Evaluation of fosmidomycin analogs as inhibitors of the Synechocystis sp. PCC6803 1-deoxy-D-xylulose 5-phosphate reductoisomerase

Article abstract of DOI:10.1016/j.bmc.2005.11.012

Analogs of the antibiotic fosmidomycin, an inhibitor of the methylerythritol phosphate pathway to isoprenoids, were synthesized and evaluated against the recombinant Synechocystis sp. PCC6803 1-deoxy-D-xylulose 5-phosphate reductoisomerase (DXR). Fosfoxac

Full text of DOI:10.1016/j.bmc.2005.11.012

Bioorganic & Medicinal Chemistry 14 (2006) 2375–2385
Evaluation of fosmidomycin analogs as inhibitors of the
Synechocystis sp. PCC6803 1-deoxy-D-xylulose
5-phosphate reductoisomerase
Youn-Hi Woo, Roberta P. M. Fernandes and Philip J. Proteau*
Department of Pharmaceutical Sciences, College of Pharmacy, Pharmacy Bldg. Rm. 203, Oregon State University,
Corvallis, OR 97331-3507, USA
Received 21 September 2005; revised 5 November 2005; accepted 8 November 2005
Available online 28 November 2005
Abstract—Analogs of the antibiotic fosmidomycin, an inhibitor of the methylerythritol phosphate pathway to isoprenoids, were
synthesized and evaluated against the recombinant Synechocystis sp. PCC6803 1-deoxy-D-xylulose 5-phosphate reductoisomerase
(DXR). Fosfoxacin, the phosphate analog of fosmidomycin, and its acetyl congener were found to be more potent inhibitors of
DXR than fosmidomycin.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
2-C-methyl-D-erythritol 4-phosphate (MEP, 3), is cata-
lyzed by 1-deoxy-D-xylulose 5-phosphate reductoisom-
erase (DXR, EC 1.1.1.267; Fig. 1).9–11 In this step,
DXR catalyzes the isomerization of DXP into the inter-
mediate, 2-C-methyl-D-erythrose 4-phosphate (2), which
is subsequently reduced to MEP (3).12 The isomerization
requires a divalent metal ion and the reduction requires
the cofactor NADPH. Shortly after the discovery of
DXR, it was reported that the natural product fosmido-
mycin (4) was a potent inhibitor of the enzyme, validat-
Isoprenoids are one of the largest natural product fam-
ilies with more than 35,000 reported structurally diverse
compounds.1 Despite their amazing structural diversity,
all isoprenoids are derived from the common five carbon
building blocks, isopentenyl diphosphate (IPP) and/or
dimethylallyl diphosphate (DMAPP).2 The biosynthesis
of IPP was long believed to occur only through the mev-
alonate (MVA) pathway, which utilizes acetyl CoA as a
starting molecule. The subsequent revelation of the
methylerythritol phosphate (MEP) pathway, however,
has forever changed our understanding of isoprenoid
biosynthesis.1,3 Because the MEP pathway is dominant
in eubacteria and the malaria parasite Plasmodium
falciparum, but absent in animals,4,5 inhibition of the
pathway holds promise for the development of a new
class of antibiotics with minimal toxicity to humans.6–8
O
CH3
DXR
Mg2+
H3C OH
O
O
P
O
P
HO
O O-
O O-
O-
O-
OH
H OH
DXP (1)
2-C-Methylerythrose
4-phosphate (2)
The first committed step of the MEP pathway, the con-
version of 1-deoxy-D-xylulose 5-phosphate (DXP, 1) to
DXR
O
H3C OH
P
O O-
Keywords: Isoprenoid biosynthesis; Fosmidomycin; Reductoisomer-
ase; Methylerythritol phosphate pathway.
O-
NADPH
Mg2+
OH OH
Abbreviations: DXR, 1-deoxy-D-xylulose-5-phosphate reductoisomer-
ase; DXP, 1-deoxy-D-xylulose 5-phosphate; IPP, isopentenyl diphos-
phate; DMAPP dimethylallyl diphosphate; MEP, methylerythritol
phosphate.
MEP (3)
Figure 1. The conversion of 1-deoxy-D-xylulose 5-phosphate (DXP, 1)
to 2-C-methyl-D-erythritol 4-phosphate (MEP, 3) by deoxyxylulose
5-phosphate reductoisomerase (DXR).
*
Corresponding author. Tel.: +1 541 737 5776; fax: +1541 737 3999;
e-mail: phil.proteau@oregonstate.edu
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.11.012

2376
Y.-H. Woo et al. / Bioorg. Med. Chem. 14 (2006) 2375–2385
ing the MEP pathway as a target for antibacterial
agents.13 Fosmidomycin and its close relative
FR900098 (5) were subsequently found to be effective
at curing malaria in a mouse model system.14 This work
stimulated the development of fosmidomycin as a
human antimalarial treatment and clinical trials are
currently in progress.15–17
the recombinant DXR from the cyanobacterium Syn-
echocystis sp. PCC6803. Prior work in our laboratory
has focused on the Synechocystis DXR21,22 and these
studies provide complementary results to research con-
ducted on the Escherichia coli enzyme. Evaluation of
DXR inhibition by the fosmidomycin analogs should
aid in defining the structural requirements for the
design of potent DXR inhibitors in conjunction with
the X-ray structural data23–27 and QSAR studies28 that
are available.
Due to the potential for fosmidomycin as an antima-
larial agent, many groups have conducted research
on fosmidomycin and its mechanism of action.
Recently several analogs of fosmidomycin have been
reported for their inhibition of DXR or as MEP path-
way inhibitors.18–20 Prompted by the publication of
these articles, we would like to report our investiga-
tions in this area. Several analogs of fosmidomycin
have been synthesized and evaluated as inhibitors of
2. Results
Structurally, fosmidomycin contains a N-formyl, N-hy-
droxyl moiety (hydroxamate) at one end of the molecule
with a phosphonate as a polar group, whereas DXP (1)
and 2 have an a-hydroxycarbonyl group at one end and
a phosphate group at the other end. Previous attempts
to shorten the length of the propylene spacer
compromised antibacterial activity,29 so we decided to
focus on changing the hydroxamate and polar phospho-
nate groups, rather than altering the spacer. The polar
phosphonate head group of 4/5 was modified to a phos-
phate (6, 7), a carboxylate (8), and a sulfamate (9). The
N-formyl, N-hydroxyl moiety was substituted with N-
formyl, N-methyl, and N-acetyl, N-methyl groups (10,
11) or with the reversed hydroxamate (12). Figure 2
summarizes the analogs used in this study. Other inves-
tigators have also prepared analogs that replace the
hydroxamate or phosphonate group with various
groups.18–20,30,31
OH
N
O
OH
N
O
P
O
P
O
O OH
OH
OH
OH
R
R
R = H, fosfoxacin (6)
R = CH3, 7
R = H, fosmidomycin (4)
R = CH3, FR900098 (5)
O
S
O
OH
N
O
OH
N
O
O
O
OH
O
NH2
H
R
H
8
9
The syntheses of the analogs utilized a combination of
known synthetic transformations and the synthetic
routes are illustrated in Schemes 1–4. Because the main
goal was to prepare the analogs for subsequent analysis
as DXR inhibitors, rather than to maximize yields, fo-
cused efforts were not made to optimize the individual
steps. During the course of our research, syntheses of
two of the inhibitors, the carboxylate 8 and the reversed
hydroxamate 12, have been published.18,31
O
O
CH3
N
O
P
OH
HO
P
N
H
OH
OH
OH
12
R = H, 10
R = CH3, 11
Figure 2. Fosmidomycin and its analogs.
a, b
c, d
HO
TBSO
(BnO)HN
OH
OP(O)(OBn)2
14
OP(O)(OBn)2
15
13
OBn
N
OH
O
O
O
O
N
P
g
OP(O)(OBn)2
e
f
O OH
OH
H
H
16
6
OBn
N
OH
N
O
O
g
P
OP(O)(OBn)2
O OH
CH3
OH
CH3
17
7
Scheme 1. Synthesis of fosfoxacin (6) and its acetyl congener 7. Reagents: (a) TBSCl, DMF, TEA, 99%; (b) 1-H-tetrazole, (BnO)2PN(i-Pr)2,
t-BuOOH, 79%; (c) Dowex 50, H+, 87%; (d) Tf2O, 2,6-lutidine; NH2OBn, 83%; (e) 4-formyl-2-methyl-1,3,4-thiadiazole-5-thione (FMT), CH2Cl2;
(f) Ac2O, pyridine; (g) 10% Pd/C, H2; 95% for (e + g) and 95% for (f + g).

