubated with DXR during 2 min in the
presence of NADPH. Residual activity was measured after initiating the enzymatic reaction
by addition of DXP. The IC50 values are listed in Table 1.
Table 1
Inhibition of E. coli and M. smegmatis H6-DXRs by fosmidomycin 1a, FR-900098 2a,
phophonohydroxamic analogs 3a and 4a, fosfoxacin 1b and its analogs 2b-4b.
E. coli DXR
M. smegmatis DXR
Compounds
IC50 (nM)
IC50 (nM)
510*
320*
1480*
410*
2100
279
Fosmidomycin 1a
42
4
FR-900098 2a
3a
170
48
4a
Fosfoxacin 1b
342
77
2b
3b
2600
46
9200
233
4b
* Values obtained from reference.20
Phosphates 2b and 4b are effective inhibitors against E. coli DXR as well as fosmidomycin
1a with similar IC50 values (IC50 = 46 nM and 77 nM vs IC50 = 42 nM) but they are 12 to 20-
fold less potent inhibitors than FR-900098 2a. The IC50 values for the fosfoxacin 1b (IC50 =
342 nM) and the N-H analog 3b (IC50 = 2600 nM) indicate that these compounds are poor
inhibitors of DXR in contrast to the results obtained with Synechocystis sp. PCC6803 for
fosfoxacin (Ki = 19 nM) and its acetylated analog 2b (Ki = 2 nM), which were found to be
more potent than fosmidomycin (Ki = 57 nM).15 Concerning the DXR of M. smegmatis,
phosphate compounds 2b and 4b, with IC50 values of 279 and 233 nM respectively, are the
6

most potent inhibitors. The phosphonate compounds fosmidomycin 1a, FR-900098 2a and N-
H phosphonohydroxamate 4a with respectively IC50 values of 510 nM, 320 nM and 410 nM
are slightly less efficient. Fosfoxacin and the N-H analog 3b are poor inhibitors with IC50
values 8 and 30-fold lower than the methylated derivatives 2b and 4b indicating that the
presence of the methyl group in compounds 2b and 4b has clearly a positive effect on the
affinity of the enzyme for these compounds.
2.3. Growth inhibition of E. coli and M. smegmatis with compounds 1-4
The antimicrobial activity of phosphate compounds 1b-4b was determined using the paper
disc diffusion method and was compared with the antimicrobial activity of their phosphonate
homologs 1a-4a. The diameters of the inhibition zone are given with respect to the amount of
inhibitor deposited on the disc (Table 2).
Table 2
E. coli growth inhibition with compounds 1-4 ((Supporting information Fig. S2)
Amount of
Diameter of the
Compounds
compound/disc (nmoles)
growth inhibition zone (mm)
Fosmidomycin 1a
2
34
32
FR-900098 2a
2
3a
400
80
12
4a
30
Fosfoxacin 1b
160
100
400
160
24
2b
3b
4b
27
<10
20
Fosmidomycin which is, with FR-900098, the most efficient inhibitor of E. coli growth was
used as a positive control and the reference. The other DXR inhibitors had a lower
effectiveness to inhibit the bacterial growth as larger amount had to be added to observe
growth inhibition zones ((Supporting information Fig. S2). Clearly there is no direct relation
between the capacity to inhibit the DXR and that to stop the bacterial growth; for compound
7

4a, a 40-fold larger amount was deposited on the paper disc to observe a diameter similar to
that of fosmidomycin. The N-H phosphonate 3a and N-H phosphate 3b compounds were
almost ineffective. As a general rule, the phosphate compounds were less efficient than the
phosphonate homologs.
All DXR inhibitors were tested on a fosmidomycin resistant strain of E. coli (FosR).10 Except
compound 4a, none of the inhibitors was able to affect bacterial growth (Supporting
information Fig. S3).
Concerning M. smegmatis, as anticipated, none of the compounds was able to inhibit its
growth, probably the consequence of a lack of uptake (data not shown).20,35
3. Discussion
Over a large variety of natural and synthetic derivatives of the phosphono retrohydroxamic
acid fosmidomycin, only two phosphate analogs, the natural fosfoxacin 1b and its synthetic
N-acetylated derivative 2b have been reported in the literature. This is not that surprising as
phosphate compounds are rarely exploited as potential antibiotics due to their limited lack of
cell penetration and their sensitivity towards the hydrolytic activity of phosphatases.
