Amycolamicin: A novel broad-spectrum antibiotic inhibiting bacterial topoisomerase

Article abstract of DOI:10.1002/chem.201202645

The abuse of antibacterial drugs imposes a selection pressure on bacteria that has driven the evolution of multidrug resistance in many pathogens. Our efforts to discover novel classes of antibiotics to combat these pathogens resulted in the discovery of amycolamicin (AMM). The absolute structure of AMM was determined by NMR spectroscopy, X-ray analysis, chemical degradation, and modification of its functional groups. AMM consists of trans-decalin, tetramic acid, two unusual sugars (amycolose and amykitanose), and dichloropyrrole carboxylic acid. The pyranose ring named as amykitanose undergoes anomerization in methanol. AMM is a potent and broad-spectrum antibiotic against Gram-positive pathogenic bacteria by inhibiting DNA gyrase and bacterial topoisomerase IV. The target of AMM has been proved to be the DNA gyrase B subunit and its binding mode to DNA gyrase is different from those of novobiocin and coumermycin, the known DNA gyrase inhibitors. Equilibrium structure: Amycolamicin (AMM) was discovered as a new antibiotic and its absolute structure was determined. The pyranose ring named as amykitanose undergoes anomerization in methanol. AMM is effective against methicillin-resistant Staphylococcus aureus and its strains, which are resistant to known DNA gyrase inhibitors. AMM is a potent and selective inhibitor of DNA gyrase and it does not inhibit human topoisomerase II (see figure). Copyright

Full text of DOI:10.1002/chem.201202645

DOI: 10.1002/chem.201202645
Amycolamicin: A Novel Broad-Spectrum Antibiotic Inhibiting Bacterial
Topoisomerase
Ryuichi Sawa,[a] Yoshiaki Takahashi,[b] Hideki Hashizume,[a] Kazushige Sasaki,[b]
Yoshimasa Ishizaki,[a] Maya Umekita,[a] Masaki Hatano,[a] Hikaru Abe,[a]
Takumi Watanabe,[a] Naoko Kinoshita,[a] Yoshiko Homma,[a] Chigusa Hayashi,[a]
Kunio Inoue,[a] Syunichi Ohba,[c] Toru Masuda,[c] Masayuki Arakawa,[a]
Yoshihiko Kobayashi,[b] Masa Hamada,[a] Masayuki Igarashi,[a] Hayamitsu Adachi,*[a]
Yoshio Nishimura,[a] and Yuzuru Akamatsu[a]
Abstract: The abuse of antibacterial
drugs imposes a selection pressure on
bacteria that has driven the evolution
of multidrug resistance in many patho-
gens. Our efforts to discover novel
classes of antibiotics to combat these
pathogens resulted in the discovery of
amycolamicin (AMM). The absolute
structure of AMM was determined by
NMR spectroscopy, X-ray analysis,
chemical degradation, and modification
of its functional groups. AMM consists
of trans-decalin, tetramic acid, two un-
usual sugars (amycolose and amykita-
nose), and dichloropyrrole carboxylic
acid. The pyranose ring named as amy-
kitanose undergoes anomerization in
methanol. AMM is
broad-spectrum antibiotic
a
potent and
against
Gram-positive pathogenic bacteria by
inhibiting DNA gyrase and bacterial
topoisomerase IV. The target of AMM
has been proved to be the DNA gyrase
B subunit and its binding mode to
DNA gyrase is different from those of
novobiocin and coumermycin, the
known DNA gyrase inhibitors.
Keywords: antibiotics · DNA repli-
cation · natural products · structure
elucidation
Introduction
faecium (VRE) have been a widespread problem in nosoco-
mial infections.[2,3] Community-acquired pneumonia is one
of the most serious infections caused by Streptococcus pneu-
moniae and Haemophilus influenzae.[4] Drug-resistant strains
are also increasingly isolated from patients and are known
as penicillin-resistant S. pneumoniae (PRSP),[5–7] macrolide-
resistant S. pneumoniae (MRSP), and beta-lactamase-nega-
tive ampicillin-resistant (BLNAR) and beta-lactamase-posi-
tive amoxicillin-clavulanate-resistant (BLPACR) strains of
H. influenzae.[8]
In order to combat the growing number of drug-resistant
pathogens, synthetic tailoring of natural antibiotics has been
widely carried out to create a next generation of antibiotics.
New generation antibiotics are often designed to inhibit the
activity of pathogens that have gained resistance to the pre-
vious generation ones. For example, second- and third-gen-
eration quinolones like ciprofloxacin and levofloxacin are
far more active against the resistant strains of pathogen-mu-
tated DNA gyrase and topoisomerase IV by appropriate
functionalization of the quinolone scaffold.[9,10] Improve-
ments of existing scaffolds are a good short-term strategy
for refilling the antibiotic pipeline.
