Growth inhibition of Mycobacterium smegmatis by prodrugs of deoxyxylulose phosphate reducto-isomerase inhibitors, promising anti-mycobacterial agents

Article abstract of DOI:10.1016/j.ejmech.2012.02.031

Since Mycobacterium tuberculosis sets up several multiple anti-tuberculosis drug resistance mechanisms, development of new drugs with innovative target is urgent. The methylerythritol phosphate pathway (MEP) involved in the biosynthesis of essential metab

Full text of DOI:10.1016/j.ejmech.2012.02.031

European Journal of Medicinal Chemistry 51 (2012) 277e285
Contents lists available at SciVerse ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Short communication
Growth inhibition of Mycobacterium smegmatis by prodrugs of deoxyxylulose
phosphate reducto-isomerase inhibitors, promising anti-mycobacterial agents
*
*
Sarah Ponaire, Catherine Zinglé, Denis Tritsch, Catherine Grosdemange-Billiard , Michel Rohmer
Université de Strasbourg/CNRS, UMR 7177, Institut Le Bel, 4 rue Blaise Pascal, 67081 Strasbourg, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Since Mycobacterium tuberculosis sets up several multiple anti-tuberculosis drug resistance mechanisms,
development of new drugs with innovative target is urgent. The methylerythritol phosphate pathway
(MEP) involved in the biosynthesis of essential metabolites for the survival of mycobacteria, represents
such a target. Fosmidomycin 1a and FR900098 1b, two inhibitors of DXR, do not affect the viability of M.
tuberculosis cells, due to a lack of uptake. To overcome the absence of the mycobacterial cell wall crossing
of these compounds, we synthesized and tested the inhibition potency of acyloxymethyl phosphonate
esters as prodrugs of fosmidomycin 1a, FR900098 1b and their analogs 2a and 2b on Mycobacterium
smegmatis. Only the prodrugs 4be6b inhibit the bacterial growth and could be effective anti-
mycobacterial agents.
Received 2 February 2012
Received in revised form
15 February 2012
Accepted 16 February 2012
Available online 25 February 2012
Keywords:
Isoprenoid biosynthesis
Prodrug synthesis
Deoxyxylulose 5-phosphate reducto-
isomerase
Ó 2012 Elsevier Masson SAS. All rights reserved.
Mycobacterium
1. Introduction
Isopentenyl diphosphate (IPP) and dimethylallyl diphosphate
(DMAPP), the universal precursors of all isoprenoids are synthe-
Tuberculosis is one of the major infectious diseases: around nine
million new cases are annually estimated with approaching two
million human deaths [1,2]. Mycobacterium tuberculosis, respon-
sible for tuberculosis, owing to its unique cell wall organization, can
be considered as a fortress. The development of new drugs to fight
this bacterium is a difficult challenge. The first problem encoun-
tered by anti-bacterial agents is the crossing of the cell wall. In
addition, this Mycobacterium sets up several mechanisms of resis-
tance to major anti-tuberculosis drugs [3]. As the continuing
increase of tuberculosis risk is in large part due to the development
of antibiotic resistant strains, it is urgent to find other targets for the
development of new anti-tuberculosis drugs than those presently
used. Isoprenoid biosynthesis in Mycobacterium species represents
such a target in this context. Isoprenoids are found in all living
organisms. A number of them including bactoprenyl diphosphate
required for the biosynthesis of peptidoglycan, a major cell wall
component [4], and the prenyl side-chains of menaquinones
involved in electron transport chains is present in mycobacteria
and represents essential metabolites for their survival [5].
sized via two pathways: the mevalonate pathway, which is present
in animals and humans, fungi and the cytosol of plant cells [6], and
the 2-C-methyl-
D-erythritol 4-phosphate (MEP) pathway, which
has been found in plant plastids, unicellular green algae, apicom-
plexan parasites and in most bacteria including many pathogens
such as Mycobacterium spp [7].
The determination of the nucleotide sequences of the genomic
DNA of Mycobacterium spp. confirmed the presence of the sole MEP
pathway in this genus of bacteria. Accordingly, all enzymes of the
MEP pathway can be considered as potential targets for anti-
tuberculosis drug development [8]. The knowledge of the DNA
sequence also allows obtaining and studying the recombinant
enzymes of this pathway.
The 1-deoxy-D-xylulose 5-phosphate reducto-isomerase (DXR),
the second enzyme of the MEP pathway [9] (Scheme 1) is inhibited
by fosmidomycin (1a) and FR900098 (1b), two natural antibiotics
[10], (Scheme 2) which consequently block the growth of many
bacteria possessing the MEP pathway. Owing to the rapid emer-
gence of resistance observed in the case of fosmidomycin and the
fast elimination of the antibiotic in the urine [11], several groups
attempted to synthesize other DXR inhibitors with improved
pharmacokinetic and pharmacodynamic properties. The structures
of these new molecules are all inspired from those of the antibiotics
1a and 1b, possessing on the one hand a phosphonate or another
* Corresponding authors.
E-mail addresses: grosdemange@unistra.fr (C. Grosdemange-Billiard),
mirohmer@unistra.fr (M. Rohmer).
0223-5234/$ e see front matter Ó 2012 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2012.02.031

278
S. Ponaire et al. / European Journal of Medicinal Chemistry 51 (2012) 277e285
Scheme 1. Biosynthesis of IPP and DMAPP. Second step of MEP pathway: conversion of 1-deoxy-D-xylulose 5-phosphate into 4-methyl-D-erythritol 4-phosphate by the deoxy-
xylulose phosphate reducto-isomerase.
isosteric moiety fitting into the phosphate-recognition site and on
the other side a chelating group tightly binding the metal cation
required for enzyme activity [12].
In this context, we synthesized and tested on the DXR of
Escherichia coli two compounds (Scheme 2, 2a and 2b) where the
the cells, stop the bacterial growth by inhibition of the DXR. The fact
that such prodrugs of FR900098 are able to enhance in vivo and
in vitro anti-malarial activity strengthens the relevance of this
approach [19]. In this work, we synthesized and tested the inhibi-
tion potency of acyloxymethyl phosphonate esters as prodrugs of
fosmidomycin, FR900098 and their analogs 2a and 2b on the non-
pathogenic, fast growing Mycobacterium smegmatis. It is expected
that the parent compounds are released from such prodrugs
intracellularly in two steps: the hydrolysis of the acyl group by an
esterase is followed by a rapid and spontaneous hydrolysis of the
resulting hemiacetal into the free inhibitor and formaldehyde [20].
Finally, the DXR of M. smegmatis was cloned to check whether this
enzyme is inhibited by fosmidomycin and its analogs. Such lipo-
philic fosmidomycin and FR900098 prodrugs have been recently
described and tested as anti-bacterial agents [21].
metal cation-binding moiety is
a hydroxamic acid group.
Compound 2b inhibits the E. coli DXR much like fosmidomycin and
has significant anti-bacterial activity against wild type E. coli and
even against an E. coli mutant resistant to fosmidomycin [13]. In the
case of recombinant M. tuberculosis DXR, fosmidomycin 1a and
FR900098 1b were shown to inhibit the enzyme. The two published
IC50 values differed, probably depending on different experimental
conditions: 80 nM and 310 nM for 1a [11,14] and 160 nM for 1b [15].
Fosmidomycin does not inhibit the growth or affect the viability of
M. tuberculosis cells. This lack of effect was interpreted in terms of
a lack of uptake of the compound by the bacterium [16]. This fact is
not too surprising. The hydrophobic cell wall of mycobacteria is an
efficient barrier that prevents passive transport and makes the
bacteria naturally resistant to most antibiotics. Mycobacteria cells
are surrounded by a hydrophobic, waxy cell wall. Accordingly,
hydrophilic compounds cross slowly this cell wall, probably via
porins that are rare in mycobacteria, whereas lipophilic molecules
penetrate by diffusion through the cell wall [17]. For that reason,
anti-bacterial agents with lipophilic characteristics are more active
against mycobacteria. In this context, less polar FR900098 analogs
in which either the phosphonate or the hydroxamate moiety have
been replaced by alternative acidic or metal coordinating groups
have been prepared, but had no effect on M. tuberculosis growth
[15,18].
2. Results
2.1. Synthesis of the prodrugs
The propyloxymethyl ester prodrugs 3a and 3b of fosmidomycin
(1a) and FR900098 (1b) were synthesized according to a previously
described method [19,22]. The synthesis of the prodrugs 4e6 was
achieved starting from the phosphonate 8, which is readily acces-
sible by nucleophilic substitution of the commercially available
ethyl 4-bromobutyrate with the NaH generated anion of diethyl
phosphite (Scheme 3) [1
2]. The O-benzyl hydroxamate derivative
9a was obtained in one step from 8 using O-benzylhydroxylamine
hydrochloride in presence of LiHMDS as described by Gissot et al.
[23]. The methyl group was introduced by reaction of 9a with NaH
followed by addition of methyl iodide. Deprotection of the phos-
phonate group using bromotrimethylsilane led to the acid 10 and
subsequent alkylation of the crude phosphonic acids with the
appropriate chloromethyl esters.
A way to overcome the absence of mycobacterial cell wall
crossing of fosmidomycin and its analogs is to mask the polar
phosphonate moiety with hydrophobic groups. Owing to their
hydrophobic character, such prodrugs may penetrate in the
bacterial cells and, after enzymic conversion into the parent drug in
Scheme 2. Chemical structures of the tested compounds. Fosmidomycin 1a, FR900098 1b, phosphono-hydroxamic acids 2a and 2b and their acyloxymethyl phosphonate esters
prodrugs 3, 4, 5 and 6 (a: R1 ¼ H; b: R1 ¼ CH3).