However, fosfoxacin, as well as its synthetic N-acetylated derivative, were shown to inhibit
the growth of several bacteria,14 and to be more potent inhibitors of Synechocystis sp.
PCC6803 DXR than phosphonate homologs 1a and 2a.15 Such promising results prompted us
to evaluate the efficiency of fosfoxacin 1b and its derivatives 2b-4b against E. coli and M.
smegmatis DXRs. Except for compound 4b, the IC50 values of phosphate derivatives are
noticeably higher than those of the corresponding phosphonates, with IC50 values
approximately 10 fold higher, suggesting that the binding of these compounds to the enzyme
is significantly weaker. These results are quite surprising as such a phosphate group should
better fit into a phosphate binding site than a phosphonate group. Indeed, DXR has a lower
catalytic efficiency for the phosphonate DXP analog (kcat/Km = 0.7 min-1µM-1) than for DXP,
the natural substrate (kcat/Km = 24.7 min-1µM-1).21 Likewise, in the case of triose phosphate
isomerase, which possesses, as DXR, a flexible loop closing the active site, the
phosphoglycolohydroxamate inhibitor (Ki = 3 µM) was shown to be a better inhibitor than its
phosphonate analog (Ki = 159 µM).22 Even if a phosphonate group is considered to be an
isosteric phosphate group, some differences such as the pKa and the C-O-P/C-C-P bond
8

angles might impact their binding in the active site. The pKa2 value of a phosphonate (7.6)
being higher than that of a phosphate (6.4), indicates that the phosphate should be in dianionic
form at the pH of the enzymatic assay, whereas the phosphonate should be predominantly in
singly ionized form. This latter form is considered to be less favorable for an efficient
binding. In addition, the C-O-P angle (118.7°) is wider than the C-C-P angle (112.1°)23
resulting in a different setting of the inhibitors in the active site. This leads to modifications of
the electrostatic and van der Waals interactions and consequently in an increase or a decrease
of the affinity for the inhibitors. Moreover, the modification of the three-carbon spacer by
replacing a methylene group by an oxygen atom such as in the β-oxa isosters of fosmidomycin
and compound 3a, was reported to lead to a noticeable drop of the IC50 values for the E. coli
enzyme while those of the 2a and 4a derivatives are still unchanged.24
Another contributing factor that may be involved in interpreting these results is the presence
of a flexible loop in the DXR that has an essential role in catalysis and inhibition, shielding
from the ligands the bulk solvent in the active site. 25,26 Therefore, any small but significant
conformational changes (e.g disruption of hydrogen bound network, additional hydrogen
bounds, new interactions between an inhibitor and the loop …) initiated by the binding of the
substrate or an inhibitor, such as fosmidomycin vs fosfoxacin, might alter the closure of the
loop over the active site leading to a more or less tight [DXR-NADPH-inhibitor] ternary
complex.
However, it should be noted that the presence of the methyl group in compounds 2b and 4b as
well as in the phosphonate homologs 2a and 4a increases the inhibitory activity against E. coli
and M. smegmatis DXRs. Based on crystal structure analyses of Plasmodium falciparum and
M. tuberculosis DXRs, van der Waals interactions between the methyl group of the 4a and 2a
with the indole ring of the highly conserved tryptophan residue (P. falciparum, Trp 296; M.
tuberculosis, Trp 203) from the flexible loop region might occur with the consequence of the
better activity of N-methylated compared to N-H compounds.27,28
Compared to the phosphonate compounds, the efficiency of phosphates on the growth
inhibition of E. coli was much lower. Several reasons may be invoked: i) the lower DXR
inhibition efficiency and/or ii) the cell wall/plasma membrane crossing of the phosphate esters
and/or iii) the inactivation by hydrolysis of the phosphate group by phosphatases.
