Synthesis and evaluation of α,β-unsaturated α-aryl-substituted fosmidomycin analogues as DXR inhibitors

Article abstract of DOI:10.1016/j.bmcl.2007.06.026

Fosmidomycin, which acts through inhibition of 1-deoxy-d-xylulose phosphate reductoisomerase (DXR) in the non-mevalonate pathway, represents a valuable recent addition to the armamentarium against uncomplicated malaria. In this paper, we describe the synt

Full text of DOI:10.1016/j.bmcl.2007.06.026

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4920–4923
Synthesis and evaluation of a,b-unsaturated a-aryl-substituted
fosmidomycin analogues as DXR inhibitors
Vincent Devreux,a,b Jochen Wiesner,c Hassan Jomaa,c Johan Van der Eyckenb
and Serge Van Calenbergha,*
aLaboratory for Medicinal Chemistry (FFW), Ghent University, Harelbekestraat 72, B-9000 Gent, Belgium
bLaboratory of Organic and Bioorganic Synthesis, Department of Organic Chemistry, Faculty of Sciences, Ghent University,
Krijgslaan 281 (S.4), B-9000 Gent, Belgium
cJustus-Liebig-Universita¨t Giessen, Institut fu¨r Klinische Chemie und Pathobiochemie, Gaffkystrasse 11, D-35392 Giessen, Germany
Received 7 May 2007; revised 6 June 2007; accepted 7 June 2007
Available online 10 June 2007
Abstract—Fosmidomycin, which acts through inhibition of 1-deoxy-D-xylulose phosphate reductoisomerase (DXR) in the non-
mevalonate pathway, represents a valuable recent addition to the armamentarium against uncomplicated malaria. In this paper,
we describe the synthesis and biological evaluation of E- and Z-a,b-unsaturated a-aryl-substituted analogues of FR900098, a fosm-
idomycin congener, utilizing a Stille or a Suzuki coupling to introduce the aryl group. In contrast with our expectations based on the
promising activity earlier observed for several a-substituted fosmidomycin analogues, all synthesized analogues exhibited much low-
er binding affinity for DXR than fosmidomycin.
ꢀ 2007 Elsevier Ltd. All rights reserved.
With over 500 million clinical cases each year, malaria
still remains a major threat in the world, as more than
two million people succumb from the disease each year.
The high prevalence of the disease is attributed to the
difficulties in vector control, as well as the increasing
resistance of Plasmodium falciparum, the main causative
agent of the disease, towards the commonly used anti-
malarials such as chloroquine. Therefore, new antima-
larial drugs acting on alternative, yet unexplored
biochemical pathways are urgently needed.
nate pathway. In recent clinical trials fosmidomycin
combinations with clindamycin3 or artesunate4 were
found very efficient in curing P. falciparum uncompli-
cated malaria (Fig. 1). FR900098, the acetyl congener
of fosmidomycin, was found to be twice as active as
fosmidomycin, both in vitro and in a Plasmodium vinckei
mouse model.2
For structure–activity relationship analyses of fosmido-
mycin derivatives, various alterations have been made,
mainly addressing the retrohydroxamate and phospho-
nate moieties.5–9 Comparatively few modifications of
the carbon spacer have been explored. Hence, we re-
cently focused on such type of modifications (Fig. 1).
Recently, the discovery of the mevalonate-independent
pathway for isoprenoid biosynthesis opened the way
for new therapeutics to cure malaria, as this alternative
pathway is absent in humans. Two groups simulta-
neously discovered that fosmidomycin effectively inhibits
Incorporation of aryl functionalities in a-position of the
phosphonate of fosmidomycin (e.g., 1) or FR900098 re-
sulted in a markedly enhanced in vitro antiplasmodial
activity, which proved to be associated to the electron
withdrawing properties of the aryl substituents.10,11
1-deoxy-D-xylulose
5-phosphate
reductoisomerase
(DXR).1,2 This essential enzyme converts 1-deoxy-D-
xylulose 5-phosphate (DOXP) to 2C-methyl-D-erithrytol
phosphate (MEP), the second step in the non-mevalo-
Based on a reported prodrug approach,6 Kurz et al. sys-
tematically investigated the effect of introduction of
different substituents in a-position of the bis(pivaloyl-
oxymethyl)esters of fosmidomycin or FR900098.12–14
Introduction of an a-methyl or a-phenyl substituent
Keywords: Fosmidomycin; Non-mevalonate pathway; DOXP reduc-
toisomerase inhibitor; Antimalarial; Antibiotic.