S. Ponaire et al. / European Journal of Medicinal Chemistry 51 (2012) 277e285
279
O
O
P
O
O
P
OEt
R1
d
O
b
a
N
OEt
Br
OEt
EtO
OEt
EtO
OBn
R1= H 9a
8
c
R1= Me 9b
O
O
P
O
O
O
P
O
O
P
O
e
R1
R1
f
R1
R2
2
N
O
O
R2
2
N
O
O
N
OH
OH
OH
OBn
OBn
R1= H 10a
R2= n-Pr
R = t-Bu
R2= n-Pr
11a-b
12a-b
4a-b
2
R1= Me 10b
R2= t-Bu 5a-b
R2= Ph 6a-b
R2= Ph 13a-b
Scheme 3. Synthesis of prodrugs 4, 5 and 6 (a: R1 ¼ H; b: R1 ¼ CH3). Reagents and conditions. a) (EtO)2P(O)H, NaH, THF, 84%; b) BnONH2$HCl, LiHMDS, THF, ꢀ78 ꢁC, 69%; c) MeI,
NaH, THF, 85%; d) 1) TMSBr, CH2Cl2, 0 ꢁC; 2) H2O, THF; e) chloroacyloxyalkyl/arylester, Et3N, DMF, 70 ꢁC, 11a (26%), 11b (45%), 12 (21%), 12b (39%), 13a (35%), 13b (40%); f) H2, Pd/C,
EtOH, 4a (95%), 4b (94%), 5a (94%), 5b (92%), 6a (96%), 6b (93%).
Benzyl deprotection was performed by catalytic hydrogenolysis
with palladium over charcoal at atmospheric pressure and room
temperature. Prodrugs 4e6 were obtained from the O-benzylated
precursors 11e13 in quantitative yields and required no further
purification.
with His-tagged DXR of E. coli [13], fosmidomycin and compound
2b inhibit the DXR of M. smegmatis at about the same level.
FR900098 seems to be a little more efficient while 2a, as observed
with the enzyme of E. coli, is substantially less effective. The
comparison of IC50 values is delicate as they depend on the
experimental conditions: concentration of the substrates, pre-
incubation or not of the inhibitor with the enzyme, nature of the
cation (Mg2þ or Mn2þ), use of His-tagged enzyme or not. In spite of
these restrictions, the DXR of M. smegmatis seems to have less
affinity for the tested inhibitors than the E. coli enzyme as the IC50
values are about tenfold higher than those reported for the E. coli
DXR. Concerning the DXR of M. tuberculosis, IC50 values of 80 nM
[11] and 310 nM [14] are given for fosmidomycin. Compounds 3b
and 4b were tested on DXR of E. coli and M. smegmatis but no
2.2. M. smegmatis DXR sequence
The amino acid sequence obtained from the nucleotide
sequence of the DXR from M. smegmatis DSM43756 (ATCC 43756)
was compared to the known sequence of the enzyme from
M. smegmatis ATCC 700084 (Swiss-Prot entry A0QVH7) (Fig. S1 in
supporting information). Two mutations were detected: Ala-282
and Asp-310 (numbering of DXR from M. smegmatis MC2 155
ATCC 700084) were respectively replaced by Pro and Glu. All
sequenced clones had the two mutations, indicating that they are
effective in the presently investigated strain. The sequence identity
with the DXR from M. tuberculosis (Swiss-Prot entry P64012) is
around 76%.
inhibition was found for a 1 mM concentration.
2.4. Growth inhibition of M. smegmatis with prodrugs (3e6)
Growth inhibition of M. smegmatis induced by the synthesized
prodrugs (3e6) was evaluated by the paper disc diffusion method.
Isoniazid is an effective anti-tuberculosis agent inhibiting the
biosynthesis of mycolic acids, which are essential component of the
mycobacterial cell wall. It is considered as a prodrug that requires
activation by an endogenous catalase peroxidase for inhibiting the
enoyleacyl carrier protein [27]. This anti-bacterial agent was
accordingly utilized as reference compound for a positive M.
smegmatis growth inhibition test.
As previously mentioned in the case of M. tuberculosis [10,14,15]
the free DXR inhibitors fosmidomycin 1a and FR900098 1b do not
block the growth of M. smegmatis. The same behavior was observed
in the presence of the synthetic DXR inhibitors 2a and 2b.
2.3. Kinetic properties of recombinant His-tagged DXR of
M. smegmatis
As observed by others in the case of DXR from M. tuberculosis,
the enzyme of M. smegmatis loses its activity during the dialysis
step when glycerol and NaCl are omitted in the incubation buffer.
The Km for DXP of the M. smegmatis DXR, determined with MgCl2 as
metal cofactor, is 212
the His-tagged DXR from E. coli [13]. Different values of Km for the
DXR of M. tuberculosis are reported in the literature: 42 M [24] for
the native enzyme, 47 M [11] and 340 M [25] for His-tagged
mM, about two fold higher than that found for
m
m
m
enzymes. The discrepancy may be due to the assay conditions
and/or to the instability of the enzyme.
As fosmidomycin and its derivatives are slow-binding inhibitors
[13,26], the residual activities were measured after a pre-incuba-
tion of the DXR in presence of the inhibitors during 2 min. The
enzymatic reaction was initiated by addition of DXP.
The IC50 values (mean values from two independent determi-
nations) were determined by plotting the residual activity versus
the Log of the concentration of inhibitor. Initial rates of the enzy-
matic reaction were used to determine the IC50 values for fosmi-
domycin, FR900098 and inhibitors 2a and 2b (Table 1). As shown
Table 1
Inhibition of Mycobacterium smegmatis DXR. Assays were performed in 50 mM Tris/
HCl buffer (pH 7.5) containing 3 mM MgCl2 and 2 mM DTT. The concentrations of
NADPH and DXP were 0.15 and 0.5 mM respectively. The enzymatic reaction was
initiated by adding DXP after pre-incubation of the enzyme with the inhibitor in the
presence of NADPH for 2 min at 37 ꢁC.
Inhibitors
IC50 (nM)
Fosmidomycin 1a
FR900098 1b
2a
2b
510 ꢂ 30
320 ꢂ 20
1480 ꢂ 40
410 ꢂ 30

280
S. Ponaire et al. / European Journal of Medicinal Chemistry 51 (2012) 277e285
All these DXR inhibitors are most probably unable to cross the
[13]. Fosmidomycin is carried into the E. coli cell via the transport
system of glycerophosphate and at a lower level of glucose 6-
phosphate [30]. In the absence of such transporters in mycobac-
teria, none of the DXR inhibitors is capable of penetrating into the
cell or at a too low concentration. A fosmidomycin resistant E. coli
strain was obtained by mutation of the glycerophosphate trans-
porter. It remained, however, sensitive to the DXR inhibitor 2b,
suggesting that this compound may be transported into the cell
via another transporter (A. Hemmerlin, D. Tritsch, unpublished
results). The mycobacterial cell wall is an extremely efficient
impermeable barrier protecting the cell from passive diffusion of
toxic compounds such as antibiotics. It is responsible for the natural
resistance of mycobacteria toward most common antibiotics and
chemotherapeutic agents [31]. Small hydrophilic compounds
diffuse through porins, which are water-filled protein channels
[32]. The permeation of hydrophilic compounds is low due to the
limited number of porins in the mycobacterial outer membrane and
to their longer size as compared to those of other bacteria. A
hydrophilic free DXR inhibitor is thus probably unable to penetrate
into the cell through those porins or only at such low rates that its
intracellular concentration is too low to efficiently inhibit the DXR.
This permeation problem encountered with small hydrophilic
enzyme inhibitors may be overcome by converting them into
lipophilic prodrugs. Acyloxymethyl phosphonate esters represent
interesting candidates. Once they have penetrated into the cells,
hydrolysis of the ester group of the phosphonate masking group by
an endogenous esterase releases a hemiketal, which cleaves
spontaneously into the free DXR inhibitor and formaldehyde [20].
This concept is validated. The N-methylated prodrugs 4b, 5b and 6b
induced significant growth inhibition of M. smegmatis, with the
prodrug 4b with an n-propyl chain in the masking group being by
far the most powerful growth inhibitor. The prodrugs 5b and 6b
with respectively t-butyl and a phenyl group showed some growth
inhibition on a M. tuberculosis strain with minimal inhibiting
cell wall and were therefore converted into less polar and more
lipophilic acyloxymethyl phosphonate esters 3e6. The first
attempts were performed with the 2a and 2b derivatives 4e6. None
of the prodrugs 4a, 5a and 6a based on the non N-methylated
inhibitor 2a inhibited the bacterial growth when 400 nmol were
deposed on the disc. In contrast, in the presence of the N-methyl-
ated prodrugs 4b, 5b and 6b of inhibitor 2b, clear growth inhibition
zones were observed around the impregnated paper discs (Fig. S2
in supporting information). Like isoniazid, these compounds are
able to inhibit the bacterial growth. The diameters of the growth
inhibition zones depended on the nature of the phosphonate
masking acyl chains (Table 2).
The n-propyloxymethyl esters prodrug 4b being the most
effective, the similar prodrugs of fosmidomycin and FR900098
bearing the same n-propyl containing masking group, 3a and 3b
respectively, were synthesized. Surprisingly, neither the prodrug of
fosmidomycin 3a, nor that of FR900098 3b blocked the growth of
M. smegmatis while fosmidomycin 1a and FR900098 1b inhibited
like the hydroxamate 2b the DXR of this bacterium (Table 1).
2.5. Resazurin viability test
Bacterial viability can be estimated by a colorimetric resazurin
test performed in a liquid medium [28]. A color change from blue to
pink indicates reduction of resazurin and is correlated with
bacterial growth and viability of the cells. After an incubation of
15 min, a culture containing no inhibitor turned already pink. In the
presence of isoniazid at a 10
pink in less than 1 h. At a 20
m
M concentration, the color became
mM concentration and above, the blue
color lasted at least 5 h, but the color of the culture became after
12 h purplish, suggesting that some bacterial cells were perhaps
still surviving. In the presence of the n-propyl prodrug 4b, a similar
phenomenon was observed. A 250
had little influence on the growth of the bacteria, and the color
changed to pink within 15 min. From a minimal 500 M concen-
mM concentration of the prodrug
concentration in the 50e100 mg/mL and 25e100 mg/mL range [23].
