From a total synthesis of cepabactin and its 3:1 ferric complex to the isolation of a 1:1:1 mixed complex between iron (III), cepabactin and pyochelin

Article abstract of DOI:10.1016/j.bmcl.2005.01.034

A novel and straightforward total synthesis of cepabactin and its iron (III) complex is described. The latter compound was compared and identified to that obtained from the cultures of Burkholderia cepacia. On treatment of the growth medium of two different strains of B. cepacia with ferric chloride, we have isolated and characterized an unexpected mixed complex of iron (III), cepabactin and pyochelin.

Full text of DOI:10.1016/j.bmcl.2005.01.034

Bioorganic & Medicinal Chemistry Letters 15 (2005) 1721–1724
From a total synthesis of cepabactin and its 3:1 ferric complex
to the isolation of a 1:1:1 mixed complex between iron (III),
cepabactin and pyochelin
*
Cedric Klumpp, Alain Burger, Gaetan L. Mislin and Mohamed A. Abdallah
*
´
¨
´
´
´
´
Metaux et Microorganismes: Chimie, Biologie et Applications, Departement des Recepteurs et Proteines Membranaires
´ ´
(CNRS UPR 9050), Ecole Superieure de Biotechnologie de Strasbourg, Boulevard Sebastien Brant, F-67400 Illkirch, France
Received 17 November 2004; revised 6 January 2005; accepted 13 January 2005
Abstract—A novel and straightforward total synthesis of cepabactin and its iron (III) complex is described. The latter compound
was compared and identified to that obtained from the cultures of Burkholderia cepacia. On treatment of the growth medium of two
different strains of B. cepacia with ferric chloride, we have isolated and characterized an unexpected mixed complex of iron (III),
cepabactin and pyochelin.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Iron is one of the most essential elements for aerobic life.
Although extremely abundant in earth crust, in aerobic
conditions and at physiological pH, iron (III) yields
insoluble polymeric oxides in aqueous media. When
grown in iron deficient conditions, microorganisms syn-
thesize and excrete into extracellular medium, small
molecules called siderophores in order to strongly chel-
ate iron (III).1 In Gram-negative bacteria, the ferrisidero-
phore is translocated through the membranes into the
cytoplasm by a multiproteic system.2 Burkholderia (ex-
Pseudomonas) cepacia3 is an emerging opportunistic
pathogen responsible of the Ôcepacia syndromeÕ, charac-
terized by severe lung infections, often lethal especially
for cystic fibrosis patients.4 In the last decade, this bac-
terium increased its resistance against most of the anti-
biotics currently used. In a programme to develop new
therapeutic strategies against B. cepacia, we have fo-
cused our attention on iron uptake systems in this bac-
terium. B. cepacia excretes mainly three siderophores:
cepabactin 1,5 pyochelin 26 and ornibactins 3,7 in strain
depending proportions (Fig. 1).8
In due course to our studies of iron uptake in Pseudomo-
nas aeruginosa,9 we have described a few years ago a ste-
reocontrolled synthesis of pyochelin 2.10 This protocol
gave us an access to a panel of analogues currently
under investigation for their biological activities.11 Sim-
ilarly we wish to develop a comparable strategy based
on cepabactin 1, which would be applied to B. cepacia.
In this respect we describe in this paper, a new and
straightforward synthesis of cepabactin and its iron
(III) complex. As a matter of comparison we have also
isolated the natural ferricepabactin from B. cepacia
OH
H3CO
COOH
H
N
N
N
O
S
S
OH
H
2
1
OH
NH2
O
O
H
H
N
N
N
H
NH
O
O
N
COOH
HO
O
N
HO
Keywords: Burkholderia cepacia; Siderophore; Cepabactin; Pyochelin;
Microwave-assisted reaction.
OH
OH
NH2
*
Corresponding authors. Tel.: +33 3 9024 4727; fax: +33 3 9024
4829; e-mail: mislin@esbs.u-strasbg.fr
Present address: Laboratoire de Chimie Bioorganique, UMR 6001
O
R
R = CH3, C3H7, C5H11
3
´
CNRS, Universite de Nice Sophia Antipolis, Parc Valrose, F-06108
Nice Cedex 2, France.
Figure 1. The main siderophores of B. cepacia.
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.01.034

1722
C. Klumpp et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1721–1724
cultures. After treatment of the culture medium with fer-
ric chloride, during the purification process, an unprec-
edented type of mixed complex between iron (III),
cepabactin and pyochelin was isolated and is described
as well in this article.