*
Corresponding author. Tel.: +32 9 264 81 24; fax: +32 9 264 81 46;
e-mail: serge.vancalenbergh@ugent.be
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.06.026

V. Devreux et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4920–4923
4921
Scheme 1. Retrosynthetic approach towards analogues 3 and 4.
of 5 (Scheme 2) was recently reported as part of a diver-
gent synthetic route towards a-substituted fosmidomy-
cin analogues.16 Due to the inherent stereoselectivity
of the Stille-coupling, only the Z-isomers were obtained.
Further elaboration to FR900098 analogues 3a–e in-
volved a BCl3-assisted removal of the benzyl protecting
group from 8a to 8e. Subsequent cleavage of the phos-
phonate ester groups with TMSBr followed by purifica-
tion by reversed phase HPLC yielded the Z-configured
analogues 3a–e in low to moderate yields.17
Figure 1. Structures of fosmidomycin, FR900098 and analogues under
study.
afforded analogues, which exhibited antiplasmodial
activities that came close to that of the FR900098 pro-
drug, while a 3,4-difluorophenyl-substituted analogue
was slightly more potent.12
A Suzuki-based strategy was chosen for the synthesis of
the E-a,b-unsaturated analogues 4a,f–g (Scheme 3).
First Z-a-bromo-1-propenyl phosphonate (11) was syn-
Consistent with our findings, a-aryl-substituted fosmi-
domycin analogues were generally superior to their
FR900098 homologues. The introduction of an ethyl,
propyl, isopropyl, dimethyl and hydroxymethyl group
was associated with a considerable drop in antimalarial
activity.12 Also a-arylmethyl13 or phenylethyl14 ana-
logues failed to surpass the activities of the a-aryl
prodrugs.
Furthermore, rigidification of the carbon spacer by the
introduction of a cyclopropane15 (as in 2) or cyclopen-
tane11 ring indicated a preferred trans geometry for the
substituents on these rings for binding DXR.
The scope of the present work is to combine both fea-
tures, that is, rigidification of the carbon spacer
(through incorporation of an a,b-unsaturated bond)
and introduction of an a-aryl substituent. As the a,b-
double bond could occur both in the E- or in the Z-con-
figuration, these analogues might provide insight into
the preferred binding conformation of saturated
a-substituted fosmidomycin analogues.
Scheme 2. Reagents and conditions: (a) see Ref. 16; (b) ArAI,
Pd2dba3 Æ CHCl3, (2-furyl)3P, CuI, NMP, rt; (c) BCl3, CH2Cl2,
ꢀ50 ꢁC; (d) TMSBr, MeCN, rt; C-18 RP HPLC.
The retrosynthetic approaches towards the synthesis of
the Z- and E-configured analogues of FR900098 are
briefly depicted in Scheme 1. The synthesis of the Z-ana-
logues 3a–e is based on a palladium(0)-catalysed Stille-
coupling on organotin derivative 5,16 while the synthesis
of the E-configured analogues 4a,f–g features a Suzuki-
type cross-coupling on vinylic bromide 6 as a key-step.
Scheme 3. Reagents and conditions: (a) see Ref. 18; (b) NBS,
Ph(COO)2, CCl4, reflux; (c) NaI, acetone, rt; (d) BnONHBoc,
NaH,DMF, rt.