Although rich in lipids, the mycobacterial cell wall represents also
a strong barrier for hydrophobic molecules. For penetration, such
molecules are expected to cross the cell wall through their lipid
domain, but the influx is limited by the low fluidity of the lipid
phase due to the presence of mycolic acids [33]. This low fluidity of
the lipid domain may explain the lower efficacy of prodrugs 5b and
6b with a bulky pivaloyl or a rigid phenyl group as compared to
compound 4b, which presents with its n-propyl chain the smallest
and most flexible phosphonate masking group.
m
tration, the growth is strongly inhibited. The color remained blue
for at least 5 h and tended after to become also slowly purplish. The
resazurin test confirms thus that the prodrug 4b is capable of
inhibiting the growth of M. smegmatis. It is, however, less efficient
than isoniazid on the growth of M. smegmatis as the minimal
inhibitory concentration (MIC) is between 250 and 500
contrast with the value between 10 and 20 M (or 1.4e2.8
for isoniazid, i.e. close to the value reported in the literature for
mM for 4b in
m
mg/mL)
The negative results obtained on the growth of M. smegmatis
with all other prodrugs are quite puzzling. The rather similar non-
methylated analogs 4a, 5a and 6a do not inhibit the growth of this
bacterium at the highest tested dose (400 nmol/disk). Due to the
structural similarity of the non-methylated prodrugs 4a, 5a and 6a
and their N-methylated homologs 4b, 5b and 6b, it is likely that the
non-methylated prodrugs penetrate in the bacterial cells and are
converted into the corresponding DXR inhibitor 2a. The absence of
growth inhibition with the non N-methylated compounds may
result from a too low intracellular concentration of the DXR
inhibitor 2a, which is in addition three to four fold less effective
than its N-methylated homolog 2b, to efficiently inhibit the DXR
and accordingly the bacterial growth. Surprisingly, the fosmido-
mycin and FR900098 prodrugs 3a and 3b had also no effect on the
growth of M. smegmatis at the highest tested dose (400 nmol/disk).
This is in sharp contrast with the anti-bacterial activity of prodrug
2b against M. smegmatis. The explanation for this discrepancy is not
obvious as solubility and diffusion are expected to be rather similar
for all prodrugs. The absence of growth inhibition may be ascribed
to a lack of penetration of compounds 3a and 3b. In this case, the
prodrugs may not cross the M. smegmatis cell wall by passive
M. tuberculosis (7 ꢂ 2
mg/mL) [29].
3. Discussion
Fosmidomycin 1a, FR900098 1b and the phosphono-
hydroxamate 2b are efficient inhibitors of the DXR of mycobac-
teria. They do not interfere with the growth of M. tuberculosis for
the first two ones [17] and of M. smegmatis for all three inhibitors,
whereas they easily inhibit the growth of the Gram-negative E. coli
Table 2
Growth inhibition of Mycobacterium smegmatis. The antimicrobial activity of isoni-
azid (INH) and DXR inhibitor prodrugs 4b, 5b and 6b was determined using the
paper disc diffusion method (see Fig. S2).
Tested compounds
Amount of
Diameter of the
growth inhibition zone (mm)
prodrug/disc (nmol)
Isoniazid
15
400
400
400
29
23
15
11
4b
5b
6b

S. Ponaire et al. / European Journal of Medicinal Chemistry 51 (2012) 277e285
281
diffusion, a more or less specific transporter allowing the influx of
prodrugs 4b, 5b and 6b. Another explanation could be that the
fosmidomycin and FR900098 prodrugs or their parent compounds
are expelled outside the cell by efflux pumps. Finally, the hypoth-
esis that fosmidomycin and FR900098, generated inside the cell,
are inactivated by the bacteria cannot be set aside [34]. At this stage
it cannot even be excluded that the free DXR inhibitor 2b has also
another target than the DXR, the inhibition of which would lead to
the arrest of bacterial growth.
When the prodrugs were tested on E. coli, none of them had
a noticeable anti-bacterial activity, even at the highest tested
amount (400 nmol/disk). Either the compounds most probably do
not cross the cell wall of this bacterium, or E. coli does not have the
esterase required to cleave the ester bond in the phosphonate
masking groups of the prodrugs.
Small hydrophilic DXR inhibitors do not affect the growth of M.
smegmatis. This problem was overcome with at least one of these
inhibitors 2b, which presents a reverse hydroxamic moiety as
compared to that of FR900098. Masking the polar phosphonate
group with acyloxymethyl esters resulted in prodrugs that effec-
tively inhibit the growth of M. smegmatis. This approach solves
apparently the problem of the impermeability of the mycobacterial
cell wall toward enzyme inhibitors and will be applied to the
growth inhibition of M. tuberculosis.
(300 MHz, CDCl3):
d
¼ 1.19 (t, 3H, J ¼ 7.1 Hz, OCH2CH3), 1,26 (t, 6H,
J ¼ 7.1 Hz, POCH2CH3), 1.69e1.89 (m, 4H, PCH2CH2), 2.33 (t, 2H,
J ¼ 7.2 Hz, CH2CO), 4.04 (m, 6H, OCH2, POCH2); 13C NMR (75.5 MHz,
CDCl3):
d
¼ 14.2, 16.3 (d, J ¼ 6.0 Hz), 18.1 (d, J ¼ 4.3 Hz), 24.9
(d, J ¼ 141.6 Hz), 34.4 (d, J ¼ 15.5 Hz), 60.4, 61.4 (d, J ¼ 6.4 Hz), 172.5;
31P NMR (121.5 MHz, CDCl3):
d
¼ 32.3; HRMS (EIþ) m/z calcd for
C10H21O5PNa [M þ Na]þ 275.1019, found 275.1025.
4.1.2. Diethyl 4-(benzyloxyamino)-4-oxobutylphosphonate (9a)
A 1 M solution of LiHMDS in THF (14 mL, 13.9 mmol) was added
to a solution of O-benzylhydroxylamine hydrochloride (2.2 g,
14 mmol) in THF (20 mL) at 0 ꢁC for 30 min. A solution of (8) (3.5 g,
13.9 mmol) in THF (10 mL) was then added dropwise to the cold
reaction mixture that was allowed to warm up at room tempera-
ture and stirred overnight. The reaction was quenched with
a saturated aqueous NH4Cl solution and extracted several times
with EtOAc. After evaporation to dryness of the combined organic
layers, the residue was purified by flash chromatography (EtOAc)
yielding (9a) as a pale yellow oil (3.0 g, 93.0 mmol, 65%, Rf ¼ 0.30,
EtOAc); 1H NMR (300 MHz, CDCl3):
d
¼ 1.28 (t, 6H, J ¼ 7.0 Hz, CH3),
1.66e1.98 (m, 4H, PCH2CH2), 2.20 (t, 2H, J ¼ 6.6 Hz, CH2CO),
3.94e4.13 (m, 4H, OCH2CH3), 4.84 (s, 2H, CH2Ph), 7.35 (m, 5H, Ph),
9.12 (br s, 1H, NH); 13C NMR (75.5 MHz, CDCl3):
d
¼ 16.4 (d,
J ¼ 6.0 Hz), 18.4 (d, J ¼ 4.5 Hz), 23.9 (d, J ¼ 140.5 Hz), 32.6 (d,
J ¼ 12.1 Hz), 61.7 (d, J ¼ 6.4 Hz), 78.1, 128.5, 128.6, 129.1, 135.3, 170.1;
4. Experimental section
31P NMR (121.5 MHz, CDCl3):
d
¼ 32.5; HRMS (EIþ) m/z calcd for
C15H24NO5PNa [M þ Na]þ 352.1284, found 352.1263.
All non-aqueous reactions were run in dry solvents under an
argon atmosphere. All reagents and solvents were reagent grade.
After solvent extraction of an aqueous phase, the organic layers
were combined, dried over anhydrous sodium sulfate, filtered and
evaporated to dryness. DXP was obtained by chemical synthesis as
previously described [35]. Fosmidomycin was obtained from Dr R. J.
Eilers (Monsanto, St Louis, MO, U.S.A.). Isoniazid (INH) was
purchased from Acros Organics. Flash chromatography was per-
formed on Merck silica gel 60 (230e400 mesh) with the solvent
system as indicated. TLC plates were revealed by spraying with an
ethanolic solution of p-anisaldehyde (2.5%), sulfuric acid (3.5%) and
acetic acid (1.6%) or with an ethanolic solution of phosphomolybdic
acid (20%) followed by heating. 1H NMR experiments were per-
formed in CDCl3, CD3OD, D2O or DMSO-d6 with CHCl3
4.1.3. Diethyl 4-[benzyloxy(methyl)amino]-4-oxobutylphosphonate
(9b)
To a solution of (9a) (450 mg, 1.4 mmol) in THF (26 mL) were
successively added at 0 ꢁC sodium hydride (36 mg, 1.5 mmol) and
methyl iodide (130 mL, 2.0 mmol). The mixture was stirred over-
night at room temperature, quenched with a saturated solution of
NH4Cl and extracted several times with AcOEt. After evaporation to
dryness, the residue was purified by flash chromatography (EtOAc)
yielding (9b) as a pale brownish oil (400 mg, 1.2 mmol, 85%,
Rf ¼ 0.17, EtOAc); 1H NMR (300 MHz, CDCl3):
d
¼ 1.21 (t, 3H,
J ¼ 7.5 Hz, CH3), 1,60e1.90 (m, 4H, PCH2CH2), 2.40 (t, 2H, J ¼ 6.5 Hz,
CH2CO), 3.10 (s, 3H, CH3N), 3.90e4.07 (m, 4H, CH2CH3), 4.74 (s, 2H,
CH2Ph), 7.30 (m, 5H, Ph); 13C NMR (75.5 MHz, CDCl3):
d
¼ 16.4
(
d
¼ 7.26 ppm), DHO (
d
¼ 4.65 ppm), CHD2OD (
d
¼ 3.31 ppm),
(d, J ¼ 6.0 Hz), 17.6 (d, J ¼ 4.7 Hz), 25.0 (d, J ¼ 140.8 Hz), 32.3
(d, J ¼ 15.9 Hz), 61.5 (d, J ¼ 6.4 Hz), 76.2, 128.7, 129.0, 129.3, 134.5,
CHD2SOCD3 (
CD3OD, D2O or DMSO-d6 with CDCl3
d
¼ 2.50 ppm) as internal reference, 13C NMR in CDCl3,
(
d
¼ 77.0 ppm), CD3OD
173.4; 31P NMR (121.5 MHz, CDCl3):
d
¼ 32.5; HRMS (EIþ) m/z calcd
(d
¼ 49.0 ppm), DMSO-d6 (
d
¼ 39.5 ppm) as internal reference, and
for C16H26NO5PNa [M þ Na]þ: 366.144, found 366.139.
31P NMR in CDCl3, CD3OD, DMSO-d6 with H3PO4
(
d
¼ 0 ppm) as
external reference. Most of the hydroxamates are present as two Z
and E conformers in equilibrium. If only one signal is described, it is
common to all conformers. The evaluation of the relative amount of
the conformers was made by integration of the N-CH3 and/or the
CH2eCON proton signals.
