Design, synthesis, and X-ray crystallographic studies of α-aryl substituted fosmidomycin analogues as inhibitors of mycobacterium tuberculosis 1-deoxy-d-xylulose 5-phosphate reductoisomerase

Article abstract of DOI:10.1021/jm2000085

The natural antibiotic fosmidomycin acts via inhibition of 1-deoxy-d-xylulose 5-phosphate reductoisomerase (DXR), an essential enzyme in the non-mevalonate pathway of isoprenoid biosynthesis. Fosmidomycin is active on Mycobacterium tuberculosis DXR (MtDXR

Full text of DOI:10.1021/jm2000085

ARTICLE
pubs.acs.org/jmc
Design, Synthesis, and X-ray Crystallographic Studies of r-Aryl
Substituted Fosmidomycin Analogues as Inhibitors of Mycobacterium
tuberculosis 1-Deoxy-D-xylulose 5-Phosphate Reductoisomerase†
Mounir Andaloussi,‡,# Lena M. Henriksson,§,# Anna Wieckowska,‡ Martin Lindh,‡ Christofer Bj€orkelid,§
)
Anna M. Larsson,§ Surisetti Suresh,‡ Harini Iyer,^ Bachally R. Srinivasa,^ Terese Bergfors,§ Torsten Unge,§
Sherry L. Mowbray,||,§ Mats Larhed,‡ T. Alwyn Jones,§ and Anders Karlꢀen*,‡
‡Organic Pharmaceutical Chemistry, Department of Medicinal Chemistry, Uppsala University, Biomedical Center, Box 574, SE-751 23
Uppsala, Sweden
§Department of Cell and Molecular Biology, Uppsala University, Biomedical Center, Box 596, SE-751 24 Uppsala, Sweden
Department of Molecular Biology, Swedish University of Agricultural Sciences, Biomedical Center, Box 590,
SE-751 24 Uppsala, Sweden
^AstraZeneca India Private Limited, Bellary Road, Hebbal, Bangalore 560024, India
S Supporting Information
b
ABSTRACT: The natural antibiotic fosmidomycin acts via inhibition of 1-deoxy-
D-xylulose 5-phosphate reductoisomerase (DXR), an essential enzyme in the non-
mevalonate pathway of isoprenoid biosynthesis. Fosmidomycin is active on
Mycobacterium tuberculosis DXR (MtDXR), but it lacks antibacterial activity
probably because of poor uptake. R-Aryl substituted fosmidomycin analogues
have more favorable physicochemical properties and are also more active in
inhibiting malaria parasite growth. We have solved crystal structures of MtDXR in
complex with 3,4-dichlorophenyl substituted fosmidomycin analogues; these show important differences compared to our
previously described forsmidomycinꢀDXR complex. Our best inhibitor has an IC50 = 0.15 μM on MtDXR but still lacked activity
in a mycobacterial growth assay (MIC > 32 μg/mL). The combined results, however, provide insights into how DXR accommodates
the new inhibitors and serve as an excellent starting point for the design of other novel and more potent inhibitors, particularly
against pathogens where uptake is less of a problem, such as the malaria parasite.
’ INTRODUCTION
Bacillus subtilis.7 In the second step, 1-deoxy-D-xylulose 5-phos-
phate reductoisomerase (DXR, also referred to as IspC; EC
1.1.1.267) catalyzes the NADPH-dependent rearrangement and
reduction of 1-deoxy-D-xylulose 5-phosphate (DXP) to form
2-C-methyl-D-erythritol 4-phosphate (MEP), a reaction that also
requires the presence of a divalent cation such as Mg2+, Co2+, or
Mn2+.8 Knockouts of the dxr gene in Escherichia coli are lethal,9
and the essentiality of the M. tuberculosis dxr gene for growth in
vitro has also been demonstrated.10
The natural antibiotic fosmidomycin (Figure 1) is a known
inhibitor of the non-mevalonate pathway in plants and
bacteria.12,13 The compound has been shown to efficiently
inhibit DXRs from E. coli14 (EcDXR) and the malaria parasite
Plasmodium falciparum11 (PfDXR), and the activities of various
fosmidomycin analogues on these two enzymes seem to be well
correlated.15 Furthermore, fosmidomycin has antibacterial activ-
ity on E. coli10,14 and can inhibit the growth of P. falciparum in cell
culture.11,16 The acetyl derivative of fosmidomycin, 3-(N-hydro-
xyacetamido)-1-propylphosphonic acid 2 (FR900098, Figure 1),
has also been evaluated in many studies and shown to be twice as
Tuberculosis, one of the oldest diseases known, is caused by an
infection with the bacterium Mycobacterium tuberculosis. The
seriousness of tuberculosis is underlined by the fact that the
World Health Organization (WHO, http://www.who.org) in
1993 took the unprecedented step of declaring the disease a
global emergency. The WHO estimates that M. tuberculosis
currently infects one-third of the world’s population and caused
1.7 million deaths in 2009. The search for new drugs and the
identification of suitable new drug targets have become even
more urgent because of the emergence of drug-resistant and
multidrug-resistant strains.
Isopentenyl diphosphate (IPP), the precursor of the highly
diversified group of essential isoprenoids,1 is synthesized through
the non-mevalonate pathway in plants, protozoa, green algae,
and many bacteria,2ꢀ4 starting from pyruvate and D-glyceralde-
hyde 3-phosphate. In other eukaryotes as well as archaea,5 IPP is
instead formed through the classical mevalonate pathway,6
starting from acetyl-CoA. The different routes used for IPP
synthesis suggest that all enzymes within the non-mevalonate
pathway are potentially interesting targets for new drugs against
many pathogens, including M. tuberculosis. Indeed, studies have
shown that all the enzymes within this pathway are essential in
Received: January 4, 2011
Published: June 16, 2011
r
2011 American Chemical Society
4964
dx.doi.org/10.1021/jm2000085 J. Med. Chem. 2011, 54, 4964–4976
|

Journal of Medicinal Chemistry
ARTICLE
Table 1. Structures and Inhibition Data for the 3,4-Dichlo-
rophenyl-Substituted Fosmidomycin Analoguesa
Figure 1. Structures of fosmidomycin, FR900098,11 and the 3,4-di-
chlorophenyl-substituted fosmidomycin analogue.
IC50 (μM)
active as fosmidomycin against P. falciparum in vitro and in the
P. vinckei mouse model.11 A number of clinical studies have
demonstrated that fosmidomycin in combination with clinda-
mycin has efficacy and good tolerability in the treatment of
P. falciparum malaria.17ꢀ19
compd
R1
R2
R3
MtDXR
>100
EcDXR27
8a
8b
9a
9b
9c
9d
H
Et
Et
H
Et
H
Et
Et
Et
H
H
H
H
Nt
Nt
CH3
H
>100
0.15 ( 0.02
38 ( 18
0.059 ( 0.020
Nt
In 2005, Dhiman et al.20 showed that fosmidomycin inhibits
M. tuberculosis DXR (MtDXR) with an IC50 of 0.31 μM but has
no effect on M. tuberculosis cell growth. The antibacterial
activity of fosmidomycin on E. coli has been shown to rely on
a cAMP-dependent glycerol 3-phosphate transporter that al-
lows uptake in that organism21 but that seems to be lacking in
M. tuberculosis. The absence of activity on M. tuberculosis is not
due to the presence of exporters or to modification of fosmi-
domycin inside the cell.10 Since the lack of uptake of this
compound into the mycobacterial cell most likely results from
its polar character, it would be of interest to explore analogues
of fosmidomycin with modified hydrophobic/hydrophilic
properties as potential antimycobacterial drugs, similar to the
approach followed in the pursuit of more bioavailable antima-
larial agents. These include modifications of the phosphonate
group that yield phosphonate prodrugs expected to enhance
oral availability.22,23 Different acyl group substituents have also
been prepared, as well as modifications of the hydroxamate
group and of the three-carbon spacer (see, for example, refs
24ꢀ26 and other work cited therein). Although such analogues
have not been evaluated on MtDXR activity or in M. tuberculosis
whole cell growth assays, the SAR obtained from the published
studies can be used as a starting point for the development of
MtDXR inhibitors. The conserved nature of the DXR active site
suggests that newly synthesized inhibitors may show broad-
spectrum activity against a range of pathogens.
H
CH3
CH3
0.7 ( 0.1
22 ( 7
0.119 ( 0.019
Nt
fosmidomycin (1)
0.08 ( 0.02
0.16 ( 0.03
0.030 ( 0.008
0.030 ( 0.008
2
a Nt = not tested.
and on M. tuberculosis cell growth. In parallel studies, we
determined crystal structures of MtDXR in complex with two
such analogues, which show some similarities to, but also
important differences from, fosmidomycin binding. These
structures provided new insights for the design and synthesis
of additional novel inhibitors.
’ RESULTS AND DISCUSSION
Synthesis and Biochemical Evaluation of 8a, 8b, and
9aꢀd. We prepared both formyl (9a) and acetyl derivatives
(9c) of the CR-substituted analogues. Furthermore we prepared
the monoethyl (9b, 9d) and diethyl phosphonate esters (8a, 8b)
of these analogues (Table 1). The compounds could be synthe-
sized as racemic mixtures in good yield according to published
procedures30 and as described below (Scheme 1). The inhibitory
capacity of the analogues was evaluated in a spectrophotometric
assay, in which the MtDXR-catalyzed NADPH-dependent re-
arrangement and reduction of DXP to form MEP are monitored
at 340 nm (see Experimental Section). The compounds were
also tested for activity against the growth of M. tuberculosis strain
H37Rv in a microplate Alamar blue assay.
Fosmidomycin and 2 were included as reference compounds
in this study. The IC50 of 2 has to our knowledge not previously
been reported for MtDXR; here we show it to be half as potent as
fosmidomycin. The IC50 values for some of these inhibitors on
EcDXR have been published27 and are included for comparison
in Table 1. Compounds 9a and 9c had IC50 values of 0.15 and
0.7 μM, respectively, on MtDXR. These can be compared with
IC50 values of 0.06 and 0.12 μM on EcDXR;27 inhibition data are
not available for these compounds on PfDXR. The IC50 values of
the monoethyl esters 9b and 9d were significantly poorer, 38 and
22 μM, respectively, while the diesters 8a and 8b lacked activity
(IC50 > 100 μM). This is consistent with previous findings for
EcDXR; molecules with two charges on the phosphonate moiety
are the most active, while activity falls when one of the charges is
removed, and the uncharged molecule is inactive.29 The reasons
for the progressive decrease in activity are probably steric as well
No analogues have yet been synthesized that have been
shown to be significantly more potent than fosmidomycin or 2
on either EcDXR or PfDXR. Therefore, it was interesting when
Haemers et al.27 showed that several R-aryl-substituted fosmi-
domycin analogues were more effective than fosmidomycin in
inhibiting the malaria parasite’s growth. The 3,4-dichlorophe-
nyl-substituted analogue (Figure 1) was the most potent in the
series. The authors speculated that the improved in vitro
antimalarial activity could be due to either an improved inter-
action with PfDXR (compared to EcDXR) or the electronic and
lipophilic properties of the substituent or alternatively that the
aromatic ring in the CR-position facilitated entry into the
parasite cells. Recently, it was suggested25 that the electron
withdrawing properties of the 3,4 dichlorophenyl group lower
the pKa of the phosphonate group, thereby favoring the doubly
ionized form that seems to be beneficial for activity.28,29 The
improved activity against the parasite prompted us to first
resynthesize this compound (9a) and its acetyl derivative
(9c), as well as their monoethyl and diethyl phosphonate esters,
and to evaluate the activity of the various analogues on MtDXR
4965
dx.doi.org/10.1021/jm2000085 |J. Med. Chem. 2011, 54, 4964–4976