The synthesis of compounds 8a–e in five steps from
THP-protected propargyl alcohol 7 via Stille coupling

4922
V. Devreux et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4920–4923
thesized from diethyl phosphite (10) in two steps.18 The
required Z-configuration was assigned based on the
Table 1. IC50 values of the synthesized FR900098 analogues against
recombinant E. coli DXR
1
3JPH coupling constant (14.2 Hz) in the H NMR spec-
Compound
R
IC50 E. coli DXR (lM)
trum, which is in accordance with the coupling constant
typically found for a vinylic proton located cis to a phos-
phonate.19 Radical allylic bromination afforded com-
pound 12 in moderate yield. Prior conversion of the
allylic bromide 12 into iodide 13 via Finkelstein reaction
was required due to the low reactivity of compound 12
in the substitution reaction to yield key intermediate 6.
Conversely, when this substitution reaction was applied
to iodide 13, protected allylic hydroxylamine 6 was
obtained in an adequate yield.
Fosmidomycin
FR900098
0.034
0.032
0.059
>30
>30
34
1
3a
3b
3c
3d
3e
4a
4f
Ph
(3-NO2)Ph
(4-NO2)Ph
2-Thienyl
3-Thienyl
Ph
42
>30
5.7
(3,4-Cl2)Ph
(4-CN)Ph
Br
5.5
16
4g
19
0.45
Optimization of the Suzuki-coupling with a series of
arylboronic acids (Scheme 4) proved to be trouble-
some, as poor to moderate yields were obtained for
14a,f–g. Furthermore, every analogue required differ-
ent, specific conditions for the reaction to occur, in
contrast to the generally applicable Stille-conditions
described above.
allowed for the synthesis of the Z-a,b-unsaturated a-
bromo-substituted analogue 19.
Because of the difficulties associated with the handling
of P. falciparum DXR, we investigated the ability of
the synthesized FR900098 analogues 3a–e, 4a,f–g and
19 to inhibit the highly homologous Escherichia coli iso-
zyme. The conversion of DOXP to MEP by the enzyme
was determined in an assay based on the NADPH
dependency of the reaction and the results are summa-
rized in Table 1.21
To complete the synthesis of analogues 4a,f–g, the Boc-
protecting group of 14a,f–g was first removed, followed
by in situ acetylation of the free amine to yield the N-
benzyloxyacetamides 15a,f–g. Subsequent removal of
the benzyl protecting group with BCl3 and cleavage of
the phosphonate esters yielded the final E-configured
a,b-unsaturated, a-aryl substituted analogues 4a,f–g,
which were purified via Whatman CF-11 cellulose chro-
matography, resulting in much improved yields as com-
pared to RP-chromatographic purification used for
3a–e.19,20 Furthermore, application of the same depro-
tection sequence directly on the vinylic bromide 6 also
Compared to fosmidomycin and FR900098, the Z-con-
figured unsaturated analogues 3a–e were found to be
much weaker inhibitors of E. coli DXR, which was antic-
ipated considering the previously found preferred trans-
orientation of the phosphonate and retrohydroxamate
functionalities in cyclopropyl and cyclopentyl fosmido-
mycin analogues.11,15 Unexpectedly, in contrast with
the active trans-substituted cyclopropyl derivative 2, also
the E-configured a-aryl substituted FR900098 analogues
4a,f–g displayed poor inhibitory activity towards E. coli
DXR. Apparently, the combination of an a-aryl substi-
tuent and an a,b-unsaturated bond constrains the rota-
tional freedom in a way that is unfavourable for
binding to the active site of DXR. It should be noted that
the relative conformation of two trans-substituents on a
cyclopropane ring diverges from that of the trans-substit-
uents in the E-configured analogues 4a,f–g. Remarkably,
the a-bromo derivative 19, capable of undergoing elec-
tronical interactions with the target enzyme, performs
much better than analogues 3a–e and 4a,f–g.
In conclusion, two different divergent procedures for the
preparation of both Z- and E-unsaturated a-aryl-substi-
tuted FR900098 analogues 3 and 4 were developed
using, respectively, a Stille and a Suzuki coupling as a
key step. Unfortunately, with the exception of vinylic
bromide 19, none of the synthesized compounds exhib-
ited submicromolar DXR inhibitory activity.
