Vanchrobactin: absolute configuration and total synthesis

Article abstract of DOI:10.1016/j.tetlet.2007.02.125

The stereochemistry of vanchrobactin, a siderophore produced by the bacterial fish pathogenVibrio anguillarum serotype O2, was elucidated by chiral capillary electrophoresis analysis and total synthesis as N-[N′-(2,3-dihydroxybenzoyl)-d-arginyl]-l-serine.

Full text of DOI:10.1016/j.tetlet.2007.02.125

Tetrahedron Letters 48 (2007) 3021–3024
Vanchrobactin: absolute configuration and total synthesis
Raquel G. Soengas,a Cristina Anta,b Alfonso Espada,b Rosa M. Nieto,a Marta Larrosa,a
a
a,
*
´
´
Jaime Rodrıguez and Carlos Jimenez
aDepartamento de Quı´mica Fundamental, Facultad de Ciencias, Universidad de A Corun˜a, A Corun˜a E-15071, Spain
bAnalytical Technologies DCR&T Alcobendas, Lilly S.A., Avda. de la Industria 30, E-28108-Alcobendas/Madrid, Spain
Received 23 January 2007; revised 21 February 2007; accepted 26 February 2007
Available online 1 March 2007
Abstract—The stereochemistry of vanchrobactin, a siderophore produced by the bacterial fish pathogenVibrio anguillarum serotype
O2, was elucidated by chiral capillary electrophoresis analysis and total synthesis as N-[N0-(2,3-dihydroxybenzoyl)-D-arginyl]-L-
serine.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Iron is the fourth most abundant element in the earth’s
crust and is an essential nutrient for most microorgan-
isms. However, the low solubility of Fe(III) in aqueous
media (10ꢀ8 M at neutral pH) prompted many bacteria
to develop specialised mechanisms for its acquisition.
Iron uptake is generally mediated by siderophores,
which are low molecular weight (300–2000 Da) high-
affinity chelators produced by microorganisms to
sequester Fe3+. These molecules, which are generally
excreted into the culture medium, are able to strongly
chelate in a specific manner to solubilize and deliver
Fe3+ into the cells.1
firmed not only that arginine and serine are the amino
acids incorporated in vanchrobactin but also the assem-
bly order of the three components (DHBA, arginine,
and serine).5 However, the stereochemistry of vanchro-
bactin could not be elucidated by spectroscopic analysis.
In order to confirm the proposed planar structure of
vanchrobactin and to determine its absolute configura-
tion, a synthetic sample was required. A practical meth-
od for the synthesis of vanchrobactin would allow the
development of new strategies to interfere with the
iron-acquisition mechanisms of V. anguillarum or
approaches that could provide ‘trojan horses’6 for the
development of new rationally designed antibacterial
agents against vibriosis.7
In the course of our studies into siderophores from mar-
ine microorganisms, vanchrobactin was isolated from
iron-deficient cultures in the CM9 medium of the bacte-
rial fish pathogen V. anguillarum serotype O2 strain
RV22.2 This pathogen is the causative agent of vibriosis,
an extremely fatal hemorrhagic septicaemia that results
in considerable economic losses in aquaculture farming
worldwide.3 It is known that the ability to scavenge iron
through the utilization of siderophores is a key factor in
the virulence of this fish pathogen.4 The planar structure
of vanchrobactin was established by extensive NMR
studies on the natural sample. The identification and
characterization of the genes and enzymes involved in
the biosynthesis of vanchrobactin in strain RV22, con-
Herein we report a short and efficient method for the
synthesis of vanchrobactin and the elucidation of its
absolute configuration by chiral capillary electrophore-
sis analysis. The configuration was unequivocally deter-
mined by comparison of the natural product with a
synthetic sample.
Keywords: Vanchrobactin; Marine siderophore; Vibriosis; Vibrio
anguillarum; Iron uptake; Synthesis; Absolute configuration.
*
Corresponding author. Tel.: +34 981 167000; fax: +34 981 167065;
e-mail: carlosjg@udc.es
0040-4039/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.02.125