Nevertheless, the fact that the phosphate derivatives inhibit bacterial growth of E. coli shows
that these compounds enter the bacteria. The hydrophilic character of the compounds suggests
that they do not passively diffuse across the cell membrane. A crossing via a transporter is
accordingly privileged, particularly via the glycerol 3-phosphate (GlpT)29,30 and the hexose 6-
9

phosphate (UhpT) transporters.31 These transporters allow the penetration of essential
nutrients (glycerol 3-phosphate and glucose 6-phosphate) but also hydrophilic antibiotics like
fosfomycin, a phosphonate antibiotic which inhibits the UDP-N-acetylglucosamine
enoylpyruvyl transferase (MurA), an enzyme implied in the biosynthesis of peptidoglycan.32
The import of the molecules is driven by the efflux of inorganic phosphate.33 As already
observed for fosmidomycin,10 many resistant clones grew in the growth inhibition zone when
compounds 1b-4b were used. Some resistant bacteria were picked off from the inhibition
zone and tested for cross-resistance with fosfomycin. All were resistant to this antibiotic
indicating that a lack of uptake of the inhibitor via the GlpT and/or the UhpT is probably the
cause.34 The lack of uptake by mutation in the transporter system is one of the multiple causes
for resistance to antibiotics. A fosmidomycin resistant strain in Escherichia coli (FosR) was
genetically characterized and was shown to contain a transposon insertion mutation in the
gene coding for the GlpT transporter.10 Sequencing of the UhpT gene showed that it is also
mutated as a transposon was inserted (Fig. S1 in supporting information). In this strain the
GlpT/UhpT transporters are therefore inoperative and the fact that all phosphate derivatives
were unable to prevent the growth of the bacteria, suggests that these compounds enter into
the bacteria via these transporters. Moreover, this strain remains sensitive to compound 4a
implying that the latter follows besides a different way to penetrate into the bacteria since it
was less efficient to inhibit the FosR strain than the wild type (Supporting information Fig.
S4).
In the present work, we pointed out that the efficiency of a phosphate inhibitor compared to
its phosphonate homolog might be notably different towards DXRs isolated from different
microorganisms. Even if the phosphate derivatives were previously shown to be more potent
inhibitors of Synechocystis DXR than their phosphonate analogs, this conclusion is not valid
for all DXRs, as we have shown that they are less efficient against E. coli and M. smegmatis. .
It is not obvious to rationalize these results as several changes in van der Waals and
electrostatic forces at the anchoring and chelating sites as well as in the closure of the
flexible loop might affect the binding of an inhibitor in the DXR active site, preventing or
privileging its fixation in the DXR active site. It is important to note that the presence of the
methyl group on the hydroxamate/retrohydroxamate group increases the inhibitory activity
against DXR isolated from E. coli and M. smegmatis DXR as well as their phosphonate
homologues.Even if the phosphate compounds are less efficient compared to the
phosphonates on the growth inhibition of E. coli, they can enter into the bacteria via the
GlpT/UhpT transporters. The tested DXR inhibitors cannot cross the bacterial wall of M.
10

smegmatis and to circumvent the lack of uptake of these compounds, we prepared prodrugs
where the phosphate group is masked to render them less hydrophilic to allow the crossing of
the bacterial envelope. Moreover such a masking should also protect the inhibitors against
hydrolysis by phosphatases (Munier et al., unpublished results).
4. Experimental Section
4.1. General methods
All non-aqueous reactions were run in oven-dried glassware under an argon atmosphere,
using dry solvents. Commercial grade reagents were purchased from Sigma-Aldrich or Acros
Organics and used without further purification. Petroleum ether 40-60 °C (Sigma-Aldrich)
was used for purification Flash chromatography was performed on silica gel 60 230–400
mesh with the solvent system as indicated. TLC plates were revealed under UV light (254
nM) and/or by spraying with an ethanolic solution of phosphomolybdic acid (20%) or an
ethanolic solution of potassium permanganate followed by heating.