Scheme 4. Reagents and conditions: (a) PhB(OH)2, Pd(PPh3)4,
Na2CO3, DME, H2O, 80 ꢁC; (b) (4-CN)PhB(OH)2, Pd(PPh3)4,
Na2CO3, THF, H2O, 80 ꢁC; (c) (3,4-Cl2)PhB(OH)2, Pd(PPh3)4,
Na2CO3, dioxane, H2O, 80 ꢁC; (d) (i)—TFA, CH2Cl2, 0 ꢁC;
(ii)—Ac2O, pyridine, rt; (e) BCl3, CH2Cl2, ꢀ50 ꢁC; (f) TMSBr, MeCN,
rt; NH4OH(aq); CF-11 cellulose chromatography.
Acknowledgments
This study was supported by grants from the European
Commission (Antimal, LSHP-CT-2005-018834) and
INTAS (03-51-4077).

V. Devreux et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4920–4923
4923
(dt, 1H, J = 40.0 and 5.5 Hz, @CHACH2N), 7.32 (s, 5H,
References and notes
arom. H); 13C NMR (75.47 MHz; D2O)
d 19.3
3
(s, ACH3), 48.4 (d, JPC = 6.1 Hz, @CHACH2N), 127.2
(s, arom. @CH), 128.1 (d, JPC = 3.8 Hz, 2· arom.
@CH), 128.3 (s, 2· arom. @CH), 136.3 (d,
1. Kuzuyama, T.; Shimizu, T.; Takahashi, S.; Seto, H.
Tetrahedron Lett. 1998, 39, 7913.
2. Zeidler, J.; Schwender, J.; Muller, C.; Wiesner, J.; Weide-
meyer, C.; Beck, E.; Jomaa, H. Z. Naturforsch. C 1998, 53,
980.
3. 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. Antimicrob. Agents Chemother. 2006, 50, 3749.
4. 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, 2713.
5. Woo, Y. H.; Fernandes, R. P. M.; Proteau, P. J. Bioorg.
Med. Chem. 2006, 14, 2375.
6. Ortmann, R.; Wiesner, J.; Reichenberg, A.; Henschker,
D.; Beck, E.; Jomaa, H.; Schlitzer, M. Bioorg. Med. Chem.
Lett. 2003, 13, 2163.
7. Ortmann, R.; Wiesner, J.; Reichenberg, A.; Henschker,
D.; Beck, E.; Jomaa, H.; Schlitzer, M. Arch. Pharm. 2005,
338, 305.
8. Reichenberg, A.; Wiesner, J.; Weidemeyer, C.; Dreiseidler,
E.; Sanderbrand, S.; Altincicek, B.; Beck, E.; Schlitzer, M.;
Jomaa, H. Bioorg. Med. Chem. Lett. 2001, 11, 833.
9. Kuntz, L.; Tritsch, D.; Grosdemange-Billiard, C.; Hem-
merlin, A.; Willem, A.; Bach, T. J.; Rohmer, M. Biochem.
J. 2005, 386, 127.
2
2JPC = 7.7 Hz, @CHACH2N), 142.1 (d, JPC = 11.0 Hz,
1
arom. @C), 142.6 (d, JPC = 164.1 Hz, PAC@CH), 173.5
(s, NAC@O); 31P NMR (121.50 MHz, D2O) d 8.9 ppm;
HR-MS (ESI-MS) [MꢀH]ꢀ found, 270.0535; calcd,
270.0531.
18. Kobayashi, Y.; William, A. D. Adv. Synth. Catal. 2004,
346, 1749.
19. Xiong, Z. C.; Huang, X. J. Chem. Res. 2003, 372.
20. Typical procedure for the preparation and purification of
the final ammonium phosphonates demonstrated for 4a:
To a solution of 16a (72 mg, 0.219 mmol) in dry MeCN
(2.2 mL) was added dropwise TMSBr (290 lL,
3.67 mmol) at rt and the mixture was stirred for 24 h.