Journal of Medicinal Chemistry
ARTICLE
Scheme 1. Synthesis of r-Aryl Substituted Fosmidomycin Analogues 9aꢀja
a Reagents and conditions: (i) Pd(OAc)2, acrolein, dmphen, p-benzoquinone, 100 ꢀC, microwave; (ii) a) TEP, phenol, 100 ꢀC; b) 2 M HCl, acetone,
100 ꢀC; (iii) 3-pyridine boronic acid, Pd(P-t-Bu3)2, K2CO3, H2O/DME, 130 ꢀC; (iv) 2-thiophene boronic acid, Pd(OAc)2, [(t-Bu)3PH]BF4, H2O/
DME 100 ꢀC; (v) (a) O-benzylhydroxylamine, pyridine, EtOH, room temperature; (b) NaCNBH3, MeOH, HCl, room temperature;
(vi) carbonyldiimidazole, HCOOH, DCM or triethylamine, acetyl chloride, DCM, room temperature; (vii) Zn(CN)2, Pd(OAc)2, [(t-Bu)3PH]BF4,
DMF, microwave, 140 ꢀC; (viii) H2, 10% Pd/C or BCl3, DCM, ꢀ50 ꢀC (for 8c); (ix) TMSBr, DCM, room temperature.
as electrostatic; our earlier structural results showed that the
phosphonate group interacts with a cluster of MtDXR side chains
and with a number of water molecules that lead out to the solvent
continuum.31 It was encouraging, however, to see that there is
room for at least one ethyl group in the MtDXR phosphonate
binding site (9b and 9d) that can be used to increase the general
lipophilicity of such inhibitors. Because the predicted lipophili-
city of these compounds is higher than for fosmidomycin
(ClogP(fosmidomycin) = ꢀ1.8 and ClogP(9a) = 1.2) and since
they are active on the parasite assay, it was anticipated that they
could show some activity on M. tuberculosis cell growth. How-
ever, all six compounds had MIC > 32 μg/mL.
MtDXR Crystal Structures. Crystallographic studies were
initiated in parallel to reveal the mode of binding of the
fosmidomycin analogues to MtDXR and so act as a framework
for the design of more potent analogues with more promising
biological properties. All structural investigations to date have
shown that DXR exists as a homodimer. Each subunit consists of
three domains, an N-terminal NADPH binding domain, a
C-terminal R-helical domain, and a catalytic domain that also
provides much of the dimer interface. DXRs, in general, show
structural differences due to rigid-body domain motions com-
bined with the flexibility of an active site flap. We have previously
described structures of MtDXR produced under two different
crystallization conditions. In our first study, by use of an enzyme
truncated by 20 residues at the C-terminus,32 both magnesium
sulfate and fosmidomycin were present in the crystallization, but
only a sulfate ion could be located in the DXP binding site where
fosmidomycin binds. After truncating the C-terminus by an
additional four residues, we produced new (better and more
reproducible) crystals in ammonium sulfate, which allowed us to
study the binding of fosmidomycin, NADPH, and the manganese
ion.31 Since the high concentrations of sulfate might prevent
weak inhibitors from binding in the active site, we here sought
new crystallization conditions for the shorter construct that did
not contain either sulfate or phosphate ions (see Supporting
Information); this produced a third crystal form. Our second and
third crystal forms both have the symmetry of space group P21
but with different unit cell parameters and are indicated in the
text by MtDXRb or MtDXRc. In all of our crystal forms, MtDXR
exists as a dimer in the crystallographic asymmetric unit, with the
subunits showing some differences. In both forms b and c, we
have been able to bind fosmidomycin and its analogues to the
substrate-binding site of the A chain but not the B chain.
Furthermore, NADPH is better defined in the cofactor binding
site of the A chain, which is also more closed.31
We have solved the structures of the apo-enzyme in both
MtDXRb and MtDXRc crystal forms. The former represents the
4966
dx.doi.org/10.1021/jm2000085 |J. Med. Chem. 2011, 54, 4964–4976

Journal of Medicinal Chemistry
ARTICLE
Figure 2. Stereoviews of the MtDXR complexes made in the program O34 and rendered in Molray.35 The SIGMAA-weighted36 |Fo| ꢀ |Fc| omit maps
for and around each ligand were contoured at 2.0 and 3.5 multiples of the root-mean-square (rms) value of the respective map, given in parentheses:
(A) MtDXRc-9a (rms 0.052 e/Å3); (B) MtDXRb-9a-NADPH (rms, 0.062 e/Å3); (C) MtDXRb-9c (rms 0.067 e/Å3); (D) active site of 2JCZ (side
chains and fosmidomycin in gold), with superpositioning of inhibitors from MtDXRc-9a (pink), MtDXRb-9a-NADPH (maroon), and MtDXRb-9c
(cyan). Water molecules lining a hydrated cavity are shown as small red spheres. In all panels, the Mn2+ ion is shown in magenta.
4967
dx.doi.org/10.1021/jm2000085 |J. Med. Chem. 2011, 54, 4964–4976