The NMR spectra were recorded on a BRUKER Avance 300 (1H-NMR: 300 MHz; 13C-NMR,
75.5 MHz; 31P-NMR 121.5 MHz; 19F-NMR 282.4 MHz) or a BRUKER Avance 400 (1H-
NMR: 400 MHz; 13C-NMR, 100.6 MHz; 31P-NMR 162 MHz) or a BRUKER Avance 500
1
(1H-NMR: 500 MHz; 13C-NMR, 125.8 MHz). H-NMR experiments were performed in
CDCl3, D2O, CD3OD in CDCl3 with CHCl3 (δ= 7.26 ppm), DHO (δ= 4.79 ppm), CD2HOD (δ
= 3.31 ppm) as internal references. 13C-NMR experiments were performed in CDCl3 with
CDCl3 (δ= 77.23 ppm), CD3OD (δ= 49.0 ppm) as internal references. For 31P-NMR reference,
the spectrometer had an external reference, corresponding to 80 % phosphoric acid in D2O (δ
= 0 ppm). The chemical shifts (δ) are expressed in ppm. s, d, t, q, or bs are abbreviations for
multiplicity correspond to singulet, doublet, triplet and quadruplet or broad singulet. J-
couplings are exposed in Hz.Negative or positive-mode electrospray MS were performed on a
Bruker Daltonics microTOF spectrometer (Bruker Daltonik GmbH, Bremen, Germany)
equipped with an orthogonal electrospray (ESI) interface. Calibration was performed using a
solution of 10 mM sodium formate. Sample solutions were introduced into the spectrometer
source with a syringe pump (Harvard type 55 1111: Harvard Apparatus Inc., South Natick,
MA, USA) with a flow rate of 5 μL min−1.The compounds 1b (fosfoxacin) and 2b were
synthesized according to Woo et al.15
11

The cloning and purification of His-tagged DXR (H-DXR) of E. coli and M. smegmatis were
previously described.11,20 DXP was prepared according to Meyer et al.36
4.2. Synthesis of phosphate derivatives 3-4
4.2.1. N-(Benzyloxy)-3-hydroxypropanamide (6a)
A stirred suspension of O-benzylhydroxylamine hydrochloride (2.63 g, 16 mmol) in dry THF
(20 mL) at -78 °C was treated with a 1 M solution of LiHMDS (65 mL, 65 mmol). After 1 h,
a solution of β-propiolactone 5 (1 mL, 16 mmol) in dry THF (2 mL) was added dropwise at -
78 °C. The reaction mixture was stirred at room temperature overnight. The reaction mixture
was cooled at -78 °C then a saturated aqueous solution of NH4Cl (50 mL) was slowly added
and warmed to room temperature. The aqueous layer was extracted with ethyl acetate (3x70
mL), saturated with NaCl and extracted with diethyl ether (3x100 mL). The organic layers
were collected and dried over Na2SO4, filtered and concentrated under reduced pressure. The
crude product was dissolved in THF and treated with TBAF.3H2O (7,65 g, 24 mmol). After
total consumption of starting material, the solvent was removed under reduced pressure. The
crude product was purified by flash chromatography (EtOAc ꢀ EtOAc/MeOH 90:10) to give
the compound 6a (2,49 g, 80 %) as a white solid and a mixture of two conformers Z and E in
a 7:3 ratio respectively. Rf = 0.19 (EtOAc/petroleum ether, 7:3).1H NMR (500 MHz, CDCl3):
δ(ppm) = 2.33 (7/10 of 2H, bs, CH2CO), 2.50 (3/10 of 2H, bs, CH2CO), 2.63 (7/10 of 1H, bs,
OH), 2,81 (3/10 of 1H, bs, OH), 3.86 (2H, td, 3J = 5.5 Hz, 3J = 5.5 Hz, CH2OH), 4.83 (3/10 of
2H, bs, OCH2Ph), 4.94 (7/10 of 2H, bs, OCH2Ph), 7.39 (5H, s, CHAr), 7.87 (3/10 of 1H, bs,
NH), 8.31 (7/10 of 1H, bs, NH).13C NMR (125.8 MHz, CDCl3): δ(ppm) = 34.0 (CH2CO),
35.7 (CH2CO), 58.1 (CH2OH), 58.7 (CH2OH), 78.5 (OCH2Ph), 79.7 (OCH2Ph), 128.9 (CHAr),
129.1 (CHAr), 129.5 (CHAr), 135.3 (CAr), 135.4 (CAr), 170.2 (CO). MS (EI)+ m/z calcd for
C10H13NO3Na [M+Na]+ : 218.08, found 218.08.