The emergence and widespread frequency of multiple drug-
resistant bacteria have made the treatment of bacterial in-
fectious diseases increasingly difficult. In nosocomial infec-
tions, methicillin-resistant Staphylococcus aureus (MRSA) is
one of the most problematic pathogens where vancomycin
(VAN) is used as the last resort. However, overuse of VAN
has enhanced the widespread development of vancomycin-
intermediate-resistant S. aureus.[1] Moreover, VAN-resistant
S. aureus (VRSA) and VAN-resistant Enterococcus faecalis/
[a] Dr. R. Sawa, Dr. H. Hashizume, Dr. Y. Ishizaki, M. Umekita,
Dr. M. Hatano, H. Abe, Dr. T. Watanabe, N. Kinoshita, Y. Homma,
C. Hayashi, K. Inoue, Dr. M. Arakawa, Dr. M. Hamada,
Dr. M. Igarashi, Dr. H. Adachi, Dr. Y. Nishimura, Dr. Y. Akamatsu
Institute of Microbial Chemistry (BIKAKEN), Tokyo
3-14-23, Kamiosaki, Shinagawa-ku
Tokyo 141-0021 (Japan)
Fax : (+81)3-3441-7589
E-mail: adachih@bikaken.or.jp
[b] Dr. Y. Takahashi, Dr. K. Sasaki, Dr. Y. Kobayashi
Institute of Microbial Chemistry (BIKAKEN), Hiyoshi
3-34-17, Ida, Nakahara-ku, Kawasaki-shi
Kanagawa 211-0035 (Japan)
On the other hand, discovery of new antibiotic scaffolds
from natural products would be a more sustainable way to
overcome resistant strains, leading to unprecedented struc-
ture diversity and a greater chance for biological activity.
Unfortunately, development and approval of new antibiotics
from natural products has not caught up with the emergence
[c] S. Ohba, T. Masuda
Institute of Microbial Chemistry (BIKAKEN), Numazu
18-24, Miyamoto, Numazu-shi, Shizuoka 410-0301 (Japan)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201202645.
15772
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 15772 – 15781

FULL PAPER
of drug-resistant pathogens. As well, the discovery of new
scaffolds has been decreasing year by year due to an in-
crease in known scaffold rediscovery.
In order to effectively eliminate the rediscovery of known
scaffolds during antibiotic screening of microbial metabo-
lites, we used many types of clinical isolates, passaged-resist-
ant strains, and genetically modified strains as test organisms
to observe the cross-resistance. In the course of our screen-
ing, we discovered a new broad-spectrum antibiotic named
amycolamicin (AMM, Figure 1) from the culture broth of
liquid culture for four days and 80 L of the fermentation
broth were separated into the mycelial cake (2.5 kg) and the
supernatant. AMM was purified by successive steps of ex-
traction from the mycelia followed by gel filtration and
silica gel chromatography. Active fractions were collected
and concentrated in vacuo to give a pale orange solid
(827 mg). A small amount of the solid (54 mg) was purified
by reversed-phase HPLC developed with a linear gradient
of 50% aqueous MeOH to 100% MeOH for 75 min. The
fractions containing pure AMM analyzed by HPLC were
collected to give a white solid. Finally the solid was suspend-
ed in ethyl acetate and washed with phosphate buffer
(pH 7.0) and aqueous HCl solution to give a free form of
AMM (34 mg).
Structural features and absolute structure of amycolamicin:
AMM is soluble in CHCl3, MeOH, DMSO, and acetone, but
insoluble in H2O. The specific rotation of AMM fluctuated
depending on the isolation conditions. AMM isolated by
HPLC showed
a
specific rotation of [a]2D0 =À20.68 in
MeOH, whereas it exhibited [a]2D3 =À36.58 after washing an
ethyl acetate solution of AMM with phosphate buffer and
aqueous HCl solution. The UV spectrum of AMM has a
characteristic absorption maxima at l=280 nm (loge=4.45)
in acidic MeOH and has a hypsochromic shift to l=248
(loge=4.30) and 277 nm (loge=4.45) in alkaline MeOH.
The molecular formula of AMM was found to be
C44H60Cl2N4O14 by HRESI-MS. The 1H NMR spectrum of
AMM showed major and minor species in a ratio of 2:1 in
CDCl3 (Figure S2 in the Supporting Information). The spec-
tra in [D4]MeOH gave simplified NMR signals originating
from a single species, but not all signals of the 1H and
13C NMR spectra were observed (Figures S11 and S12 in the
Supporting Information). Furthermore, new species ap-
peared gradually over time as shown in Figure S14 in the
Supporting Information.
Figure 1. Structure of AMM.
the soil actinomycete, Amycolatopsis sp. MK575-fF5 and pa-
tented AMM in 2009. The chemical structure of AMM was
determined by NMR spectroscopy, chemical degradation, X-
ray analysis, and modification of its functional groups.
AMM has a unique structure consisting of trans-decalin, tet-
ramic acid, two unusual sugars (amycolose and amykita-
nose), and 3, 4-dichloro-1H-pyrrole-2-carboxylic acid. AMM
shows potent antibacterial activity against Gram-positive
bacteria including PRSP, MRSA, and VRE and some
Gram-negative bacteria, including BLNAR and BLPACR
strains of H. influenzae.
Decatromicin B,[11] pyrrolosporin A,[12] and kibdelomy-
cin,[13] which are related to AMM, all show potent antibacte-
rial activity against pathogenic Gram-positive bacteria.
These antibiotics possess a common pyrrolamide-sugar-deca-
lin ring scaffold. In particular, the pyrrolamides have recent-
ly been identified as a new class of antibacterial agents that
inhibit the activity of DNA gyrase, resulting in inhibition of
DNA synthesis leading to bacterial cell death.[14]
The characteristic structure and promising antibacterial
activities of AMM have led us to clarify the absolute struc-
ture and the mode of action of AMM in order to encourage
further drug development and design of new AMM deriva-
tives. Herein, we describe the isolation, absolute structure,
structural isomerization, biological activities, and mode of
action of AMM.
Various NMR spectra in CDCl3 were extensively analyzed
for structural determination purposes because they had
sharper signals and some exchangeable protons were availa-
ble for structural analysis even in a mixture of major and
minor forms originating from the tautomerization of the tet-
1
ramic acid moiety. The H and 13C NMR data of the major
form are summarized in Table 1. The 13C NMR and DEPT
spectra revealed that AMM has 44 carbon atoms, comprised
of twelve quaternary sp2 carbon atoms, including carbonyl
atoms, one quaternary sp3 carbon atom, two sp2 methine
carbon atoms, sixteen sp3 methine carbon atoms, one sp2
methylene carbon atom, three sp3 methylene carbon atoms,
and nine methyl carbon atoms.