Journal of Medicinal Chemistry
ARTICLE
The phosphonate group of each makes interactions with the
backbone nitrogen of Ser177, as well as the side chains of
Ser177, Ser213 (in one of its two possible conformations),
Asn218, Lys219, and several waters. The hydroxamate group is
bound in a very similar manner in each of the analogue
complexes, coordinating the manganese ion and overlapping
with two of the three metal-bound waters in apo-MtDXRb;
interactions between the terminal oxygen and the side chains of
Asp151 and Ser152 are also present. A sixth metal-coordinating
group, the third water in apo-MtDXRb, is absent in the com-
plexes because of lack of space. The dichlorophenyl ring shows
some variability, including a 180ꢀ flip; electron density is
weakest for the ring, which may arise from heterogeneity
concerning the ring flip, from some small population of bound
R-enantiomer, or both. Finally, the fosmidomycin propyl back-
bone atoms assume a very similar conformation in the three
analogue structures.
Figure 3. Interactions in the active site of MtDXRb-9a-NADPH.
Protein carbon atoms are shown in gold, while those of the inhibitor
9a are maroon. Water molecules are shown as small red spheres, and the
Mn2+ ion is shown in magenta. Dark gray bubbles show the observed
interactions.
Unexpectedly, the backbone conformation seen for the R-aryl
fosmidomycin analogues differs from that observed for fosmido-
mycin itself, with two of the three torsion angles, adopting
different rotamers (Figure 2D). Interestingly, our modeling of
the R-enantiomers of 9a and 9c suggests that they would adopt
backbone conformations more similar to fosmidomycin when
binding to the active site, because of steric constraints. However,
until the optically pure enantiomers of 9a and 9c have been
prepared and evaluated, it is not possible to say if both enantio-
mers are active or if the binding is stereoselective. Large changes
have been described previously in a particular loop near the active
site of both MtDXR31 and EcDXR.33 In the present case, the most
dramatic change is the fact that this flap containing Trp203 is
closed (and ordered) in the fosmidomycin complex of MtDXR
but open in the case of the analogues described here (residues 199
to ∼204 are disordered in the analogue complex structures).
This difference arises because the dichlorophenyl ring of the
analogues would clash with the indole ring of Trp203 if the active
site flap were to adopt the closed conformation (Figure 2D).
Thus, a hydrogen bond between His200 and the phosphonate
group, as well as the interaction of the indole ring of Trp203 with
the fosmidomycin backbone, is lost in the analogues. Further, the
interaction of the hydroxamate group with the backbone nitro-
gen of residue 152 in fosmidomycin is lost, and only the
interaction of the side chain hydroxyl group of Ser152 is
maintained in the analogue complexes. This is a consequence
of fosmidomycin lying ∼0.5 Å deeper in the MtDXR active site
than the analogues. The shift also avoids clashes that would occur
between the dichlorophenyl ring and Pro265. The p-chlorine
atom and parts of the phenyl ring are partially accessible to
solvent, wedged into a depression between three ordered and
one disordered loop (containing residues 179, 245, 265 and 203,
respectively). In the first complex structure that we solved,
MtDXRc-9a, we were concerned because this edge of the
dichlorophenyl ring was in contact with the active site flap from
a symmetry-related copy of a B subunit (Thr202 CG2ꢀCL atom
contacts of 3.6 Å). To ensure that this had not produced a
crystallographic artifact, we obtained complexes that were crys-
tallized under other conditions, as described above. In the
MtDXRb-9a-NADPH and MtDXRb-9c structures, there are no
stabilizing interactions with symmetry related molecules. How-
ever, the 180ꢀ ring flip that we observe in MtDXRc-9a may
indeed represent an artifact of crystal packing that allows the
m-chlorine to pack against the symmetry molecule, instead of being
buried as in the other structures (Figure 2). The extra methyl
highest resolution that we have yet achieved for this enzyme, and
the latter allows us to evaluate structural changes due to the new
unit cell as well as to the removal of sulfate ions from the
crystallization conditions. In the apo-MtDXRb A subunit, a
sulfate is located in the DXP/MEP-binding site, making five
interactions with protein hydrogen bond donors (Ser177 OG,
Ser177 N, Ser213 OG, Asn218 ND2, Lys219 NZ) and four water
molecules. In apo-MtDXRc, the sulfate ion is replaced by a pair of
water molecules forming three hydrogen bond interactions
(Asn218 ND2 for one, Ser177 N and Lys219 NZ for the other,
while the Ser213 side chain rotates to form an interaction with
209 O).
Crystals of three complexes could be produced that were
suitable for structural work: MtDXRb-9a-NADPH, MtDXRc-9a,
and MtDXRb-9c. The A subunits of MtDXRb structures can be
aligned with the MtDXRb-fosmidomycin-NADPH structure31
(2JCZ) with rmsd of ∼0.3 Å over ∼375 CR atoms, excluding the
large structural variations in the flap. The MtDXRc A subunits
superimpose with rmsd of 0.7 Å for 359 CRs, while A subunits of
the different crystal forms superimpose with pairwise rmsd in the
range of 0.6ꢀ1.0 Å over ∼360 CR atoms. The larger deviations
are the result of rigid body shifts in the C-terminal R-helical
domain. Unless otherwise indicated, the electron density for the
main chain of all five structures is of good quality, with the
exception of the N-terminal His6-tag, the first 10 residues, the
active site flap of the A subunit (residues A198ꢀA208), and
residues A69ꢀA78 in the MtDXRc structures. The density for
the cofactor in the MtDXRb-9a-NADPH structure is weaker than
we observed in the equivalent fosmidomycin ternary complex
(PDB code 2JCZ).31 In particular, the NADPH in the B subunit
is best defined for the phosphate groups and is otherwise poorly
defined.
Binding of Fosmidomycin Analogues. The two fosmido-
mycin analogues bind in a very similar manner in the DXP/MEP
site of the A chain, both in the presence and absence of NADPH
(Figure 2). The detailed interactions to one of them (MtDXRb-
9a-NADPH) are shown in Figure 3. As for the inhibition studies,
a racemic mixture was used in the crystallographic work. Our
crystallographic results indicate that 9a and 9c bind primarily, if
not exclusively, as the S-enantiomers (see Supporting Information).
4968
dx.doi.org/10.1021/jm2000085 |J. Med. Chem. 2011, 54, 4964–4976

Journal of Medicinal Chemistry
ARTICLE
analogues, since these are generally more stable and have
similar potency (e.g., comparing pairs of compounds in
Table 1).
DXR is known to be a challenging protein to use in structure-
based drug design, primarily because of the flexible flap and the
strong substrate/inhibitorꢀmetal interactions.24 As a control
experiment, we redocked fosmidomycin to the MtDXRb-fosmi-
domycin-NADPH structure31 (PDB entry 2JCZ) and redocked
9a to the MtDXRb-9a-NADPH structure, with NADPH and the
manganese ion included in the complex, as described in the
methods section. Each inhibitor was randomly perturbed, and
after redocking, the rmsd values between the best-scored docked
pose and the X-ray structure were 0.47 and 0.33 Å, respectively
(heavy atoms superimposed). Encouraged by these results, we
proceeded with this protocol in docking studies to evaluate an
expanded set of potential inhibitors. A total of 441 synthetically
feasible compounds were built based on the assumption that
ortho substituents could be introduced using a Suzuki reaction
starting from boronic acids and the phenyl halide. These were
docked to the MtDXRb-9a-NADPH crystal structure in order to
evaluate their fit to the substrate-binding pocket of the enzyme.
Two compounds, one incorporating a pyridine substituent (9h)
and one a thiophene substituent (9j), were chosen from these
studies (see Table 2). The top-scoring docking pose of both
these compounds places the ortho substituent in the hydrated
cavity with the backbone adopting almost exactly the same
conformations as in the MtDXRb-9a-NADPH crystal structure.
However, the phenyl ring was slightly reoriented. The nitrogen in
the pyridine substituent was able to hydrogen-bond to His248.
The o-bromo substituted compound 9f, which can be readily
accessible from the common intermediate 7d, was also evaluated.
Additionally, compounds incorporating a nitrile (9g), hydro-
xymethyl (9k), and hydroxyethyl group (9n) were suggested,
with the intent of introducing a more hydrophilic group in the
pocket that could displace one of the water molecules while
keeping the hydrogen bonding pattern intact. Docking of these
compounds into the MtDXRb X-ray structure, after removal of
the four water molecules, indicated that at least 9n could
participate in the water-mediated hydrogen bonding interaction
with His248 (Figure 4B). The methoxymethyl 9m and methyl 9l
compounds were intermediates in the synthetic route leading to
the other compounds and were also evaluated. For comparison
we also prepared the nonsubstituted CR-phenyl analogue (9o).
This compound was previously shown to have an IC50 of 0.3 μM
on EcDXR.27 Finally, we prepared the cyclic phosphonate ester
(12), a rigidified analogue that was reachable using the same
synthetic route that led to 9k.
Figure 4. (A) Crystal structure of MtDXRb-9a-NADPH showing the
hydrated cavity, visualized in PyMOL.37 The protein surface is shown in
blue, compound 9a is in purple, the four waters in the hydrated cavity are
in red, Mn2+ is in black, and the nicotinamide moiety of the NADPH
cofactor is in green. (B) Compound 9n docked in to the crystal structure
of MtDXRb-9a-NADPH. The same hydrated cavity as in panel A is
shown. Compound 9n is colored yellow; coloring is otherwise as in
panel A. Cavity-lining amino acid residues His248 and Ser152 are
included in the panel. Possible hydrogen bonds are shown with
dotted lines.
group of 9c, compared to 9a, is directed away from the
manganese ion toward the space occupied by the indole ring of
Trp203 in the closed MtDXRb-fosmidomycin-NADPH struc-
ture. The precise details of this potential interaction may account
for the 2-fold loss in activity of 2 compared to fosmidomycin that
we observe for MtDXR, as opposed to EcDXR, where the
activities are identical.
Structure-Based Design To Investigate the Hydrated
Cavity. One obvious conclusion from the structural work is
that the opening of the active-site flap in the analogue-bound
structures creates a large solvent-exposed region that could be
explored using CR-substitutions other than the 3,4-dichloro-
phenyl ring seen in our complex structures. Indeed, Haemers
et al.27 showed that analogues of 2 representing a variety of para
and meta substitutions of the phenyl ring could be accommo-
dated by EcDXR with only modest differences in potency. We
have previously described a hydrated cavity close to the DXP/
MEP-binding site that is lined with the side chains of conserved
amino acids.31 It became apparent that this cavity could be
reached from the ortho position of the 3,4-dichlorophenyl
ring of 9a (Figure 4A), which we attempted to exploit to
optimize the binding of this structural class. For synthetic
reasons, we prepared derivatives based on the CR-substituted
phenyl ring instead of the 3,4-dichlorophenyl moiety. Further-
more, we decided to prepare the acetyl instead of the formyl
Synthesis of Compounds 8a, 8b, 9aꢀo, and 12. The synthesis
of the desired compounds is outlined in Schemes 1 and 2 (see also
Supporting Information). The route is a modification of a previously
reported method.27 It starts with the synthesis of R,β-unsaturated
aldehydes 4a and 4b, prepared from boronic acids 3a and 3b and
acrolein in an oxidative Heck reaction (Scheme 1).38 Aldehydes 4c
and 4d were instead obtained from aryl halides and acrolein diethyl
acetal using a palladium(0)-catalyzed MizorokiꢀHeck reaction
(Scheme 2).39,40 1,4-Addition of triethyl phosphite to compounds
4aꢀd in the presence of phenol resulted in formation of the acetal
intermediates, which were then hydrolyzed to aldehydes 5aꢀd.
Next compound 5b was reacted with 3-pyridine boronic acid and
2-thiophene boronic acid in a microwave-assisted Suzuki coupling.41
By use of the Pd(P-t-Bu3)2 catalyst in the reaction with 3-pyridine
boronic acid and the Pd(OAc)2/[(t-Bu3)HP]BF4 combination in
4969
dx.doi.org/10.1021/jm2000085 |J. Med. Chem. 2011, 54, 4964–4976