4.1.4. Diethyl 3-oxopropylphosphonate [19b]
A solution of diethyl(3,3-diethoxypropyl) phosphonate (1 g,
3.7 mmol) in an aqueous 2 M HCl solution (10 mL) was stirred at
room temperature overnight. The reaction mixture was extracted
several times with CHCl3. After evaporation to dryness, the residue
was purified by flash chromatography (EtOAc) affording diethyl
3-oxopropylphosphonate as a colorless oil (600 mg, 3.1 mmol, 83%,
4.1. Synthesis of prodrugs
Rf ¼ 0.15, EtOAc); 1H NMR (300 MHz, CDCl3):
d
¼ 1.29 (t, 3H,
4.1.1. Ethyl 4-(diethoxyphosphoryl)butanoate (8)
J ¼ 7.5 Hz, CH3), 1.95e2.06 (m, 2H, PCH2), 2.70e2.79 (m, 2H,
Sodium hydride (575 mg, 24 mmol) was added to a solution of
diethyl phosphite (3 g, 22 mmol) in THF (100 mL) at 0 ꢁC. The
reaction mixture was stirred for 20 min at room temperature. A
solution of ethyl bromobutyrate (4.70 mL, 33 mmol) was added
dropwise and the reaction mixture was stirred for 48 h. The reac-
tion was quenched with a saturated solution of NH4Cl and extracted
with diethyl ether (3ꢃ15 mL). The solvent was removed under
reduce pressure. The residue was purified by flash chromatography
(EtOAc/petroleum ether, 7:3) affording (8) as a colorless oil (4.6 g,
18 mmol, 84%, Rf ¼ 0.12, EtOAc/petroleum ether, 7:3); 1H NMR
CH2CO), 4.01e4.12 (m, 4H, CH2CH3), 9.76 (t, 1H, J ¼ 1.2 Hz, CHO); 13
C
NMR (75.5 MHz, CDCl3):
d
¼
16.4 (d,
J
¼
6.0 Hz), 18.0
(d, J ¼ 145.9 Hz), 36.9 (d, J ¼ 4.1 Hz), 61.8 (d, J ¼ 6.4 Hz), 199.2
(d, J ¼ 15.3 Hz); 31P NMR (121.5 MHz, CDCl3):
d
¼ 31.9; HRMS (EIþ)
m/z calcd for C7H15O4PNa [M þ Na]þ 217.0600, found 217.0577.
4.1.5. Diethyl 3-(benzyloxyamino)-propylphosphonate [19b]
To a solution of diethyl 3-oxopropylphosphonate (200 mg,
1.0 mmol) in MeOH (8 mL) were added at room temperature in one
portion O-benzylhydroxylamine hydrochloride (8.2 g, 52 mmol)

282
S. Ponaire et al. / European Journal of Medicinal Chemistry 51 (2012) 277e285
and sodium cyanoborohydride (650 mg, 10 mmol). The reaction
mixture was stirred overnight and then evaporated to dryness. The
crude was dissolved in H2O, the pH of the solution adjusted to
pH ¼ 10, and the solution extracted several times with CH2Cl2. After
evaporation to dryness, the residue was purified by flash chroma-
tography (EtOAc to EtOAc/MeOH 9:1) giving diethyl 3-(benzylox-
yamino)-propylphosphonate as a colorless oil (225 mg, 0.75 mmol,
at room temperature for 5 min. Acyloxymethyl chloride (10 eq.)
was added, and the mixture was stirred at 70 ꢁC for 5 h. Another
3 eq. of Et3N and 5 eq. of acyloxymethyl chloride were added and
the mixture was stirred overnight at room temperature. Et2O was
added to the reaction mixture and the solution was successively
washed with H2O, with a saturated solution of NaHCO3 and again
with H2O. The resulting organic layer was dried, filtered and
evaporated to dryness and the residue purified by flash chroma-
tography on silica gel (Et2O).
73%, Rf ¼ 0.20 EtOAc); 1H NMR (300 MHz, CDCl3):
d
¼ 1.20 (t, 3H,
J ¼ 7.5 Hz, CH3), 1.61e1.80 (m, 4H, PCH2CH2), 2.84 (t, 2H, J ¼ 7.5 Hz,
CH2NH), 3.91e4.02 (m, 4H, OCH2CH3), 4.57 (s, 2H, CH2Ph), 5.14 (br s,
1H, NH), 7.21e7.23 (m, 5H, Ph); 13C NMR (75.5 MHz, CDCl3):
d
¼ 16.3
4.1.9. Di(butyloxymethyl)[3-((N-benzyloxy)(N-formyl)amino)]
propylphosphonate
(d, J ¼ 6.2 Hz), 20.4 (d, J ¼ 5.0 Hz), 23.0 (d, J ¼ 142.1 Hz), 51.8
(d, J ¼ 16.2 Hz), 61.2 (d, J ¼ 6.2 Hz), 76.0,127.5, 127.6, 128.1, 137.7; 31P
Yellowish oil, 358 mg, 0.76 mmol, 70%, Rf ¼ 0.14, (Et2O); 1H NMR
NMR (121.5 MHz, CDCl3):
d
¼ 33.3; HRMS (EIþ) m/z calcd for
(300 MHz, DMSO-d6):
d
¼ 0.87 (t, 6H, J ¼ 7.5 Hz, CH3CH2), 1.54
C14H25NO4P [M þ H]þ 302.1516, found 302.1486.
(qui, 4H, J ¼ 7.5 Hz, CH2CH3), 1.70e1.90 (m, 4H, PCH2CH2), 2.34
(t, 4H, J ¼ 7.5 Hz, CH2CO), 3.53e3.61 (m, 2H, CH2N), 4.89 (s, 2H,
CH2Ph), 5.55 (dd, 2H, J ¼ 5.4, 9.3 Hz, OCH2O), 5.60 (dd, 2H, J ¼ 5.7,
9.9 Hz, OCH2O), 7.41 (m, 5H, Ph), 7.96 (br s, 2/5 of 1H, CHO), 8.22 (br
4.1.6. Diethyl 3-[(N-acetyl)(N-benzyloxy)amino]propylphosphonate
[19b]
To a solution of diethyl 3-(benzyloxyamino)-propylphosphonate
(200 mg, 0.7 mmol) in acetic anhydride (2.6 mL) was added drop-
s, 3/5 of 1H, CHO); 13C NMR (75.5 MHz, DMSO-d6):
d
¼ 13.2, 17.5,
19.4, 21.9 (d, J ¼ 141.5 Hz), 30.4, 34.9, 43.9, 64.9, 80.9 (d, J ¼ 6.2 Hz),
wise pyridine (133
mL) at room temperature. The reaction mixture
128.4, 128.7, 129.5, 134.6, 162.7, 171.5; 31P NMR (121.5 MHz, DMSO-
was stirred overnight and then evaporated to dryness. The residue
was purified by flash chromatography (EtOAc to EtOAc/MeOH, 9:1)
affording diethyl 3-[(N-acetyl)(N-benzyloxy)amino]propylphosph-
onate as a colorless oil (195 mg, 0.6 mmol, 86%, Rf ¼ 0.16, EtOAc); 1H
d6):
d
¼ 33.2; HRMS (EIþ) m/z calcd for C21H32NO9PNa [M þ Na]þ
496.1707, found 496.1643.
4.1.10. Di(butyloxymethyl) 3-[(acetyl)(benzyloxy)amino]
propylphosphonate
NMR (300 MHz, CDCl3):
d
¼ 1.23 (t, 6H, J ¼ 7.5 Hz, CH3), 1.60e1.71
(m, 2H, eCH2e), 1.80e1.96 (m, 2H, PCH2), 2.02 (s, 3H, CH3), 3.64
(t, 2H, J ¼ 7.5 Hz, CH2N), 3.95e4.06 (m, 4H, OCH2CH3), 4.75 (s, 2H,
Pale yellow oil, 53 mg, 0.11 mmol, 46%, Rf ¼ 0.12, (Et2O); 1H NMR
(300 MHz, CDCl3):
d
¼ 0.91 (t, 6H, J ¼ 7.5 Hz, CH3CH2), 1.60 (sex, 4H,
CH2Ph), 7.30 (s, 5H, Ph); 13C NMR (75.5 MHz, CDCl3):
d
¼ 16.5
J ¼ 7.5 Hz, CH2CH2CH3), 1.74e1.96 (m, 4H, PCH2CH2), 2.04 (s, 3H,
CH3CO), 2.29 (t, 4H, J ¼ 7.5 Hz, CH2CO), 3.65 (t, 2H, J ¼ 7.5 Hz, CH2N),
4.77 (s, 2H, CH2Ph), 5.61 (br s, 2H, OCH2O), 5.65 (br s, 2H, OCH2O),
(d, J ¼ 6.2 Hz), 20.3 (d, J ¼ 4.3 Hz), 20.5, 23.0 (d, J ¼ 142.7 Hz), 45.6
(d, J ¼ 18.6 Hz), 61.7 (d, J ¼ 6.8 Hz), 76.4, 128.7, 128.9, 129.2, 134.4,
172.4; 31P NMR (121.5 MHz, CDCl3):
d
¼ 32.5; HRMS (EIþ) m/z calcd
7.34 (s, 5H, Ph); 13C NMR (75.5 MHz, CDCl3):
d
¼ 13.7, 18.1, 19.9
for C16H26NO5PNa [M þ Na]þ 366.1441, found 366.1406.
(d, J ¼ 5.0 Hz), 24.0 (d, J ¼ 142.1 Hz), 30.5, 35.9, 45.5 (d, J ¼ 26.0 Hz),
76.6, 81.2 (d, J ¼ 6.9 Hz), 128.9, 129.2, 129.4, 134.4, 172.1; 31P NMR
4.1.7. Diethyl [3-((N-benzyloxy)(N-formyl)amino)]
propylphosphonate [19b]
(121.5 MHz, CDCl3):
d
¼
33.5; HRMS (EIþ) m/z calcd for
C22H34NO9PNa [M þ Na]þ 510.186, found 510.185.