The structure of AMM was elucidated mainly by using
two-dimensional NMR spectroscopy. The proton signals
without C–H correlations in the 1H–13C HMQC spectrum
were assigned to two NH groups (d=6.64 (NH-9’) and
9.73 ppm (NH-16’)) and one NH2 group (d=4.79 ppm
Results
1
(NH2-10’’)), confirmed by H–15N HMQC spectroscopy.
Fermentation and isolation of amycolamicin: The producing
organism, Amycolatopsis sp. MK575-fF5, was grown in
The 1H–1H COSY spectrum showed four partial struc-
tures, amycolose (I),[15] decalin (II), the isopropyl group of
Chem. Eur. J. 2012, 18, 15772 – 15781
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15773

H. Adachi et al.
Table 1. 1H and 13C NMR data of AMM measured in CDCl3.[a]
Position
dC [ppm]
dH [ppm]
Multiplet, J [Hz]
Position
dC [ppm]
dH [ppm]
Multiplet, J [Hz]
À
À
À
1
2
3
4
4a
5
6
7
8
149.7
34.3
33.9
77.7
48.0
125.9
130.9
32.2
42.4
37.3
191.4
102.5
175.6
71.2
193.5
30.5
15.8
17.9
105.2
17.9
95.4
C=
5’
6’
7’
8’
70.5
18.1
52.6
16.2
CH
3.59
1.32
4.41
1.30
6.64
dq 6.2, 9.1
d 6.1
quintet 7.0
d 7.0
À
CH2
À
À
À
À
2.12, 2.32
1.32, 2.24
3.59
1.90
6.01
5.63
2.60
4.10
2.34
m, m
m, m
CH3
À
À
CH2
À
CH
CH
À
À
CH3
dt 4.7, 10.3
td 10.3, 1.5
dt 10.0, 1.5
ddd 2.5, 4.2, 10.0
m
À
À
CH
NH-9’
10’
11’
12’
13’
14’
15’
NH-16’
1’’
2’’
3’’
4’’
5’’
d 7.1
À
À
=CH
161.4
117.4
112.4
111.0
128.8
11.4
C=
À
À
=CH
C=
À
À
À
CH
C=
À
À
À
CH
dd 6.3, 12.0
m
C=
À
À
À
8a
9
CH
C=
À
À
CH3
C=
2.28
9.73
5.01
4.27
5.93
5.03
4.40
1.41
3.31
s
brs
À
10
11
13
14
15
16
17
18
19
1’
2’
C=
À
À
À
À
C=
76.2
74.3
67.8
68.6
70.5
13.9
57.4
170.0
21.1
155.2
CH
d 8.5
dd 3.2, 8.5
t 3.3
dd 3.3, 6.1
m
À
À
À
À
CH
3.79
d 3.0
CH
À
À
À
C=
CH
À
À
À
CH
2.23
0.94
1.19
4.39, 4.67
0.89
4.90
1.54
1.83
m
d 6.9
d 7.0
brs, brs
d 7.2
dd 1.8, 9.2
dd 9.2, 12.9
d 12.9
CH
À
CH3
À
À
CH
À
CH3
À
CH3
6’’
7’’
8’’
9’’
d 6.8
s
À
À
=CH2
CH
À
À
CH3
À
CH
C=
À
À
CH3
2.17
4.79
s
À
CH2
À
À
C=
36.9
10’’
À À
3’
4’
76.5
74.1
C
NH2-10’’
brs
À
À
CH
3.17
d 9.0
[a] Chemical shifts are given in [ppm] with tetramethylsilane as internal standard.
tetramic acid (III), and a pyranose named as amykitanose
(IV) as shown by the bold lines in Figure 2A.
suggested a tetramic acid moiety attached to the C-8 (d=
42.4 ppm) of the decalin ring. The chemical shifts of the tet-
ramic acid moiety were in good agreement with those of the
corresponding signals of a related compound.[16,17]
The long range couplings from H-1’’ (d=5.01 ppm) to C-
5’’ (d=70.5 ppm), from H-5’’ (d=4.40 ppm) to C-1’’ (d=
76.2 ppm), from H-2’’ (d=4.27 ppm) to C-7’’ (d=57.4 ppm)
for the methoxy group, from H-9’’ (d=2.17 ppm) to C-8’’
(d=170.0 ppm), from H-3’’ (d=5.93 ppm) to C-8’’ for the
acetoxy group connected to C-3’’ (d=67.8 ppm) of the oxy-
methine, from the amino proton of NH2-10’’ (d=4.79 ppm)
to C-10’’ (d=155.2 ppm), from H-4’’ (d=5.03 ppm) to C-10’’
The connectivities of these partial structures were estab-
lished by HMBC spectroscopy. The C–H long range cou-
plings from the exo-methylene of H-18a (d=4.39 ppm) and
H-18b (d=4.67 ppm) to C-1 (d=149.7 ppm), C-2 (d=
34.3 ppm), and C-8a (d=37.3 ppm) revealed the presence of
the decalin ring. The long range couplings from the H-8
(d=4.10 ppm) of the decalin ring to the characteristic
carbon signals at C-9 (d=191.4 ppm) and C-10 (d=
102.5 ppm) and from H-13 (d=3.79 ppm) of the isopropyl
group to C-11 (d=175.6 ppm) and C-14 (d=193.5 ppm),
Figure 2. Structure determination of AMM. A) Key correlations of AMM obtained by 1H–1H COSY and HMBC spectroscopy. B) Key correlations of
AMM obtained by ROEs and coupling constants. C) Dd values of 6a and 6b for the determination of the stereochemistry at C-4 with a modified Mosh-
erꢁs method.