Journal of Medicinal Chemistry
ARTICLE
Table 2. Inhibition of MtDXR by R-Aryl-Substituted Fosmidomycin Analoguesa
a Nt = not tested.
the reaction with 2-thiophene boronic acid, compounds 5e and 5f
were obtained in satisfactory yields (35% and 41%, respectively).
The bromine 7d was also used to prepare the corresponding nitrile
derivative 7e via a direct Pd-catalyzed microwave-assisted transfor-
mation using Zn(CN)2.42 In the reaction with O-benzylhydroxyla-
mine and the subsequent reduction with sodium cyanoborohydride
and hydrochloric acid, aldehydes 5aꢀf were transformed into
benzyloxyamines 6aꢀf, which were then formylated (7a, 7c) or
acetylated (7b, 7dꢀi). In order to obtain hydroxamates 8a,b and
8dꢀj, the O-benzyl protecting group was removed in compounds
7aꢀi using Pd-catalyzed (10% Pd/C) hydrogenation under atmo-
spheric hydrogen pressure. Depending on the starting material,
different conditions were used for this reaction. To avoid debromi-
nation, milder conditions (ice-bath, Na2CO3, THF)43 were used for
compound 7d, while compound 7g required acidic conditions (2 M
HCl, EtOH)44 because of the thiophene substituent. When com-
pound 7d was subjected to Pd-catalyzed hydrogenation without any
acid or base additive, both the O-benzyl group and aryl bromide
4970
dx.doi.org/10.1021/jm2000085 |J. Med. Chem. 2011, 54, 4964–4976

Journal of Medicinal Chemistry
ARTICLE
Scheme 2. Synthesis of r-Aryl Substituted Fosmidomycin Analogues 9kꢀo and 12a
a Reagents and conditions: (i) Pd(OAc)2, acrolein diethyl acetal, TBAA, K2CO3, KCl, DMF, 90 ꢀC, microwave, 2 M HCl, reflux or Pd(OAc)2, acrolein,
dmphen, p-benzoquinone, 100 ꢀC, microwave (for 4b); (ii) (a) TEP, phenol, 100 ꢀC, (b) 2 M HCl, acetone, 100 ꢀC; (iii) (a) O-benzylhydroxylamine,
pyridine, EtOH, room temperature; (b) NaCNBH3, MeOH, HCl, room temperature; (iv) triethylamine, acetyl chloride, DCM, room temperature; (v)
H2, 10% Pd/C; (vi) TMSBr, DCM, room temperature; (vii) TMSBr, DCM, room temperature, MeOH, room temperature (from 8h to 9m); (viii) NaH,
THF, room temperature.
underwent hydrogenolysis to obtain 8k. The nonsymmetric diben-
zyl derivative, 7h, furnished two products in the hydrogenation
reaction, that is, R-o-hydroxymethylphenyl and R-o-methylphenyl
derivatives (8h and 8i). In the case of 8c the benzyl group was
deprotected using 1 M BCl3. The final step in the synthesis was
hydrolysis of the phosphonate esters by bromotrimethylsilane in
dichloromethane (DCM). 9e partly undergoes deformylation un-
der aqueous conditions. Interestingly, when compound 8h was
reacted with TMSBr, in addition to the ester hydrolysis, we also
observed the formation of the o-bromomethyl derivative which was
transformed to 9m by reaction with MeOH.45 We used the same
approach to prepare 12. Compound 7h was reacted with TMSBr to
give 10, which after treatment with NaH gave the bicyclic com-
pound 11 that after hydrogenation gave 12. Incomplete hydrolysis
of esters 8a, 8b, and 8f, afforded the mono ethylphosphonate esters
9b, 9d, and 9i (Scheme 1). All final compounds were purified by
reverse phase HPLC using gradient elution of MeOH or MeCN
with 0.05% HCOOH in water.
we determined the IC50 values. First of all it is interesting to
compare the IC50 values of 9c and 9o, which show that the
chlorine atoms contribute some 10-fold in potency, comparable
to what was found for these compounds on EcDXR.27 The
bromo compound 9e had an IC50 of 5.6 μM, which indicates that
substitution in the ortho position is tolerated by the enzyme.
However, the activity of the acetyl analogue (9f) dropped ∼40-
fold. This is a distinctly larger loss of activity than the 5-fold
potency difference seen with the 3,4-dichlorophenyl substituted
9a/9c, while 9d actually has a better IC50 than 9b (Table 1). A
5-fold potency difference had also been seen in the EcDXR
enzyme assay for the corresponding 4-MeO-phenyl substituted
analogues,27 suggesting that there may be differences in how
these compounds are interacting with the enzymes. Modeling the
ortho substituted bromo compound 9f based on the structure of
the 9c complex introduces close contact clashes of the bromine
atom with the OE2 atom of Glu153 or with the phosphonate
group of the ligand. If the formidomycin backbone is kept fixed,
these contacts can only be relieved by a dihedral rotation of the
phenyl ring which eventually produces a clash between the acetyl
group and the ring. No such clash would exist in the formyl
Biochemical Evaluation of 9eꢀo and 12. The different
analogues were evaluated for activity on MtDXR (Table 2).
For compounds that had more than 35% inhibition at 100 μM,
4971
dx.doi.org/10.1021/jm2000085 |J. Med. Chem. 2011, 54, 4964–4976

Journal of Medicinal Chemistry
ARTICLE
derivative and would therefore explain the observed inhibition
data. On the basis of this reasoning, substitutions at the ortho
position are more likely to result in poorer inhibition for the
acetyl derivative compared to the equivalent formyl compound.
Indeed compounds 9gꢀj were only weakly potent, having
20ꢀ30% inhibition at 100 μM, showing that these substituents
did not improve binding. We then tried to replace one of the
waters identified in the hydrated cavity of the X-ray structure
with a water-mimicking group. This strategy has previously
been used to increase the binding affinity of ligands, presumably
because of a favorable increase in entropy associated with the
release of the water molecule into the bulk solvent.46 Docking
studies, after removal of the enzyme-bound water molecules,
show that the hydroxyl group of at least 9n is able to replace one
of the water molecules in the hydrated cavity (Figure 4B).
However, the low activity of this compound (IC50 = 150 μM)
indicates that it cannot effectively compete with the interac-
tions between the water molecule and the enzyme (although 9n
is indeed more active than 9k, which has a smaller ortho
substitution). A cyclic phosphonate ester (12) was also pre-
pared but lacked activity. Unfortunately, we were unable to
produce crystal structures of these complexes to evaluate the
accuracy of our docking and modeling experiments and to give
direct experimental evidence for how the different compounds
interact with MtDXR. All compounds (9eꢀo and 12) were
evaluated for their activity against the M. tuberculosis strain
H37Rv and shown to have MIC > 32 μg/mL.
which becomes disordered in the analogue structures. A protocol
was established that could successfully redock these structures to
the active site. A conserved hydrated cavity close to the binding
site was explored in an effort to find compounds that would fill
the cavity, perhaps displacing one of the bound water molecules.
This strategy did not lead to more potent compounds, and we
conclude that the structureꢀproperty relationships of fosmido-
mycin need to be further explored to obtain antimycobacterial
activity for this structural class. However, the combined results
provide key insights into how DXR responds to the binding of
new inhibitors, as well as how the inhibitors themselves respond
to the protein. The highly conserved DXP/MEP binding site of
DXR32 means that even if we are ultimately unsuccessful in
producing compounds with antitubercular activity, we now have
the basis for the design of novel, more potent inhibitors against
other pathogens, such as the malaria parasite.
’ EXPERIMENTAL SECTION
Chemistry. The microwave reactions were performed in a Smith
synthesizer producing controlled irradiation at 2450 MHz with a power
of 0ꢀ300 W. The reaction temperature was determined using the built-
in online IR sensor. Flash column chromatography was performed on
Merck silica gel 60 (40ꢀ63 μm). Analytical thin layer chromatography
was done using aluminum sheets precoated with silica gel 60 F254
.
Analytical RPLC-MS was performed on a Gilson HPLC system with a
Finnigan AQA quadrupole mass spectrometer with detection by UV
(DAD) using an Onyx Monolithic C18 column (50 mm ꢁ 4.6 mm).
Alternatively, a Gilson HPLC system was used with a Finnigan
Thermoquest MSQ quadrupole mass spectrometer (ESI+) and a
SEDERE ELSD (Sedex 55) detector, equipped with an Onyx Monolithic
C18 column (50 mm ꢁ 4.6 mm). For both systems a H2O/CH3CN
or H2O/CH3OH gradient with 0.05% HCOOH was used as the mobile
phase at a flow rate of 4 mL/min. Preparative RP-HPLC was per-
formed on a system equipped with a Zorbax SB-C8 column (150 mm ꢁ
21.2 mm) using a H2O/CH3CN gradient with 0.1% CF3COOH or
0.05% HCOOH as mobile phase at a flow rate of 5 mL/min. Purity of the
final compounds was determined by RP-HPLC on a Dionex UltiMate
3000 binary analytical LC system using Kinetex C18 2.6 μm, 3.0 mm ꢁ
50 mm column with a H2O/CH3CN gradient with 0.05% HCOOH and
UV detection at 214 nm. All the compounds showed purity above 95%.
1H and 13C NMR spectra were recorded on Varian Mercury Plus
’ CONCLUSIONS
Fosmidomycin should serve as an excellent starting point for
the development of antitubercular drugs. Clinical trials con-
ducted with fosmidomycin in combination with clindamycin
have produced good results in the treatment of acute uncompli-
cated malaria, showing that PfDXR is a druggable target.47,48
Fosmidomycin is also highly active on MtDXR but unfortunately
lacks activity on M. tuberculosis whole cells, probably because of
poor uptake. Therefore, one of the main challenges is to prepare
modified analogues of fosmidomycin that can cross the myco-
bacterial cell wall. It has previously been shown that introduction
of a 3,4-dichlorophenyl group in the CR-position relative to the
phosphonate group produces analogues that have a higher in
vitro antimalarial activity than fosmidomycin. We resynthesized
this compound and showed that it was potent on MtDXR (9a,
IC50 = 0.15 μM) but that it still lacked activity on M. tuberculosis
whole cells (MIC > 32 μg/mL). Additional analogues were
synthesized, which showed lower potency against the enzyme,
and again, no activity was observed on mycobacterial cells. X-ray
crystallographic studies were initiated in parallel, and five
MtDXR structures were solved, representing the apo protein as
well as complexes with 9a and 9c, under two distinct crystal-
lization conditions. The overall geometry of these analogues
when bound to MtDXR is very similar. The interactions at the
phosphonate group are also very similar to those seen in the
fosmidomycinꢀenzyme complex. However, the propyl back-
bone adopts a different conformation in the analogues compared
to fosmidomycin. Fosmidomycin lies deeper in the active site.
Although the hydroxamate groups show similar coordination to
the metal ion in all cases, the analogues show a different set of
hydrogen bonding interactions with protein, compared to fos-
midomycin. Furthermore, the R-substituted 3,4-dichlorphenyl
ring displaces the indole ring of Trp203 from the active site flap,
1
1
instruments: H at 399.9 MHz and 13C at 100.6 MHz or H at 399.8
MHz and 13C at 100.5 MHz. The chemical shifts for 1H NMR and 13
C
NMR are referenced to TMS via residual solvent signals (1H, CDCl3 at
7.26 ppm, CD3OD at 3.31 ppm; 13C, CDCl3 at 77.16 ppm, CD3OD at
49.00 ppm). 31P spectra were recorded on a Varian Mercury 300 Plus
instrument at 121.4 or 162 MHz, and the chemical shifts are referenced
to 85% H3PO4 which was used as an external standard. Molecular
masses (HR-ESI-MS) were determined on a Micromass Q-Tof2 mass
spectrometer equipped with an electrospray ion source. All final
products were obtained as racemic mixtures. 3-(3,4-Dichlorophe-
nyl)acrylaldehyde (4a), 3-(2-bromophenyl)acrylaldehyde (4b), diethyl
(1-(3,4-dichlorophenyl)-3-oxopropyl)phosphonate (5a), diethyl (1-(3,
4-dichlorophenyl)-3-(N-hydroxyformamido)propyl)phosphonate (8a), di-
ethyl (1-(3,4-dichlorophenyl)-3-(N-hydroxyacetamido)propyl)phosphonate
(8b), 3-(N-hydroxyformamido)-1-(3,4-dichlorophenyl)propylphosphonic
acid (9a), 3-(N-hydroxyacetamido)-1-(3,4-dichlorophenyl)propylpho-
sphonic acid (9c), and (3-(N-hydroxyacetamido)-1-phenyl)propyl-
phosphonic acid (9o) were previously reported.30,49
Materials. All reagents were purchased from commercial suppliers
and used without further purification. Dichloromethane (DCM) and
tetrahydrofuran (THF) were distilled under nitrogen immediately
4972
dx.doi.org/10.1021/jm2000085 |J. Med. Chem. 2011, 54, 4964–4976