In a round bottomed flask were added at room temperature
formic acid (3.15 mL, 83 mmol) and acetic anhydride (1.6 mL,
17 mmol). After 30 min, the reaction mixture was cooled down to
4.1.11. Di(butyloxymethyl) 4-(benzyloxyamino)-4-
oxobutylphosphonate (11a)
0
ꢁC, and a solution of diethyl 3-oxopropylphosphonate acid
Colorless oil, mixture of two E and Z conformers in a 60:40 ratio,
(500 mg, 1.66 mmol) in THF (3.3 mL) was added dropwise. The
reaction mixture was stirred for 10 min at 0 ꢁC and was allowed to
warm up at room temperature over a 2 h period. The reaction
mixture was diluted with EtOAc and successively washed with H2O,
a 0.1 M aqueous KOH solution and H2O. The organic layer was dried,
filtered and evaporated to dryness. The residue was purified by
flash chromatography (EtOAc) affording diethyl [3-((N-benzy-
75 mg, 0.16 mmol, 26%, Rf ¼ 0.52, (Et2O); 1H NMR (300 MHz, CDCl3):
d
¼ 0.92 (t, 6H, J ¼ 7.4 Hz, CH3CH2), 1.62 (sex, 4H, J ¼ 7.4 Hz,
CH2CH2CH3), 1.76e2.03 (m, 4H, PCH2CH2), 2.16 (t, 2/5 of 2H,
J ¼ 6.0 Hz, CH2CON), 2.32 (t, 4H, J ¼ 7.5 Hz, CH2CO), 2.45 (t, 3/5 of
2H, J ¼ 5.4 Hz, CH2CON), 4.88 (s, 2/5 of 2H, CH2Ph), 4.90 (s, 3/5 of
2H, CH2Ph), 5.60 (m, 4H, OCH2O), 7.33 (m, 5H, Ph), 8.97 (br s, 1H,
NH); 13C NMR (75.5 MHz, CDCl3):
d
¼ 13.7, 16.9, 18.1, 24.9
loxy)(N-formyl)amino)]propylphosphonate as
a
colorless oil
(d, J ¼ 140.9 Hz), 25.6 (d, J ¼ 140.9 Hz), 32.4 (d, J ¼ 21.1 Hz), 32.3
(d, J ¼ 21.6 Hz), 35.9, 78.3, 81.2 (d, J ¼ 6.3 Hz), 128.9, 129.3, 129.6,
134.7, 134.6, 169.7, 169.8, 172.2, 172.3; 31P NMR (121.5 MHz, CDCl3):
(390 mg, 1.2 mmol, 71%, Rf ¼ 0.10, EtOAc); 1H NMR (300 MHz,
CDCl3):
d
¼ 1.22 (t, 6H, J ¼ 7.5 Hz, CH3), 1.59e1.69 (m, 2H, CH2),
1.81e1.94 (m, 2H, PCH2), 3.23e3.56 (m, 2H, CH2N), 3.94e4.05
(m, 4H, CH2CH3), 4.76e4.88 (m, 2H, CH2Ph), 7.29 (s, 5H, Ph),
d
¼ 33.5, 33.7; HRMS (EIþ) m/z calcd for C21H32NO9PNa [M þ Na]þ
496.171, found 496.168.
7.90e8.12 (br s, 1H, CHO); 13C NMR (75.5 MHz, CDCl3):
d
¼ 16.3
(d, J ¼ 6.2 Hz), 20.2 (d, J ¼ 15.6 Hz), 22.8 (d, J ¼ 141.5 Hz), 44.1
(d, J ¼ 19.3 Hz), 61.5 (d, J ¼ 6.8 Hz), 77.6, 128.6, 129.0, 129.3, 134.1,
4.1.12. Di(butyloxymethyl) 4-[benzyloxy(methyl)amino]-4-
oxobutylphosphonate (11b)
163.0; 31P NMR (121.5 MHz, CDCl3):
d
¼ 32.1.
Colorless oil, 110 mg, 0.22 mmol, 45%, Rf ¼ 0.55, (Et2O); 1H NMR
(300 MHz, CDCl3):
d
¼ 0.90 (t, 6H, J ¼ 7.3 Hz, CH3CH2), 1.61 (sex, 4H,
4.1.8. General procedure to introduce the acyloxyalkyl/phenyl group
To a solution of the diethyl phosphonate in CH2Cl2 (3 mL/mmol)
was added dropwise at 0 ꢁC bromotrimethylsilane (5 eq.). The
reaction mixture was stirred at 0 ꢁC for 1 h and then overnight at
room temperature. The reaction was quenched by addition of H2O
(0.3 mL/mmol), and the mixture was stirred for 5 min and evapo-
rated to dryness. The residue was dissolved in DMF (6 mL/mmol),
and after addition of Et3N (3 eq.), the reaction mixture was stirred
J ¼ 7.5 Hz, CH2CH2CH3), 1.76e1.90 (m, 4H, PCH2CH2), 2.30 (t, 4H,
J ¼ 7.5 Hz, CH2CO), 2.42 (t, 2H, J ¼ 6.5 Hz, CH2CON), 3.13 (s, 3H,
CH3N), 4.77 (s, 2H, CH2Ph), 5.58 (s, 2H, OCH2O), 5.62 (s, 2H, OCH2O),
7.33 (m, 5H, Ph); 13C NMR (75 MHz, CDCl3):
d
¼ 13.6, 17.1
(d, J ¼ 4.3 Hz), 18.1, 25.8 (d, J ¼ 140.8 Hz), 32.1 (d, J ¼ 15.5 Hz), 35.8,
76.3, 81.1 (d, J ¼ 6.2 Hz), 128.8, 129.1, 129.4, 135.6, 172.1, 173.9; 31P
NMR (121.5 MHz, CDCl3):
d
¼ 33.9; HRMS (EIþ) m/z calcd for
C22H34NO9PNa [M þ Na]þ 510.186, found 510.185.

S. Ponaire et al. / European Journal of Medicinal Chemistry 51 (2012) 277e285
283
4.1.13. Di(pivaloyloxymethyl) 4-(benzyloxyamino)-4-
oxobutylphosphonate (12a)
(m, 2H, CH2N), 5.55 (dd, 2H, J ¼ 5.4, 7.8 Hz, OCH2O), 5.60 (dd, 2H,
J ¼ 5.4, 8.4 Hz, OCH2O), 7.88 (s,1/2H of 1H, CHO), 8.24 (s,1/2H of 1H,
CHO), 9.58 (s, 1/2H of 1H, OH), 10.05 (s, 1/2H of 1H, OH); 13C NMR
Colorless oil, mixture of two E and Z conformers in a 70:30 ratio,
66 mg, 0.13 mmol, 21%, Rf ¼ 0.53, (Et2O); 1H NMR (300 MHz, CDCl3):
(75.5 MHz, DMSO-d6):
d
¼ 13.2, 17.5, 19.1 (d, J ¼ 3.7 Hz), 19.8
d
¼ 1.19 (s, 18H, CH3), 1.78e1.93 (m, 4H, PCH2CH2), 2.17 (br s, 3/10 of
(d, J ¼ 3.7 Hz), 22.3 (d, J ¼ 139.6 Hz), 22.9 (d, J ¼ 139.6 Hz), 35.0, 46.0
(d, J ¼ 20.5 Hz), 49.0 (d, J ¼ 19.2 Hz), 80.9 (d, J ¼ 6.2 Hz), 157.3, 161.9,
2H, CH2CON), 2.46 (t, 7/10 of 2H, J ¼ 6.0 Hz, CH2CON), 4.91 (s, 2H,
CH2Ph), 5.60 (s, 2H, OCH2O), 5.64 (s, 2H, OCH2O), 7.35 (m, 5H, Ph),
171.6; 31P NMR (121.5 MHz, DMSO-d6):
d
(ppm) ¼ 33.5, 33.6; HRMS
8.89 (br s, 1H, NH); 13C NMR (75 MHz, CDCl3):
d
¼ 17.0, 18.3, 25.7
(EIþ) m/z calcd for C14H26NO9PNa [M þ Na]þ 406.1237, found
(d, J ¼ 147.2 Hz), 25.8 (d, J ¼ 147.1 Hz), 27.0, 32.3 (d, J ¼ 12.7 Hz),
38.9, 78.3, 81.5 (d, J ¼ 6.3 Hz), 128.9, 129.2, 129.7, 134.6, 175.3, 175.4,
406.1217.
177.1, 177.2; 31P NMR (121.5 MHz, CDCl3):
d
¼ 33.4, 33.6; HRMS (EIþ)
4.1.19. Di(butyloxymethyl) 3-[(N-acetyl)(N-hydroxy)amino]
m/z calcd for C23H36NO9PNa [M þ Na]þ: 524.202, found 524.199.
propylphosphonate (3b)
Colorless oil, 30 mg, 75.5
CD3OD):
m
mol, 92%; 1H NMR (300 MHz,
4.1.14. Di(pivaloyloxymethyl) 4-[benzyloxy(methyl)amino]-4-
oxobutylphosphonate (12b)
d
¼ 0.97 (t, 6H, J ¼ 7.5 Hz, CH3CH2), 1.69 (sex, 4H, J ¼ 7.5 Hz,
CH2CH2CH3), 1.82e1.97 (m, 4H, PCH2CH2), 2.10 (s, 3H, CH3CO), 2.40
(t, 4H, J ¼ 7.5 Hz, CH2CO), 3.66 (t, 2H, J ¼ 6.0 Hz, CH2N), 5.64 (dd, 2H,
J ¼ 5.4, 8.7 Hz, OCH2O), 5.68 (dd, 2H, J ¼ 5.4, 9.0 Hz, OCH2O); 13C
Colorless oil, 80 mg, 0.15 mmol, 39%, Rf ¼ 0.56, (Et2O); 1H NMR
(300 MHz, CDCl3):
d
¼ 1.16 (s, 18H, CH3), 1.75e1.87 (m, 4H,
PCH2CH2), 2.41 (t, 2H, J ¼ 6.3 Hz, CH2CON), 3.13 (s, 3H, NCH3), 4.77
NMR (75.5 MHz, CD3OD):
d
¼ 14.0, 19.2, 20.4, 20.8 (d, J ¼ 4.4 Hz),
(s, 2H, CH2Ph), 5.58 (s, 2H, OCH2O), 5.62 (s, 2H, OCH2O), 7.32 (s, 5H,
24.5 (d, J ¼ 142.1 Hz), 31.0, 36.6, 82.7 (d, J ¼ 6.8 Hz), 173.5, 174.1; 31P
Ph); 13C NMR (75 MHz, CDCl3):
d
¼ 17.2 (d, J ¼ 4.9 Hz), 25.9
NMR (121.5 MHz, CD3OD):
d
¼ 34.1; HRMS (EIþ) m/z calcd for
(d, J ¼ 140.2 Hz), 26.9, 32.1 (d, J ¼ 16.1 Hz), 38.8, 76.3, 81.4
C15H28NO9PNa [M þ Na]þ 420.139 found 420.134.
