Synthesis of brasilibactin A and confirmation of absolute configuration of β-hydroxy acid fragment

Article abstract of DOI:10.1016/j.tetlet.2007.09.112

A synthesis of brasilibactin A, a cytotoxic siderophore from the actinomycete of Nocardia brasiliensis, and three unnatural diastereomers of the natural product is described. Four possible diastereomers of the β-hydroxy acid fragment were prepared via asy

Full text of DOI:10.1016/j.tetlet.2007.09.112

Available online at www.sciencedirect.com
Tetrahedron Letters 48 (2007) 8104–8107
Synthesis of brasilibactin A and confirmation of
absolute configuration of b-hydroxy acid fragment
Yongcheng Ying and Jiyong Hong*
Department of Chemistry, Duke University, Durham, NC 27708, USA
Received 21 August 2007; revised 6 September 2007; accepted 18 September 2007
Available online 21 September 2007
Abstract—A synthesis of brasilibactin A, a cytotoxic siderophore from the actinomycete of Nocardia brasiliensis, and three unnat-
ural diastereomers of the natural product is described. Four possible diastereomers of the b-hydroxy acid fragment were prepared
via asymmetric aldol reactions and used to synthesize brasilibactin A and its diastereomers. Careful analysis of 1H NMR data con-
firmed that brasilibactin A possesses the 17S,18R absolute stereochemistry.
Ó 2007 Elsevier Ltd. All rights reserved.
Brasilibactin A (1, Fig. 1) is a potent cytotoxic sidero-
phore isolated from the actinomycete of Nocardia brasil-
iensis IFM 0995 by Tsuda et al.1 The structure and
stereochemistry of brasilibactin A was determined by
extensive spectroscopic and chemical analysis.1 Brasili-
bactin A (1) possesses a nearly identical molecular
nucleus to many mycobactin-type siderophores, which
includes a hydroxamic acid, an N-hydroxyformamide,
and a 2-(2-hydroxyphenyl)-D2-1,3-oxazoline which all
serve as iron-chelating components.2 Mitchell et al.
recently reported the first total synthesis of 1 and revised
the originally assigned anti-configuration1 of the
b-hydroxy acid fragment in 1 as syn-configuration by
synthesizing two of the four possible diastereomers of
the b-hydroxy acid fragment.3 Brasilibactin A (1) was
reported to exhibit potent cytotoxicity against murine
leukemia L1210 and human epidermoid carcinoma KB
cells (IC50, 0.02 and 0.04 lg/mL, respectively) and to
cause a concentration-dependent increase in the cas-
pase-3 activity in HL60 cells.1 However, the molecular
mechanism of brasilibactin A (1) has yet to be estab-
lished. We undertook the total synthesis of brasilibactin
A (1) and its diastereomers to study the effect of config-
uration of the b-hydroxy acid fragment on biological
activity (e.g., Fe3+-chelation capability, cytotoxicity).
Herein, we report a synthesis of brasilibactin A (1)
and three unnatural diastereomers as well as an unam-
biguous confirmation of the absolute configuration of
the b-hydroxy acid fragment of the natural product.
Scheme 1 describes our approach to the synthesis of bra-
silibactin A and its diastereomers (1–4) via coupling of
four fragments (5–11). Disconnection of an ester and
two amide bonds of brasilibactin A yields four frag-
ments, an oxazoline 5, an N-hydroxyformamide 6, a
b-hydroxy acid (7–10), and a cyclic hydroxamic acid
11. The oxazoline 5 could be prepared from a direct con-
densation of a derivative of salicylic acid 12 and D-serine
benzyl ester (13). The N-hydroxyformamide 6 and the
cyclic hydroxamic acid 11 could be derived from
Na-Cbz-D-lysine (14) via methodologies previously
established by Miller et al. We anticipated that highly
stereoselective aldol reactions of hexadecanal (19) with
N-propanoyl-oxazolidinone and O-propanoyl-norephe-
drine could be used to prepare the four possible syn-
and anti-diastereomers of the b-hydroxy acid (7–10).
Figure 1. Structure of brasilibactin A.
Keywords: Brasilibactin A; Siderophore; Absolute configuration;
b-Hydroxy acid.
Corresponding author. Tel.: +1 919 660 1545; fax: +1 919 660
As outlined in Scheme 2, oxazoline 5 was synthesized
from commercially available methyl salicylate 20.2a Pro-
tection of 20 with BnBr4 and subsequent hydrolysis
*
1605; e-mail: jiyong.hong@duke.edu
0040-4039/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.09.112