Y.-H. Woo et al. / Bioorg. Med. Chem. 14 (2006) 2375–2385
2377
and introduction of the protected phosphate group in
14 proceeded smoothly, but displacement of an activat-
ed hydroxyl group by the protected hydroxylamine re-
quired several attempts. No success was achieved with
the common leaving groups such as methanesulfonyl
(MsO-), p-toluenesulfonyl (TsO-), or via a Mitsunobu-
type substitution. After this survey of several methods,
transformation to the triflate with trifluoromethanesulf-
onic anhydride (Tf2O) followed by displacement with
O-benyzlhydroxylamine resulted in the formation of 15
with a reasonable yield (>80%). Intermediate 15 was
N-formylated with 4-formyl-2-methyl-1,3,4-thiadiazole-
5-thione (FMT) and both protecting groups were effi-
ciently removed by catalytic hydrogenation to provide
fosfoxacin (6). The phosphate analog (7) of FR900098
also was prepared from 15 in two steps.
O
O
a
Br
Ts(BnO)N
OEt
OEt
18
19
O
OH
HN
O
OH
N
b
c
O
OH
OH
H
20
8
Scheme 2. Synthesis of carboxylate analog 8. Reagents: (a) NaH,
HN(Ts)OBn, 99%; (b) HCl/AcOH, 1:1, 80%; (c) acetylformic anhy-
dride, 72%.
O
a, b
c, d
TBSO
HO
e
S
OH
O
O
NH2
The synthesis of the carboxylic acid analog 8 started
with the introduction of N-Ts-O-benzyl-hydroxylamine
to 4-bromoethylbutanoate (18) by a modification of a
procedure previously used for the synthesis of fosmido-
mycin (Scheme 2).33 The tosyl, benzyl, and ester-protect-
ing groups of 19 were removed by treatment with acid.
Formylation with the mixed acetylformic anhydride
produced the desired compound 8. During our studies
there was a report on the syntheses of analog 8 and sev-
eral related carboxylic acid analogs of fosmidomycin.31
These carboxylic acid analogues, however, were not
analyzed for inhibition of DXR.
O
22
21
OH
O N
O
OBn
O N
S
O
S
O
O
NH2
O
NH2
H
H
9
23
Scheme 3. Synthesis of sulfamate analog 9. Reagents: (a) ClSO2NH2,
DMA, TEA; (b) AG-50, MeOH, 50% for two steps; (c) Tf2O, 2,6-
lutidine, BnONH2, 80%; (d) FMT, CH2Cl2, 90%; (e) 10% Pd/C,
MeOH, 50%.
The synthesis of the sulfamate analog 9 began with the
monoprotected ethylene glycol 21, which was sulfamoy-
lated with sulfamoyl chloride (Scheme 3).34 The interme-
diate 22, formed by removal of the silyl protecting
group, was converted into 23 by displacement of the tri-
flate with the protected hydroxylamine. Formylation
with FMT and final removal of the benzyl protecting
group by catalytic hydrogenation provided the sulfa-
mate analog 9.
CH3
TsN
c
a, b
P(O)(OEt)2
CH3
Br
Br
25
24
CH3
HN
d or e
O
N
P(O)(OH)2
P(O)(OH)2
R
The N-methyl fosmidomycin and FR900098 analogs 10
and 11 were prepared by following a synthetic route
analogous to an early synthesis of fosmidomycin,
replacing N-tosyl-O-benzylhydroxylamine with N-meth-
yl-N-tosylamine (Scheme 4).35 The bromo groups of 1,3-
dibromopropane (24) were displaced sequentially with
N-methyl-N-tosylamine and the anion of diethylphosph-
ite to provide 25. Acidic deprotection provided the com-
mon intermediate 26, which was converted to both the
N-formyl and N-acetyl derivatives. The N-acylation step
in both cases proceeded in unacceptably low yields, but
sufficient material was obtained for evaluation as DXR
inhibitors.
26
R = H (10)
R = CH3 (11)
Scheme 4. Synthesis of N-methyl analogs 10 and 11. Reagents: (a)
NaH, NH(Ts)CH3, 50%; (b) NaH, HP(O)(OEt)2, 70%; (c) HCl:AcOH,
1:1, 70%; (d) acetylformic anhydride, 15%; (e) Ac2O, pyridine, 13%.
Originally, the phosphate analog offosmidomycin (6) was
selected for comparison of the binding effects of the phos-
phonate in fosmidomycin versus a phosphate, as in the
substrate DXP. Surprisingly, a literature search on the
phosphate analog revealed that this compound is actually
a known natural product named fosfoxacin (6).32 Fosfox-
acin was isolated from the filtrate of Pseudomonas fluores-
cens PK-52 in 1990 and was shown to have antibacterial
activity. Despite this prior report and the structural simi-
larity to fosmidomycin, no further testing of this com-
pound for inhibition of DXR has been reported.
The synthesis of the reversed hydroxamic acid derivative
12 used a synthetic route that was essentially the same as
the route recently published for this same compound
prepared during studies with the E. coli DXR.18 Analyt-
ical data for all intermediates and the final product 12
matched well with the published data.18
In spite of the simple structures, the syntheses of the
phosphate analogs 6 and 7 posed some challenges
(Scheme 1). Monoprotection of ethylene glycol (13)
All final compounds were tested against the recombi-
nant Synechocystis DXR for inhibition. In the initial

2378
Y.-H. Woo et al. / Bioorg. Med. Chem. 14 (2006) 2375–2385
characterization of fosmidomycin as an inhibitor of the
E. coli DXR, it was reported that fosmidomycin showed
a mixed (competitive and non-competitive with DXP)
inhibition pattern with a Ki value of 37 nM.13 Fosmido-
mycin and its analogs also exhibited mixed inhibition
patterns with the Synechocystis DXR, with varying Ki
values (Table 1). The initial Ki value for fosmidomycin
(4) was found to be 58 nM, while fosfoxacin (6) and
its acetyl congener 7 have somewhat lower Ki values of
19 and 2 nM, respectively. The carboxylate and sulfa-
mate derivatives of fosmidomycin had much higher Ki
values of 240 lM and 2.8 mM, indicating the impor-
tance of the phosphonate or phosphate group for high
affinity to DXR. The N-methyl analogs 10 and 11 were
weak inhibitors, with Ki values in the low mM range.
The reversed hydroxamic acid analog 12 was an effective
inhibitor of DXR, although the Ki of 4 lM is approxi-
mately 70-fold higher than the Ki for fosmidomycin.
fosmidomycin (58 nM) and the reversed hydroxamate
(4 lM) both underestimate the Kꢀi values for these inhib-
itors (4 nM and 2.4 lM, respectively). The Ki (900 nM)
and Kꢀi (0.9 nM) values for FR900098 indicate that this
compound is a very potent inhibitor of the Synechocystis
DXR. A second method, which relies upon pre-incuba-
tion with the inhibitor prior to addition of the substrate,
was also used to estimate the Kꢀi .38 Using this approach,
DXR was pre-incubated with the various concentrations
of the inhibitors for 10 minutes prior to addition of sub-
strate. The Kꢀi s were then determined by plotting the
inhibitor concentration versus 1/v.38 The inhibition con-
stants are shown in Table 2. As can be seen from the ta-
ble, the Kꢀi values determined in this manner for
fosmidomycin and FR900098 are almost the same as
the values obtained from the progress curves. For the
hydroxamate analog, however, the Kꢀi value is almost
7-fold lower.
Although the initial report characterized fosmidomycin
as a mixed inhibitor of E. coli DXR, further studies
demonstrated that fosmidomycin is actually a slow,
tight-binding inhibitor of DXR.36 Because of this deter-
mination of fosmidomycin as a slow, tight-binding
inhibitor of DXR, we re-examined the inhibition of
the Synechocystis DXR with three inhibitors, fosmido-
mycin (4), FR900098 (5), and the reversed hydroxamate
12. According to this mechanism of inhibition, two dif-
ferent inhibition constants can be obtained. Ki describes
the formation of the EI complex, while Kꢀi represents
conversion to the tightly bound complex and can be
considered the overall inhibition constant.37 Analyses
of the progress curves for the three inhibitors indicated
slow, tight-binding inhibition. The two inhibition con-
stants are presented in Table 2 for each inhibitor. The
original mixed inhibition constants determined for
3. Discussion
Fosmidomycin was the first reported inhibitor of DXR
and today remains one of the most potent inhibitors
of this enzyme. Although initially described as a mixed
inhibitor of the E. coli DXR13 and as a competitive
inhibitor of the Zymomonas mobilis DXR,39 subsequent
work has shown that fosmidomycin is a slow, tight-
binding inhibitor of DXR, with an initial phase of inhi-
bition that is competitive with DXP and a second phase
of inhibition that is non-competitive.36 A conformation-
al change in the enzyme has been proposed for this sec-
ond phase and a recent X-ray crystal structure of the E.