15774
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 15772 – 15781

Amycolamicin: A Broad-Spectrum Antibiotic
FULL PAPER
for the carbamoyl group, from H-1’’ to C-11 and the chemi-
cal shift of C-1’’ (d=76.2 ppm) established the linkage of 3-
O-acetyl-4-O-carbamoyl-6-deoxy-2-O-methyl-hexopyranose
to the N-12 of the tetramic acid by an N-glycosidic linkage.
The structure of the amycolose moiety was also elucidat-
ed. The correlations from H-1’ (d=4.90 ppm) to C-5’ (d=
70.5 ppm) and from both H-2’eq (d=1.83 ppm) and H-5’
(d=3.59 ppm) to C-3’ (d=76.5 ppm), and from both H-2’eq
and H-4’ (d=3.17 ppm) to C-7’ (d=52.6 ppm) established
the amycolose moiety as described in the previous report.[15]
Furthermore, the long range couplings observed from NH-9’
(d=6.64 ppm) to C-10’ (d=161.4 ppm) and C-11’ (d=
117.4 ppm), from NH-16’ (d=9.73 ppm) to C-11’, C-12’ (d=
112.4 ppm), C-13’ (d=111.0 ppm), and C-14’ (d=
128.8 ppm), from H-15’ (d=2.28 ppm) to C-13’ and C-14’ es-
tablished that 5-methyl-1H-pyrrole-2-carboxylic acid was at-
tached to the aminoethyl group of the amycolose by an
amide linkage. Observation of the cross peaks from H-1’ to
C-4 (d=77.7 ppm) of the oxymethine and from H-4 (d=
3.59 ppm) to C-1’ (d=95.4 ppm) indicated the connection of
amycolose to the decalin ring by an O-glycosidic linkage.
All of these correlation data and the fragment ion from
HRESI-MS data (m/z calcd for [C14H19Cl2N2O4ÀH2O]+:
331.0611; found: 331.0616) also suggested that the remaining
chlorine atoms should be attached to C-12’ and C-13’ on the
pyrrole ring as shown in Figure S1 in the Supporting Infor-
mation. The planar structure of AMM was determined as
shown in Figure 2A.
The relative stereochemistry of the decalin ring and the
amykitanose moiety of AMM was successfully elucidated by
the analyses of ROE spectra (Figures S9 and S10 in the Sup-
porting Information). ROEs were observed between H-4
and H-8a, H-4a (d=1.90 ppm) and H-8, H-8a (d=2.34 ppm)
and H-19 (d=0.89 ppm), H-18a and H-2, H-18b and H-8,
hence establishing a trans-decalin skeleton as shown in Fig-
ure 2B. For amykitanose with bulky groups at C-2’’, C-3’’,
and C-4’’, the correlation between H-1’’ and H-6’’ (d=
1.41 ppm), H-2’’ and H-3’’, H-2’’ and H-4’’, H-4’’ and H-5’’,
respectively in the 1D ROE spectra together with proton–
3
3
3
3
proton couplings of J1,2 =8.5, J2,3 =3.2, J3,4 =3.3, J4,5 =6.1,
3
and J5,6 =6.8 Hz confirmed its structure as 3-O-acetyl-4-O-
carbamoyl-6-deoxy-2-methyl-b-d/a-l-talopyranoside
as
shown in Figure 2B.
The absolute stereochemistry of AMM was clarified by a
combination of additional NMR analysis and a degradation
study (Figure 3). Acidic methanolysis of AMM with concen-
trated aqueous HCl in MeOH at 508C for one hour gave
the aglycon 2 and an anomeric mixture of the methyl glyco-
side of the amycolose-pyrrolecarboxamide (3a and 3b).[15]
The anomers 3a and 3b were separated and 3b was crystal-
lized from MeOH to give a colorless crystal. The crystal was
subjected to X-ray crystallographic analysis and the absolute
Figure 3. Degradation study of AMM. Reagents and conditions: a) aqueous solution of HCl (10m), MeOH; b) K2CO3 (0.2m), MeOH; c) 2,2-dimethoxy-
propane, p-toluenesulfonic acid, DMF; d) (R)-MTPA-Cl (MTPA=methoxy(trifluoromethyl)phenylacetic acid), Et3N, 4-dimethylaminopyridine
(DMAP), CH2Cl2; e) (S)-MTPA-Cl, Et3N, DMAP, CH2Cl2; f) 10% HCl/MeOH, 608C; g) NaIO4, aqueous solution of HCl (6m).
Chem. Eur. J. 2012, 18, 15772 – 15781
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15775

H. Adachi et al.
structure of the amycolose portion of 3b was determined to
be (2R,3R,4R,6R)-4-((S)-1-aminoethyl)-6-methoxy-2-ethylte-
trahydro-2H-pyran-3,4-diol as shown in Figure S18 in the
Supporting Information.
compound 4 with 2, 2-dimethoxypropane in DMF in the
presence of p-toluenesulfonic acid gave a separable anome-
ric mixture of the 3’’,4’’-O-isopropylidene derivatives 5a and
5b.
Compound 5a was subjected to the modified Mosherꢁs
method.[18] Both the (Æ)-MTPA esters (6a and 6b) were
prepared and analyzed by 1H NMR spectroscopy (Figur-
es S23 and S25 in the Supporting Information). The Dd
values of the corresponding protons indicated that the abso-
lute configuration at C-4 is (R). Thus, the stereochemistry of
the five asymmetric centers of the trans-decalin ring was de-
termined to be (4R), (4aR), (7S), (8S), and (8aR) as shown
in Figure 2C.