(d, J ¼ 6.2 Hz), 128.8, 129.1, 129.4, 134.5, 173.9, 176.9; 31P NMR
(121.5 MHz, CDCl3):
d
(ppm) ¼ 33.6; HRMS (EIþ) m/z calcd for
4.1.20. Di(butyloxomethyl) 4-[(hydroxy)amino]-4-
oxobutylphosphonate (4a)
C24H36NO9PNa [M þ Na]þ 538.222, found 538.226.
Colorless oil, 135 mg, 0.27 mmol, 95%, Rf ¼ 0.25, (EtOAc); 1H
4.1.15. Di(benzyloxymethyl) 4-[benzyloxy(methyl)amino]-4-
oxobutylphosphonate (13a)
NMR (300 MHz, CD3OD):
d
¼ 0.97 (t, 6H, J ¼ 7.4 Hz, CH3CH2), 1.67
(sex, 4H, J ¼ 7.4 Hz, CH2CH2CH3), 1.79e2.01 (m, 4H, PCH2CH2), 2.31
Colorless oil, 29 mg, 0.06 mmol, 35%, Rf ¼ 0.52, (Et2O); 1H NMR
(t, 2H, J ¼ 7.1 Hz, CH2CON), 2.40 (t, 4H, J ¼ 7.2 Hz, CH2CO), 5.64 (dd,
(300 MHz, CDCl3):
d
¼ 1.82e1.96 (m, 4H, PCH2CH2), 2.42 (t, 2H,
2H, J ¼ 4.2, 9.0 Hz, OCH2O), 5.69 (dd, 2H, J ¼ 4.3, 9.3 Hz, OCH2O); 13
C
J ¼ 6.5 Hz, CH2CON), 4.75 (s, 2H, CH2Ph), 5.86 (dd, 2H, J ¼ 5.1, 9.6 Hz,
OCH2O), 5.92 (dd, 2H, J ¼ 5.1, 10.8 Hz, OCH2O), 7.35 (m, 5H, Ph), 7.38
(pseudo t, 4H, J ¼ 7.8 Hz, m-CH), 7.54 (pseudo t, 2H, J ¼ 1, 7.2 Hz, p-
CH), 8.01 (dd, 4H, J ¼ 1, 8.1 Hz, o-CH), 8.90 (br s, 1H, NH); 13C NMR
NMR (75 MHz, CD3OD):
d
¼ 14.0, 19.2, 19.5 (d, J ¼ 5.2 Hz), 26.5
(d, J ¼ 140.9 Hz), 36.3 (d, J ¼ 16.7 Hz), 36.6, 82.7 (d, J ¼ 6.9 Hz),173.6,
177.6; 31P NMR (121.5 MHz, CD3OD):
d
¼ 33.9; HRMS (EIꢀ) m/z calcd
for C14H25NO9P [M ꢀ Hþ] 382.126 found 382.122.
(75 MHz, CDCl3):
d
¼ 17.4 (d, J ¼ 4.3 Hz), 25.7 (d, J ¼ 141.4 Hz), 34.0
(d, J ¼ 16.1 Hz), 76.5, 81.9 (d, J ¼ 6.8 Hz), 128.8, 129.2, 129.3, 129.7,
4.1.21. Di(butyloxomethyl) 4-[(hydroxy(methyl)amino]-4-
oxobutylphosphonate (4b)
130.2, 134.0, 134.1, 165.1, 171.7; 31P NMR (121.5 MHz, CDCl3):
d
(ppm) ¼ 34.6.
Colorless oil, 69 mg, 0.17 mmol, 94%, Rf ¼ 0.23, (EtOAc); 1H NMR
(300 MHz, CD3OD):
d
¼ 0.97 (t, 6H, J ¼ 7.4 Hz, CH3CH2), 1.67 (sex,
4.1.16. Di(benzyloxymethyl) 4-[benzyloxy(methyl)amino]-4-
oxobutylphosphonate (13b)
4H, J ¼ 7.4 Hz, CH2CH2CH3),1.78e2.02 (m, 4H, PCH2CH2), 2.40 (t, 4H,
J ¼ 7.4 Hz, CH2CO), 2.58 (t, 2H, J ¼ 7.1 Hz, CH2N), 3.19 (s, 3H, NCH3),
5.64 (dd, 2H, J ¼ 5.4, 8.7 Hz, OCH2O), 5.67 (dd, 2H, J ¼ 5.4, 8.7 Hz,
Colorless oil, 95 mg, 0.17 mmol, 40%, Rf ¼ 0.50, (Et2O); 1H NMR
(300 MHz, CDCl3):
d
¼ 1.82e1.96 (m, 4H, PCH2CH2), 2.41 (t, 2H,
OCH2O); 13C NMR (75.5 MHz, CD3OD):
d
¼ 14.0, 18.6 (d, J ¼ 4.4 Hz),
J ¼ 6.5 Hz, CH2CON), 3.10 (s, 3H, CH3N), 4.73 (s, 2H, CH2Ph), 5.86
(dd, 2H, J ¼ 5.1, 9.6 Hz, OCH2O), 5.91 (dd, 2H, J ¼ 5.1, 10.0 Hz,
OCH2O), 7.31 (m, 5H, Ph), 7.38 (pseudo t, 4H, J ¼ 7.8 Hz, m-CH), 7.54
(pseudo t, 2H, J ¼ 7.2 Hz, p-CH), 8.00 (dd, 4H, J ¼ 0.9, 8.1 Hz, o-CH);
19.2, 26.5 (d, J ¼ 140.8 Hz), 33.2 (d, J ¼ 17.4 Hz), 36.4, 36.6, 82.7
(d, J ¼ 6.2 Hz), 173.5, 174.7; 31P NMR (121.5 MHz, CD3OD):
d
(ppm) ¼ 34.1; HRMS (EIþ) m/z calcd for C15H28NO9PNa [M þ Na]þ
420.1394, found 420.1360.
13C NMR (75.5 MHz, CDCl3):
d
¼ 17.5 (d, J ¼ 4.6 Hz), 26.2
(d, J ¼ 140.3 Hz), 32.3 (d, J ¼ 15.5 Hz), 33.8, 76.5, 82.1 (d, J ¼ 6.3 Hz),
4.1.22. Di(pivaloyloxomethyl) 4-[(hydroxy)amino]-4-
oxobutylphosphonate (5a)
129.0, 129.1, 129.4, 129.7, 130.4, 134.2, 134.9, 165.3, 174.1; 31P NMR
(121.5 MHz, CDCl3):
d
¼
34.3; HRMS (EIþ) m/z calcd for
Colorless oil, 39 mg, 0.09 mmol, 94%, Rf ¼ 0.21, (EtOAc); 1H NMR
C28H30NO9PNa [M þ Na]þ 578.1550, found 578.1516.
(300 MHz, CD3OD):
d
¼ 1.25 (s, 18H, CH3), 1.81e2.30 (m, 4H,
PCH2CH2), 2.33 (t, 2H, J ¼ 7.1 Hz, CH2CON), 5.67 (dd, 2H, J ¼ 5.4,
4.1.17. General procedure for the catalytic hydrogenation
The O-benzylated hydroxamate was hydrogenolyzed in absolute
methanol (4 mL/mmol) for 1 h at room temperature over 10%
palladium on charcoal. The catalyst was removed by filtration, and
the reaction mixture was evaporated to dryness to give the desired
compound without purification.
7.2 Hz, OCH2O), 5.71 (dd, 2H, J ¼ 5.4, 7.2 Hz, OCH2O); 13C NMR
(75 MHz, CD3OD):
d
¼ 19.6 (d, J ¼ 5.1 Hz), 26.5 (d, J ¼ 141.4 Hz), 27.4,
36.3 (d, J ¼ 17.2 Hz), 39.9, 83.0 (d, J ¼ 6.9 Hz), 175.1, 177.4, 178.3; 31P
NMR (121.5 MHz, CD3OD):
d
(ppm) ¼ 33.8; HRMS (EIꢀ) m/z calcd for
C16H29NO9P [M ꢀ Hþ] 410.157, found 410.156.
4.1.23. Di(pivaloyloxomethyl) 4-[(hydroxy(methyl)amino]-4-
oxobutylphosphonate (5b)
4.1.18. Di(butyloxymethyl) 3-[(N-formyl)(N-hydroxy)amino]
propylphosphonate (3a)
Mixture of two E and Z conformers in a 90:10 ratio, colorless oil,
Mixture of two E and Z conformers in a 50:50 ratio, yellowish oil,
51 mg, 0.12 mmol, 92%, Rf ¼ 0.26, (EtOAc); 1H NMR (300 MHz,
76 mg, 0.19 mmol, 90%; 1H NMR (300 MHz, DMSO-d6):
d
¼ 0.89
CD3OD):
d
¼ 1.24 (s, 18H, CH3), 1.79e2.02 (m, 4H, PCH2CH2), 2.28
(t, 6H, J ¼ 7.5 Hz, CH3CH2), 1.57 (sex, 4H, J ¼ 7.5 Hz, CH2CH2CH3),
(t, 9/10 of 2H, J ¼ 7.1 Hz, CH2CON), 2.58 (t, 1/10 of 2H, J ¼ 6.9 Hz,
1.66e1.90 (m, 4H, PCH2CH2), 2.36 (t, 4H, J ¼ 7.5 Hz, CH2CO), 3.46
CH2CON)*, 2.70 (s, 9/10 of 3H, NCH3), 2.71 (s,1/10 of 3H, NCH3), 5.65

284
S. Ponaire et al. / European Journal of Medicinal Chemistry 51 (2012) 277e285
(dd, 2H, J ¼ 5.4, 7.5 Hz, OCH2O), 5.69 (dd, 2H, J ¼ 5.4, 7.2 Hz, OCH2O);
13C NMR (75.5 MHz, CD3OD):
GCATCCGG-30) and the reverse primer (50-CCCAAGCTTGG-
GAATTCCTCATTTGGTGACGAGCCCTCTTC-30). NdeI (simple under-
lining) restriction site was introduced into the forward primer; EcoRI
(single underlining) and HindIII (double underlining) restriction
sites were introduced into the reverse primer. The primers were
drawn according to the sequence of M. smegmatis str. MC2 155
(accession number CP000480). PCR amplification was performed in
d
¼ 19.8 (d, J ¼ 5.2 Hz), 26.4, 26.6
(d, J ¼ 140.3 Hz), 27.4, 33.2 (d, J ¼ 16.7 Hz), 36.8 (d, J ¼ 17.2 Hz), 39.9,
83.0 (d, J ¼ 6.3 Hz), 175.2, 175.3, 178.2; 31P NMR (121.5 MHz,
CD3OD):
d
¼ 33.8, 34.1; HRMS (EIþ) m/z calcd for C17H32NO9PNa
[M þ Na]þ: 448.171, found 448.167.