coli DXR with fosmidomycin and NADPH bound at
the active site is proposed to represent this closed con-
formation of the enzyme.27
A variety of natural and synthetic analogs of fosmidomy-
cin have been reported over the years, including the nat-
ural product FR900098, a compound which is a more
potent inhibitor of the P. falciparum DXR than fosmido-
mycin.14 In addition to fosmidomycin and FR900098,
two other natural phosphonate compounds, FR-32863
and FR-33289, have been characterized, one with a dou-
ble bond in the carbon backbone and one with a hydrox-
yl in the chain.40 Of these natural phosphonates,
fosmidomycin and FR-33289 were typically more potent
as antibacterial agents than FR-900098, with FR-32863
having minimal antibacterial activity.41 No reports have
been presented for FR-32863 or FR-33289 as DXR
inhibitors. The length of the carbon chain connecting
the hydroxamic acid group of fosmidomycin with the
phosphonate also is critical for optimal antibacterial
activity. Analogs with a two-carbon chain are not effec-
tive as antibacterial agents, suggesting a poorer inhibi-
tion of DXR.29 The same is true for methyl phosphinic
acid analogs of fosmidomycin and FR-900098, which
also do not have antibacterial effects.29
Table 1. Ki values for fosmidomycin analogs with Synechocystis DXR
Compound
Ki (lM)a,b
4
6
0.057
0.019
0.002
240
7
8
9
2800
5000
3600
4
10
11
12
a Inhibition assays utilized at least four different concentrations of each
analog, in triplicate.
b All compounds displayed mixed inhibition as determined from
Lineweaver–Burk plots.
Table 2. Slow, tight-binding inhibition constants for inhibitors of
Synechocystis DXR determined either by estimation from progress
curves or by pre-incubation studies
Inhibitor
Progress curves
Pre-incubation
Ki (nM) Kꢀi (nM) Kꢀi ðnMÞ
For comparison with the inhibition values with the
E. coli enzyme,18,36 fosmidomycin and FR900098 were
both examined as inhibitors of the Synechocystis
DXR. An initial evaluation of fosmidomycin showed
Fosmidomycin (4)
FR900098 (5)
Hydroxamate analog 12 5000
900
4
0.9
4
500
2
350
2400

Y.-H. Woo et al. / Bioorg. Med. Chem. 14 (2006) 2375–2385
2379
mixed inhibition of DXR with a Ki = 57 nM. A later
examination as a slow, tight-binding inhibitor led to
values of Ki = 900 nM and Kꢀi = 4 nM. Both of these
values compare favorably with the published values
for the E. coli enzyme (Ki = 215 or 40 nM and
Kꢀi ¼ 21 or 10 nM). The FR900098 analog was slightly
more potent than fosmidomycin, with a Ki value about
2-fold less and a Kꢀi value 2- to 4-fold less. The more po-
tent nature of FR900098 is consistent with the previous-
ly observed inhibition of the P. falciparum DXR by
FR900098.14
the Synechocystis DXR. Compound 12 also displayed
slow, tight-binding inhibition to the Synechocystis
DXR, but the Ki (5 lM) and Kꢀi (2.4 lM) values deter-
mined from the progress curves were much higher than
those observed with the E. coli enzyme. Using a pre-in-
cubation approach, however, produced a lower Kꢀi value
of 350 nM. This lower Kꢀi value with the pre-incubation
studies suggests that the conformational change associ-
ated with the slow, tight-binding inhibition by the
hydroxamate takes longer than the change associated
with fosmidomycin. The progress curves for the four-
minute assay for 12 apparently led to this underestima-
tion of Kꢀi . Even the lower Kꢀi value is approximately 5-
fold higher than the Kꢀi value for 12 with the E. coli
DXR. The nature of this difference in inhibitor potency
is as yet unclear, but a comparison of the sequence align-
ments reveals that all of the active site residues shown to
act directly with either fosmidomycin or DXP in the E.
coli crystal structures are completely conserved in the
Synechocystis enzyme. Apparently subtle differences in
how these residues align at the active sites of the E. coli
and Synechocystis DXRs lead to a greater difference in
inhibition between fosmidomycin and its reversed
hydroxamate derivative in the Synechocystis DXR.
The phosphate analogs of fosmidomycin and
FR900098, fosfoxacin and the acetyl congener, were
determined to be more potent mixed inhibitors of the
Synechocystis DXR than fosmidomycin, with the acetyl
congener of fosfoxacin having a Ki of 2 nM. Although
the Ki and Kꢀi values for slow, tight-binding inhibition
were not determined for these compounds, it is likely
that the Kꢀi values would be less than the mixed inhibi-
tion Ki values, indicating that these phosphate analogs
are the most potent inhibitors reported of DXR to date.
Despite its low Ki value and its associated antibacterial
activity,32 the development of fosfoxacin as a new anti-
bacterial or antimalarial agent is likely to be limited by
its unfavorable molecular properties. The phosphate
group of fosfoxacin, which contributes significantly to
the low Ki value, would be subjected to cleavage by
phosphatases, rapidly inactivating this compound as a
DXR inhibitor.
The broad range of inhibition potencies of these analogs
provides information about the structural requirement
for the design of better inhibitors. At the initiation of
this project, it had been proposed that fosmidomycin
might be mimicking the intermediate 2-C-methyleryth-
rose-4-phosphate (2) because each had a formyl carbon
and phosphorous atom separated by five chemical
bonds.13 However, the first X-ray structure of the E. coli
DXR with fosmidomycin bound suggested that fosmi-
domycin binds in a position more similar to the sub-
strate DXP than to the intermediate.25 The analog
studies add to the knowledge from the structural studies.
Potent inhibitors should contain either a phosphate or
phosphonate for high affinity, although the metabolic
liability of the phosphate group argues for use of a phos-
phonate. The hydroxamate group is also highly sensitive
to changes, with even minor alterations leading to a de-
crease in inhibition. These results along with prior work
on fosmidomycin inhibition suggest that it will be diffi-
cult to improve upon the inhibition qualities of fosmido-
mycin and FR900098.
The other two changes to the phosphonate head group,
substitution with either a carboxylate or a sulfamate, led
to large increases in the Ki values. The carboxylate ana-
log had a Ki of 240 lM, while the sulfamate group had a
Ki approximately 12-fold higher than that of 8. The
presence of two ionizable groups on the phosphonate
and phosphate must play key roles in the high potency
of inhibition. The phosphonate (2nd pKa ꢁ 8) and the
phosphate (2nd pKa ꢁ 7)42 groups can potentially bind
as mono- or dianions, while the carboxylate can only
be a monoanion and the sulfamate is neutral. Even if
the phosphonate or phosphate binds as a monoanion,
the remaining ionizable group could be involved in a
key hydrogen bond. In the crystal structures, the pH
of the crystallization solution was 5–6, so both phos-
phate and phosphonate groups would be expected to
be monoanions.
4. Experimental
The removal of the N-hydroxy group would clearly be
expected to greatly decrease the potency of the analogs,
because the compounds would no longer be able to che-
late the divalent metal ion. The N-methyl analogs clearly
demonstrate this phenomenon. With the absence of a
group capable of chelating the metal ion, the inhibition
constants increase to millimolar concentrations.
Nuclear magnetic resonance (NMR) spectra were
recorded on Bruker AM 400, AC 300 or DRX 600
instruments. 1H and 13C NMR chemical shifts were
reported as d using the residual solvent signal as an
internal standard (CHCl3 at 7.26 ppm for 1H NMR
and 77.00 ppm for the centerline of the CDCl3 triplet
for 13C NMR samples). Either tert-butanol (31.21 ppm
for the CH3) or MeOH-d4 (49.15 ppm for the centerline
of the quintet) was added as an internal standard for the
samples in D2O for referencing 13C NMR spectra. Phos-
phoric acid (85%) was used as an external standard for
31P NMR spectra (0.00 ppm). Low- and high-resolution
chemical ionization and fast atom bombardment mass
Initial analysis of the reversed hydroxamate analog of
fosmidomycin resulted in mixed inhibition with a Ki of
4 lM. The recent report of this same analog as a slow,
tight-binding inhibitor of the E. coli DXR with a
Ki = 169 nM and a Kꢀi ¼ 68 nM,18 however, led to a
re-examination of the nature of the inhibition of 12 with

2380
Y.-H. Woo et al. / Bioorg. Med. Chem. 14 (2006) 2375–2385
spectra (CIMS and FABMS) were obtained on a Kratos
MS50TC spectrometer. Electrospray mass spectra were
obtained on a PerkinElmer SCIEX API3 spectrometer.