Y. Ying, J. Hong / Tetrahedron Letters 48 (2007) 8104–8107
8105
Scheme 1. Retrosynthetic analysis of brasilibactin A.
Scheme 3. Synthesis of N-hydroxyformamide 6 and cyclic hydroxamic
acid 11. Reagents and conditions: (a) SOCl2, MeOH, rt, 24 h; then
dimethyldioxirane, acetone, À78 °C, 15 min, 58% for two steps; (b) (i)
NH2OHÆHCl, MeOH, 40 °C, 12 min; (ii) HCO2H, EDC, CH2Cl2, 0 °C,
2 h; then i-Pr2NEt, MeOH, rt, 48 h, 70% for two steps; (iii) SEMCl,
i-Pr2NEt, DMAP, CH2Cl2, rt, 24 h, 80%; (c) aq. LiOH, THF, 0 °C,
˚
30 min, rt, 1 h, 90%; (d) PhCHO, KOH, MeOH, 3 A MS, rt, 24 h, then
m-CPBA, MeOH, 0 °C, 2 h; (e) TFA, CH2Cl2, rt, 1 h; then PhCHO, rt,
24 h, 50% for two steps; (f) (i) NH2OHÆHCl, MeOH, 60 °C, 20 min; (ii)
EDC, HOAt, NaHCO3, CH3CN/DMF (7:2), rt, 48 h; (iii) TBDPSCl,
imidazole, DMF, 35 °C, 24 h, 40% for three steps; (g) 10% Pd–C, H2,
MeOH, rt, 2 h, >90%.
Scheme 2. Synthesis of oxazoline 5. Reagents and conditions: (a) (i)
BnBr, K2CO3, KI, THF, rt, 24 h; (ii) KOH, MeOH, H2O, 64% for two
steps; (b) 13, EDC, Et3N, CH2Cl2, rt, 24 h, 64%; (c) Burgess reagent,
THF, reflux, 30 min, 70%; (d) 10% Pd–C, H2, MeOH, rt, 2 h, >90%.
the corresponding nitrone 26.7c,d Since hydrolysis of
26 resulted in the formation of a hydroxylamine, addi-
tion of benzaldehyde to the reaction mixture improved
the yield of the reaction (50% for two steps).7c Treat-
ment of 26 with NH2OHÆHCl, cyclization of the corre-
sponding hydroxylamine under standard conditions
(EDC, HOAt, NaHCO3), and protection of the cyclic
hydroxamic acid with TBDPSCl afforded 27.2a,7c Final
deprotection of the Cbz protecting group in 27
under conventional conditions (H2/Pd–C) provided 11
(>90%).
(KOH, MeOH)5 provided 12 (64% for two steps). A
coupling of 12 to D-serine benzyl ester (13)6 followed
by treatment of 21 with Burgess reagent provided 22.2a
Final deprotection of the Bn protecting groups in 22 un-
der conventional conditions (H2/Pd–C) completed the
synthesis of 5 (>90%).
The N-hydroxyformamide 6 and the cyclic hydroxamic
acid 11 were prepared following the procedures estab-
lished by Miller et al.7 As shown in Scheme 3, Na-
Cbz-D-lysine (14) was converted to the corresponding
methyl ester followed by oxidation with dimethyldioxi-
rane in acetone to give nitrone 23.7a Treatment of 23
with NH2OHÆHCl, coupling with formic acid, protec-
tion with SEMCl, and hydrolysis under basic conditions
completed the synthesis of N-hydroxyformamide 6.7b
Na-Cbz-D-lysine (14) was treated with benzaldehyde
under basic conditions followed by oxidation with
m-CPBA and TFA-promoted isomerization to provide
Synthesis of four possible diastereomers of the
b-hydroxy acid fragment (7–10) was achieved by
employing the highly stereoselective syn-aldol reactions
of the N-propanoyl-oxazolidinones (15 and 16)8 and
anti-aldol reactions of the O-propanoyl-norephedrines
(17 and 18)9 (Scheme 4). The aldol reactions of hexade-
canal (19)10 with 15 and 16 in the presence of n-Bu2-
BOTf and i-Pr2NEt provided the desired syn-aldol
adducts 28 (85%) and 29 (85%), respectively, as a single