H3CO
N
O
O
H3CO
O
O
OCH3
N
Fe3+
O
O
i, ii, iii
N
O
Br
Ohta et al. described the first total synthesis of cepabac-
tin 1.12 Unfortunately this protocol, which gave 1 only
in poor yield, is based on an expensive commercially
available starting material.13 Therefore our synthesis
starts with the bromination of commercial 3-hydroxy-
2-methylpyridine 4 in alkaline medium according to a
procedure reported by Hassberg and Gerlach.14
Although under these conditions the yield of the bromi-
nated product 5 was low (22%), this latter could be eas-
ily isolated from a complex mixture by selective
precipitation. The hydroxyl function of the substituted
pyridine 5 was then methylated using either trimethylsil-
yldiazomethane or iodomethane. Both protocols led to
the expected compound 6 with approximately the same
yield, 91% and 95%, respectively.15 The oxidation of
the pyridine nitrogen atom was performed using
mCPBA in refluxing chloroform leading to the expected
N-oxide 7 isolated in 84% yield.16 Other oxidizing sys-
tems like potassium perborate in acetic acid or aqueous
hydrogen peroxide in trifluoroacetic acid appeared to be
poorly efficient (Scheme 1).
N
7
8
OCH3
Scheme 2. Reagents and conditions: (i) 10% NaOHaq, l-waves; (ii)
HCl; (iii) FeCl3.
ium under iron depleted conditions.18 When the super-
natant was complemented with ferric chloride, the
corresponding metal (III)–siderophore complexes were
extracted with CH2Cl2. Under these conditions, two dif-
ferent complexes, an orange and a purple one, were iso-
lated and purified by chromatography.
The orange complex showed two maxima in UV at
445 nm
(e = 4880 MÀ1 cmÀ1
)
and
at
314 nm
(e = 14800 MÀ1 cmÀ1) in methanol suggesting that this
compound is the ferricepabactin according to previously
reported spectrophotometric properties.5 This was con-
firmed by the following mass spectrometry experiments.
Thus, high resolution mass measurements performed on
this complex gave a mass of 518.0860 as expected for
iron (III)–cepabactin complex (calcd mass: 518.0862).
The electron impact spectra gave a main signal at m/z
364 corresponding to a fragment with two molecules
of ligand bound to one metallic ion. In a last experi-
ment, submitted to electrospray, the orange complex
gave two major signals at m/z 541 and at m/z 1059 cor-
responding, respectively, to M+Na+ and 2M+Na+. Sub-
mitted to the same UV and mass spectrometry
measurements, the synthetic complex 8 proved to be
identical to the natural ferricepabactin.
In the last step of our protocol, the halogen/hydroxide
exchange generated the expected cepabactin 1. Among
the different conditions tested, the best results were ob-
tained when the reaction was performed in a 10% aque-
ous sodium hydroxide solution using microwave assisted
reaction. The crude mixture was then acidified and con-
tinuously extracted with chloroform yielding crude
cepabactin 1. If ferric chloride was added at this stage,
after extraction and purification, the expected ferrice-
pabactin 8 was isolated in 58% yield.17 Without micro-
waves, the same reaction gave 8 only in 27% yield
after one week with a classical reflux (Scheme 2).
In UV, the purple complex dissolved in methanol
showed three maxima, respectively, at 469 nm
(e = 2400 MÀ1 cmÀ1), 398 nm (e = 2700 MÀ1 cmÀ1) and
319 nm (e = 9900 MÀ1 cmÀ1). Using FAB mass spec-
trometry measurements this complex gave a signal at
m/z 534 and some of the corresponding clusters. By add-
ing dilute acid to the sample, this signal disappeared giv-
ing rise to a signal at m/z 519 corresponding to the
ferricepabactin and a signal at m/z 364 corresponding
to the fragment already observed above (iron chelated
by two molecules of cepabactin). These signals were
accompanied by a signal at m/z 325 corresponding to
free pyochelin (M+H+). These results proved that this
purple complex contains iron (III), cepabactin and
pyochelin. After calibration, MS electrospray gave two
signals at m/z 533.037 and m/z 555.019 and mainly some
of the corresponding adducts. These calculations agreed
with a 1:1:1 stoichiometry between iron (III), cepabactin
and pyochelin in the complex 9 (calcd mass for M+H+:
533.038 and for M+Na+: 555.019).