Journal of Medicinal Chemistry
ARTICLE
before use. For DCM, calcium hydride was used as the drying agent.
For THF, sodium/benzophenone ketyl was used as the drying agent.
(E)-3-(2-Bromophenyl)acrylaldehyde (4b)49. A test tube was
charged with Pd(OAc)2 (0.007 g, 0.031 mmol), 2,9-dimethyl-1,10-
phenanthroline (neocuproine, dmphen) (0.006 g, 0.031 mmol) and
acetonitrile (2 mL), and the mixture was stirred for 30 min at room
temperature. A 2ꢀ5 mL microwave-transparent process vial was
charged with acrolein (1 mL, 15.50 mmol), 2-bromophenylboronic acid
(0.50 g, 3.10 mmol), p-benzoquinone (0.17 g, 1.55 mmol), and
acetonitrile (1 mL). The content of the test tube was added to the
process vial, which thereafter was capped and exposed to microwave
heating for 30 min at 100 ꢀC. After the mixture was cooled to room
temperature, solvent was removed. The mixture was diluted with 0.1 M
NaOH and extracted with DCM. Organic layers were then dried with
anhydrous MgSO4. The solvent was evaporated and the obtained crude
product purified by column chromatography on silica gel (isohexane/
(0.13 g, 0.27 mmol) was cooled to ꢀ50 ꢀC and 1 M BCl3 in hexane (1.1 mL,
1.10 mmol) was added dropwise under N2 atmosphere. The contents were
stirred at ꢀ50 ꢀC for 1 h and quenched with aqueous saturated NaHCO3
solution. The mixture was extracted with DCM (2 ꢁ 50 mL). The combined
organic layers were dried and filtered, then concentrated. The crude product
was purified using preparative RP LCꢀMS system with gradient elution
(50ꢀ100% of acetonitrile in 0.05% aqueous formic acid) to yield 73% 8c. 1H
NMR (400 MHz, CDCl3) δ8.33 and 7.64 (s, 1H), 7.52 (m, 2H), 7.32 (q, J=
7.09 Hz, 1H), 7.11 (m, 1H), 4.09 (m, 2H), 3.79 (m, 4H), 3.38 (m, 1H), 2.57
(m, 1H), 2.22 (m, 1H), 1.28 (m, 3H), 1.09 (m, 3H). MS (ESI+): m/z 394
(M + H+), 396 (M + 2 + H+). Formula: C14H21BrNO5P.
3-(N-Hydroxyformamido)-1-(2-bromophenyl)propylpho-
sphonic Acid (9e). To a solution of phosphonate diethyl ester 8c (75
mg, 0.19 mmol) in dry DCM (2 mL) was added TMSBr (0.1 mL, 0.76
mmol) dropwise under N2 at 10 ꢀC and stirred at room temperature.
After 6 h the volatiles were removed under vacuum and the resulting
crude product was purified using preparative RP LCꢀMS system with
gradient elution (0ꢀ20% of acetonitrile in 0.05% aqueous formic acid).
Compound 9e was obtained in 33% yield. 1H NMR (400 MHz, DMSO-
d6) δ 8.15 and 7.66 (s, 1H), 7.54 (m, 2H), 7.36 (m, 1H), 7.15 (m, 1H),
3.48 (m, 1H), 3.30 (m, 1H), 3.12 (m, 1H), 2.34 (m, 1H), 2.01 (m, 1H);
31P NMR (162 MHz, DMSO-d6) δ 23.13, 23.10 (major and minor
isomer). MS (ESI+): m/z 338 (M + H+), 340 (M + 2 + H+). Formula:
C10H13BrNO5P. HRMS: m/z found [M + H]+ 337.9791, C10H14-
BrNO5P requires 337.9793.
Inhibition Assay. Inhibition of MtDXR activity was measured in
a spectrophotometric assay8,31,50 by monitoring the NADPH-depen-
dent rearrangement and reduction of DXP to form MEP, using the
absorption of NADPH at 340 nm. Assay reactions had a final volume
of 50 μL and contained 50 mM HEPESꢀNaOH, pH 7.5, 100 mM
NaCl, 1.5 mM MnCl2, 0.2 mM NADPH, 0.048 μM MtDXR, 0.2 μM
DXP, and inhibitory compound at various concentrations. For com-
parison, the Km values are 7.2 μM for NADPH and 340 μM for DXP.32
Initial screening for inhibition was performed with an inhibitor
concentration of 100 μM. IC50 measurements were performed using
six reactions with inhibitor concentrations ranging between 0.01 and
1000 μM. Reactions were initiated by adding DXP and followed
simultaneously in a 96-well plate (UV-Star, Greiner) at 22 ꢀC with a
spectrophotometer (Envision 2140 multilabel reader, PerkinElmer).
Absorbance at 340 nm was measured every 5 s during a 500 s period.
The slope of the linear phase of each reaction was used to calculate the
initial velocity. This was compared to the velocity of the uninhibited
reaction and used to calculate enzyme activity. Enzyme activities were
plotted against the corresponding inhibitor concentration, and data
points were fitted to eq 1,
1
ethyl acetate, 4/1) to yield 0.42 g (64%) of 4b. H NMR (400 MHz,
CDCl3) δ 9.69 (d, J = 7.8 Hz, 1H), 7.83 (d, J = 15.6 Hz, 1H), 7.60 (m,
1H), 7.28 (m, 2H), 7.31 (m, 1H), 6.61 (dd, J = 7.8, 15.6 Hz, 1H); 13C
NMR (100 MHz, CDCl3) δ 193.8, 150.9, 137.8, 133.9, 132.4, 130.9,
128.3, 128.2, 126.0. MS (ESI+): m/z 211 (M + H+), 213 (M + 2 + H+).
Formula: C9H7BrO.
Diethyl 1-(2-Bromophenyl)-3-oxopropylphosphonate (5b).
A solution of aldehyde 4b (0.42 g, 2.00 mmol), phenol (0.49 g, 5.20 mmol),
and triethyl phosphite (0.39 g, 2.40 mmol) was stirred at 100 ꢀC for 2 h.
Then the mixture was evaporated under vacuum and the obtained product
was refluxed for 24 h in a mixture of acetone (17 mL), water (3 mL), and 2
M HCl (8 mL). After that time, the mixture was cooled to room
temperature and extracted with diethyl ether (3 ꢁ 50 mL). The combined
organic layers were dried over anhydrous MgSO4, filtered, and concentrated
under vacuum. The crude product was purified by silica gel column
chromatography using ethyl acetate as eluent to yield 0.50 g (72%) of 5b.
1H NMR (400 MHz, CDCl3) δ9.62 (d, J = 1.6 Hz, 1H), 7.57 (d, J=8.8 Hz,
1H), 7.59 (m, 1H), 7.28 (t, J = 7.4 Hz, 1H), 7.10 (m, 1H), 4.35 (m, 1H),
4.10 (m, 2H), 3.85 (m, 2H), 3.15 (m, 2H), 3.05 (m, 2H), 1.20 (t, J=7.0 Hz,
6H); 13C NMR (100 MHz, CDCl3) δ 198.7 (d, JCꢀP = 15.5 Hz), 135.3
(d, JCꢀP = 5.9 Hz), 133.4, 129.9 (d, JCꢀP = 4.4 Hz), 129.3, 128.1, 126.0
(d, JCꢀP = 9.6 Hz), 63.3 (d, JCꢀP = 6.6 Hz), 62.7 (d, JCꢀP = 6.6 Hz), 44.7 (d,
JCꢀP = 2.2 Hz), 36.7 (d, JCꢀP = 141.5 Hz), 16.6 (d, JCꢀP = 5.9 Hz), 16.4 (d,
JCꢀP = 5.9 Hz). MS (ESI+): m/z 349 (M + H+), 351 (M + 2 + H+).
Formula: C13H18BrO4P.
Diethyl 1-(2-Bromophenyl)-3-(N-hydroxyformamido)propyl-
phosphonate (8c). A mixture of 5b (0.47 g, 1.36 mmol) and O-
benzylhydroxylamine hydrochloride (0.21 g, 1.36 mmol) in a solution of
pyridine/ethanol, 1/1 (5.6 mL), was stirred at room temperature. After 4 h,
TLC (100% ethyl acetate) indicated that the reaction was completed. The
reaction mixture was evaporated and thereafter coevaporated with toluene 3
times. The obtained crude product was dissolved in methanol (20 mL)
together with NaBH3CN (0.26 g, 4.10 mmol), and the solution was stirred at
room temperature for 30 min. After this time the solution was cooled to 0 ꢀC
and 37% HCl (2 mL) was slowly added. The mixture was then allowed to
warm to room temperature, and NaBH3CN (0.05 g, 0.90 mmol) was added;
the resulting mixture was stirred for an additional 2 h. After the reaction was
completed (TLC, ethyl acetate/methanol, 95/5), the mixture was basified by
the addition of 10% NaOH (pH 12) and extracted with DCM. The
combined organic layers were dried, filtered, and evaporated. The crude
product was purified on a silica gel column (DCM/methanol 99/1) to obtain
6b in quantitative yield. Compound 6b (0.27 g, 0.60 mmol) dissolved in
DCM (2 mL) was added to a prepared solution of formic acid (3 mL) and
1,10-carbonyldiimidazole (0.49 g, 3.00 mmol) in DCM (5 mL). After 40 h,
60 mL of water was added, and the mixture was extracted with DCM (2 ꢁ
60 mL). The combined organic layers were dried, filtered, and concentrated.
The crude product was purified by column chromatography on silica gel
(EtOAc) to yield 79% of 7c. Next a DCM solution (5 mL) of formamide 7c
Hi ꢀ Lo
Y ¼ Lo +
ð1Þ
X
1 +
IC50
where Hi is the estimated highest enzyme activity at zero inhibitor
concentration, Lo is the estimated lowest enzyme activity at infinite
inhibitor concentration, X is the concentration of inhibitor, and Y is
the measured enzyme activity. IC50 values presented are the average of
three independent experiments.
Antimycobacterial Activity. The antimycobacterial activity of
the compounds was determined using a resazurin dye based assay in a
96-well V-bottomed plate format as described by Marcel et al.51 Serially
diluted compound solution was added to log phase culture of M.
tuberculosis H37Rv (ATCC no. 27294). The growth of the bacteria
was monitored over 14 days with spectrophotometric readings at 575
and 610 nm. The growth curve and MIC values were computed using
XL-fit (Excel, Microsoft Corp.).
Crystallographic Work. Crystallization trials were performed at
22 ꢀC by the hanging-drop vapor-diffusion method. Drop volumes
4973
dx.doi.org/10.1021/jm2000085 |J. Med. Chem. 2011, 54, 4964–4976