8106
Y. Ying, J. Hong / Tetrahedron Letters 48 (2007) 8104–8107
Scheme 4. Synthesis of b-hydroxy acid (7–10). Reagents and condi-
tions: (a) (i) (n-Bu)2BOTf, i-Pr2NEt, CH2Cl2, 0 °C, 30 min; (ii) 19,
CH2Cl2, À78 °C, 2 h, 85%; (b) 30% H2O2, LiOH, THF/H2O (4:1),
0 °C, 1 h, 90%; (c) (i) (c-hex)2BOTf, Et3N, CH2Cl2, 0 °C, 30 min; (ii)
19, CH2Cl2, À78 °C, 2 h, 80%; (d) LiOH, THF/MeOH/H2O (2:3:2), rt,
4 days, 85%.
Scheme 5. Synthesis of 1–4. Reagents and conditions: (a) EDC,
CH2Cl2, rt, 24 h, 64–68%; (b) 6, DCC, DMAP, toluene, rt, 48 h, 76–
89%; (c) 10% Pd–C, H2, MeOH, rt, 2 h; then 5, EDC, CH2Cl2, rt, 24 h,
48–57%; (d) TFA, CH2Cl2, rt, 1.5 h, 68–73%.
diastereomer. Hydrolysis of the b-hydroxy amides (28
and 29) under conventional conditions (30% H2O2,
LiOH) provided the corresponding b-hydroxy acids (7
and 8) in 90% yield, respectively. In a similar manner,
the aldol reactions of hexadecanal (19) with the chiral
derivatives of norephedrine (17 and 18) in the presence
of c-Hex2BOTf and Et3N followed by the hydrolysis
of the corresponding esters (30 and 31) were utilized
for the preparation of the anti-b-hydroxy acids (9 and
10).
TBDPS protecting groups in 40 by treatment with
TFA afforded 1 in 68% yield.3,7b The diastereomers of
brasilibactin A (2–4) were also prepared in the same
manner. 1H NMR data for 1–4 were carefully compared
with the authentic material (Table 1). The chemical
shifts and coupling patterns of 1, in particular those of
H-14 (d 4.25, m), H-17 (d 4.90, dt, J = 2.8, 8.8 Hz), H-
18 (d 2.62, m), and H-20 (d 8.13, d, J = 7.2 Hz), were
identical to those of the natural product.1,3 The compar-
The synthesis of brasilibactin A (1) and its diastereomers
(2–4) began with coupling 7–10 and 11 as shown in
Scheme 5. EDC-coupling2a of 7 to 11 followed by subse-
quent coupling3,7b of 32 to 6 provided 36. Alternative
methods for the formation of the ester bond (DEAD
or DIAD, PPh3; 2,4,6-trichlorobenzoyl chloride, NEt3,
then DMAP; EDC, DMAP) failed to react with this
sterically hindered substrate. Deprotection of the Cbz
protecting group in 36 and coupling of the correspond-
ing amine to 5 completed the synthesis of the protected
depsipeptide (40).3 Final deprotection of the SEM and
ison unambiguously showed that brasilibactin
A
possesses the 17S,18R absolute stereochemistry, which
is consistent with the assignment by Mitchell et al.3
In summary, we completed a synthesis of brasilibactin A
(1), a structurally and biologically interesting linear
depsipeptide, and its unnatural diastereomers (2–4).
The convergent synthetic strategy should be broadly
applicable to the synthesis of a diverse set of analogs
Table 1. Comparison of 1H NMR data for 1–4 with natural brasilibactin Aa
H no.
Natural 11
1 (17S,18R)
2 (17R,18S)
3 (17S,18S)
4 (17R,18R)
(17S,18R) in Ref. 3
(17R,18S) in Ref. 3
18
14
21
9
17
20
2.62
4.25
4.44
4.47
4.90
8.11
2.62
4.25
4.44
4.47
4.90
8.13
2.66
4.21
4.46
4.48
4.95
8.15
2.70
4.23
4.40
4.48
4.93
7.88
2.74
4.15
4.42
4.51
4.95
7.93
2.61
4.24
4.42
4.47
4.90
8.15
2.66
4.22
4.47
4.46
4.95
8.16
a Chemical shifts (ppm) in DMSO-d6.