To confirm the structure of 8 we decided to compare its
properties with the natural molecule. For this purpose
two different B. cepacia strains (ATCC 17754 and
ATCC 25416) were grown in succinate containing med-
HO
HO
H3CO
ii or iii
N
i
N
Br
N
6
Br
Br
4
5
iv
H3CO
N
O
7
Scheme 1. Reagents and conditions: (i) Br2, NaOHaq, 20 ꢁC (yield:
22%); (ii) TMSCHN2, MeOH, CH2Cl2, 20 ꢁC (yield: 91%); (iii) CH3I,
K2CO3, acetone, reflux (yield: 95%); (iv) mCPBA, CHCl3, reflux (yield:
84%).
These mass spectrometry experiments are quite consis-
tent with the low stability of the pyochelin metal com-

C. Klumpp et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1721–1724
3. Burkholder, W. Phytopathology 1950, 40, 115–118.
1723
OCH3
N
4. (a) Govan, J. R.; Hughes, J. E.; Vandamme, P. J. Med.
Microbiol. 1996, 45, 395–407; (b) Meghdas, I.; Loiez, C.;
Baida, N.; Dabboussi, F.; Hamze, M.; Husson, M.-O.;
O
´
Izard, D. Arch. Pediatrie 2004, 11, 360–366.
O
5. Barker, W. R.; Callaghan, C.; Hill, L.; Noble, D. J.
Antibiot. 1979, 32, 1096–1102; (a) Itoh, J.; Miyadoh, S.;
Takahasi, S.; Amano, S.; Ezaki, N.; Yamada, Y. J.
Antibiot. 1979, 32, 1089–1095; (b) Itoh, J.; Amano, S.;
Ogawa, Y.; Kodama, Y.; Ezaki, N.; Yamada, Y. J.
Antibiot. 1980, 33, 377–382; (c) Meyer, J.-M.; Hohnadel,
O
Fe3+
O
O
N
N
S
S
´
D.; Halle, F. J. Gen. Microbiol. 1989, 135, 1479–1487.
9
6. (a) Cox, C. D.; Graham, R. J. Bacteriol. 1979, 137, 357–
364; (b) Liu, P. V.; Shokrani, F. Infect. Immunol. 1978, 22,
878–890; (c) Cox, C. D.; Rinehart, K. L.; Moore, M. L.;
Cook, J. C. Proc. Natl. Acad. Sci. U.S.A. 1981, 78, 4256–
4260.
Figure 2. The 1:1:1 complex of iron (III), cepabactin and pyochelin.
7. (a) Stephan, H.; Freund, S.; Beck, W.; Jung, G.; Meyer,
J.-M.; Winkelmann, G. Biometals 1993, 6, 93–100; (b)
Sokol, P. A.; Darling, P.; Lewenza, S.; Corbett, C. R.;
Kooi, C. D. Infect. Immunol. 2000, 68, 6554–6560.
8. Darling, P.; Chan, M.; Cox, A. D.; Sokol, P. A. Infect.
Immunol. 1998, 66, 874–877.
9. For selected recent publications of our group: (a) Hen-
nard, C.; Truong, Q. C.; Desnottes, J. F.; Paris, J. M.;
Moreau, N. J.; Abdallah, M. A. J. Med. Chem. 2001, 44,
2139–2151; (b) Schalk, I. J.; Abdallah, M. A.; Pattus, F.
Biochemistry 2002, 41, 1663–1671; (c) Schalk, I. J.;
Abdallah, M. A.; Pattus, F. Biochem. Soc. Trans. 2002,
30, 702–705; (d) Folschweiller, N.; Gallay, J.; Vincent, M.;
Abdallah, M. A.; Pattus, F.; Schalk, I. J. Biochemistry
2002, 41, 14591–14601.
plexes in solution and show that the most stable com-
plexes involve the presence of cepabactin either as a
3:1 iron complex 8 or as a mixed complex 9. Neverthe-
less this latter seems somehow less stable than the for-
mer since it tends to equilibrate to it. Finally in such a
mixed complex 9, pyochelin apparently contributes to
the complexation giving four of the six dentates neces-
sary to achieve an octahedral ferric complex. Some very
recent results based on the tridimensional structure of
the specific outer membrane receptor FptA loaded with
ferripyochelin seem to confirm this observation.20 Taken
together these results support the structure shown above
for the complex 9 (Fig. 2).
10. (a) Zamri, A.; Abdallah, M. A. Tetrahedron 2000, 56, 249–
256; (b) Zamri, A.; Abdallah, M. A. Tetrahedron 2000, 56,
9397; (c) For other examples of pyochelin synthesis,
mutasynthesis or biosynthesis, see: Ino, A.; Murabayashi,
A. Tetrahedron 2001, 57, 1897–1902; (d) Rinehart, K. L.;
Staley, A. L.; Wilson, S. R.; Ankenbauer, R. G.; Cox, C.