The solvents were removed under reduced pressure and
remaining traces of TMSBr were removed under high
vacuum (0.05 mbar). The residual oil was dissolved in
2 mL of Type I water and the pH of the mixture was
adjusted to 8–9 with a 5% NH4OH solution. The
solution was lyophilized and the residual solid was
purified by Whatman CF-11 cellulose column chroma-
tography [MeCN/NH4OH (aq, 1 M): 4:1]. The fractions
were assayed using cellulose TLC and the spots were
visualized under UV-light (365 nm) after dipping in a
pinacryptol yellow solution (0.1% in H2O) and drying
the plate under a stream of hot air. The appropriate
fractions were lyophilized, yielding 32 mg of a white
hygroscopic solid (54%). Spectral data for 4a: 1H NMR
(300.13 MHz; D2O) d 1.83 (s, 3H, minor ACH3), 2.02
(s, 3H, major ACH3), 4.94 (dd, 2H, J = 6.4 and 2.7 Hz,
@CHACH2N), 6.35 (dt, 1H, J = 20.6 and 6.6 Hz, major
@CHACH2N), 6.42–6.52 (m, 1H, minor @CHACH2N),
7.22 (app dt, 2H, J = 8.0 and 1.6 Hz, arom. H), 7.31–
7.40 (m, 3H, arom. H); 13C NMR (75.47 MHz; D2O) d
10. Haemers, T.; Wiesner, J.; Van Poecke, S.; Goeman, J.;
Henschker, D.; Beck, E.; Jomaa, H.; Van Calenbergh, S.
Bioorg. Med. Chem. Lett. 2006, 16, 1888.
11. Haemers, T.; Wiesner, J.; Busson, R.; Jomaa, H.; Van
Calenbergh, S. Eur. J. Org. Chem. 2006, 3856.
12. Kurz, T.; Schluter, K.; Kaula, U.; Bergmann, B.; Walter,
¨
R. D.; Geffken, D. Bioorg. Med. Chem. 2006, 14, 5121.
13. Schluter, K.; Walter, R. D.; Bergmann, B.; Kurz, T. Eur.
¨
J. Med. Chem. 2006, 41, 1385.
3
19.2 (s, ACH3), 46.7 (d, JPC = 18.6 Hz, @CHACH2N),
5
127.5 (d, arom. @CH, JPC = 1.7 Hz), 128.4 (s, 2· arom.
14. Kurz, T.; Behrendt, C.; Kaula, U.; Bergmann, B.; Walter,
R. D. Aust. J. Chem. 2007, 60, 154.
15. Devreux, V.; Wiesner, J.; Goeman, J. L.; Van der Eycken,
J.; Jomaa, H.; Van Calenbergh, S. J. Med. Chem. 2006, 49,
2656.
@CH), 129.0 (d, JPC = 4.4 Hz, 2· arom. @CH), 132.9
2
2
(d, JPC = 9.9 Hz, arom. @C), 136.3 (d, JPC = 8.8 Hz,
1
@CHACH2N), 142.7 (d, JPC = 168.0 Hz, PAC@CH),
173.7 (s, NAC@O); 31P NMR (121.50 MHz; D2O) d
10.5 and 10.8 ppm (minor and major isomer); HR-MS
(ESI-MS) [Mꢀ2· NH4þ+H+]ꢀfound, 270.0535; calcd,
270.0531.
16. Devreux, V.; Wiesner, J.; Jomaa, H.; Rozenski, J.; Van der
Eycken, J.; Van Calenbergh, S. J. Org. Chem. 2007, 72,
3783.
17. Spectral data for 3a: 1H NMR (300.13 MHz; D2O) d
2.09 (s, 3H, ACH3), 4.74–4.76 (m, 2H, ANCH2), 5.94
21. Kuzuyama, T.; Takahashi, S.; Watanabe, H.; Seto, H.
Tetrahedron Lett. 1998, 39, 4509.