4.2.2. N-(Benzyloxy)-3-hydroxy-N-methylpropanamide (6b)
To a suspension of compound 6a (552 mg, 2.83 mmol), anhydrous K2CO3 (611 mg, 4.42
mmol) and anhydrous acetone (22 mL) was added iodomethane (0.90 mL, 14.5 mmol). The
resulting mixture was refluxed overnight. The suspension was filtered and acetone was
12

removed under reduced pressure. The resulting oil was dissolved in diethyl ether (25 mL),
washed with water (25 mL) and with a 10 % aqueous solution of Na2S2O3 (2 x 25 mL). The
aqueous layer was saturated with NaCl and extracted with diethyl ether (2 x 25 mL). The
organic layers were collected, dried over Na2SO4, filtered and evaporated. The crude product
was purified by flash chromatography (EtOAc/petroleum ether 7:3 ꢀ EtOAc) to give 6b as a
yellow oil (480 mg, 81 %) as a sole E conformer. Rf = 0.43 (EtOAc/petroleum ether, 7:3). 1H
3
3
NMR (400 MHz, CDCl3): δ(ppm) = 2.62 (2H, t, J = 5.2 Hz, CH2CO), 3.05 (1H, t, J = 6.0
Hz, OH), 3.22 (3H, s, NCH3), 3.82 (2H, q, 3J = 5.7 Hz, CH2OH), 4.84 (2H, s, OCH2Ph), 7.38
(5H, m, CHAr).13C NMR (125.8 MHz, CDCl3): δ(ppm) = 33.4 (NCH3), 34.4 (CH2CO), 58.5
(CH2OH), 76.5 (OCH2Ph), 128.9 (CHAr), 129.3 (CHAr), 129.5 (CHAr), 134.3 (CAr), 174.8
(CO). MS (EI)+: m/z calcd for C11H15NO3Na [M+Na]+ : 232.09, found 232.09.
4.2.3. General procedure for phosphorylation of the alcohol group
To a solution of alcohol (1 equiv.) in dry DCM (48 mL/mmol) was successively added 1H-
tetrazole (0.45 M, 2 equiv.) and dibenzyl N,N-diisopropylphosphoramidite (1.8 equiv.). After
the total consumption of the starting material (TLC), the reaction mixture was cooled at 0°C
then treated with 3-chloroperbenzoic acid (2.2 equiv.), which was slowly added as solid in
small portions. The reaction mixture was warmed to room temperature and stirred overnight.
DCM was added and the organic layer was washed with 10 % aqueous solution of Na2S2O5,
saturated aqueous solution of NaHCO3, and brine. The organic layer was dried over Na2SO4,
filtered and evaporated. The crude product was purified by flash chromatography.
4.2.3.1. Dibenzyl (3-((benzyloxy)amino)-3-oxopropyl) phosphate (7a). After purification
by flash chromatography (EtOAc/petroleum ether 9:1 to EtOAc) the product was obtained as
a colorless oil (450 mg, 60 %) and as a mixture of two conformers Z and E in a 8:2 ratio
respectively. Rf = 0.43 (EtOAc/petroleum ether, 9:1) 1H NMR (500 MHz, CDCl3): δ(ppm) =
2.32 (8/10 of 2H, bs, CH2CO), 2.70 (2/10 of 2H, bs, , CH2CO), 4.25 (2H, bs, CH2OP), 4.86
(2H, bs, OCH2Ph), 4.99-5.06 (4H, m, POCH2Ph), 7.33-7.35 (15H, m, CHAr), 7.72 (2/10 of
1H, bs, NH), 8.342 (8/10 of 1H, bs, NH). 13C NMR (125.8 MHz, CDCl3): δ(ppm) = 34.7
(CH2CO), 62.9 (CH2OP), 63.9 (CH2OP), 69.5 (POCH2Ph), 69.7 (POCH2Ph), 78.4 (OCH2Ph),
79.6 (OCH2Ph), 128.3 (CHAr), 128.8 (CHAr), 129.4 (CHAr), 135.4 (CAr), 135.7 (CAr), 167.5
(CO).31P NMR (121.5 MHz, CDCl3): δ (ppm) = -1.07, -1.13 HR MS (EI)+: m/ calcd for
C24H26NPO6Na [M+Na]+ : 478.1390, found 478.1339.