4.1.24. Di(benzyloxymethyl) 4-[(hydroxy)amino]-4-
oxobutylphosphonate (6a)
a reaction mixture (20 mL), which included of each primer (20 pmol),
template DNA (0.5 pmol), 3% DMSO and 2ꢃ PhusionTM High Fidelity
PCR Master system (Finnzymes, Finland) using a step cycle (30
cycles) program of 98 ꢁC for 0.1 min, 60 ꢁC for 0.05 min and 72 ꢁC for
0.22 min. The purified PCR product was poly(dA) tailed with TAQ
polymerase before cloning into pGEMT-easy (Promega) and
sequencing. The dxr gene was subcloned in an engineered pBAD
plasmid [13]. The insert obtained after digestion with NdeI and
EcoRI restriction enzymes was ligated into the mutated vector
digested with the same enzymes. The resulting plasmid,
pBAD-DXRmyco designed for obtaining an N-terminal His-tagged
enzyme was transformed into XL1-blue E. coli cells (Stratagene
Europe).
Mixture of two E and Z conformers in a 60:40 ratio, colorless oil,
216 mg, 0.45 mmol, 96%, Rf ¼ 0.23, (EtOAc); 1H NMR (300 MHz,
CD3OD):
d
¼ 1.78e2.05 (m, 4H, PCH2CH2), 2.21 (t, 2/3 of 2H,
J ¼ 7.2 Hz, CH2CON), 2.59 (t, 1/3 of 2H, J ¼ 6.9 Hz, CH2CON),
5.87e5.99 (m, 4H, OCH2O), 7.47 (pseudo t, 4H, J ¼ 8.0 Hz, m-CH),
7.61 (dd, 2H, J ¼ 1.1, 7.8 Hz, p-CH), 8.00 (dd, 4H, J ¼ 1.5, 8.1 Hz, o-CH);
13C NMR (75 MHz, CD3OD):
d
¼ 17.9 (d, J ¼ 5.6 Hz), 27.3
(d, J ¼ 142.5 Hz), 26.6 (d, J ¼ 139.7 Hz), 32.8 (d, J ¼ 16.8 Hz), 81.8
(d, J ¼ 6.8 Hz), 127.5, 128.4, 129.5, 133.7, 164.8; 31P NMR (121.5 MHz,
CD3OD):
d
¼ 33.9, 34.2; HRMS (EIþ) m/z calcd for C20H22NO9PNa
[M þ Na]þ 474.092, found 474.087.
4.1.25. Di(benzyloxymethyl) 4-[hydroxy(methyl)amino]-4-
oxobutylphosphonate (6b)
4.2.4. Overexpression and purification of mycobacterial DXR
For overexpression of the dxr gene, bacteria were grown at
30 ꢁC in LuriaeBertani medium containing ampicillin
(100
by adding L-arabinose (0.05%, w/v). After additional growth for 4 h,
cells were harvested by centrifugation and broken by powdering in
a mortar in the presence of liquid N2 in a 50 mM Tris/HCl buffer
(pH 8) containing 250 mM NaCl and 5 mM 2-mercaptoethanol. The
recombinant protein was purified using Ni2þ-spin columns
(Qiagen). The columns were washed with the same buffer, con-
taining 10 mM, 50 mM and 100 mM imidazole for each respective
Mixture of two E and Z conformers in a 60:40 ratio, colorless oil,
59 mg, 0.13 mmol, 93%, Rf ¼ 0.26, (EtOAc); 1H NMR (300 MHz,
m
g mLꢀ1). Induction was started at the mid-exponential phase
CD3OD):
d
¼ 1.81e2.08 (m, 4H, PCH2CH2), 2.23 (t, 2/3 of 2H,
J ¼ 7.2 Hz, CH2CON), 2.53 (t, 1/3 of 2H, J ¼ 6.9 Hz, CH2CON), 2.64
(s, 2/3 of 3H, CH3N), 3.14 (s, 1/3 of 3H, CH3N), 5.87e5.99 (m, 4H,
OCH2O), 7.47 (pseudo t, 4H, J ¼ 8.0 Hz, m-CH), 7.63 (pseudo td, 2H,
J ¼ 1.1, 7.8 Hz, p-CH), 8.02 (dd, 4H, J ¼ 1.5, 8.1 Hz, o-CH); 13C NMR
(75.5 MHz, CD3OD):
d
¼ 18.6 (d, J ¼ 4.0 Hz), 19.6 (d, J ¼ 5.2 Hz), 26.5
(d, J ¼ 140.3 Hz), 26.6 (d, J ¼ 139.7 Hz), 33.1 (d, J ¼ 17.1 Hz), 36.3, 36.7
(d, J ¼ 16.7 Hz), 83.4 (d, J ¼ 6.9 Hz), 129.9, 130.1, 131.1, 135.2, 166.3,
wash (600
500 mM and 750 mM imidazole in the buffer in three successive
steps (150 L). The fractions with highest DXR activity were pooled
mL). Mycobacterial DXR was eluted with 300 mM,
174.6, 175.2; 31P NMR (121.5 MHz, CD3OD):
d
¼ 34.3, 34.6; HRMS
(EIþ) m/z calcd for C21H24NO9PNa [M þ Na]þ 488.108 found 488.102.
m
and dialyzed against a 50 mM Tris/HCl buffer (pH 7.5) containing
100 mM NaCl, 10% glycerol and 2 mM dithiothreitol (DTT) by
repeated centrifugal ultrafiltration with Centricon 30 concentra-
tors (Millipore). The protein concentration was determined by the
method of Bradford, with BSA as the standard [38]. A molecular
mass of 41 kDa for the DXR subunit was utilized in kinetic
calculations.
4.2. Molecular biology general procedures
4.2.1. Microorganisms and culture conditions
M. smegmatis DSM43756 (ATCC 19420) was obtained from the
Deutsche Sammlung von Microorganismen und Zellkulturen
GmbH (Braunschweig, Germany). Bacteria were grown aerobically
at 37 ꢁC in a liquid medium containing 0.4% yeast extract, 1% malt
extract, 0.2% CaCO3 and 0.4%
dishes containing 0.05% yeast extract, 0.05% malt extract, 0.1%
CaCO3 and 0.1% -glucose and 1.6% agar [36].
D
-glucose. They were plated on Petri
4.2.5. Determination of the enzymatic activity
The DXR enzymatic activity was determined at 37 ꢁC in 50 mM
Tris/HCl buffer, (pH 7.5) containing 3 mM MgCl2 and 2 mM DTT. The
concentrations of NADPH and DXP were 0.15 and 0.5 mM respec-
tively. Initial rates were measured by following the decrease of the
absorbance at 340 nm due to the oxidation of NADPH into NADPþ
(Uvikon 933; Kontron Instruments). The kinetic constants (Vmax
and Km) were calculated from a double-reciprocal plot of enzymatic
rate versus DXP concentration (0.05e0.5 mM). The reaction was
D
4.2.2. Isolation of mycobacterial genomic DNA
Genomic DNA was isolated essentially as described [37]. Cells
from a 50 mL overnight culture were harvested by centrifugation.
The pellet was resuspended in solution I (1 mL, 25% sucrose, 50 mM
Tris/HCl pH 8.0, 50 mM EDTA, 500
mg lysozyme) and incubated
overnight at 37 ꢁC. Solution II (1 mL, 100 mM Tris/HCl pH 8, SDS 1%,
initiated by adding the His-tagged DXR (2.25 mg).
400 mg proteinase K) was then added. After incubation for 4 h at
55 ꢁC, DNA was extracted with phenolechloroform and precipi-
4.2.6. DXR inhibition assay
tated by ethanol. The precipitate was dissolved in 10 mM Tris/HCl,
The IC50 concentrations were determined for each compound.
The enzymatic reaction was initiated by adding DXP after pre-
incubation of the enzyme with the inhibitor in the presence of
NADPH for 2 min at 37 ꢁC. Assays were performed in 50 mM Tris/
HCl buffer (pH 7.5) containing 3 mM MgCl2 and 2 mM DTT. The
concentrations of NADPH and DXP were 0.15 and 0.5 mM respec-
tively. The reaction medium contained inhibitors at various
1 mM EDTA pH 8 buffer (500 mL) and a ribonuclease solution (10 mL,
10 mg/mL) was added. After incubation of 15 min at 37 ꢁC, insoluble
was discarded by centrifugation. Genomic DNA was extracted with
phenol-chloroform and precipitated by ethanol. It was redissolved
in 10 mM Tris/HCl, 1 mM EDTA buffer (pH 8, 100 mL).
4.2.3. Cloning of mycobacterial DXR
concentrations, depending on the inhibitor: from 0.1 to 0.6
1a, 1b and 2b, from 0.2 to 2 M for 2a. Prodrugs 3b and 4b were
tested at a concentration of 1 M.
mM for
The gene coding for DXR was amplified from the genomic DNA
m
m
using the forward primer (50-GGGAATTCCATATGACCACCTCAGCC-

S. Ponaire et al. / European Journal of Medicinal Chemistry 51 (2012) 277e285
285
[8] (a) H. Eoh, A.C. Brown, L. Buetow, W.N. Hunter, T. Parish, D. Kaur, P.J. Brennan,
D.C. Crick, J. Bacteriol. 189 (2007) 8922e8927;
4.2.7. Antimicrobial activity of prodrugs
The antimicrobial activity against M. smegmatis was determined
using the paper disc diffusion method. Plates (10 cm-diameter)
(b) H. Eoh, P.J. Brennan, D.C. Crick, Tuberculosis 89 (2009) 1e11;
(c) P. Narayanasamy, H. Eoh, P.J. Brennan, D.C. Crick, Chem. Biol. 17 (2010)
117e122.
were inoculated with bacteria (120 mL) of a tenfold-diluted over-
[9] T. Kuzuyama, S. Takahashi, M. Takagi, H. Seto, J. Biol. Chem. 275 (2000)
19928e19932.