Journal of Medicinal Chemistry
ARTICLE
typically consisted of 2ꢀ4 μL of protein solution (2.8ꢀ4.4 mg/mL in
the final concentration buffer,31 6ꢀ12 mM MnCl2, 10 mM
dithiothreitol) and 2ꢀ4 μL of screening solution. Drops were equili-
brated against a reservoir of 1 mL of screening solution. For cocrystal-
lization trials with either the inhibitor alone or the inhibitor and
NADPH, the protein solution was mixed with at least a 10ꢁ molar
excess of ligands immediately before setup. Compounds 9a and 9c were
dissolved in 10% methanol and 100% DMSO, respectively.
and structural studies. This material is available free of charge via
the Internet at http://pubs.acs.org.
Accession Codes
†Coordinates and structure factor data have been deposited at
the Protein Data Bank with entry codes 2Y1D (MtDXRc-9a),
2Y1F (MtDXRb-9a-NADPH), 2Y1G (MtDXRb-9c), 2Y1E
(MtDXRb), and 2Y1C (MtDXRc).
Our previously reported crystals of MtDXR in complex with
fosmidomycin31 (indicated by MtDXRb) were obtained with a screen-
ing solution consisting of 25% (w/v) PEG 3350, 0.2 M ammonium
sulfate, and 0.1 M Bis-Tris, pH 5.7ꢀ5.9. Three of our new structures
were grown from these same conditions. They include the apo
structures of MtDXRb, MtDXRb-9a-NADPH, and MtDXRb-9c. Apo-
crystals of the protein were used to seed the cocrystallization trials.
Crystals appeared within a few days and grew to average dimensions of
0.2 mm ꢁ 0.1 mm ꢁ 0.03 mm in 1ꢀ2 weeks. For data collection, the
crystals were flash-cooled in liquid nitrogen after a brief soak in a
cryosolution consisting of the screening solution in 25% glycerol and
75 mM NaCl. DMSO, dithiothreitol, cofactors (Mn2+, NADPH), and
ligands (compounds 9a, 9c) were also included, where appropriate, in
the respective cryosolutions.
To avoid the presence of a sulfate (or phosphate) group in the DXP-
binding site, a search for alternative crystallization conditions was
undertaken. As above, drops were seeded 24 h after setup. Crystals of
apo-MtDXR appeared in 17ꢀ20% (w/v) PEG 3350, 0.2 M sodium
formate, pH 7.2 (conditions indicated by MtDXRc). Cocrystals of
MtDXRc-9a were also obtained with these new conditions. The crystals
grew within 1ꢀ2 weeks to dimensions similar as those reported above.
Cryosolutions were prepared in the same manner but substituting
sodium formate for the ammonium sulfate. Soaking times varied from
a few seconds to 12 min.
Details of the data collection and crystallographic refinement are
provided in the Supporting Information. Briefly, two distinct apo
structures were solved and refined at resolutions of 1.90 and 1.65 Å,
with crystallographic R-factors of 19.8% and 18.7%, respectively. A
complex of MtDXR with compound 9a alone was refined at 2.05 Å
resolution (R-factor 21.2%), while a 9a complex with bound NADPH
was refined at 1.96 Å resolution (R-factor 17.4%). A complex with
compound 9c at 1.95 Å resolution (R-factor 18.5%) was also obtained.
Docking Study. Docking was performed using Glide52 in XP mode
with default settings. All inhibitors were docked to the A chain of
MtDXRb-9a. In order to prepare the enzyme for docking, the protein
preparation wizard in Maestro53 was used with default settings. Manual
corrections were made by deleting zero-order bonds to the metal,
correcting the bond order of NADPH, and deleting all water molecules.
Furthermore, the protonation state of His248 was adjusted so that the
protonated nitrogen was oriented toward the inhibitor. The missing side
chains of the active-site flap A199-A204 were added to the protein model
in an open conformation. The docked compounds do not directly
interact with the flexible flap, and the position of the open flap is
therefore not considered to be of significance for the docking results.
The docking site was defined around 9a using the “dock ligands of
similar size” setting, and three docking poses were saved for each ligand.
On the basis of the Suzuki reaction, 441 synthetically feasible
compounds were created by joining compound 9a with different boronic
acids using the software Legion,54 Tripos. Ligprep55 was used to
generate different ionization states, tautomers, and stereoisomers for
each compound.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: +46-18-4714293. Fax: +46-18-4714474. E-mail: anders.
karlen@orgfarm.uu.se.
Author Contributions
#M. Andaloussi and L. M. Henriksson contributed equally to
this work.
’ ACKNOWLEDGMENT
We thank Dr. Aleh Yahorau, Department of Pharmaceutical
Biosciences, Uppsala University, Sweden, for conducting HRMS
analyses. We also acknowledge the Swedish Foundation for
Strategic Research (SSF), the Swedish Research Council (VR),
and the EU Sixth Framework Program NM4TB CT:018 923 for
financial support.
’ ABBREVIATIONS USED
DCM, dichloromethane; DMF, dimethylformamide; DME, di-
methoxyethane; DXP, 1-deoxy-D-xylulose 5-phosphate; DXR,
1-deoxy-D-xylulose 5-phosphate reductoisomerase; EcDXR, DXR
from Escherichia coli; IPP, isopentenyl diphosphate; MEP, 2-C-
methyl-D-erythritol 4-phosphate; MtDXR, DXR from Mycobac-
terium tuberculosis; MIC, minimum inhibitory concentration;
PDB, Protein Data Bank; PfDXR, DXR from Plasmodium
falciparum; rmsd, root mean square deviation; RP HPLC,
reversed phase high performance liquid chromatography; SAR,
structureꢀactivity relationship; THF, tetrahydrofuran; TLC,
thin layer chromatography; TMSBr, trimethylsilyl bromide;
WHO, World Health Organization
’ REFERENCES
(1) Sacchettini, J. C.; Poulter, C. D. Creating isoprenoid diversity.
Science 1997, 277, 1788–1789.
(2) Lichtenthaler, H. K. The 1-deoxy-D-xylulose-5-phosphate path-
way of isoprenoid biosynthesis in plants. Annu. Rev. Plant Physiol. Plant
Mol. Biol. 1999, 50, 47–65.
(3) Rohmer, M. The discovery of a mevalonate-independent path-
way for isoprenoid biosynthesis in bacteria, algae and higher plants. Nat.
Prod. Rep. 1999, 16, 565–574.
(4) Rohmer, M.; Knani, M.; Simonin, P.; Sutter, B.; Sahm, H.
Isoprenoid biosynthesis in bacteria: a novel pathway for the early steps
leading to isopentenyl diphosphate. Biochem. J. 1993, 295, 517–524.
(5) Boucher, Y.; Doolittle, W. F. The role of lateral gene transfer in
the evolution of isoprenoid biosynthesis pathways. Mol. Microbiol. 2000,
37, 703–716.
(6) Spurgeon, S. L.; Porter, J. W. Biosynthesis of Isoprenoid Com-
pounds; Wiley: New York, 1981; pp 1ꢀ46.
(7) Kobayashi, K.; Ehrlich, S. D.; Albertini, A.; Amati, G.; Andersen,
K. K.; Arnaud, M.; Asai, K.; Ashikaga, S.; Aymerich, S.; Bessieres, P.;
Boland, F.; Brignell, S. C.; Bron, S.; Bunai, K.; Chapuis, J.; Christiansen,
L. C.; Danchin, A.; Debarbouille, M.; Dervyn, E.; Deuerling, E.; Devine,
K.; Devine, S. K.; Dreesen, O.; Errington, J.; Fillinger, S.; Foster, S. J.;
’ ASSOCIATED CONTENT
S
Supporting Information. Additionalexperimentaldetails,
b
spectroscopic data, cloning, protein expression and purification,
4974
dx.doi.org/10.1021/jm2000085 |J. Med. Chem. 2011, 54, 4964–4976