13

4.2.3.2. Dibenzyl (3-((benzyloxy)(methyl)amino)-3-oxopropyl) phosphate (7b). After
purification by flash chromatography (EtOAc/petroleum ether, 5:5 to EtOAc/Petroleum ether
8:2) the product was obtained as a colorless oil (666 mg, 85 %) as the sole conformer E. Rf =
3
0.35 (EtOAc/petroleum ethe,r 5:5).1H NMR (300 MHz, CDCl3): δ(ppm) = 2.73 (2H, t, J =
3
3
6.6 Hz, CH2CO), 3.18 (3H, s, NCH3), 4.31 (2H, td, J = 6.7 Hz, J = 6.7 Hz, CH2OP), 4.78
2
2
(2H, s, OCH2Ph), 5.03 (2H, dd, J = 12.4 Hz, 3JP-H = 7.9 Hz, POCH2Ph), 5.06 (2H, dd, J =
12.4 Hz, 3JP-H = 7.9 Hz, POCH2Ph), 7.31-738 (15H, m, CHAr).13C NMR (125.8 MHz, CDCl3):
δ(ppm) = 33.1 (d, 3J = 6.5 Hz, CH2CO), 33.6 (NCH3), 63.5 (d, 2J = 5.3 Hz, CH2OP), 69.4 (d,
2J = 5.3 Hz, POCH2Ph), 76.6 (OCH2Ph), 128.2 (CHAr), 128.6 (CHAr), 128.7 (CHAr), 128.9
3
(CHAr), 129.3 (CHAr), 129.5 (CHAr), 134.4 (CAr), 136.0 (d, J = 7.0 Hz, CArCH2OP), 171.6
(CO).31P NMR (121.5 MHz, CDCl3): δ (ppm) = -1.18 HR MS (EI)+: m/z calcd for
C25H28NPO6Na [M+Na]+ : 492.1546, found 492.1522.
4.2.4. General procedure for catalytic hydrogenolysis:
The protected alcohol (1 equiv.), palladium on charcoal (10 % mol) in absolute ethanol (22
mL/mmol) was hydrogenolyzed at room temperature and at atmospheric pressure under an
atmosphere of H2. After the total consumption of starting material (TLC) the catalyst was
removed by filtration over celite and the reaction mixture was evaporated to dryness to give
the desired compound without purification.
4.2.4.1. 3-(Hydroxyamino)-3-oxopropyl dihydrogen phosphate (3b). The product was
obtained without purification, as colorless oil, (19 mg, 99 %) and as a sole conformer Z. 1H
NMR (500 MHz, CD3OD): δ(ppm) = 2.45 (2H, t, 3J = 5.7 Hz, CH2CO), 4.21 (2H, td, 3J = 5.9
3
3
Hz, J = 5.9 Hz, CH2OP). 13C NMR (125.8 MHz, CD3eOD): δ(ppm) = 35.2 (d, J = 7.8 Hz,
CH2CO), 63.6 (d, 2J = 4.2 Hz, CH2OP), 169.8 (CO). 31P NMR (121.5 MHz, CD3OD): δ(ppm)
= 0.0. HR MS (EI)+: m/z calcd for C3H7NPO6 [M-H]+ : 184.0005, found 184.0011.
4.2.4.2. 3-(Hydroxy(methyl)amino)-3-oxopropyl dihydrogen phosphate (4b). The product
was obtained without purification, as colorless oil, (11.5 mg, 56 %) and as a sole conformer E.
3
1H NMR (500 MHz, CD3OD): δ (ppm) = 2.87 (2H, t, J = 6.2 Hz, CH2CO), 3.21 (3H, s,
3
3
NCH3), 4.23 (2H, td, J = 6.2 Hz, J = 6.1 Hz, CH2OP). 13C NMR (125.8 MHz, CD3OD): δ
3
2
(ppm) = 34.4 (d, J = 7.0 Hz, CH2CO), 36.3 (NCH3), 63.5 (d, J = 5.2 Hz, CH2OP), 172.8
14

(CO). 31P NMR (121.5 MHz, CD3OD): δ (ppm) = 0.24. HR MS (EI)+: m/z calcd for
C4H9NPO6 [M-H]+ : 198.0162, found 198.0169.
4.3. His-tagged DXR activity11
The assays were performed at 37°C in a 50 mM Tris/HCl buffer pH 7.5 containing 3 mM
MgCl2 and 2 mM DTT. The concentrations of DXP and NADPH were 480 µM and 160 µM
respectively. The decrease of absorbance at 340 nm due to NADPH oxidation was monitored
to determine the initial rates. The retained values were the average of at least two
measurements. The relative average deviation must be lower than 4%.
4.4. Inhibition of His-tagged DXR
H-DXR was pre-incubated during 2 min in the presence of the inhibitors (1-4) at different
concentrations and NADPH. DXP was then added to measure the residual activity. The
inhibitory potential of the tested compounds was quantified by determining the IC50 values.
They were obtained by plotting the percentage of residual activity versus the Log of inhibitor
concentration.
4.5. Bacterial growth inhibition
The antimicrobial activity of each inhibitor of H-DXR was tested by the paper disc diffusion
method on Escherichia coli XL1 Blue growing in a LB-agar medium. The paper discs (6 mm)
were impregnated with the different compounds at indicated amount. The Petri dishes were
incubated overnight at 37°C.