[10] (a) M. Okuhara, Y. Kuroda, T. Goto, M. Okamoto, H. Terano, M. Kohsaka,
H. Aoki, H. Imanaka, J. Antibiot. 33 (1980) 13e17;
(b) M. Okuhara, Y. Kuroda, T. Goto, M. Okamoto, H. Terano, M. Kohsaka,
H. Aoki, H. Imanaka, J. Antibiot. 33 (1980) 24e28.
[11] R.K. Dhiman, M.L. Schaeffer, A.M. Bailey, C.A. Testa, H. Scherman, D.C. Crick,
J. Bacteriol. 187 (2005) 8395e8402.
night culture. Prodrugs, not soluble in water, were dissolved at
a concentration of 100 mM in DMSO. Fosmidomycin 1a, 2a and 2b
were dissolved in water at the same concentration, INH at 7.5 mM.
Except INH (2
mL), paper discs (Durieux no. 268, diameter 6 mm)
were impregnated with each solution (4
dishes. Growth inhibition was examined after 24 h incubation at
37 ꢁC.
mL) and placed on Petri
[12] C. Zinglé, L. Kuntz, D. Tritsch, C. Grosdemange-Billiard, M. Rohmer, J. Org.
Chem. 75 (2010) 3203e3207 and references therein.
[13] L. Kuntz, D. Tritsch, C. Grosdemange-Billiard, A. Hemmerlin, A. Willem,
T.J. Bach, M. Rohmer, Biochem. J. 386 (2005) 127e135.
[14] L.M. Henriksson, T. Unge, J. Carlsson, J. Åqvist, S.L. Mowbray, T.A. Jones, J. Biol.
Chem. 282 (2007) 19905e19916.
[15] M. Andaloussi, L.M. Henriksson, A. Wieckowska, M. Lindh, C. Björkelid,
A.M. Larsson, S. Suresh, H. Iyer, B.R. Srinivasa, T. Bergfors, T. Unge,
S.L. Mowbray, M. Larhed, T.A. Jones, A. Karlén, J. Med. Chem. 54 (2011)
4964e4976.
[16] A.C. Brown, T. Parish, BMC Microbiol. 8 (2008) 78.
[17] (a) J. Trias, V. Jarlier, R. Benz, Science 258 (1992) 1479e1481;
(b) H. Engelhardt, C. Heinz, M. Niederweis, J. Biol. Chem. 277 (2002)
37567e37572.
[18] M. Andaloussi, M. Lindh, C. Björkelid, S. Suresh, A. Wieckowska, H. Iyer,
A. Karlén, M. Larhed, Bioorg. Med. Chem. Lett. 21 (2011) 5403e5407.
[19] (a) T. Verbrugghen, P. Cos, L. Maes, S. Van Calenbergh, J. Med. Chem. 53 (2010)
5342e5346;
(b) R. Ortmann, J. Wiesner, A. Reichenberg, D. Henschker, E. Beck, H. Jomaa,
M. Schlitzer, Bioorg. Med. Chem. Lett. 13 (2003) 2163e2166.
[20] J.P. Krise, V.J. Stella, Adv. Drug Deliv. Rev. 19 (1996) 287e310.
[21] E. Uh, E.R. Jackson, G. San Jose, M. Maddox, R.E. Lee, R.E. Lee, H.I. Boshoff,
C.S. Dowd, Bioorg. Med. Chem. Lett. 21 (2011) 6973e6976.
[22] (a) A. Reichenberg, J. Wiesner, C. Weidemeyer, E. Dreiseidler, S. Sanderbrand,
B. Altincicek, E. Beck, M. Schlitzer, H. Jomaa, Bioorg. Med. Chem. 11 (2001)
833e835;
4.2.8. Resazurin test [28]
A fresh preculture of M. smegmatis was diluted to McFarland
standard 0.5 (OD600 w0.132). The OD was measured against a blank
containing the growth medium to take in account the turbidity due
to the presence of CaCO3 in suspension in it. The inoculums (2 mL)
were prepared by adding the bacterial suspension (0.2 mL) to the
growth medium (1.8 mL) containing the inhibitors isoniazid (INH)
or 4b. INH is soluble in water and the stock solution (5 mM) was
filter sterilized. The final concentration of INH was 10e40 mM, that
of 4b 250e1000 mM. The tubes were then incubated during 24 h at
37 ꢁC under agitation. A growth control containing no antibiotic
and a blank without bacteria were included in the test.
The resazurin viability test was performed in 96-well plates. To
fresh medium (150
were performed in duplicate. A 0.01% filter sterilized resazurin
solution (30 L) was then added. Sterile water (200 L) was added
mL) was added the inoculum (50 mL). Assays
m
m
to all perimeter wells to avoid evaporation during the incubation.
Plates were covered, sealed with ParafilmÒ and incubated at 37 ꢁC.
The evolution of the color was visually followed during 12 h.
(b) T. Kurz, K. Schlüter, U. Kaula, B. Bergmann, R.D. Walter, D. Geffken, Bioorg.
Med. Chem. 14 (2006) 5121e5135;
(c) K. Schlüter, R.D. Walter, B. Bergmann, T. Kurz, Eur. J. Med. Chem. 41 (2006)
1385e1397;
(d) T. Kurz, C. Behrendt, M. Pein, U. Kaula, B. Bergmann, R.D. Walter, Arch.
Pharm. (Weinheim) 340 (2007) 661e666;
Acknowledgments
The authors are indebted to Jean-Daniel Sauer, Maurice Coppé
and Dr. Lionel Allouche for running the NMR spectra and Hélène
Nierengarten, for measuring the mass spectra. This work was
supported by a grant from the “Agence Nationale de la Recherche”
(Grant Nb ANR-06-BLAN-0291-01).
(e) J. Wiesner, R. Ortmann, H. Jomaa, M. Schlitzer, Arch. Pharm. (Weinheim)
340 (2007) 667e669;
(f) T. Haemers, J. Wiesner, D. Giessmann, T. Verbrugghen, U. Hillaert,
R. Ortmann, H. Jomaa, A. Link, M. Schlitzer, S. Van Calenbergh, Bioorg. Med.
Chem. 16 (2008) 3361e3371.
[23] A. Gissot, A. Volonterio, M. Zanda, J. Org. Chem. 70 (2005) 6925e6928.
[24] A. Argyrou, J.S. Blanchard, Biochemistry 43 (2004) 4375e4384.
[25] L.M. Henriksson, C. Björkelid, S.L. Mowbray, T. Unge, Acta Crystallogr., Sect. D
Biol. Crystallogr. 62 (2006) 807e813.
[26] A.T. Koppisch, D.T. Fox, B.S. Blagg, C.D. Poulter, Biochemistry 41 (2002)
236e243.
[27] T. Scior, I.M. Morales, S.J.G. Eisele, D. Domeyer, S. Laufer, Arch. Pharm.
(Weinheim) 335 (2002) 511e525.
[28] A. Martin, M. Camacho, F. Portaels, J.C. Palomino, Antimicrob. Agents Che-
mother. 47 (2003) 3616e3619.
Appendix A. Supplementary material
Supplementary data related to this article can be found online at
doi:10.1016/j.ejmech.2012.02.031.
References
[29] L.M. McMurry, P.F. McDermott, S.B. Levy, Antimicrob. Agents Chemother. 43
(1999) 711e713.
[30] Y. Sakamoto, S. Furukawa, H. Ogihara, M. Yamasaki Biosci, Biotechnol. Bio-
chem. 67 (2003) 2030e2033.
[31] (a) P.J. Brennan, Tuberculosis 83 (2003) 91e97;
[1] C. Dye, Lancet 367 (2006) 938e940.
[2] C. Dye, K. Lönnroth, E. Jaramillo, B.G. Williams, M. Raviglione, Bull. World
Health Organ. 87 (2009) 683e691.
(b) P.J. Brennan, H. Nikaido, Annu. Rev. Biochem. 64 (1995) 29e63;
(c) V. Jarlier, H. Nikaido, FEMS Microbiol. Lett. 123 (1994) 11e18.
[32] M. Niederweis, Mol. Microbiol. 49 (2003) 1167e1177.
[33] J. Liu, C.E. Barry III, G.S. Besra, H. Nikaido, J. Biol. Chem. 271 (1996)
29545e29551.
[3] S.H. Gillespie, Antimicrob. Agents Chemother. 46 (2002) 267e274.
[4] B.A. Wolucka, M.R. McNeil, E. Hoffmann, T. Chojnacki, P.J. Brennan, J. Biol.
Chem. 269 (1994) 23328e23335.
[5] M.D. Collins, T. Pirouz, M. Goodfellow, D.E. Minnikin, J. Gen. Microbiol. 100
(1977) 221e230.
[34] (a) P.E. Almeida da Silva, A. Von Groll, A. Martin, J.C. Palomino, FEMS Immunol.
Med. Microbiol. 63 (2011) 1e9;
(b) G.E. Louw, R.M. Warren, N.C. Gey van Pittius, C.R.E. McEvoy, P.D. Van
Helden, T.C. Victor, Antimicrob. Agents Chemother. 53 (2009) 3181e3189.
[35] O. Meyer, J.-F. Hoeffler, C. Grosdemange-Billiard, M. Rohmer, Tetrahedron 60
(2004) 12157e12166.
[6] (a) N. Qureshi, J.W. Porter, in: J.W. Porter, S.L. Spurgeon (Eds.), Biosyn-
thesis of Isoprenoid Compounds, vol. 1, John Wiley and Sons, New York,
1981, pp. 1e46;
(b) I. Buhaescu, H. Izzedine, Clin. Biochem. 40 (2007) 575e584.
[7] (a) S. Rosa-Putra, A. Disch, J.M. Bravo, M. Rohmer, FEMS Microbiol. Lett. 164
(1998) 169e175;
[36] S.R. Putra, A. Disch, J.M. Bravo, M. Rohmer, FEMS Microbiol. Lett. 164
(1998) 169.
[37] V. Pelicic, J.-M. Reyrat, B. Gicquel, Mol. Microbiol. 20 (1996) 919e925.
[38] M.M. Bradford, Anal. Biochem. 72 (1976) 248e254.
(b) M. Rohmer, Nat. Prod. Rep. 16 (1999) 565e574;
(c) M. Rohmer, C. Grosdemange-Billiard, M. Seemann, D. Tritsch, Curr. Opin.
Invest. Drugs 5 (2004) 154e162;
(d) M. Rohmer, Pure Appl. Chem. 79 (2007) 739e751.