Silica gel (Merck, grade 60, 220–400 mesh) was used
for flash chromatography. The solvents used for the syn-
thetic procedures were dried before use as commonly
recommended.43 Commercial grade reagents and the
starting materials were purchased from Sigma–Aldrich
and used without further purification. Acetylformic
anhydride was prepared as described.44 The water used
in the experiments was purified through a Milli Q water
purification system (Millipore). The cultured bacterial
cells were disrupted by sonication with a XL2000 Micro-
sonTM (Misonix) ultrasonic cell disruptor. The bacterial
culture samples were centrifuged in a Beckman JI-HS or
Beckman J2-HS centrifuge at the temperature as indi-
cated. Unless stated otherwise, synthetic reactions were
performed in oven-dried glassware under a positive pres-
sure of argon.
(121 MHz, CDCl3) d (ppm) 0.16 (br s); HRCIMS
calcd for C22H34O5PSi: 437.19132. Found: 437.19076
(M+H).
4.3. Phosphoric acid, dibenzyl 2-hydroxyethyl ester
Compound 14 (50 mg, 0.1 mmol) was dissolved in meth-
anol prior to the addition of 1 g Dowexꢂ 50Wx8-100
(H+ form) resin. This heterogeneous mixture was stirred
for 12 h at rt. The resin was removed by filtration and
the filtrate was concentrated to a pale yellow oil, which
was purified by flash chromatography (5% MeOH/
CHCl3) to provide the product as a colorless oil
(32 mg, 87%). IR (neat) 3582, 1230 cmꢂ1
;
1H NMR
(300 MHz, CDCl3) d (ppm) 7.37–7.34 (m, 10 H), 5.06
(d, JHP = 8.5 Hz, 2H), 5.05 (d, JHP = 8.4 Hz, 2H),
2
3
3
4.08 (dt, J = 4.4 Hz, JHP = 9.4 Hz, 2H), 3.73 (dt,
4
J = 4.4 Hz, JHP = 0.9 Hz, 2H), 2.64 (br s, 1H); 13C
2
NMR (75 MHz, CDCl3) d (ppm) 62.01 (d, JCP
=
5.3 Hz), 69.59 (d, 2JCP = 5.6 Hz), 69.82 (d,
3JCP = 6.1 Hz), 128.02, 128.62, 128.68, 135.61 (d,
3JCP = 6.5 Hz); 31P NMR (121 MHz, CDCl3) d (ppm)
1.19; HRFABMS calcd for C16H20O5P: 323.10484.
Found: 323.10496 (M+H).
4.1. 2-(tert-Butyldimethylsilyloxy)-ethanol45
To a solution of ethylene glycol (13; 1 g, 16.1 mmol), tri-
ethylamine (TEA, 2.5 mL, 24.1 mmol), and DMAP
(20 mg, 0.16 mmol) in CH2Cl2 (120 mL) was added
TBDMSCl (2.91 g, 19.3 mmol) in 5 portions over 1 h
at 0 ꢁC. The resulting heterogeneous reaction mixture
was gradually warmed to rt. The mixture was stirred
for 12 h before dilution with water and CH2Cl2. The
organic layer was washed successively with solutions
of saturated aq NaHCO3, saturated aq NH4Cl, water
and brine, dried over MgSO4, filtered, and concentrated
in vacuo. The pale yellow oil was purified by vacuum
chromatography46 (SiO2, EtOAc/hexanes, 1:3) to give
a colorless oil (2.9 g, 99%).
4.4. Phosphoric acid, dibenzyl 2-benzyloxyamino-ethyl
ester (15)
To a solution of phosphoric acid, dibenzyl 2-hydroxy-
ethyl ester (67 mg, 0.21 mmol) in CH2Cl2 (25 mL) was
added 2,6-lutidine (30 lL, 0.25 mmol) under an argon
atmosphere. The solution was cooled to ꢂ78 ꢁC before
adding trifluoromethane sulfonic anhydride (39 lL,
0.21 mmol) dropwise. After the reaction mixture was
stirred at ꢂ78 ꢁC for 30 min,O-benzyl hydroxylamine
(39 mg, 0.31 mmol) in CH2Cl2 was added dropwise.
The solution was stirred for 1 h at ꢂ78 ꢁC, warmed to
rt, and stirred for another 2 h. The reaction mixture
was diluted with CH2Cl2 and washed with saturated
aqueous solutions of NH4Cl, NaHCO3, water and brine,
dried over MgSO4, filtered, and concentrated to a pale
yellow oil in vacuo. The resulting residue was purified
by flash chromatography (EtOAc/hexanes, 2:1) to yield
the product as a colorless oil (74 mg, 83%): IR (neat)
4.2. Phosphoric acid, dibenzyl 2-(tert-butyldimethylsilyl-
oxy) ethyl ester (14)
To a solution of 2-(tert-butyldimethylsilyloxy)-ethanol
(100 mg, 0.6 mmol) in CH2Cl2 (30 mL) were added 1-
H-tetrazole (80 mg, 1.2 mmol) and dibenzyl diisopropyl
phosphoramidite (390 lL, 1.2 mmol). The resulting
solution was stirred for 4 h at rt under argon at rt before
tert-butyl hydroperoxide (70% solution, 0.5 mL) was
added dropwise at 0 ꢁC. The reaction mixture was
warmed to rt and stirred for 8 h. The mixture was dilut-
ed with CH2Cl2, and the organic layer was washed with
water twice, and the combined aqueous layers were
re-extracted with CH2Cl2 (5 mL · 3). The combined
organic layer was washed with a saturated aq solution
of sodium bicarbonate, water and brine, dried with
MgSO4, and concentrated in vacuo. The concentrate
was purified by flash chromatography (EtOAc/hexanes,
1:3) to afford the desired product (196 mg, 79%). IR
(neat) 1251, 1007, 765 cmꢂ1; 1H NMR (300 MHz, CDCl3)
d (ppm) 7.31–7.37 (m, 10H), 5.05 (d, 3JHP = 8.0 Hz, 2H),
1
1253, 1011, 732 cmꢂ1; H NMR (400 MHz, CDCl3) d
3
(ppm) 7.31–7.38 (m, 15H), 5.04 (d, JHP = 8.3 Hz, 2H),
3
5.03 (d, JHP = 8.3 Hz, 2H), 4.67 (s, 2H), 4.15 (dt,
3
J = 5.1 Hz, JHP = 7.2 Hz, 2H), 3.07 (dt, J = 5.1Hz,
4JHP = 1.2 Hz, 2H); 13C NMR (100 MHz, CDCl3) d
(ppm) 137.75, 135.76 (d, JCP = 6.2 Hz), 128.56,
128.32, 127.94, 127.82, 128.47, 128.36, 76.04, 69.35 (d,
3
2
2JCP = 5.4 Hz), 64.25 (d, JCP = 5.8 Hz), 51.68 (d,
3JCP = 7.3 Hz); 31P NMR (121 MHz, CDCl3) d (ppm)
0.38; HRFABMS calcd for C23H27O5NP: 428.16269.
Found: 428.16377 (M+H).
4.5. Phosphoric acid, dibenzyl 2-(N-benzyloxy-N-formyl-
amino)-ethyl ester (16)
3
3
5.04 (d, JHP = 7.9 Hz, 2H), 4.04 (dt, JHP = 7.1 Hz,
J = 5.2 Hz, 2H) 3.76 (dt, J = 5.2 Hz, 4JHP = 1.1 Hz, 2H),
0.88 (s, 9H), 0.04 (s, 6H); 13C NMR (75 MHz, CDCl3) d
To a solution of 15 (74 mg, 1.17 mmol) in CH2Cl2
(20 mL) was added FMT47 (50 mg, 0.31 mmol) at rt.
The resulting solution was stirred for 12 h. The reac-
tion mixture was washed with water, and the aqueous
3
(ppm) 135.88 (d, JCP = 6.9 Hz), 128.53, 128.45, 127.88,
2
69.18 (d, JCP = 5.5 Hz), 68.61 (d, JCP = 5.9 Hz), 62.10
2
3
(d, JCP = 7.9 Hz), 25.81, 18.29, ꢂ5.37; 31P NMR

Y.-H. Woo et al. / Bioorg. Med. Chem. 14 (2006) 2375–2385
2381
layer was re-extracted with CH2Cl2 (5 mL · 2). The
combined organic layers were washed with water and
brine, dried over MgSO4, filtered, and concentrated
to a yellow film. The crude residue was purified by
flash chromatography (EtOAc/hexanes, 1:2) to afford
a colorless film (77 mg, 98%): IR (neat) 1689, 1280,
3.88; HRFABMS calcd for C3H7NO6P: 184.00110.
Found: 184.00101 (MꢂH).