1
The H and 13C NMR spectra of compound 2 were com-
plicated spectra similar to the ones of AMM in [D4]MeOH
or [D6]acetone (Figures S16 and S17 in the Supporting Infor-
mation). This was due to the tautomeric feature of the tetra-
mic acid and anomerization of the C-1’’ of the amykitanose.
The NMR spectrum of compound 2 in [D6]acetone showed
the presence of two tautomers of the tetramic acid and
anomers of the amykitanose at C-1’’, whereas in [D4]MeOH
only an anomeric mixture was observed (a/b=1:1).
A similar phenomenon of anomerization was also ob-
served in a solution of AMM in MeOH. The initial simpli-
fied H NMR spectrum of the amykitanose moiety gradually
changed over time as shown in Figures S11 and S14 in the
Supporting Information, and this was also supported by LC-
MS analysis (Figure 4A). The analyses of the 1D ROE spec-
Methanolysis of compound 4 under acidic conditions at
608C gave anomeric isomers (7a and 7b) of methyl glyco-
side. The absolute stereochemistry of the sugar was deter-
mined by comparing its specific rotation with that of a syn-
thetic compound prepared from the methyl 3,4-O-isopropy-
lidene-a-l-fucopyranoside (8)[19,20] through four steps as
shown in Scheme S1 in the Supporting Information. Oxida-
tion of compound 8 with pyridinium dichlorochromate gave
the 2-ulose 9 (92%), which was reduced with LiAlH4 to give
the a-l-talopyranoside 10 (major) and its epimer 11
(minor). Methylation of compound 10 with methyl iodide
gave the 2-O-methyl derivative 12 (96%). Deacetonation of
12 with an aqueous solution of acetic acid (80%) afforded
the methyl 6-deoxy-2-O-methyl-a-l-talopyranoside 13
1
1
(99%). H and 13C NMR spectra and specific rotation of the
natural methyl glycoside 7a ([a]2D4 =À74.78) were in fair
agreement with those of the synthetic compound 13 ([a]2D2 =
À758). Therefore, the amykitanose moiety of AMM was de-
termined to be the N-linked 3-O-acetyl-4-O-carbamoyl-6-
deoxy-2-O-methyl-a-l-talopyranoside, which is
a novel
sugar. Thus, the stereochemistry at the C-5’’ of the amykita-
nose moiety of AMM[15] was revised.
The stereochemistry at the C-13 in the five-membered
ring of the tetramic acid moiety was determined by using a
known method.[21] AMM was degraded by periodate oxida-
tion and acid hydrolysis to obtain valine, whose absolute
structure was determined to be (S) by the advanced Mar-
feyꢁs method[22] (see Figure S35 in the Supporting Informa-
tion). All of these data helped to establish the absolute
structure of AMM as shown in Figure 1.
Figure 4. Behavior of a solution of AMM in MeOH. A) LC-MS analyses
of a solution of AMM in MeOH. Each chart shows an extracted ion
chromatograph of a protonated molecule (m/z 939.3406–939.3594) of
AMM. B) 3J
ACHTUNGTRENNUNG(H,H) values and ROE analyses of AMM in [D4]MeOH.
tra of newly observed additional signals of AMM in
[D4]MeOH showed ROE correlations between H-1’’ (d=
5.54 ppm) and H-3’’ (d=5.12 ppm), H-1’’ and H-5’’ (d=
3.92 ppm), H-2’’ (d=3.78 ppm) and H-7’’ (d=3.56 ppm),
and H-3’’ and H-5’’, suggesting the presence of an equilibri-
um of a and b anomers as shown in Figure 4B (see also Fig-
ures S13 and S15 in the Supporting Information).
In order to avoid the complexity of the NMR spectra due
to anomerization of the sugar moiety of 2, the 3’’,4’’-di-O-
acyl groups were replaced by an O-isopropylidene group to
yield a simplified 1H NMR spectrum of compound 5a in
[D5]pyridine. Treatment of compound 2 with a 0.2m solution
of K2CO3 in MeOH at 608C for one hour gave an anomeric
mixture of the 3’’,4’’-di-O-deacylaglycon 4. Treatment of
Amycolamicin exhibits potent antibacterial activity against
Gram-positive and some Gram-negative bacteria: AMM
showed very potent antibacterial activity against methicillin-
sensitive Staphylococcus aureus (MSSA), methicillin-resist-
ant S. aureus (MRSA), and quinolone-resistant S. aureus
(QRSA) with minimum inhibitory concentrations (MICs) of
0.125–2, 0.125–1, and 0.25–1 mgmLÀ1, respectively, as shown
in Table S2 in the Supporting Information. AMM also had
potent antibacterial activity against other Gram-positive
pathogenic bacteria such as Enterococcus faecalis, E. faeci-
um, and Streptococcus pneumoniae, including their drug-re-
sistant strains such as vancomycin-resistant enterococci
(VRE) and penicillin-resistant S. pneumoniae (PRSP), with
15776
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 15772 – 15781

Amycolamicin: A Broad-Spectrum Antibiotic
FULL PAPER
Table 2. Antimicrobial activity of AMM.