D. J. Org. Chem. 1995, 60, 2786–2791; (e) Ankenbauer, R.
G.; Toyokuni, T.; Staley, A.; Rinehart, K. L., Jr.; Cox, C.
D. J. Bacteriol. 1988, 170, 5344–5351.
11. (a) Zamri, A.; Schalk, I. J.; Pattus, F.; Abdallah, M. A.
Bioorg. Med. Chem. Lett. 2003, 13, 1147–1150; (b) Mislin,
G. L.; Burger, A.; Abdallah, M. A. Tetrahedron 2004, 60,
12139–12145.
In conclusion, we have described in this paper a novel
and straightforward microwave assisted synthesis of
cepabactin 1, a siderophore of B. cepacia and of its iron
(III) complex 8. A strong point is that this protocol will
give an easy access to cepabactin analogues with a diver-
sity point located on position 5 using alkylation reac-
tions on the key compound 5. These types of
analogues are, to the best of our knowledge, only poorly
documented in the literature and are currently under
investigation in our group. In addition we have isolated
and characterized a new type of mixed complex 9
formed by two siderophores of B. cepacia, pyochelin
and cepabactin, and iron (III). The occurrence and sta-
bility of such a complex raises the question of the species
that effectively transport iron into bacterial cells.
12. Ohta, A.; Takahashi, N.; Shirokoma, Y.; Yuasa, K.
Heterocycles 1990, 30, 875–884.
13. The starting material was the 3-methoxy-2-(1H)-pyridone.
The synthesis of the latter compound starting from the less
costly 1,2-dihydroxypyridine is described: Hanessian, S.;
Saavedra, O. M.; Mascitti, V.; Marterer, W.; Oehrlein, R.;
Mak, C.-P. Tetrahedron 2001, 57, 3267–3280, but in our
hands this protocol led only to an unworkable mixture of
compounds.
Acknowledgements
We wish to thank the Centre National de la Recherche
Scientifique (CNRS) and the association Vaincre la
Mucoviscidose for financial support. We also wish to
thank Dr. Martine Schmitt, Dr. Jean-Jacques Bourgui-
14. Hasseberg, H.-A.; Gerlach, H. Liebigs Ann. Chem. 1989,
255–261.
15. 6-Bromo-3-methoxy-2-methylpyridine 6: a solution of 5
(332 mg, 1.77 mmol), CH3I (132 lL, 301 mg, 2.13 mmol),
K2CO3 (489 mg, 3.54 mmol) in acetone (15 mL) was
refluxed 16 h. The mixture was cooled to room tempera-
ture, filtered through a Celite pad and adsorbed on silica
before being purified on a silica gel column (20 g SiO2,
hexane/Et2O: 2:1) leading to the expected methylether 6
(338 mg, 1.68 mmol, yield: 95%) isolated as a white solid.
Mp 63–65 ꢁC; Rf 0.55 (hexane/Et2O: 2:1); 1H NMR
(300 MHz, CDCl3): d 2.49 (s, 3H), 3.86 (s, 3H), 6.71 (d,
J = 9.1 Hz, 1H), 7.43 (d, J = 9.1 Hz, 1H); 13C NMR
(75 MHz, CDCl3): d 11.8, 56.4, 108.5, 124.6, 125.8, 142.3,
154.8. MS (ES+) m/z 202 (M+H+, 100), 203 (7), 204 (100),
´
gnon and Dr. Gilbert Schlewer (Faculte de Pharmacie,
Illkirch, France) for their kind help during microwave
experiments.
References and notes
1. For a review, see: Abdallah, M. A.; Pattus, F. J. Chin.
Chem. Soc. 2000, 47, 1–20.
2. For a recent review on iron uptake in P. aeruginosa, see:
Poole, K.; McKay, G. A. Front. Biosci. 2003, 661–686.

1724
C. Klumpp et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1721–1724
205 (7); IR (neat, cmÀ1): 3078, 3011, 2926, 2845,
liquid–liquid extraction with CHCl3. The resulting deep
red solution was dried over Na2SO4, filtered, and evapo-
rated under reduced pressure. The crude (54 mg) was
dissolved in a minimum amount of CH2Cl2 and hexane
was dropwise added until a red precipitate appears. This
precipitate was filtered off and dilution in MeOH induced
the formation of a white precipitate. The latter was
eliminated by filtration and the resulting deep orange
filtrate was evaporated under reduced pressure. The final
purification could be performed on Sephadex LH 20.19
The expected ferricepabactin 8 (13 mg, yield: 58%) was
isolated as orange-red crystals. Analytical and spectro-
photometric data were consistent with those described
previously for the natural ferrisiderophore.5,19
2541, 2019, 1870, 1633, 1576, 1447, 1427, 1387, 1364,
1305, 1280, 1252, 1198, 1176, 1124, 1011, 970, 940, 931,
873, 820, 809, 749, 727. Anal. Calcd for C7H8BrNO:
C, 41.61; H, 3.99; N, 6.93. Found: C, 41.93; H, 4.21; N,
6.75.