Journal of Medicinal Chemistry
ARTICLE
Fujita, Y.; Galizzi, A.; Gardan, R.; Eschevins, C.; Fukushima, T.; Haga,
K.; Harwood, C. R.; Hecker, M.; Hosoya, D.; Hullo, M. F.; Kakeshita, H.;
Karamata, D.; Kasahara, Y.; Kawamura, F.; Koga, K.; Koski, P.; Kuwana,
R.; Imamura, D.; Ishimaru, M.; Ishikawa, S.; Ishio, I.; Le Coq, D.;
Masson, A.; Mauel, C.; Meima, R.; Mellado, R. P.; Moir, A.; Moriya, S.;
Nagakawa, E.; Nanamiya, H.; Nakai, S.; Nygaard, P.; Ogura, M.;
Ohanan, T.; O’Reilly, M.; O’Rourke, M.; Pragai, Z.; Pooley, H. M.;
Rapoport, G.; Rawlins, J. P.; Rivas, L. A.; Rivolta, C.; Sadaie, A.; Sadaie,
Y.; Sarvas, M.; Sato, T.; Saxild, H. H.; Scanlan, E.; Schumann, W.;
Seegers, J.; Sekiguchi, J.; Sekowska, A.; Seror, S. J.; Simon, M.; Stragier,
P.; Studer, R.; Takamatsu, H.; Tanaka, T.; Takeuchi, M.; Thomaides,
H. B.; Vagner, V.; van Dijl, J. M.; Watabe, K.; Wipat, A.; Yamamoto, H.;
Yamamoto, M.; Yamamoto, Y.; Yamane, K.; Yata, K.; Yoshida, K.;
Yoshikawa, H.; Zuber, U.; Ogasawara, N. Essential Bacillus subtilis genes.
Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 4678–4683.
(8) Takahashi, S.; Kuzuyama, T.; Watanabe, H.; Seto, H. A 1-deoxy-
D-xylulose 5-phosphate reductoisomerase catalyzing the formation of
2-C-methyl-D-erythritol 4-phosphate in an alternative nonmevalonate
pathway for terpenoid biosynthesis. Proc. Natl. Acad. Sci. U.S.A. 1998,
95, 9879–9884.
(9) Rodriguez-Concepcion, M.; Campos, N.; Lois, L. M.; Maldonado,
C.; Hoeffler, J. F.; Grosdemange-Billiard, C.; Rohmer, M.; Boronat, A.
Genetic evidence of branching in the isoprenoid pathway for the production
of isopentenyl diphosphate and dimethylallyl diphosphate in Escherichia coli.
FEBS Lett. 2000, 473, 328–332.
(10) Brown, A. C.; Parish, T. Dxr is essential in Mycobacterium
tuberculosis and fosmidomycin resistance is due to a lack of uptake. BMC
Microbiol. 2008, 8, 78.
(11) Jomaa, H.; Wiesner, J.; Sanderbrand, S.; Altincicek, B.;
Weidemeyer, C.; Hintz, M.; Turbachova, I.; Eberl, M.; Zeidler, J.;
Lichtenthaler, H. K.; Soldati, D.; Beck, E. Inhibitors of the nonmevalonate
pathway of isoprenoid biosynthesis as antimalarial drugs. Science 1999, 285,
1573–1576.
(12) Shigi, Y. Inhibition of bacterial isoprenoid synthesis by fosmi-
domycin, a phosphonic acid-containing antibiotic. J. Antimicrob. Che-
mother. 1989, 24, 131–145.
(13) Zeidler, J.; Schwender, J.; Mueller, C.; Wiesner, J.; Weidemeyer,
C.; Beck, E.; Jomaa, H.; Lichtenthaler, H. K. Inhibition of the non-
mevalonate 1-deoxy-D-xylulose-5-phosphate pathway of plant isopre-
noid biosynthesis by fosmidomycin. Z. Naturforsch., C 1998, 53,
980–986.
the treatment of Plasmodium falciparum malaria. Antimicrob. Agents Che-
mother. 2007, 51, 1869–1871.
(20) Dhiman, R. K.; Schaeffer, M. L.; Bailey, A. M.; Testa, C. A.;
Scherman, H.; Crick, D. C. 1-Deoxy-D-xylulose 5-phosphate reducto-
isomerase (IspC) from Mycobacterium tuberculosis: towards understand-
ing mycobacterial resistance to fosmidomycin. J. Bacteriol. 2005, 187,
8395–8402.
(21) Sakamoto, Y.; Furukawa, S.; Ogihara, H.; Yamasaki, M. Fosmi-
domycin resistance in adenylate cyclase deficient (cya) mutants of
Escherichia coli. Biosci., Biotechnol., Biochem. 2003, 67, 2030–2033.
(22) Reichenberg, A.; Wiesner, J.; Weidemeyer, C.; Dreiseidler, E.;
Sanderbrand, S.; Altincicek, B.; Beck, E.; Schlitzer, M.; Jomaa, H. Diaryl
ester prodrugs of FR900098 with improved in vivo antimalarial activity.
Bioorg. Med. Chem. Lett. 2001, 11, 833–835.
(23) Ortmann, R.; Wiesner, J.; Reichenberg, A.; Henschker, D.;
Beck, E.; Jomaa, H.; Schlitzer, M. Acyloxyalkyl ester prodrugs of
FR900098 with improved in vivo anti-malarial activity. Bioorg. Med.
Chem. Lett. 2003, 13, 2163–2166.
(24) Silber, K.; Heidler, P.; Kurz, T.; Klebe, G. AFMoC enhances
predictivity of 3D QSAR: a case study with DOXP-reductoisomerase.
J. Med. Chem. 2005, 48, 3547–3563.
(25) Verbrugghen, T.; Cos, P.; Maes, L.; Van Calenbergh, S. Synthesis
and evaluation of R-halogenated analogues of 3-(acetylhydroxyamino)-
propylphosphonic acid (FR900098) as antimalarials. J. Med. Chem. 2010,
53, 5342–5346.
(26) Zinglꢀe, C.; Kuntz, L.; Tritsch, D.; Grosdemange-Billiard, C.;
Rohmer, M. Isoprenoid biosynthesis via the methylerythritol phosphate
pathway: structural variations around phosphonate anchor and spacer of
fosmidomycin, a potent inhibitor of deoxyxylulose phosphate reducto-
isomerase. J. Org. Chem. 2010, 75, 3203–3207.
(27) Haemers, T.; Wiesner, J.; Van Poecke, S.; Goeman, J.; Henschker,
D.; Beck, E.; Jomaa, H.; Van Calenbergh, S. Synthesis of R-substituted
fosmidomycin analogues as highly potent Plasmodium falciparum growth
inhibitors. Bioorg. Med. Chem. Lett. 2006, 16, 1888–1891.
(28) Woo, Y. H.; Fernandes, R. P. M.; Proteau, P. J. Evaluation of
fosmidomycin analogs as inhibitors of the Synechocystis sp. PCC6803
1-deoxy-D-xylulose 5-phosphate reductoisomerase. Bioorg. Med. Chem.
2006, 14, 2375–2385.
(29) Perruchon, J.; Ortmann, R.; Altenkamper, M.; Silber, K.; Wiesner,
J.; Jomaa, H.; Klebe, G.; Schlitzer, M. Studies addressing the importance of
charge in the binding of fosmidomycin-like molecules to deoxyxylulosepho-
sphate reductoisomerase. ChemMedChem 2008, 3, 1232–1241.
(30) Haemers, T.; Wiesner, J.; Busson, R.; Jomaa, H.; Van Calenbergh,
S. Synthesis of R-aryl-substituted and conformationally restricted
fosmidomycin analogues as promising antimalarials. Eur. J. Org.
Chem. 2006, 3856–3863.
(31) Henriksson, L. M.; Unge, T.; Carlsson, J.; Åqvist, J.; Mowbray,
S. L.; Jones, T. A. Structures of Mycobacterium tuberculosis 1-deoxy-D-
xylulose-5-phosphate reductoisomerase provide new insights into cata-
lysis. J. Biol. Chem. 2007, 282, 19905–19916.
(14) Kuzuyama, T.; Shimizu, T.; Takahashi, S.; Seto, H. Fosmido-
mycin, a specific inhibitor of 1-deoxy-D-xylulose 5-phosphate reductoi-
somerase in the nonmevalonate pathway for terpenoid biosynthesis.
Tetrahedron Lett. 1998, 39, 7913–7916.
(15) Giessmann, D.; Heidler, P.; Haemers, T.; Van Calenbergh, S.;
Reichenberg, A.; Jomaa, H.; Weidemeyer, C.; Sanderbrand, S.; Wiesner,
J.; Link, A. Towards new antimalarial drugs: synthesis of non-hydrolyz-
able phosphate mimics as feed for a predictive QSAR study on 1-deoxy-
D-xylulose-5-phosphate reductoisomerase inhibitors. Chem. Biodiversity
2008, 5, 643–656.
(32) Henriksson, L. M.; Bj€orkelid, C.; Mowbray, S. L.; Unge, T. The
1.9 Å resolution structure of Mycobacterium tuberculosis 1-deoxy-D-
xylulose 5-phosphate reductoisomerase, a potential drug target. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 2006, 62, 807–813.
(16) Wiesner, J.; Henschker, D.; Hutchinson, D. B.; Beck, E.; Jomaa,
H. In vitro and in vivo synergy of fosmidomycin, a novel antimalarial
drug, with clindamycin. Antimicrob. Agents Chemother. 2002, 46,
2889–2894.
(33) Reuter, K.; Sanderbrand, S.; Jomaa, H.; Wiesner, J.; Steinbrecher,
I.; Beck, E.; Hintz, M.; Klebe, G.; Stubbs, M. T. Crystal structure of
1-deoxy-D-xylulose-5-phosphate reductoisomerase, a crucial enzyme in the
non-mevalonate pathway of isoprenoid biosynthesis. J. Biol. Chem. 2002,
277, 5378–5384.
(17) Borrmann, S.; Adegnika, A. A.; Matsiegui, P. B.; Issifou, S.;
Schindler, A.; Mawili-Mboumba, D. P.; Baranek, T.; Wiesner, J.; Jomaa,
H.; Kremsner, P. G. Fosmidomycinꢀclindamycin for Plasmodium
falciparum infections in African children. J. Infect. Dis. 2004, 189,
901–908.
(34) Jones, T. A.; Zou, J. Y.; Cowan, S. W.; Kjeldgaard, M. Improved
methods for building protein models in electron density maps and the
location of errors in these models. Acta Crystallogr., Sect. A 1991, 47, 110–119.
(35) Harris, M.; Jones, T. A. Molray—A Web interface between O
and the POV-Ray ray tracer. Acta Crystallogr., Sect. D: Biol. Crystallogr.
2001, 57, 1201–1203.
(18) Borrmann, S.; Lundgren, I.; Oyakhirome, S.; Impouma, B.;
Matsiegui, P.-B.; Adegnika, A. A.; Issifou, S.; Kun, J. F. J.; Hutchinson, D.;
Wiesner, J.; Jomaa, H.; Kremsner, P. G. Fosmidomycin plus clindamycin
for treatment of pediatric patients aged 1 to 14 years with Plasmodium
falciparum malaria. Antimicrob. Agents Chemother. 2006, 50, 2713–2718.
(19) Oyakhirome, S.;Issifou, S.;Pongratz, P.;Barondi, F.;Ramharter, M.;
Kun, J. F.; Missinou, M. A.; Lell, B.; Kremsner, P. G. Randomized controlled
trial of fosmidomycinꢀclindamycin versus sulfadoxineꢀpyrimethamine in
(36) Read, R. J. Improved fourier coefficients for maps using phases from
partial structures with errors. Acta Crystallogr., Sect. A 1986, 42, 140–149.
4975
dx.doi.org/10.1021/jm2000085 |J. Med. Chem. 2011, 54, 4964–4976