Tested strain
MIC
[mgmLÀ1
Tested strain
MIC
G
]
ACHTUNGTRENNUNG ]
[mgmLÀ1
Staphylococcus aureus
St. aureus
St. aureus
St. aureus
St. aureus
FDA209P
Smith
MRSA No.5
MRSA No.17
MS16526 (MRSA)
TY-04282 (MRSA)
JCM 5803
NCTC12201 (VRE, vanA type)
NCTC12203 (VRE, vanA type)
JCM 5804
0.25
0.25
0.5
0.5
1.0
Escherichia coli
Shigella dysenteriae
Sh. flexneri
Salmonella enteritidis
Proteus vulgaris
Pr. mirabilis
Serratia marcescens
Pseudomonas aeruginosa
Klebsiella pneumoniae
NIHJ
8
8
32
32
16
8
32
8
8
JS11910
4b JS11811
1891
St. aureus
0.5
Enterococcus faecalis
Ent. faecalis
Ent. faecalis
Ent. faecium
Ent. faecium
Ent. faecium
0.25
0.25
0.25
1.0
0.5
0.5
A3
PCI602
NCTC12202 (VRE, vanA type)
NCTC12204 (VRE, vanA type)
Mycobacterium smegmatis
Candida albicans
ATCC607
3147
>64
>64
MICs of 0.25–1 mgmLÀ1 (Tables 2 and S3 in the Supporting
Information). AMM showed moderate antibacterial activity
against some pathogenic Gram-negative bacteria such as Es-
cherichia coli, Shigella dysenteriae, Proteus mirabilis, Pseu-
domonas aeruginosa, and Klebsiella pneumoniae with MICs
of 8 mgmLÀ1. AMM did not exhibit antibacterial and anti-
fungal activity against Mycobacterium smegmatis and Candi-
da albicans even at doses of 64 mgmLÀ1 (Table 2). As shown
in Table S3 in the Supporting Information, AMM also ex-
hibited potent antibacterial activity against H. influenzae, in-
cluding its drug-resistant strains such as BLNAR and
BLPACR with MICs of 0.5–2 mgmLÀ1.
Figure 5. Inhibition of the macromolecular biosynthesis by AMM in the
S. aureus Smith strain. The inhibitory percentages of the incorporation of
DNA ([methyl-3H] thymidine, solid circles), RNA ([5,6-3H] uridine, open
triangles), protein (l-[4,5-3H] leucine, open squares), fatty acid (1-14C
acetate, open circles), and cell wall ([1-3H] GlcNAc, open diamonds)
after incubation for 10 min are plotted against the drug concentration.
Amycolamicin inhibited DNA synthesis in a precursor incor-
poration study: The attractive broad-spectrum antibacterial
activity of AMM prompted us to evaluate its mode of
action. In a precursor incorporation study, AMM strongly
inhibited the biosynthesis of DNA with an IC50 of 0.3 mm,
proportional to its MIC against S. aureus Smith (Figure 5).
The effects of AMM in this assay were similar to those of
novobiocin (NOV), a DNA-synthesis inhibitor.[23] At a dose
of 0.3 mm in S. aureus Smith, the biosynthesis of RNA, pro-
teins, the cell wall, and fatty acids were not inhibited. These
results strongly indicated that AMM is a DNA-synthesis in-
hibitor.
to an unusual mobility shift of intracellular plasmid DNA,
indicating that AMM also inhibited bacterial topoisomeras-
es. On the other hand, rifamycin SV, an RNA-synthesis in-
hibitor, did not show any effects on the topology of plasmid
Amycolamicin selectively in-
hibits bacterial type II topoiso-
merases: In an in vivo plasmid
DNA topology assay, the treat-
ment of NOV led to an unusu-
al mobility shift of the plasmid
DNA in E. coli (Figure 6 and
also Figure S36 in the Support-
ing Information). NOV is
known to be a bacterial topoi-
somerase inhibitor, which can
disturb certain topoisomerase
reactions, leading to an unusu-
al topology of the plasmid
DNA detected as an unusual
mobility shift. AMM also led
Figure 6. Effect of AMM, novobiocin, and rifamycin SV on the topology of plasmid DNA in E. coli. The MIC
of each compound is circled. See also Figure S36 in the Supporting Information.
Chem. Eur. J. 2012, 18, 15772 – 15781
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15777

H. Adachi et al.
DNA even at doses over its MIC. When AMM was evaluat-
ed for its inhibitory activities against bacterial topoisomeras-
es, as expected, AMM inhibited DNA gyrase and topoiso-
merase IV (TopoIV) with IC50s of 24.4 and 6250 ngmLÀ1, re-
spectively (Table 3). Other inhibitors of bacterial topoiso-
Ser-80-Tyr and Glu-84-Lys in ParC, and Ser-84-Leu in GyrA
conferred the resistance to LVFX but did not affect the
MICs of AMM, NOV, and COM (Table 4). Mutation of
Thr-173-Ile of GyrB caused 2-, 4-, and 512-fold resistance to
AMM, NOV, and COM, respectively, and mutation of Glu-
201-Ala alone did not cause AMM resistance, whereas the
double mutation Thr-173-Ile and Glu-201-Ala caused a 4-
fold resistance to AMM and also a 16- and a >512-fold re-
sistance to NOV and COM, respectively. Mutations of Gln-
136-Glu and Ile-175-Thr, both related to COM-resistance,[13]
also impacted AMM resistance (Table 4). Although these
results indicated that the target site of AMM, NOV, and
COM could be similar, mutations of Asp-89-Gly and Arg-
144-Ile of GyrB, known to give resistance to aminocoumarin
antibiotics,[24] did not affect the AMM sensitivity, suggesting
that AMM has a mode of binding to GyrB, which is distinct
from coumarin and quinolone antibiotics.
Table 3. Inhibitory activities of AMM and related compounds against
topoisomerases.