Y. Ying, J. Hong / Tetrahedron Letters 48 (2007) 8104–8107
8107
2. (a) Hu, J.; Miller, M. J. J. Am. Chem. Soc. 1997, 119,
3462–3468; (b) Vergne, A. F.; Walz, A. J.; Miller, M. J.
Nat. Prod. Rep. 2000, 17, 99–116, and references cited
therein.
of 1. Further studies to assess their biological activity
and identify molecular targets of 1 are in progress.
3. Mitchell, J. M.; Shaw, J. T. Org. Lett. 2007, 9, 1679–
1681.
Acknowledgment
4. Lin, C.-F.; Yang, J.-S.; Chang, C.-Y.; Kuo, S.-C.; Lee,
M.-R.; Huang, L.-J. Bioorg. Med. Chem. 2005, 13, 1537–
1544.
This work was supported by Duke University.
5. Bremner, J. B.; Samosorn, S.; Ambrus, J. I. Synthesis
2004, 16, 2653–2658.
Supplementary data
6. Holden, K. G.; Mattson, M. N.; Cha, K. H.; Rapoport,
H. J. Org. Chem. 2002, 67, 5913–5918.
1
Copies of H NMR data for compounds 1–4, 7–10, 22,
24, and 27–43 and comparison of 1H NMR data for 1–4.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.tetlet.
2007.09.112.
7. (a) Hu, J.; Miller, M. J. J. Org. Chem. 1994, 59, 4858–
4861; (b) Yokokawa, F.; Izumi, K.; Omata, J.; Shioiri, T.
Tetrahedron 2000, 56, 3027–3034; (c) Walz, A. J.; Miller,
M. J. Org. Lett. 2002, 4, 2047–2050; (d) Dong, L.; Miller,
M. J. J. Org. Chem. 2002, 67, 4759–4770.
8. Evans, D. A.; Bartroli, J.; Shih, T. L. J. Am. Chem. Soc.
1981, 103, 2127–2129.
9. Abiko, A.; Liu, J.-F.; Masamune, S. J. Am. Chem. Soc.
1997, 119, 2586–2587.
10. Haldar, J.; Kondaiah, P.; Bhattacharya, S. J. Med. Chem.
2005, 48, 3823–3831.
References and notes
1. Tsuda, M.; Yamakawa, M.; Oka, S.; Tanaka, Y.; Hosh-
ino, Y.; Mikami, Y.; Sato, A.; Fujiwara, H.; Ohizumi, Y.;
Kobayashi, J. J. Nat. Prod. 2005, 68, 462–464.