3022
R. G. Soengas et al. / Tetrahedron Letters 48 (2007) 3021–3024
Firstly, we planned the synthesis of two of the four
possible stereoisomers,8 compounds 1a and 1b, using
dihydroxybenzoic acid, L-ornithine, D-serine, and L-ser-
ine as starting materials, and introducing the guanidine
functionality in a later step.
acidic hydrolysis of the tert-butoxy protecting group
afforded compounds 1a and 1b, respectively, in quanti-
tative yield as amorphous solids.
HPLC analysis of these stereoisomers using a reversed-
phase achiral column (Discoveryꢁ HS F5) showed that
compound 1b had the same retention time as the natural
product (see Fig. 1). However, chiral capillary electro-
phoresis analysis of compounds 1a and 1b on a highly
sulfated gamma cyclodextrin chiral selector, in com-
parison to the natural product, indicated that they did
not correspond to the same diastereoisomer. The two
compounds gave different migration times (see Fig. 2),
allowing us to postulate that vanchrobactin must be
the enantiomer of 1b. In order to confirm this hypo-
thesis, we carried out the synthesis of the stereoisomer
bearing D-arginine and L-serine amino acid residues
according to the synthetic strategy described previously
for compounds 1a and 1b.
The synthesis started from the commercially available
2,3-dihydroxybenzoic acid (DHBA) which, after conver-
sion into the methyl ester with MeOH and SOCl2, was
protected with benzyl bromide in the presence of potas-
sium carbonate under reflux to give 2. Subsequent
saponification with barium hydroxide in THF/H2O
(2:1) gave 2,3-dibenzyloxybenzoic acid (3) in quantita-
tive yield.9 On the other hand, Nd-Cbz-L-ornithine-
OMe (4a) was prepared by esterification of the commer-
cially available Nd-Cbz-L-ornithine.10 The coupling of
DHBA derivative 3 and Nd-Cbz-L-ornithine-OMe (4a)
using TBTU as the coupling agent,11 followed by hydro-
lysis with barium hydroxide in THF/H2O (1:1), gave
acid 5a in very good yield (Scheme 1).
In this case, Nd-Cbz-D-ornithine-OMe (4b) was prepared
from the Cbz derivative of D-ornithine copper complex,
elimination of the metal cation with EDTA and esterifi-
cation of the free acid group.14 Coupling of 4b with
DHBA derivative 3 gave compound 5b which, after
saponification, was coupled to L-serine tert-butyl ester
to afford 6c.15 Deprotection of the benzyl groups in
6c, introduction of the guanidine functionality to give
7c and acidic hydrolysis yielded 1c. Chiral capillary elec-
trophoresis analysis showed that 1c has the same migra-
tion time as the natural product and these compounds
were coeluted when a mixture was injected (see Fig. 2).
L-Serine and D-serine were converted into the corre-
sponding tert-butyl esters with AcOtBu in the presence
of a catalytic amount of HClO4.12 These compounds
were then coupled to acid 5a with TBTU to afford com-
pounds 6a and 6b in 91% and 82% yield, respectively.
Derivatives 6a and 6b were submitted to catalytic hydro-
genation to remove the benzyl protecting groups, and the
guanidine function was introduced by reaction with 1,3-
di-Boc-2-(trifluoromethanesulfonyl)guanidine13 to give
7a and 7b in 69% and 63% yield, respectively. Finally,
Scheme 1. Reagents and conditions: (i) (1) CuSO4, H2O, NaOH, (2) NaOH, dioxane, ClCbz, (3) EDTA, 1 M HCl, (4) SOCl2/MeOH (48% four
steps); (ii) (1) SOCl2/MeOH (89%), (2) BnBr/K2CO3 (99%); (iii) Ba(OH)2, THF, H2O, 50 ꢂC, 12 h, quant.; (iv) (1) Nd-Cbz-L-ornithine-OMe (4a) or
Nd-Cbz-D-ornithine-OMe (4b), TBTU, Et3N, DMF, rt, 16 h (78% and 79%, respectively); (2) Ba(OH)2, THF, H2O, 50 ꢂC, 12 h, quant; (v) (1) H2, Pd-
C, MeOH, rt, 3 h, quant.; (2) (NHBoc)2NTf, Et3N, CHCl3, rt, 1 h (69% for 7a, 63% for 7b, 61% for 7c); (vi) TFA/DCM 3:7, rt, 16 h, quant.