Test compound
IC50 [ngmLÀ1
E. coli
]
human
DNA gyrase
DNA TopoIV
TopoIIa
amycolamicin
levofloxacin
novobiocin
etoposide
24.4
97.6
97.6
>8000
6250
1562
>12500
not tested
>25000
>25000
not tested
3125
merases, such as NOV and levofloxacin (LVFX),
also inhibited the above-mentioned enzymes, but
the IC50s of NOV were at least two-fold higher
than that of AMM, whereas LVFX inhibited
TopoIV four-fold more potently than AMM and
DNA gyrase four-fold weaker than AMM (Fig-
ure S37 in the Supporting Information). With re-
spect to adverse effects of AMM on humans, the
possibility of inhibiting mammalian type II topoiso-
merase should be evaluated, because DNA gyrase
and TopoIV also belong to the type II topoisomer-
ase family. Fortunately, AMM did not inhibit
human topoisomerase II alpha even at a dose of
25000 ngmLÀ1 (Figure S38 in the Supporting Infor-
mation), indicating that AMM is a specific bacteri-
al topoisomerase inhibitor. The lethal dose of
AMM in mice was higher than 250 mgkgÀ1 when
administered subcutaneously. This favorable toxi-
cological profile might be reflected by its selective
inhibition toward bacterial type II topoisomerases.
Table 4. Resistance of genetically modified bacterial strains toward AMM versus
common topoisomerase inhibitors.
S. aureus
Description
MIC [mgmLÀ1
]
strains
AMM NOV COM
LVFX
Smith
S-TI
S-EA
S-TIEA
S-RI
S-DG
S-GSDG
S-QEIT
parental
0.25
0.5
0.25
1
0.25
0.25
0.5
2
0.5
2
0.25
4
16
8
64
0.5
0.5
0.5
0.008 0.125
0.125
0.008 0.125
GyrBThr173Ile
4
GyrBGlu201Ala
GyrBThr173Ile,Glu201Ala
GyrBArg144Ile
>4
2
0.125
0.125
GyrBAsp89Gly, KBDS[a]
GyrBGly85Ser,Asp89Gly
GyrBGln136Glu,Ile175Thr
0.008 0.125
2
2
2
0.125
0.125
0.125
S-QEITLI GyrBGln136Glu,Ile175Thr,Leu455Ile, KBDR[b]
S-SY
S-EK
S-SYSL
2
ParCSer80Tyr
0.5
0.25
0.25
0.008 0.5
0.016 0.5
ParCGlu84Lys
0.25
0.5
ParCSer80Tyr, GyrASer84Leu
0.016
4
[a] KBDS =kibdelomycin sensitive. [b] KBDR =kibdelomycin resistant.
Discussion
AMM has a unique structure consisting of a trans-decalin
skeleton with two novel sugars, that is, amycolose coupled
with 3,4-dichloro-1H-pyrrole-2-carboxylic acid by an amide
linkage and amykitanose attached to the tetramic acid by an
N-glycosidic linkage.
Keto–enol tautomerization of the tetramic acid moiety
and anomerization of the amykitanose at C-1’’ would result
in the equilibrium of the a and b forms of AMM in metha-
nol. Of these isomers, the b form could not be isolated due
to rapid conversion to the a form. We speculated that the
anomerization was caused by the acidic conditions through
an iminium intermediate of the N-glycoside linkage, which
was associated with the tautomerization of the tetramic acid
moiety and the steric effect of bulky substituents at C-2’’, C-
3’’, and C-4’’.[25]
Evaluation of possible mechanisms of resistance to amycola-
micin: Aminocoumarin antibiotics, such as NOV and cou-
mermycin A1 (COM), and quinolone antibiotics, such as
levofloxacin, are known to inhibit bacterial topoisomerases.
To evaluate if AMM acts at the same enzymatic site as the
above-mentioned antibiotics, we constructed AMM-, NOV-,
LVFX-, and COM-resistant S. aureus Smith strains by genet-
ic modification of the genome and compared their suscepti-
bilities to the above-mentioned drugs. An AMM-resistant
strain of S. aureus Smith was selected by using continuous
subculture and both the DNA gyrase (GyrA and B subunit)
and TopoIV (ParC and E subunit) genes were sequenced to
obtain the mutations governing the AMM resistance. The
obtained strain, which is resistant to AMM, had the muta-
tions Thr-173-Ile and Glu-201-Ala in GyrB. NOV-, LVFX-,
and COM-resistant strains of S. aureus Smith were con-
structed by introducing mutations known to be involved in
the resistance to the corresponding antibiotics. QRDR (qui-
nolone-resistance determinant region) mutations such as
In order to clarify the absolute chemical structure of
AMM, NMR spectra, chemical degradation, and modifica-
tion of its functional groups were performed. The chemical
degradation of AMM gave the fragments 2, 3a, and 3b. The
fragment 3b consisting of amycolose and 3,4-dichloro-1H-
15778
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 15772 – 15781

Amycolamicin: A Broad-Spectrum Antibiotic
FULL PAPER
pyrrole-2-carboxylic acid was successfully crystallized. The
structural complexity of fragment 2 due to the keto–enol
tautomerization of the tetramic acid and anomerization of
the amykitanose in AMM could be successfully avoided by
replacement of the 3’’,4’’-di-O-acyl groups with an O-isopro-
pylidene group to give the separable intermediates 5a and
5b, which have the simplified NMR spectra of a single spe-
cies in [D5]pyridine. This process enabled successful structur-
al analysis of compounds 6a and 6b by a modified Mosherꢁs
method. The stereochemistry at C-13 of the five-membered
ring of the tetramic acid moiety was determined to be (S)
configuration by degradation of AMM with periodate oxida-
tion following acid hydrolysis according to the literature.[21]
The absolute structure of the sugar 7a obtained from AMM
was clarified by comparison with the a-l-sugar 13 synthe-
sized from a-l-(À)-fucose in six steps (Scheme S1 in the
Supporting Information).
should have potent antibacterial activity against not only
Gram-positive pathogens but also Gram-negatives ones.