Journal of Medicinal Chemistry
ARTICLE
(37) DeLano, W. L. The PyMOL Molecular Graphics System; DeLano
Scientific: Palo Alto, CA, U.S., 2002.
(38) Lindh, J.; Enquist, P. A.; Pilotti, Å.; Nilsson, P.; Larhed, M.
Efficient palladium(II) catalysis under air. Base-free oxidative Heck
reactions at room temperature or with microwave heating. J. Org. Chem.
2007, 72, 7957–7962.
(39) Battistuzzi, G.; Cacchi, S.; Fabrizi, G. An efficient palladium-
catalyzed synthesis of cinnamaldehydes from acrolein diethyl acetal and
aryl iodides and bromides. Org. Lett. 2003, 5, 777–780.
(40) Larhed, M.; Hallberg, A. Microwave-promoted palladium-
catalyzed coupling reactions. J. Org. Chem. 1996, 61, 9582–9584.
€
(41) Wu, X. Y.; Ohrngren, P.; Ekegren, J. K.; Unge, J.; Unge, T.;
Wallberg, H.; Samuelsson, B.; Hallberg, A.; Larhed, M. Two-carbon-
elongated HIV-1 protease inhibitors with a tertiary-alcohol-containing
transition-state mimic. J. Med. Chem. 2008, 51, 1053–1057.
(42) Alterman, M.; Hallberg, A. Fast microwave-assisted preparation
of aryl and vinyl nitriles and the corresponding tetrazoles from organo-
halides. J. Org. Chem. 2000, 65, 7984–7989.
(43) Devreux, V.; Wiesner, J.; Jomaa, H.; Rozenski, J.; Van der
Eycken, J.; Van Calenbergh, S. Divergent strategy for the synthesis of
R-aryl-substituted fosmidomycin analogues. J. Org. Chem. 2007, 72,
3783–3789.
(44) Simpson, T. R.; Balestra, M.; Brown, D. G.; Dantzman, C. L.;
Ernst, G. E.; Frietze, W.; Holmquist, C. R.; Kang, J.; McLaren, F. M.;
Smith, R. W., Jr.; Woods, J. M. Preparation of N-Propynylthiophene-2-
carboxamides as Alpha 7 Nicotinic Acetylcholine Receptor Modulators
for the Treatment of Mental Disorders and Related Diseases.
WO2007001225 (A1) 20060626, 2007.
(45) Whalley, J. L.; Oldfield, M. F.; Botting, N. P. Synthesis of
[4-13C]-isoflavonoid phytoestrogens. Tetrahedron 2000, 56, 455–460.
(46) Garcia-Sosa, A. T.; Firth-Clark, S.; Mancera, R. L. Including
tightly-bound water molecules in de novo drug design. Exemplification
through the in silico generation of poly (ADP-ribose)polymerase
ligands. J. Chem. Inf. Model. 2005, 45, 624–633.
(47) Lell, B.; Ruangweerayut, R.; Wiesner, J.; Missinou, M. A.;
Schindler, A.; Baranek, T.; Hintz, M.; Hutchinson, D.; Jomaa, H.;
Kremsner, P. G. Fosmidomycin, a novel chemotherapeutic agent for
malaria. Antimicrob. Agents Chemother. 2003, 47, 735–738.
(48) Missinou, M. A.; Borrmann, S.; Schindler, A.; Issifou, S.;
Adegnika, A. A.; Matsiegui, P. B.; Binder, R.; Lell, B.; Wiesner, J.;
Baranek, T.; Jomaa, H.; Kremsner, P. G. Fosmidomycin for malaria.
Lancet 2002, 360, 1941–1942.
(49) Zhang, S. Y.; Tu, Y. Q.; Fan, C. A.; Yang, M.; Zhang, F. M. A
RhCl(PPh3)3/BF3 OEt2 co-promoted direct CꢀC cross-coupling of
3
alcohols at β-position with aldehydes. Tetrahedron Lett. 2009,
50, 4178–4181.
(50) Kuzuyama, T.; Takahashi, S.; Watanabe, H.; Seto, H. Direct
formation of 2-C-methyl-D-erythritol 4-phosphate from 1-deoxy-D-xy-
lulose 5-phosphate by 1-deoxy-D-xylulose 5-phosphate reductoisome-
rase, a new enzyme in the non-mevalonate pathway to isopentenyl
diphosphate. Tetrahedron Lett. 1998, 39, 4509–4512.
(51) Marcel, N.; Nahta, A.; Balganesh, M. Evaluation of killing
kinetics of anti-tuberculosis drugs on Mycobacterium tuberculosis using
a bacteriophage-based assay. Chemotherapy 2008, 54, 404–411.
(52) Glide, version 55211; Schr€odinger, LLC: New York, NY, 2009.
(53) Maestro, version 9.0; Schr€odinger, LLC: New York, NY, 2009.
(54) Legion; Tripos: St. Louis, MO, 1998.
(55) LigPrep, version 2.3; Schr€odinger, LLC: New York, NY, 2009.
4976
dx.doi.org/10.1021/jm2000085 |J. Med. Chem. 2011, 54, 4964–4976