16. 6-Bromo-3-methoxy-2-methylpyridine-1-oxide 7: a solution
of (44 mg, 0.22 mmol), mCPBA 70% (108 mg,
6
0.44 mmol) in CHCl3 (1 mL) was refluxed 18 h. The
mixture was cooled to room temperature then a 0.5 N
aqueous NaOH solution (10 mL), was added and the
mixture was vigorously stirred for 1 h. After dilution with
water (10 mL), the mixture was extracted with CHCl3
(3 · 25 mL) and the organic layers were collected, dried
over Na2SO4, and filtered. After evaporation under
18. Culture conditions for B. cepacia: the standard succinate
culture medium: Meyer, J.-M.; Abdallah, M. A. J. Gen.
Microbiol. 1978, 22, 878–890, contained no added source
of iron. B. cepacia ATCC 17754 or ATCC 25416 were
grown for 24 h at 25 ꢁC in this medium for preculture.
This culture (2 mL) were added to 500 mL of culture
medium as an inoculum, and bacteria were grown for 48 h
at 25 ꢁC.
19. Isolation and purification of the iron-complexes: 5 L of
bacterial culture grown as above were treated after
centrifugation with a 2 M FeCl3 solution (10 mL) and
the supernatant was extracted with methylene chloride
(3 · 1.0 L). The organic solution was washed with water
(3 · 0.4 L), dried, and evaporated to yield 500 mg of a
crude mixture. This was dissolved in acetonitrile/water
(70:30) and chromatographed on a Sephadex LH 20
column (900 mL) made up in the same eluent, which
separated the orange ferricepabactin complex 8 (210 mg)
from the purple mixed complex 9 (165 mg). Both com-
pounds could be easily distinguished by TLC on silica gel
eluted with MeOH/CHCl3 1:9 with respective Rf of 0.50
and 0.25.
reduced pressure, the pure N-oxide
7 was isolated
(40 mg, 0.18 mmol, yield: 84%) as a white solid. Mp
142–144 ꢁC; Rf 0.46 (CH2Cl2/MeOH: 95:5); 1H NMR
(300 MHz, CDCl3): d 2.41 (s, 3H), 3.80 (s, 3H), 6.94 (d,
J = 8.5 Hz, 1H), 7.21 (d, J = 8.5 Hz, 1H); 13C NMR
(75 MHz, CDCl3): d 19.0, 55.6, 119.4, 125.4, 130.0, 150.0,
153.5. MS (ES+) m/z 240 (M+Na+, 38), 241 (4), 242 (38),
243 (7), 277 (24), 278 (4), 279 (24), 280 (4), 457 (2M+Na+,
48), 458 (7), 459 (2M+2H++Na+, 100), 460 (17), 461 (48),
462 (7). IR (neat, cmÀ1): 3073, 3006, 2939, 2840, 2015,
1863, 1593, 1554, 1470, 1454, 1351, 1291, 1236, 1201, 1175,
1107, 1009, 930, 855, 801, 717. Anal. Calcd for
C7H8BrNO2: C, 38.56; H, 3.70; N, 6.42. Found: C,
39.02; H, 3.89; N, 6.26.
17. Cepabactin 1 and ferricepabactin 8: in an adapted reaction
tube, the N-oxide 7 (28 mg, 0.13 mmol) was dissolved in
10% aqueous solution of NaOH (4 mL). The mixture was
irradiated 20 min (200 W, 150 ꢁC, 12 bar) in a microwave
reactor. The mixture was cooled to room temperature and
the pH was adjusted to 2.0 with concentrated HCl. At this
stage, the ligand 1 could be isolated by an overnight
liquid–liquid extraction with CHCl3 whereas the ferrice-
pabactin 8 was prepared by adding a 2 M FeCl3 solution
(200 lL). The complex was then isolated by an overnight
20. Cobessi, D.; Pattus, F., private communication. See also:
Schlegel, K.; Taraz, K.; Budzikiewicz, H. Biometals 2004,
17, 409–414.