4.6. Sequence of hexose 6-phosphate transporter gene (UhpT) of E. coli XL1Blue WT
and FosR37
4.6.1. Genomic DNA preparation. Two overnight cultures (3 mL) of E. coli XL1-Blue WT
and FosR were centrifuged in 1.5 mL Eppendorf tubes. The pellets were washed with
autoclaved H2O (100 µL). After centrifugation and elimination of the supernatants, the pellets
were re-suspended in H2O (100 µL) and the suspensions were heated for 4 min in a boiling
water bath. After centrifugation the supernatants containing the genomic DNA were taken off
and stored -20°C until PCR processing.
15

4.6.2. PCR and sequence determination. The amplification of whole and partial UhpT gene
was performed by PCR using the High Fidelity PCR Master Kit (Roche) and several
combinations of primers (F1-R1, F1-R2, F2-R1). The list of primers is given in Table S1. An
aliquot of the DNA solution was diluted 50 fold and 2 to 6 µL of that dilution/50 µL of PCR
reaction mixture was tested for amplification. The conditions for thermal cycling were as
follows: denaturation of the target DNA at 94 °C for 5 min followed by 30 cycles at 94 °C for
1 min, primer annealing at 57 °C for 1 min, and primer extension at 68 °C for 1 min. The
reaction mixture was held at 68 °C for 7 min and then cooled to 4 °C. The PCR fragments
were purified after agarose electrophoresis using the NucleoSpin clean-up kit from Macherey-
Nagel (Germany). Their sequences were determined by the IBMP DNA sequencing platform
(Strasbourg, France).
Acknowledgements
M. M. acknowledges financial support from the ‘Ministère de la Recherche’ and from the
Foundation the ‘Frontier Research in Chemistry’ (Strasbourg, France). This work was
supported by a grant from the ‘Agence Nationale de la Recherche’ (Grant Nb ANR-13-BS07-
0013-01). The authors express their gratitude to Dr. L. Allouche, B. Vincent and M. Coppe
for NMR measurements.
Supplementary data
References and notes
1
Rohmer, M.; Grosdemange-Billiard, C.; Seemann, M.; Tritsch, D. Curr. Opin.
Investig. Drugs. 2004, 5, 154–162.
2
Kuzuyama, T.; Takahashi, S.; Watanabe, H.; Seto, H. Tetrahedron Lett. 1998, 39,
4509–4512.
3
4
Proteau, P. J. Bioorg. Chem. 2004, 32, 483–493.
Murkin, A. S.; Manning, K. A.; Kholodar, S. A. Bioorg. Chem. 2014, 57, 171–
185.
5
6
Okuhara, M.; Kuroda, Y.; Goto, T.; Okamoto, M.; Terano, H.; Kohsaka, M.;
Aoki, H.; Imanaka, H. J. Antibiot. 1980, 33, 13–17.
Okuhara, M.; Kuroda, Y.; Goto, T.; Okamoto, M.; Terano, H.; Kohsaka, M.;
Aoki, H.; Imanaka, H. J. Antibiot. 1980, 33, 24–28.
16

7
8
9
Y. Kuroda, M. Okuhara, T. Goto, M. Okamoto, H. Terano, M. Kohsaka, H. Aoki.
Imanaka, H. J Antibiot. 1980, 33, 29–35.
Kuzuyama, T.; Shimizu, T.; Takahashi, S.; Seto, H. Tetrahedron Lett. 1998, 39,
7913–7916.
Armstrong, C. M.; Meyers, D. J.; Imlay, L. S.; Meyers, C. F.; Odom, A. R.
Antimicrob. Agents Chemother. 2015, 59, 5511–5519.
10 Hemmerlin, A.; Tritsch, D.; Hammann, P.; Rohmer, M.; Bach, T. J. Biochimie.
2014, 99, 54–62.
11 Kuntz, L.; Tritsch, D.; Grosdemange-Billiard, C.; Hemmerlin, A.; Willem, A.;
Bach, T. J.; Rohmer, M. Biochem. J. 2005, 386, 127–135.
12 Steinbacher, S.; Kaiser, J.; Eisenreich, W.; Huber, R.; Bacher, A.; Rohdich, F. J.
Biol. Chem. 2003, 278, 18401–18407.