4.8. Phosphoric acid mono- [2-(N-acetyl-N-hydroxy-
amino)-ethyl]-ester (7)
1
1018 cmꢂ1; H NMR (300 MHz, CDCl3) d (ppm) two
A solution of 17 (62 mg, 0.13 mmol) in methanol
(10 mL) was stirred with a catalytic amount of 10%
Pd/C under a hydrogen atmosphere (1 atm) at rt over-
night. The Pd/C was removed by filtration, and the fil-
trate was concentrated to a pale yellow film. The
residue was redissolved in water, and the aqueous layer
was washed with chloroform (5 mL · 3) before being ap-
plied to a CF-11 cellulose column (2.5 · 30 cm). The de-
sired compound was eluted with a continuous gradient
of 0 to 1% (0.1% TFA/THF) in water. Five milliliter
fractions were collected, and the PMA-active fractions
were collected and freeze-dried to a white solid (24 mg,
amide rotamers were present as a 2:1 ratio: major—
8.18 (br s, 1H), 7.32–7.35 (m, 15H), 5.02 (dd,
2
3JHP = 9.8 Hz, JHH = 1.8 Hz, 4H), 4.13 (m, 2H), 3.77
(m, 2H): minor—7.87 (br s, 1H), 7.32–7.35 (m, 15H),
3
3
5.03 (d, JHP = 8.3 Hz, 2H), 5.02 (d, JHP = 8.3 Hz,
2H), 4.10 (m, 2H), 3.38 (m, 2H); 13C NMR (75 MHz,
CDCl3)
134.13, 129.48, 129.11, 128.71, 128.56, 127.98, 78.15,
d (ppm) major—163.83, 135.63, 135.58,
2
69.47 (d, JCP = 5.8 Hz), 63.14, 49.17: minor—158.79,
135.63, 135.58, 134.13, 129.48, 129.11, 128.71, 128.56,
2
127.98, 78.15, 69.25 (d, JCP = 5.3 Hz), 62.44, 45.25;
31P NMR (121 MHz, CDCl3) d (ppm) ꢂ0.03, 0.26;
HRFABMS calcd for C24H27O6NP: 456.15760. Found:
456.15872 (M+H).
93%). H NMR (600 MHz, D2O) d (ppm) two amide
1
rotamers; 2.17 (s) and 2.16 (s), 3.78–3.83 (m, 2H),
4.00–4.02 (m, 2H); 13C NMR (100 MHz, D2O) d
(ppm) 160.95 and 164.92, 60.96 (d, JCP = 2.0 Hz),
2
3
4.6. Phosphoric acid, dibenzyl 2-(N-acetyl-N-benzyloxy-
amino)-ethyl ester (17)
58.53 and 51.63 (d, JCP = 6.6 Hz), 10.84; 31P NMR
(121 MHz, D2O) d (ppm) 1.53; HRFABMS calcd for
C4H9O6NP: 198.01675. Found: 198.01695 (MꢂH).
Compound 15 (74 mg, 1.17 mmol) was dissolved in a 1:1
mixture of Ac2O and pyridine (25 mL), and stirred for
12 h at rt. The solvent was removed in vacuo and the
residue was purified by flash silica column chromatogra-
phy (EtOAc/hexanes, 1:1) to give a colorless oil (79 mg,
4.9. 4-[N-Benzyloxy-N-(toluene-4-sulfonyl)-amino]-
butyric acid ethyl ester (19)
Solid sodium (30 mg, 1.23 mmol) was added to dry eth-
anol (45 mL) and the mixture was warmed to 60 ꢁC.
After the mixture became clear, it was cooled to rt be-
fore being treated with N-benzyloxy-toluene-4-sulfon-
amide (312 mg, 1.13 mmol) in ethanol (5 mL). The
reaction mixture was stirred for 2 h at rt. To this solu-
tion was added 4-bromobutyric acid ethyl ester (18)
(200 mg, 1.02 mmol) dropwise, and the resulting solu-
tion was refluxed for 12 h. The reaction was quenched
by water and the solvent was removed in vacuo. The
resulting yellow residue was dissolved in EtOAc and
the organic layer was washed with saturated aq NH4Cl,
water and brine, dried over MgSO4, filtered, and con-
centrated to a yellow oil. The residue was purified with
flash chromatography (EtOAc/hexanes, 1:9) to give
403 mg of a colorless oil (>99%): IR (neat) 1734, 1364,
98%): IR (neat) 1672, 1280, 1018 cmꢂ1 1H NMR
;
(400 MHz, CDCl3) d (ppm) 7.32–7.37 (m, 15H), 5.02
(d, JHP = 8.2 Hz, 2H), 5.01 (d, JHP = 8.1 Hz, 2H),
3
3
3
3
4.79 (s, 2H), 4.16 (dt, JHP = 7.4 Hz, JHH = 5.6 Hz,
2H), 3.83 (br s, 2H), 2.05 (s, 3H); 13C NMR
(100 MHz, CDCl3) d (ppm) 172.93, 135.73, 135.67,
134.30, 129.28, 129.00, 128.73, 128.59, 128.39, 127.98,
76.82, 69.39 (d, 2JCP = 5.3 Hz), 63.72 (d, 3JCP = 5.4 Hz),
46.44, 20.44; 31P NMR (121 MHz, CDCl3) d (ppm) 0.04;
HRFABMS calcd for C25H29O6NP: 470.17325. Found:
470.17204 (M+H).
4.7. Phosphoric acid, mono- [2-(N-formyl-N-hydroxy-
amino)-ethyl] ester (fosfoxacin; 6)
1
A solution of 16 (60 mg, 0.13 mmol) in methanol
(10 mL) was stirred with a catalytic amount of 10%
Pd/C under a hydrogen atmosphere (1 atm) at rt over-
night. The Pd/C was removed by filtration, and the fil-
trate was concentrated to a pale yellow film. The
residue was redissolved in water (2 mL), and the aque-
ous layer was washed with chloroform (2 mL · 3) before
being purified by CF-11 cellulose column chromatogra-
phy (2.5 · 30 cm; flow rate = 1 mL/min) eluting with a
continuous gradient of 0 to 1% of 0.1% TFA/THF in
water. Five milliliter fractions were collected, and the
fractions that were PMA-active were collected and
1171 cmꢂ1; H NMR (400 MHz, CDCl3) d (ppm) 7.74
(d, J = 8.3 Hz, 2H), 7.34–7.43 (m, 5H), 7.31 (d,
J = 8.3 Hz, 2H), 5.10 (s, 2H), 4.12 (q, J = 7.1 Hz, 2H),
2.91 (br s, 2H), 2.41 (s, 3H), 2.33 (t, J = 7.43 Hz, 2H),
1.81 (p, J = 7.1 Hz, 2H), 1.25 (t, J = 7.1 Hz, 3H); 13C
NMR (100 MHz, CDCl3) d (ppm) 172.82, 144.76,
135.18, 130.02, 129.71, 129.56, 129.47, 128.51, 128.71,
79.86, 60.44, 52.61, 31.24, 22.09, 21.59, 14.17; HRCIMS
calcd for C20H26O5NS: 392.15317. Found: 392.15365
(M+H).
4.10. 4-N-Hydroxyaminobutyric acid (20)
1
freeze-dried to a colorless film (20 mg, 91%). H NMR
(400 MHz, D2O) d (ppm) two amide rotamers with 3:7
ratio: 8.32 (br s, minor) and 7.94 (br s, major), 3.93
(m, 2H), 3.64 (m, 2H) 13C NMR (100 MHz, D2O) d
A solution of 19 (345 mg, 0.89 mmol) in a 1:1 mixture of
conc. HCl and acetic acid (50 mL) was refluxed over-
night. The solvent was removed in vacuo and the residue
was dissolved in water (10 mL). The aqueous layer was
washed with chloroform and concentrated to a brown
2
(ppm) 160.42, 60.62 (d, JCP = 3.0 Hz), 51.16 (d,
3JCP = 5.0 HZ) 31P NMR (121 MHz, D2O) d (ppm)

2382
Y.-H. Woo et al. / Bioorg. Med. Chem. 14 (2006) 2375–2385
residue. The crude product was purified on a Dowex
50Wx8-100 resin, eluting with 2 N NH4OH. The frac-
tions were pooled and freeze-dried to a pale yellow film
low oil (35 mg, 50% for two steps). IR (neat) 3343, 2974,
1
2926, 2889, 1449, 1380, 1180, 1088, 1049, 881 cmꢂ1; H
NMR (300 MHz, MeOH-d4) d (ppm) 4.11 (t, J = 4.8 Hz,
2H), 3.73 (t, J = 4.8 Hz, 2H); 13C NMR (75 MHz,
MeOH-d4) d (ppm) 72.33, 61.20; HRCIMS calcd for
C2H8O4NS: 142.01740. Found: 142.01770 (M+H).