R. G. Soengas et al. / Tetrahedron Letters 48 (2007) 3021–3024
3023
residue instead of arginine.18 Interestingly, this com-
pound also has the L configuration for the serine residue,
whereas the lysine residue has a D configuration.19
In conclusion, the elucidation of the absolute stereo-
chemistry of vanchrobactin and its total synthesis have
been achieved. This has important implications for the
mechanism of vanchrobactin biosynthesis and for the
development of new rationally designed antibacterial
compounds against vibriosis. Studies on the structure-
activity relationships of vanchrobactin are currently
under way.
Figure 1. UV214nm chromatograms (from top to bottom) of 1b,
vanchrobactin, vanchrobactin + 1b (1 to 1 mixture), and 1a. Column
4.6 · 50 mm, 3 lm; mobile phase, H2O 0.05% TFA and MeCN 0.05%
TFA (90/10); flow rate, 1.0 ml/min.
Acknowledgments
This work was financially supported by the Xunta
de Galicia (PGIDIT05RMA10302PR and PGIDIT06-
PXIC103118PN) and Ministry of Science and Techno-
logy of Spain (cofunded by FEDER) (CQT2005-
00793). R.G.S. thanks the Parga Pondal Programme.
We thank Professor Manuel L. Lemos (Universidad de
Santiago de Compostela, Spain) for the biological
assays.
References and notes
1. (a) Ratledge, C.; Dover, L. G. Annu. Rev. Microbiol. 2000,
54, 881–941; (b) Wandersman, C.; Delepelaire, P. Annu.
Rev. Microbiol. 2004, 58, 611–647.
2. Soengas, R. G.; Anta, C.; Espada, A.; Paz, V.; Ares, I. R.;
´
´
Balado, M.; Rodrıguez, J.; Lemos, M. L.; Jimenez, C.
Tetrahedron Lett. 2006, 47, 7113–7116.
3. Toranzo, A. E.; Barja, J. L. Dis. Aquat. Org. 1990, 9, 73–
82.
4. Wolf, M. K.; Crosa, J. H. J. Gen. Microbiol. 1986, 132,
2949–2952.
5. Balado, M.; Osorio, C. R.; Lemos, M. L. Microbiology
2006, 152, 3517–3528.
Figure 2. UV214nm electropherograms (from top to bottom) of 1b,
vanchrobactin, 1c, vanchrobactin + 1c (1 to 1 mixture), and 1a.
Capillary dimension, 50 lm i.d., 363 lm o.d.; total capillary length,
38.5 cm; length to detector, 30 cm; running buffer, 5% gamma-HS-CD,
25 mM phosphate buffer pH 2.5; hydrodynamic injection, 60 mbars.
6. (a) Miller, M. J.; Malouin, F. Acc. Chem. Res. 1993, 26,
241–249; (b) Braun, V.; Braun, M. Curr. Opin. Microbiol.
2002, 5, 194–201.
7. Somu, R. V.; Boshoff, H.; Qiao, C.; Bennett, E. M.; Barry,
C. E., III; Aldrich, C. C. J. Med. Chem. 2006, 49, 31–34.
8. Gademann, K.; Bethuel, Y. Org. Lett. 2004, 6, 4707–4710.
9. Hu, J.; Miller, M. J. J. Am. Chem. Soc. 1997, 119, 3462–
3468.
The NMR data for the synthetic compound 1c are prac-
tically identical to those of vanchrobactin.16 The slight
chemical shift differences observed are due to the strong
pH dependence on the chemical shifts of amino acids in
D2O solution.17
The [a]D value of ꢀ16.4 (c 0.7, MeOH) of the synthetic
sample was similar to that of vanchrobactin {[a]D ꢀ13.6
(c 25 · 10ꢀ3, MeOH)}.
10. Charvat, T. T.; Lee, D. J.; Robinson, W. E.; Chamberlin,
A. R. Bioorg. Med. Chem. 2006, 14, 4552–4567.
11. Han, S-Y.; Kim, Y-A. Tetrahedron 2004, 60, 2447–2467.
12. Turner, J. J.; Sikkema, F. D.; Filippov, D. V.; van der
Marel, G. A.; van Boom, J. H. Synlett 2001, 1727–1730.
13. Feichtinger, K.; Sings, H. L.; Baker, T. J.; Matthews, K.;
Goodman, M. J. Org. Chem. 1998, 63, 8432–8439.
14. Masiukiewicz, E.; Rzeszotarska, B.; Szczerbaniewicz, J.
Org. Prep. Proced. Int. 1992, 24, 191–194.
The siderophore activity and biological function of the
synthetic compound was confirmed by testing the resto-
ration of growth of the producer cells in CM9 medium
containing the iron chelator 2,20-dipyridyl at inhibiting
concentrations. Furthermore, stereoisomers 1a and 1b
showed also growth promotion under the same condi-
tions. On the basis of these data, the absolute structure
of vanchrobactin was thus established as N-[N0-(2,3-
dihydroxybenzoyl)-D-arginyl]-L-serine.
20
1
15. Selected data for 6c: ½aꢁD +13.9 (c 1.8, CHCl3); H NMR
(300 MHz, CDCl3): d 1.45 (s, 9H, C(CH3)3), 1.79–1.90 (m,
4H, H-2 and H-3), 3.03–3.11 (m, 1H, H-1a), 3.17–3.25 (m,
1H, H-1b), 3.50 (dd, J = 8.7, 2.9 Hz, 1H, H-7a), 3.79 (dd,
J = 8.7, 2.9 Hz, 1H, H-7b), 4.59–4.70 (m, 2H, H-4 and H-
6), 5.07–5.27 (m, 6H, 3 · OCH2Ph), 6.90 (d, J = 8.4 Hz,
1H, NH), 7.13–7.46 (m, 17H, aromatics), 7.72 (dd,
J = 6.5, 2.3 Hz, 1H, aromatic), 8.51 (d, J = 7.1 Hz, 1H,
NH); 13C NMR (75 MHz, CDCl3): d 25.72 (C-2), 27.97
Chrysobactin is a siderophore isolated from the phyto-
pathogenic bacterium Erwinia chrysanthemi and is struc-
turally related to vanchrobactin but bears a lysine