13 Mac Sweeney, A.; Lange, R.; Fernandes, R. P. M.; Schulz, H.; Dale, G. E.;
Douangamath, A.; Proteau, P. J.; Oefner, C. J. Mol. Biol. 2005, 345, 115–127.
14 Katayama, N.; Tsubotani, S.; Nozaki, Y.; Harada, S.; Ono, H. J. Antibiot. 1990,
43, 238–246.
15 Woo, Y.-H.; Fernandes, R. P. M.; Proteau, P. J. Bioorg. Med. Chem. 2006, 14,
2375–2385.
16 Gissot, A.; Volonterio, A.; Zanda, M. J. Org. Chem. 2005, 70, 6925–6928.
17 Brown, D. A.; Glass, W. K.; Mageswaran, R.; Girmay, B. Magn. Reson. Chem.
1988, 26, 970–973.
18 Brown, D. A.; Glass, W. K.; Mageswaran, R.; Mohammed, S. A. Magn. Reson.
Chem. 1991, 29, 40–45.
19 Mansell, D.; Veiga, N.; Torres, J.; Etchells, L. L.; Bryce, R. A.; Kremer, C.;
Freeman, C. Tetrahedron. 2010, 66, 8949–8957.
20 Ponaire, S.; Zinglé, C.; Tritsch, D.; Grosdemange-Billiard, C.; Rohmer, M. Eur. J.
Med. Chem. 2012, 51, 277–285.
21 Meyer, O.; Grosdemange-Billiard, C.; Tritsch, D.; Rohmer, M. Org. Biomol.
Chem. 2003, 1, 4367–4372.
22 Fonvielle, M.; Mariano, S.; Therisod, M. Bioorg. Med. Chem. Lett. 2005, 15,
2906–2909.
23 Nieschalk, J.; Batsanov, A. S.; O’Hagan, D.; Howard, J. Tetrahedron. 1996, 52,
165–176.
24 Haemers, T.; Wiesner, J.; Giessmann, D.; Verbrugghen, T.; Hillaert, U.; Ortmann,
17

R.; Jomaa, H.; Link, A.; Schlitzer, M.; Van Calenbergh, S. Bioorg. Med. Chem.
2008, 16, 3361–3371.
25 Fernandes, R. P. M.; Phaosiri, C.; Proteau, P. J. Arch. Biochem. Biophys. 2005,
444, 159–164.
26 Kholodar, S. A.; Tombline, G.; Liu, J.; Tan, Z.; Allen, C. L.; Gulick, A. M.;
Murkin, A. S. Biochemistry. 2014, 53, 3423–3431.
27 Chofor R.; Sooriyaarachchi, S.; Risseeuw, M. D. P.; Bergfors, T.; Pouyez, J.;
Johny, C.; Haymond, A.; Everaert, A.; Dowd, C. S.; Maes, L.; Coenye, T.; Alex,
A.; Couch, R. D.; Jones, T. A.; Wouters, J.; Mowbray, S. L.; Van Calenbergh, S.
J. Med. Chem. 2015, 58, 2988–3001.
28 Björkelid, C.; Bergfors, T.; Unge, T.; Mowbray, S. L.; Jones, T. A. Acta Cryst. D.
2012, 68, 134–143.
29 Lemieux, M. J.; Huang Y.; Wang, D. N. Res. Microbiol. 2004, 155, 623–629.
30 Sakamoto, Y.; Furukawa, S.; Ogihara, H.; Yamasaki, M. Biosci. Biotechn. Bioch.
2003, 67, 2030–2033.
31 Island, M. D.; Wei, B. Y.; Kadner, R. J. J Bacteriol. 1992, 174, 2754–2762.
32 Tsuruoka, T.; Yamada, Y. J. Antibiot. 1975, 28, 906–911.
33 Elvin, C. M.; Hardy, C. M.; Rosenberg, H. J. Bacteriol. 1985, 161, 1054–1058.
34 Kojo, H.; Shigi, Y.; Nishida, M. J. Antibiot. 1980, 33, 44–48.
35 Brown A. C.; Parish, T. BMC Microbiol. 2008, 8, 78–86.
36 Meyer, O.; Hoeffler, J.-F.; Grosdemange-Billiard, C.; Rohmer, M. Tetrahedron.
2004, 60, 12153–12162.
37 Friedrich M. J.; Kadner, R. J. J. Bacteriol. 1987, 169, 3556–3563.
18

Graphical abstract
Inhibitory activity against E. coli and M. smegmatis DXRs
19