1
(85 mg, 81%). H NMR (400 MHz, MeOH-d4) d (ppm)
2.99 (t, J = 7.5 Hz, 2H), 2.27 (t, J = 7.4 Hz, 2H), 1.88 (p,
J = 7.5 Hz, 2H); 13C NMR (100 MHz, MeOH-d4) d
(ppm) 182.64, 40.42, 35.51, 24.73; HRCIMS calcd for
C4H10O3N: 120.06607. Found: 120.06608 (M+H).
4.14. Sulfamic acid 2-(N-benzyloxyamino)-ethyl
ester (23)
4.11. 4-(N-Formyl-N-hydroxyamino)-butyric acid (8)
To compound 22 (28 mg, 0.2 mmol) in dry CH2Cl2
(10 mL) was added 2,6-lutidine (31.9 mg, 0.3 mmol).
The solution was cooled to ꢂ78 ꢁC before adding triflu-
oromethane sulfonic anhydride (40 lL, 0.24 mmol)
dropwise. After the reaction mixture was stirred at
ꢂ78 ꢁC for 30 min,O-benzyl hydroxylamine (49 mg,
0.4 mmol) was added dropwise. The solution was stirred
for 1 h at ꢂ78 ꢁC, warmed to rt, and stirred for another
2 h. The reaction mixture was diluted with CH2Cl2 and
washed with saturated aq NaHCO3, dried over MgSO4,
filtered, and concentrated. The resulting yellow oil was
purified by flash chromatography (EtOAc/hexanes,
3:2) to yield a pale yellow oil (39 mg, 80%). IR (neat)
3343, 2974, 2926, 2889, 1449, 1380, 1180, 1088, 1049,
Compound 20 (15 mg, 0.12 mmol) was stirred in 500 ll
of acetylformic anhydride at rt for 12 h. The reaction
mixture was concentrated to a yellow brown oil and
purified on a Varian Bond Eluteꢂ C18 cartridge eluting
with gradient from 0 to 100% MeOH in H2O. The frac-
tions, which were FeCl3-active by TLC, were pooled and
freeze-dried to a pale yellow solid (13 mg, 72%). IR
1
(neat) cmꢂ1; H NMR (400 MHz, MeOH-d4) d (ppm)
1:1 mixture of amide rotamers: 7.92 and 8.25 (2s, 1H),
3.53 and 3.58 (2t, J = 6.6 Hz, 2H), 2.20–2.24 (m, 2H),
1.88–1.95 (m, 2H); 13C NMR (100 MHz, MeOH-d4) d
(ppm) 180.65, 164.19 and 159.79, 35.12 and 34.62,
24.71, 23.84; HRCIMS calcd for C5H9NO4: 147.05318.
Found: 147.05450 (M+).
881, 785 cmꢂ1
;
1H NMR (300 MHz, MeOH-d4) d
(ppm) 7.39–7.31 (m, 5H), 4.72 (s, 2H), 4.28 (t,
J = 5.3 Hz, 2H), 3.19 (t, J = 5.3 Hz, 2H); 13C NMR
(75 MHz, MeOH-d4) d (ppm) 129.47, 129.33, 128.83,
77.00, 67.52, 51.50.
4.12. Sulfamic acid 2-(tert-butyldimethylsilyloxy)-ethyl
ester
Sulfamoyl chloride was prepared as described.48 The im-
pure needle-shaped crystals were identified as the desired
product from the IR spectrum [IR (neat, cmꢂ1) 3625,
3381, 3278, 1546, 1384, 1185, 1064, 924] which matched
the reported values. To a solution of 2-(tert-butyldimeth-
ylsilyloxy)ethanol (90 mg, 0.51 mmol) in DMA (7 mL)
was added triethylamine (86 lL, 0.61 mmol). The solu-
tion was cooled in an ice bath before adding sulfamoyl
chloride (118 mg, 1.0 mmol) under a N2 atmosphere and
then stirred for 2 h at 0 ꢁC. The reaction mixture was
warmed to rt and stirred for 3 h. The reaction mixture
was diluted in EtOAc and the resulting organic layer
was washed successively with saturated aqueous solutions
of NaHCO3 and NH4Cl, water and brine. The combined
organic layer was dried over MgSO4, filtered, concentrat-
ed, and the residual TEA was removed in vacuo to afford a
pale yellow oil (60 mg). IR (neat) 2954, 2932, 1712, 1560,
4.15. Sulfamic acid 2-(N-formyl-N-hydroxyamino)-ethyl
ester (9)
To a solution of 23 (35 mg, 0.14 mmol) in CH2Cl2
(10 ml) was added FMT (45 mg, 0.28 mmol) at rt, and
the resulting solution was stirred for 12 h. The reaction
mixture was concentrated in vacuo and purified by flash
chromatography (EtOAc/hexanes, 2:1) to give a pale
yellow oil (35 mg, 90%). The residue (32 mg, 0.12 mmol)
was dissolved in methanol (10 mL) and a catalytic
amount of 10% Pd/C was added to the solution. The
heterogeneous mixture was stirred for 12 h under a
slight positive pressure of H2 gas (1 atm) at rt. The reac-
tion mixture was filtered and concentrated to a pale
yellow oil before purification on a Varian Bond Eluteꢂ
C18 column eluting with a stepwise gradient (0–70%)
methanol in H2O. The desired product was eluted with
15% methanol/H2O and the fractions were pooled and
lyophilized to give a yellow syrup (11 mg, 50%). 1H
NMR (400 MHz, MeOH-d4) d (ppm) 1:2 ratio of 8.38
& 7.98 (s), 4.33 (t, J = 5.5 Hz, 2H), 3.92 (t, J = 5.5 Hz,
2H); 13C NMR (100 MHz, MeOH-d4) d (ppm) 164.9
and 160.0, 65.95, 51.05; HRCIMS calcd for C3H8N2O5S
184.01539. Found: 184.01571 (M+).
1
1366, 1184, 936 cmꢂ1; H NMR (300 MHz, CDCl3) d
(ppm) 5.43 (br s), 4.25 (t, J = 4.8 Hz, 2H), 3.89 (t,
J = 4.8 Hz, 2H), 0.88 (s, 9H), 0.08 (s, 6H); 13C NMR
(75 MHz, CDCl3) d (ppm) 72.05, 61.48, 25.76, 18.26,
ꢂ5.44; MS (CI): m/z 256.1 (M+H).
4.13. Sulfamic acid 2-hydroxyethyl ester (22)
The crude sulfamic acid 2-(tert-butyldimethylsilyloxy)-
ethyl ester product (60 mg) in methanol (10 mL) was
treated with 0.5 g (wet weight) AG 50W X8-100 (H+
form) and stirred for 5 h at rt. The resin was removed
by filtration and the filtrate was concentrated to a yellow
oil. The crude oil was purified on a Varian Bond Eluteꢂ
C18 cartridge eluting with a gradient of methanol in
H2O (0 to 100%). The product eluted at approximately
5% methanol in H2O and was concentrated to a pale yel-
4.16. N-(3-Bromopropyl)-N-toluene-4-sulfonamide
To
a
heterogeneous mixture of NaH (120 mg,
2.75 mmol) in dry THF (100 mL) was transferred a rt
solution of N-tosyl-N-methyl amine (500 mg,
2.72 mmol) in THF (20 mL) via a cannula at 0 ꢁC.
One hour later, 1,3-dibromopropane (24; 500 mg,
2.5 mmol) was added dropwise at 0 ꢁC. The reaction

Y.-H. Woo et al. / Bioorg. Med. Chem. 14 (2006) 2375–2385
2383
mixture was stirred for 12 h at rt and refluxed for 3 h be-
fore quenching with water. The solvent was removed in
vacuo and the crude product was dissolved in EtOAc
(30 mL). The organic fraction was washed with saturat-
ed aq NH4Cl, water and brine, dried over MgSO4, fil-
tered, and concentrated to a pale yellow oil. The
residue was purified by flash chromatography (EtOAc/
hexanes, 1:9) to get a colorless oil (250 mg, 34%; 65%
based on recovered starting material). IR (neat) 1366,
4.19. [3-(N-Formyl-N-methyl-amino)-propyl]-phosphonic
acid (10)
A solution of compound 26 (30 mg, 0.19 mmol) in acet-
ylformic anhydride (10 ml) was stirred at rt for 12 h. The
solvent was removed in vacuo and the residue was puri-
fied on a Varian Bond Eluteꢂ C18 column eluting with a
stepwise gradient of MeOH (0–100%) in water. The
product eluted at 15% MeOH, along with minor impu-
rities. The fractions were pooled, concentrated, and
basified with NaHCO3 before being applied onto an
AG1-X10 (100–200 mesh, Clꢂ form) ion exchange col-
umn. The desired product eluted with 1 N NH4OH,
but a yellow contaminant remained. The concentrated
residue was purified further by C18 silica chromatogra-
phy eluting with water to give 6 mg of a white solid after
lyophilization (15%). 1H NMR (400 MHz, D2O) d
(ppm) 1:1 mixture of amide rotamers 8.39 and 7.96 (s,
1H), 3.23–3.33 (m, 2H), 1.99 and 1.96 (2 s, 3H), 1.61–
1.71 (m, 2H), 1.36–1.47 (m, 2H); 13C NMR (100 MHz,
D2O), d (ppm) 1:1 mixture of amide rotamers 165.2
1
1170 cmꢂ1; H NMR (400 MHz, CDCl3) d (ppm) 7.68
(d, J = 8.4 Hz, 2H), 7.33 (d, J = 8. 4 Hz, 2H), 3.47 (t,
J = 6.5 Hz, 2H), 3.13 (t, J = 6.7 Hz, 2H), 2.11 (dt,
J = 6.6, 6.6 Hz, 2H), 2.75 (s, 3H), 2.43 (s, 3H); 13C
NMR (100 MHz, CDCl3) d (ppm) 143.50, 134.23,
129.74, 127.45, 48.80, 35.63, 31.28, 30.14, 21.51; HR
CIMS calcd for C11H17O2N79BrS: 306.10634. Found:
306.10655 (M+).
