Structure determination and total synthesis of bottromycin A2: A potent antibiotic against MRSA and VRE

Article abstract of DOI:10.1002/anie.200804138

(Chemical Equation Presented) After 50 years of persistence the 12-membered cyclic skeleton of bottromycin A₂ (3) has been confirmed by NMR experiments (HMBC), and the configurations of two tLeu residues have been estimated by conformation anal

Full text of DOI:10.1002/anie.200804138

Communications
DOI: 10.1002/anie.200804138
Natural Products
Structure Determination and Total Synthesis of Bottromycin A2: A
Potent Antibiotic against MRSA and VRE**
Hiroyuki Shimamura, Hiroaki Gouda, Kenichiro Nagai, Tomoyasu Hirose, Maki Ichioka,
¯
Yujiro Furuya, Yutaka Kobayashi, Shuichi Hirono, Toshiaki Sunazuka,* and Satoshi Omura*
The emergence of antibiotic resistance in bacteria has been a
major and increasing threat to public health worldwide. There
is an urgent need for novel antibiotics to help solve the
problem. In the course of our search for natural products with
activity against drug-resistant bacteria, we rediscovered an
antibiotic, bottromycin A2 (1; Figure 1), which was first
mycoplasma, we found its antibacterial ability extends to
methicillin-resistant Streptococcus aureus (MRSA, MIC =
1.0 mgmLÀ1) and vancomycin-resistant Enterococci (VRE,
MIC < 1.0 mgmLÀ1) strains.[2] As for the mode of action, 1
inhibits bacterial protein synthesis by blocking the binding of
aminoacyl-tRNA to the A site on the 50S ribosome, but does
not inhibit the peptide bond formation and translocation
steps.[3] The mode of action of 1 is different from that of other
known antibiotics. Unsurprisingly, no cross-resistance of 1 to
existing drugs has been observed. Therefore, 1 may become a
promising antibiotic for treatment of bacterial infections, thus
prompting us to undertake research and development of new
bottromycin-based antibiotics.
The gross structure of bottromyin A2 has been established
through repeated revisions since its first isolation.[4–6] The
most significant structural characteristic is an amidine moiety
formed through condensation of the cyclic tetrapeptide and
linear tripeptide units.[6] However, there is no clear exper-
imental data that directly demonstrates a 12-membered cyclic
skeleton in the structure of 1, its cyclic structure was
established solely based on CIMS analysis.[5] In addition,
there is a lack of clarity concerning the configuration of tert-
leucine (tLeu 1: C18; tLeu 2: C25) and 3-(thiazol-2-yl)-b-
alanine (Thia-b-Ala-OMe: C43) residues. After degredation,
the tLeu unit(s) derived from 1 was thought to have the
l form because it was not affected by d-amino acid oxidase
and showed a slight positive Cotton effect in ORD (optical
rotation dispersion) curves.[7] However, the optical rotations
of the natural and synthetic tLeu[8] are opposite, which
suggests that the natural tLeu is more likely to have the
d form (Figure 1). In the case of Thia-b-Ala, its optical
rotation was less than expected, even though natural Thia-b-
Ala was proved to have the l form based on the experimental
rule in ORD of b-amino acids.[9] Herein, we report the
structure elucidation and total synthesis of 1 and therefore
resolve the existing structural ambiguities.
We first undertook studies to confirm the structure of
bottromycin A2 (1) by using high-resolution NMR spectros-
copy (600 MHz). Our independent assignments of the proton
and carbon signals of 1 were found to be compatible with
those report by Kaneda.[10] We performed an HMBC experi-
ment to establish the 12-membered cyclic skeleton in 1 (see
the Supporting Information). By using a long-range delay of
125 ms that corresponds to a 1/nJCH of 4 Hz, previously
unreported HMBC correlations between Gly and tLeu 1
(1NH and 23C, 2H and 23C) were clearly observed. This is the
first experimental evidence that firmly establishes the 12-
membered cyclic skeleton in 1.