3024
R. G. Soengas et al. / Tetrahedron Letters 48 (2007) 3021–3024
(C(CH3)3), 29.39 (C-3), 40.06 (C-1), 52.73, 53.09 (C-4 and
8.0 Hz, H-5 DHBA), 7.31 (dd, J = 1.5, 8.0 Hz, H-7
DHBA); 13C NMR (125 MHz, D2O): d 24.82 (C-2),
28.91 (C-3), 41.00 (C-1), 53.98 (C-4), 55.32 (C-6), 61.46
(C-7), 117.08, 119.90, 120.15, 120.21, 145.00, 147.17,
157.16 (C@N), 170.36 (CONH), 173.54 (CONH), 174.26
(CO2H). HRESIMS: m/z 398.1674 [M+H]+ (calcd for
C16H24N5O7 398.1670).
C-6), 62.06 (C-7), 71.32, 73.08, 76.08 (3 · –OCH2Ph),
81.76 (C(CH3)3), 117.32, 123.28, 124.21, 127.72, 127.90,
128.03, 128.21, 128.38, 128.49, 128.62, 128.91, 128.82,
136.20, 136.26, 146.91, 151.67, 156.43 (CONH), 165.04
(CONH), 169.26 (CONH) 171.11 (CO2C(CH3)3); HR-
ESIMS: m/z 726.3352 [M+H]+ (calcd for C41H48N3O9,
726.3385).
17. Castellanos, L.; Duque, C.; Zea, S.; Espada, A.; Rodri-
24
1
´
16. Selected data for 1c: ½aꢁD ꢀ16.4 (c 0.7, MeOH); H NMR
(500 MHz, D2O): d 1.74–2.00 (m, 4H, H-2 and H-3), 3.25
(t, J = 6.7 Hz, 2H, H-1), 3.91 (dd, J = 4.0, 11.5 Hz, 1H,
H-7a), 4.00 (dd, J = 5.0, 11.5 Hz, 1H, H-7b), 4.59 (t,
J = 4.5 Hz, 1H, H-6), 4.67 (dd, J = 5.5, 8.5 Hz, 1H, H-4),
6.90 (t, J = 8.0 Hz, 1H, H-6 DHBA), 7.11 (dd, J = 1.5,
guez, J.; Jimenez, C. Org. Lett. 2006, 8, 4967–
4970.
18. Lu, C.; Buyer, J. S.; Okonya, J. F.; Miller, M. J. BioMetals
1996, 9, 377–383.
19. Persmark, M.; Expert, D.; Neilands, J. B. J. Biol. Chem.
1989, 264, 3187–3193.