4.17. {3-[-N-Methyl-N-(toluene-4-sulfonyl)-amino]-
propyl}phosphonic acid diethyl ester (25)
3
To a heterogeneous mixture of NaH (42 mg, 1.01 mmol)
in dry THF (50 mL) was added diethyl phosphite
(126 lL, 0.98 mmol) dropwise at 0 ꢁC and stirred for
2 h. A solution of N-(3-Bromopropyl)-N-toluene-4-sul-
fonamide (250 mg, 0.82 mM, in 5 mL THF) was added
to the mixture at 0 ꢁC via a syringe. The mixture was
warmed to rt and stirred for 12 h. The solvent was re-
moved in vacuo and the residue was dissolved in EtOAc
(100 mL). The organic layer was washed with a solution
of saturated aq NH4Cl, water and brine, dried over
MgSO4, filtered, and concentrated, before being purified
by flash chromatography (3% MeOH/CH2Cl2) to yield
210 mg of a colorless oil (70%). IR (neat) 1162, 1245,
and 160.7, 51.95 (d, JCP = 18.9 Hz) and 48.72 (d,
3JCP = 19.7 Hz), 36.58 and 33.69, 25.44 (d,
JCP = 134.5 Hz), 22.22 (d, JCP = 3.0 Hz) and 21.39
2
2
(d, JCP = 3.8 Hz); 31P NMR (121 MHz, D2O) d (ppm)
26.29 and 25.87; MS (ESI): m/z 182.1 (M+H), 204.0
(M+Na).
4.20. [3-(N-Acetyl-N-methyl-amino)-propyl]-phosphonic
acid (11)
Compound 26 (30 mg, 0.19 mmol) was dissolved in a 1:1
mixture of acetic anhydride/pyridine (6 mL) and stirred
at rt for 5 h. The solvent was removed in vacuo, and the
residue was purified on a Varian Bond Eluteꢂ C18 col-
umn eluting with a stepwise gradient MeOH (0–100%)
in water. The product eluted at 15% MeOH along with
minor impurities. The fractions were pooled, concen-
trated and basified with NaHCO3, before being applied
onto an AG1-X10 (100–200 mesh, Clꢂ form) ion ex-
change column. The desired product eluted with 1 N
NH4OH, but a yellow contaminant was also present.
The concentrated residue was purified further by C18 sil-
ica chromatography eluting with water to give 5 mg of a
white solid after lyophilization (13%). 1H NMR
(400 MHz, D2O) d (ppm) 1:1 mixture of amide rotamers
3.23–3.33 (m, 2H), 2.77 and 2.92 (2s, 3H), 1.96 and 1.99
(2s, 3H), 1.61–1.71 (m, 2H), 1.36–1.47 (m, 2H); 13C
NMR (100 MHz, D2O) d (ppm) 1:1 mixture of amide
1
1350 cmꢂ1; H NMR (300 MHz, CDCl3) d (ppm) 7.65
(d, J = 8.4 Hz, 2H), 7.31 (d, J = 8.4 Hz, 2H), 4.03–4.20
(m, 4H), 3.04 (t, J = 6.3 Hz, 2H), 2.71 (s, 3H), 2.43 (s,
3H), 1.78-1.86 (m, 4H), 1.37 (t, J = 6.8 Hz, 3H); 13C
NMR (75 MHz, CDCl3) d (ppm) 143.25, 134.12,
2
129.53, 127.16, 61.42 (d, JCP = 6.4 Hz), 50.12 (d,
3JCP = 18.7 Hz), 34.52, 22.40 (d, JCP = 142.8 Hz),
21.28, 20.65 (d, 2JCP = 3.9 Hz), 16.25 (d, 3JCP = 5.8 Hz);
31P NMR (121 MHz, CDCl3)
d
(ppm) 32.55;
HRFABMS calcd for C15H27O5NPS: 364.13476.
Found: 364.13451 (M+H).
4.18. (3-N-Methylamino-propyl)-phosphonic acid (26)
Compound 25 (250 mg, 0.68 mmol) was dissolved in a
1:1 mixture of acetic acid and concd HCl (30 mL), and
refluxed overnight. The solvent was removed under re-
duced pressure and the yellow residue was purified by
cation exchange chromatography using Dowex 50X8-
100 and eluting with 1 N NH4OH. The fractions that
were stained with PMA on TLC were pooled and
3
rotamers 174.49 and 174. 37, 51.95 (d, JCP = 18.9 Hz)
and 48.72 (d, JCP = 19.7 Hz), 36.58 and 33.69, 25.44
(d, JCP = 134.2 Hz) and 25.14 (d, JCP = 134.5 Hz),
3
2
2
22.22 (d, JCP = 3.0 Hz) & 21.39 (d, JCP = 3.8 Hz),
21.29 and 20.66; 31P NMR (121 MHz, D2O) d (ppm)
25.87 and 26.29; MS (ESI): m/z 196.0 (M+1), 218.0
(M+Na).
1
freeze-dried to a white solid (72 mg, 70%). H NMR
(300 MHz, MeOH-d4) d (ppm) 2.98 (t, J = 6.1 Hz, 2H),
2.59 (s, 3H), 1.86–1.97 (m, 2H), 1.58–1.68 (m, 2H); 13C
4.21. Expression of DXR
NMR (75 MHz, MeOH-d4)
d
(ppm) 50.59 (d,
3JCP = 16.12 Hz), 33.37, 26.08 (d, JCP = 132.5 Hz),
The target enzyme, DXR from Synechocystis sp.
PCC6803, was overproduced in E. coli as a N-terminal
6· His-tagged form and purified for the inhibition
studies as described previously.21
2
21.16 (d, JCP = 3.2 Hz); 31P NMR (121 MHz, MeOH-
d4) d (ppm) 22.19; HRFABMS calcd for C4H13O3NP:
154.06331. Found: 154.06315 (M+H).

2384
Y.-H. Woo et al. / Bioorg. Med. Chem. 14 (2006) 2375–2385
4.22. Enzyme assay
References and notes
1. Rohdich, F.; Bacher, A.; Eisenreich, W. Bioorg. Chem.
2004, 32, 292.
2. Spurgeon, S. L.; Porter, J. W. In Biosynthesis of Isoprenoid
Compounds; Spurgeon, S. L., Porter, J. W., Eds.; John
Wiley & Sons: 1981, p 3.
3. Rohmer, M. Pure Appl. Chem. 2003, 75, 375.
4. Rohmer, M. In Comprehensive Natural Product Chemistry;
Cane, D. E., Ed.; Elsevier: New York: 1999; Vol. 2; p. 45.
5. Lichtenthaler, H. K. Annu. Rev. Plant Physiol. Plant Mol.
Biol. 1999, 50, 47.
Activity of the recombinant DXR and inhibition by
analogs were tested by monitoring the rate of consump-
tion of NADPH at 340 nm.10 The 500 mL total volume
for each assay consisted of 100 mM Tris–HCl (pH 7.5),
1 mM MnCl2, 0.2 mM NADPH, and varying concentra-
tions of DXP and inhibitors. Each assay was initiated by
adding DXR and the decrease in absorbance at 340 nm
was monitored for 4 min. Inhibition assays were done
with at least four different concentrations of each inhib-
itor, and each concentration was tested in triplicate.
Kinetic data were analyzed using Trinityꢂ Enzyme
Kinetics v. 1.2 software.
6. Testa, C. A.; Brown, M. J. Curr. Pharm. Biotechnol. 2003,
4, 248.
7. Rodriguez-Concepcion, M. Curr. Pharm. Des. 2004, 10,
2391.