Figure 1. Structure of bottromycin A2 and reported optical rotation of
Thia-b-Ala and tLeu.
isolated from the fermentation broth of Streptomyces bottro-
pensis by Waisvisz et al. in 1957.[1] Although 1 is known to
show antibacterial activity against Gram-positive bacteria and
[*] Dr. H. Shimamura, Prof. Dr. H. Gouda, Dr. K. Nagai, Dr. T. Hirose,
M. Ichioka, Y. Furuya, Y. Kobayashi, Prof. Dr. S. Hirono,
¯
Prof. Dr. T. Sunazuka, Prof. Dr. S. Omura
Kitasato Institute for Life Sciences and Graduate School of Infection
Control Sciences
School of Pharmacy, Kitasato University
5-9-1 Shirokane, Minato-ku, Tokyo, 108-8642 (Japan)
Fax: (+81)3-5791-6101
E-mail: sunazuka@lisci.kitasato-u.ac.jp
omuras@insti.kitasato-u.ac.jp
[**] This work was supported by a grant from the 21st Century COE
Program, Ministry of Education Culture, Sports, Science and
Technology (18790022). We also thank A. Nakagawa, C. Sakabe, N.
Sato, and Y. Kawauchi (School of Pharmacy, Kitasato University) for
analysis. MRSA=methicillin-resistant Streptococcus aureus, VRE=
vancomycin-resistant Enterococci.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200804138.
914
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 914 –917

Angewandte
Chemie
We next estimated the absolute configurations at C18 and
C25 by using a combination method of conformation analysis
with high-temperature molecular dynamics (MD) and NMR
spectroscopy (see the Supporting Information).[11] The con-
formation analysis was performed toward a simplified struc-
ture 2 differing in the configuration at C18 and C25
(Figure 2A) by using the CAMDAS (Conformational Ana-
Scheme 1. Reagents and conditions: a) MeOH, 1308C, in a sealed
tube, 3 h, 3; 89%, 4; 95%; b) NaBH4, MeOH, 08C, 30 min, 63%.
determine the configuration at C18, C25, and C43. We planed
to confirm the absolute structure by tracing the natural
degradation products 4 and 5 to our goal, the total synthesis of
1.
Determination of the configuration at C43 required
enantiomers of Thia-b-Ala that were susceptible to epimeri-
zation.[9] Our stereoselective synthesis of Thia-b-Ala-OMe 12
was performed by applying chiral sulfinamide chemistry
developed by the research groups of Davis and Elmann,[13] as
shown in Scheme 2. Condensation of (SS)-para-toluene sulfi-
namide 6 with 2-formylthiazole 7 afforded sulfinimine 8 in
90% yield.[14] The Mannich reaction of 8 with allyl methyl
malonate 9 quantitatively gave adduct 10 as a 1:1 mixture of
a-diastereomers, which was subjected to palladium-catalyzed
decarboxylation to give methyl ester 11 as a single diastereo-
mer in 76% yield.[15] Finally, removal of the sulfinyl group
afforded (+)-(43R)-12 in quantitative yield (> 99% ee). The
absolute configuration of (+)-(43R)-12 was later confirmed
by X-ray crystallographic analysis of (43R)-15.[17] The enan-
Figure 2. A) Model structure 2 shown as the 18S,25S configuration,
which was determined by the conformation analysis and NMR experi-
ments. B) Stereopairs of the superposition of the resulting two
3D structures of model structure 2. Black and grey structures have
MMFF energies of 104.5 and 115.1 kcalmolÀ1, respectively. These are
the results of the best fit of the heavy atoms. MMFF=Merck
Molecular Force Field
lyzer with Molecular Dynamics and Sampling) 2.1 program.[12]
The NMR experiments with 1 provide the total 32 requisite
constraints, which consist of 2 dihedral and 30 distance
constraints. The RMS deviations for all configurations
generated by CAMDS were subsequently calculated to
determine which configurations can adopt conformers accom-
modating the NMR constraints. The conformers that satisfied
all experimental constraints were found in only one config-
uration. The process described above led to a set of two
energy-minimized 3D structures of model structure 2, which
indicates the 18S,25S configurations (Figure 2B). We there-
fore established (18S,25S)-1 as the target of the following total
synthesis.
Prior to the start of our total synthesis, we performed a
degradation study of 1 to obtain additional information about
the configuration at C18, C25, and C43 (Scheme 1). Pyrolysis
of 1 in MeOH at 1308C was carried out in a sealed tube to
provide cyclic product 3 and dipeptide 4 in 89% and 95%
yield, respectively. In addition, treatment of 1 with excess
NaBH4 resulted in reductive cleavage of an amide bond
between the Gly and MePro units to give amino alcohol 5.
The cyclic product 3 was obtained as an inseparable mixture
of diastereomers at C25 owing to keto/enol tautomerization
at the imidazolone moiety. This tautomerization suggested
that the reported hydrolysis of 1 might cause racemization of
tLeu via the imidazolone path a). We therefore anticipated
that 4 and 5 would be useful synthetic targets to help
Scheme 2. Reagents and conditions: a) 7, Cs2CO3, CH2Cl2, 408C, 1 h,
90%; b) 9, NaHMDS, THF, À788C, 30 min, quant.; c) [Pd(PPh3)4],
HCO2H, Et3N, THF, reflux, 2 h, 76%; d) TFA, MeOH, 08C, 3 h, quant.
HMDS=hexamethyldisilazane, TFA=trifluoroacetic acid.
Angew. Chem. Int. Ed. 2009, 48, 914 –917
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
915

Communications
tiomer (À)-(43S)-12 was synthesized in the same manner
1H NMR spectroscopic data between synthetic and natural
samples to assign the configuration at C43 (Figure 3). The
1H NMR spectrum of the natural degradation product
matched that of (43R)-4, whereas the chemical shift of the
methylene group at C49 in (43S)-4 was substantially different.
using (RS)-tert-butyl sulfinamide 13.
Each enantiomer of 12 was condensed with azido acid
14[16] using PyBOP and gave azido dipeptides 15, respectively
(Scheme 3). At this stage, the absolute configuration at the
C43 position of (43R)-15 was confirmed as the R configura-
tion by X-ray crystallographic analysis (see the Supporting
Information).[17] Finally, reduction of the azido group
afforded a set of diastereomers 4. We then compared
Scheme 3. Reagents and conditions: a) 14, PyBOP, iPr2NEt, CH2Cl2,
RT, 1 h, (43R)-15, quant.; (43S)-15, quant.; b) H2, 10% Pd/C, MeOH,
RT, 1 h, (43R)-4; 97%; (43S)-4; 93%. PyBOP=benzotriazol-1-yloxytri-
pyrrolidinophosphonium hexafluorophosphate.
Figure 3. Comparison of 1H NMR spectra of diastereomers at C43 for
4.
Scheme 4. Reagents and conditions: a) 16, HBTU, iPr2NEt, CH2Cl2/DMF, RT, 2 h, quant.; b) 4n HCl/dioxane, RT, 30 min, quant.; c) 18, Hg(OTf)2,
2,6-lutidine, CH3CN, RT, 30 min, 96%; d) N2H4·H2O, EtOH, reflux, 2 h; e) 20, HBTU, iPr2NEt, CH2Cl2/DMF, RT, 2 h, 89% (over 2 steps); f) TFA/
CH2Cl2, RT, 3 h; g) 22, HBTU, iPr2NEt, CH2Cl2/DMF, RT, 2 h, quant. (over 2 steps); h) TBAF, THF, RT, 3 h; i) 4n HCl/dioxane, RT, 30 min, 68%
(over 2 steps from 23); j) Jones reagent (15 equiv), acetone, 08C, 4 h, 55% (over 2 steps from 23); k) 4n HCl/dioxane, RT, 30 min; l) EDCI·HCl
(10 equiv), iPr2NEt (4 equiv), CH2Cl2, RT, 4 h, 51% (over 2 steps). Boc=tert-butoxycarbonyl, DMF=N,N-dimethylformamide, EDCI=N-(3-
dimethylaminopropyl)-N’-ethylcarbodiimide, HBTU=O-benzotriazol-1-yl-N,N,N’,N’-tetramethyluronium hexafluorophosphate, Pht=phthaloyl,
TBAF=tetra-n-butylammonium fluoride, TBDPS=tert-butyldiphenylsilyl, Tf=trifluoromethanesulfonyl.