8. Rohdich, F.; Bacher, A.; Eisenreich, W. Biochem. Soc.
Trans. 2005, 33, 785.
9. Kuzuyama, T.; Takahashi, S.; Watanabe, H.; Seto, H.
Tetrahedron Lett. 1998, 39, 4509.
4.23. Enzyme inhibition studies (for slow-tight binding
inhibition)
10. Takahashi, S.; Kuzuyama, T.; Watanabe, H.; Seto, H.
Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 9879.
11. Proteau, P. J. Bioorg. Chem. 2004, 32, 483.
12. Hoeffler, J. F.; Tritsch, D.; Grosdemange-Billiard, C.;
Rohmer, M. Eur. J. Biochem. 2002, 269, 4446.
13. Kuzuyama, T.; Shimizu, T.; Takahashi, S.; Seto, H.
Tetrahedron Lett. 1998, 39, 7913.
Progress curves for slow, tight-binding inhibition analy-
sis were obtained by initiating the reaction by adding the
enzyme to the reaction mixture. The decrease in the
absorbance at 340 nm was monitored for 4 min. The
progress curves were fitted to:
A ¼ vst þ ðvo ꢂ vsÞð1 ꢂ eꢂktÞ=k;
14. 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.
Science 1999, 285, 1573.
15. Wiesner, J.; Borrmann, S.; Jomaa, H. Parasitol. Res. 2003,
90.
16. Borrmann, S.; Issifou, S.; Esser, G.; Adegnika, A. A.;
Ramharter, M.; Matsiegui, P. B.; Oyakhirome, S.; Mawili-
Mboumba, D. P.; Missinou, M. A.; Kun, J. F.; Jomaa, H.;
Kremsner, P. G. J. Infect. Dis. 2004, 190, 1534.
17. Borrmann, S.; Adegnika, A. A.; Moussavou, F.; Oyakhi-
rome, S.; Esser, G.; Matsiegui, P. B.; Ramharter, M.;
Lundgren, I.; Kombila, M.; Issifou, S.; Hutchinson, D.;
Wiesner, J.; Jomaa, H.; Kremsner, P. G. Antimicrob.
Agents Chemother. 2005, 49, 3749.
18. Kuntz, L.; Tritsch, D.; Grosdemange-Billiard, C.; Hem-
merlin, A.; Willem, A.; Bach, T. J.; Rohmer, M. Biochem.
J. 2005, 386, 127.
19. Courtois, M.; Mincheva, Z.; Andreu, F.; Rideau, M.;
Viaud-Massuard, M. C. J. Enzyme Inhib. Med. Chem.
2004, 19, 559.
20. Mincheva, Z.; Courtois, M.; Andreu, F.; Rideau, M.;
Viaud-Massuard, M. C. Phytochemistry 2005, 66,
1797.
where A is the absorbance at any time t, v0 and vs are the
initial and final steady-state rates, respectively, and k is
the apparent first-order rate constant for the establish-
ment of the final steady-state equilibrium. The curves
were fitted by the nonlinear least squares, using Graph-
Pad Prism software. The inhibition constants were esti-
mated from v0 and vs by nonlinear regression using the
Enzyme Kinetics program from Trinity Software.
The pre-incubation studies to determine the slow, tight-
binding inhibition were initiated by incubating the en-
zyme (3 nM) for 10 min at 37 ꢁC in 50 mM Tris–HCl
buffer, 1 mM MnCl2, 1 mg/mL BSA, 0.2 mM NADPH,
and varying concentrations of fosmidomycin,
FR900098, or the hydroxamate analog. After the incu-
bation time, DXP at 0.2 mM was added and the activity
was assayed by monitoring the decrease in absorbance
at 340 nm as described in the standard assay for DXR.
The Kꢀi s were calculated from the X-intercept of the Dix-
on plot (1/V vs [I]).38
21. Yin, X.; Proteau, P. J. Biochim. Biophys. Acta 2003, 1652,
75.
22. Phaosiri, C.; Proteau, P. J. Bioorg. Med. Chem. Lett. 2004,
14, 5309.
Acknowledgments
23. Reuter, K.; Sanderbrand, S.; Jomaa, H.; Wiesner, J.;
Steinbrecher, I.; Beck, E.; Hintz, M.; Klebe, G.; Stubbs,
M. T. J. Biol. Chem. 2002, 277, 5378.
24. Yajima, S.; Nonaka, T.; Kuzuyama, T.; Seto, H.; Ohsawa,
K. J. Biochem. 2002, 131, 313.
25. Steinbacher, S.; Kaiser, J.; Eisenreich, W.; Huber, R.;
Bacher, A.; Rohdich, F. J. Biol. Chem. 2003, 278, 18401.
26. Yajima, S.; Hara, K.; Sanders, J. M.; Yin, F.; Ohsawa, K.;
Wiesner, J. J. Am. Chem. Soc. 2004, 126, 10824.
27. Mac Sweeney, A.; Lange, R.; Fernandes, R. P.; Schulz,
H.; Dale, G. E.; Douangamath, A.; Proteau, P. J.; Oefner,
C. J. Mol. Biol. 2005, 345, 115.
We thank both Dr. Mike Schimerlik and Dr. Mark
Zabriskie for advice on conducting the slow, tight-bind-
ing inhibition studies. Financial support for this re-
search was provided by a grant from the National
Institutes of Health (R01 AI42258A). R.P.M.F. had a
PhD scholarship from CNPq (Brazil) during a portion
of this work. This publication was made possible in part
by Grant No. P30 ES00210 from the National Institutes
of Environmental Health Sciences, NIH (mass spec-
trometry). Professor Haruo Seto is thanked for provid-
ing a sample of fosmidomycin used in the early stages
of this work.
28. Silber, K.; Heidler, P.; Kurz, T.; Klebe, G. J. Med. Chem.
2005, 48, 3547.

Y.-H. Woo et al. / Bioorg. Med. Chem. 14 (2006) 2375–2385
2385
29. Hemmi, K.; Takeno, H.; Hashimoto, M.; Kamiya, T.
Chem. Pharm. Bull. 1982, 30, 111.
39. Grolle, S.; Bringer-Meyer, S.; Sahm, H. FEMS Microbiol.
Lett. 2000, 191, 131.
30. Kurz, T.; Geffken, D.; Wackendorff, C. Z. Naturforsch.
2003, 58b, 106.
31. Kurz, T.; Geffken, D.; Wackendorff, C. Z. Naturforsch.
2003, 58b, 457.
32. Katayama, N.; Tsubotani, S.; Nozaki, Y.; Harada, S.;
Ono, H. J. Antibiot. 1990, 43, 238.
40. Kuroda, Y.; Okuhara, M.; Goto, T.; Okamoto, M.;
Terano, H.; Kohsaka, M.; Aoki, H.; Imanaka, H.
J. Antibiot. 1980, 33, 29.
41. Okuhara, M.; Kuroda, Y.; Goto, T.; Okamoto, M.;
Terano, H.; Kohsaka, M.; Aoki, H.; Imanaka, H.
J. Antibiot. 1980, 33, 24.
33. Hashimoto, M.; Hemmi, K.; Takeno, H.; Kamiya, T.
Tetrahedron Lett. 1980, 21, 99.
34. Howarth, N. M.; Purohit, A.; Reed, M. J.; Potter, B. J.
J. Med. Chem. 1994, 37, 219.
42. Engel, R. Chem. Rev. 1977, 77, 349.
43. Perrin, D. D.; Armarego, W. L. F. Purification of
laboratory chemicals, 3rd ed.; Pergamon Press: New York,
1988.
35. Kamiya, T.; Hemmi, K.; Takeno, H.; Hashimoto, M.
Tetrahedron Lett. 1980, 21, 95.
44. Strazzolini, P.; Giumanini, A. G.; Cauci, S. Tetrahedron
1990, 46, 1081.
36. Koppisch, A. T.; Fox, D. T.; Blagg, B. S. J.; Poulter, C. D.
Biochemistry 2002, 41, 236.
45. McDougal, P. G.; Rico, J. G.; Oh, Y.-I.; Condon, B. D.
J. Org. Chem. 1986, 51, 3388.
37. Morrison, J. F.; Walsh, C. T. Adv. Enzymol. Relat. Areas
Mol. Biol. 1988, 61, 201.
46. OÕNeil, I. A. Synlett 1991, 661.
47. Yazawa, H.; Goto, S. Tetrahedron Lett. 1985, 26,
38. Segel, I. H. Enzyme Kinetics: Behavior and Analysis of
Rapid Equilibrium and Steady State Enzyme Systems;
Wiley: New York, 1975.
3703.
48. Peterson, E. M.; Brownell, J.; Vince, R. J. Med. Chem.
1992, 35, 3991.