916
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 914 –917

Angewandte
Chemie
Thus 1H NMR analysis indicates that the configuration at C43
of 1 is the R configuration.
b) J. M. Waisvisz, M. G. van der Hoeven, J. F. Holscher, B.
te Nijenhuis, J. Am. Chem. Soc. 1957, 79, 4522 – 4524; c) J. M.
Waisvisz, M. G. van der Hoeven, B. te Nijenhuis, J. Am. Chem.
Soc. 1957, 79, 4524 – 4527; d) J. M. Waisvisz, M. G. van derHo-
even, J. Am. Chem. Soc. 1958, 80, 383 – 385.
With the dipeptide (43R)-4 in hand, we then undertook
the synthesis of amino alcohol 5 to confirm the remaining
configurations at C18 and C25, which had been estimated by
the conformation analysis (Scheme 4). The dipeptide (43R)-4
was condensed with Boc-tLeu-OH 16, and subsequent
removal of the Boc group provided tripeptide 17. In our
total synthesis intermolecular amidine formation was a key
step, and we examined mercury-mediated condensation
between an amine and a thioamide. We found that N-
phthaloyl thioamide 18 could be smoothly coupled with 17 by
the combination of Hg(OTf)2 and 2,6-lutidine to furnish
amidine 19 in 96% yield. Removal of the phthaloyl group and
subsequent condensation with Boc-Val-OH 20 afforded 21 in
89% yield. After removal of the Boc group of 21, the resulting
amine was coupled with Boc-MePro-OH 22[18] to give
compound 23 in quantitative yield. Finally, removal of the
TBDPS and Boc groups afforded (18S,25S,43R)-5 in 68%
[2] a) N. Tanaka, T. Nishimura, S. Nakamura, H. Umezawa, J.
Antibiot. 1968, 21, 75 – 76; b) for antibacterial activity against
MRSA and VRE, see the Supporting Information.
[3] a) T. Otaka, A. Kaji, J. Biol. Chem. 1976, 251, 2299 – 2306; b) T.
Otaka, A. Kaji, FEBS Lett. 1981, 123, 173 – 176; c) T. Otaka, A.
Kaji, FEBS Lett. 1983, 153, 53 – 59.
[4] a) S. Nakamura, T. Chikaike, K. Karasawa, N. Tanaka, H.
Yonehara, H. Umezawa, J. Antibiot. Ser. A 1965, 18, 47 – 52; b) S.
Nakamura, T. Chikaike, H. Yonehara, H. Umezawa, J. Antibiot.
Ser. A 1965, 18, 60 – 61; c) S. Nakamura, T. Chikaike, H.
Yonehara, H. Umezawa, Chem. Pharm. Bull. 1965, 13, 599 –
602; d) S. Nakamura, N. Tanaka, H. Umezawa, J. Antibiot. Ser.
A 1966, 19, 10 – 12; e) S. Nakamura, H. Umezawa, Chem. Pharm.
Bull. 1966, 14, 981 – 986; f) S. Nakamura, T. Yajima, Y-C. Lin, H.
Umezawa, J. Antibiot. Ser. A 1967, 20, 1 – 5.
[5] Y. Takahashi, H. Naganawa, T. Takita, H. Umezawa, S.
Nakamura, J. Antibiot. 1976, 29, 1120 – 1123.
[6] D. Schipper, J. Antibiot. 1983, 36, 1076 – 1077.
1
yield. Optical rotation, H NMR spectroscopic, and HRMS
[7] S. Nakamura, T. Chikaike, H. Yonehara, H. Umezawa, Chem.
Pharm. Bull. 1965, 13, 599 – 602.
[8] T. Miyazawa, K. Takashima, Y. Mitsuda, T. Yamada, S. Kuwata,
H. Watanabe, Bull. Chem. Soc. Jpn. 1979, 52, 1539 – 1540.
[9] Y. Seto, K. Torii, K. Bori, K. Inabata, S. Kuwata, H. Watanabe,
Bull. Chem. Soc. Jpn. 1974, 47, 151 – 155.
data of the synthetic compound matched those of the
degradation product derived from the natural product.
From this result, we concluded that the complete structural
assignment of 1 comprises 18S, 25S, and 43R configurations.
Next, we turned to completion of the total synthesis.
Removal of the TBDPS group and subsequent Jones oxida-
tion afforded carboxylic acid 24 in 55% yield from 23. After
removal of the Boc group, and macrolactamization of 25 with
an excess amount of EDCI (10 equiv) and iPr2NEt (4.0 equiv)
in CH2Cl2 (2.0 mm) provided 1 in 51% yield. The spectro-
[10] M. Kaneda, J. Antibiot. 1992, 45, 792 – 796.
[11] This method has been successfully applied to determine the
absolute configuration of chloropeptin I and luminamicin;
a) chloropeptin I: H. Gouda, K. Matsuzaki, H. Tanaka, S.
Hirono, J. Am. Chem. Soc. 1996, 118, 13087 – 13088; b) lumina-
micin: H. Gouda, T. Sunazuka, H. Ui, M. Handa, Y. Sakoh, Y.
Iwai, S. Hirono, S. Omura, Proc. Natl. Acad. Sci. USA 2005, 102,
18286 – 18291.
1
scopic data ([a]D, H and 13C NMR, IR, and HRMS) of the
synthetic material were identical to those of the natural
product. Furthermore, the synthetic sample of 1 showed
antibacterial activity against MRSA and VRE.
[12] H. Tsujishita, S. Hirono, J. Comput.-Aided Mol. Des. 1997, 11,
305 – 315.
[13] a) F. A. Davis, R. E. Reddy, J. M. Szewczyk, J. Org. Chem. 1995,
60, 7037 – 7039; b) T. P. Tang, J. A. Ellman, J. Org. Chem. 1999,
64, 1278 – 1284.
[14] M. Nakata, S. Higashibayashi, H. Tohmiya, T. Mori, K.
Hashimoto, Synlett 2004, 457 – 460.
In conclusion, we have achieved the first total synthesis of
bottromycin A2 and confirmed its absolute structure. This
accomplishment will end any uncertainty over the structure of
the compound and offers significant promise for development
of new and urgently required antibiotics.
[15] T. Mandai, M. Imaji, H. Takada, M. Kawata, J. Nokami, J. Tsuji,
J. Org. Chem. 1989, 54, 5395 – 5397.
Received: August 22, 2008
Published online: December 29, 2008
[16] R. Dharanipragada, E. Nicolas, G. Tothm, V. J. Hruby, Tetrahe-
dron Lett. 1989, 30, 6841 – 6844.
[17] CCDC 706775 (15) contain the supplementary crystallographic
data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif..
[18] C. Flamant-Robin, Q. Wang, A. Chiaroni, N. A. Sasaki, Tetrahe-
dron 2002, 58, 10475 – 10484.
Keywords: antibiotics · drug resistance · natural products ·
structure elucidation · total synthesis
.
[1] a) J. M. Waisvisz, M. G. van der Hoeven, J. van Peppen,
W. C. M. Zwennis, J. Am. Chem. Soc. 1957, 79, 4520 – 4521;
Angew. Chem. Int. Ed. 2009, 48, 914 –917
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
917