AMM has potent antibacterial activity against various kinds
of Gram-positive drug-resistant strains, such as MRSA,
VRE, and PRSP, and some Gram-negative pathogens in-
cluding BLNAR and BLPACR strains of H. influenzae. Our
studies revealed that AMM inhibited bacterial topoisomer-
ase, DNA gyrase, and topoisomerase IV. Distinct resistant
patterns against AMM, NOV, and COM of genetically
modified strains strongly suggest that these drugs have dif-
ferent modes of binding to GyrB. This may be derived from
their structural uniqueness. Additionally, in spite of its
strong inhibition of bacterial topoisomerases, AMM did not
inhibit human topoisomerase II (even at
a dose of
25000 ngmLÀ1), which should limit its adverse effects in
humans. These observations indicate that AMM is a very
promising candidate for future drug development.
Successful determination of the entire absolute structure
of AMM was achieved by a comprehensive approach, which
combined degradation with synthesis. The absolute structure
of AMM should provide novel insight into the analysis of
the inhibitory mechanism of GyrB.
Conclusion
Screening for new antibiotics from microbial metabolites
has become more difficult for several decades due to the ex-
istence of known compounds. Under these circumstances,
we have used many types of drug-resistant bacteria from
clinical isolates, passaged resistant strains, and genetically
modified strains as test organisms for the screening. This
screening system should help to eliminate known com-
pounds and also new compounds with known modes of
action, thus selecting only new compounds with novel mode
of actions. During the course of our screening by using this
system, we have succeeded in discovering a new compound
named amycolamicin (AMM), which selectively inhibits bac-
terial topoisomerases. AMM has unique structural features
consisting of trans-decalin, tetramic acid, two unusual
sugars, that is, amykitanose and amycolose, and 3,4-di-
chloro-1H-pyrrole-2-carboxylic acid. Its absolute structure
will offer numerous opportunities for modifications for the
purpose of novel drug design.
Recently, the relative structure of kibdelomycin, which is
structurally closely related to AMM, has been elucidated by
NMR analyses in [D4]MeOH.[13] Though the specific rota-
tion [a]2D3 of AMM in methanol is similar to that of kibdelo-
mycin, the 1H and 13C NMR spectra of kibdelomycin in
[D4]MeOH are not identical with that of AMM as shown in
Figures S11 and S12 and Table S1 in the Supporting Infor-
mation (also see the Supporting Information of Ref. [13]).
1
The H chemical shift (d=4.14 ppm, dd) of H-8 for the dec-
alin moiety in AMM is clearly different from that (d=
4.33 ppm, m) of the corresponding proton in kibdelomycin.
Furthermore, all the 13C signals of kibdelomycin appear in
the spectrum, whereas the 13C NMR signals from C-9 to C-
14 of the tetramic acid moiety of AMM are not observed.
All these results suggest that kibdelomycin may be a diaster-
eomer of AMM.
We also described the antibacterial activity of AMM
against various kinds of bacteria including clinically impor-
tant pathogens such as Staphylococcus aureus, Enterococcus
faecalis/faecium, Streptococcus pneumoniae, Haemophilus
influenzae, and Pseudomonas aeruginosa as well as the anal-
ysis of the mode of action, and cross-resistance evaluations
between other bacterial topoisomerase inhibitors. Despite
vigorous efforts into the discovery and development of new
classes of antibiotics, only three new classes of drugs (fidax-
omicin, daptomycin, and linezolid) have been launched for
clinical use in the past two decades. Fidaxomicin is a
narrow-spectrum antibiotic developed for the treatment of
Clostridium difficile infection.[26] Daptomycin and linezolid
were approved as options for the treatment of Gram-posi-
tive pathogens such as S. aureus and E. faecalis/faecium and
their drug-resistant strains, that is, MRSA, VRSA, and
VRE.[27] Daptomycin and linezolid are not as active against
Gram-negative pathogens but as they are active against
Gram-positive pathogens. Therefore, ideally, a new drug
Depending on the mutations in GyrB, the strain was
slightly cross-resistant to AMM, novobiocin, and coumermy-
cin, but not to levofloxacin, which strongly suggests that the
target of AMM is the DNA gyrase B subunit, and also that
the binding modes of AMM, novobiocin, and coumermycin
to the target are different from each other. AMM exhibits
potent antibacterial activity against current problematic
drug-resistant pathogens such as MRSA, VRE, PRSP,
BLNAR, and BLPACR strains of H. influenzae, indicating
that AMM is a highly promising new drug candidate for the
treatment of bacterial infections.
Experimental Section
General: Optical rotations were recorded on
a P-1030 polarimeter
(JASCO Inc., Tokyo, Japan). UV spectra were obtained on a U2800 spec-
trophotometer (Hitachi High-Technologies Corporation, Tokyo, Japan).
Chem. Eur. J. 2012, 18, 15772 – 15781
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15779

H. Adachi et al.
NMR spectra were measured with a JNM-ECA600 spectrometer (JEOL
Resonance Inc., Tokyo, Japan) by using TMS as internal reference.
HRESI-MS spectra were obtained on a LTQ Orbitrap XL mass spec-
trometer (Thermo Fisher Scientific, San Jose, CA, USA). IR spectra
were recorded on a FT-210 Fourier transform infrared spectrometer
(Horiba Ltd., Kyoto, Japan).
Shigella sp., Salmonella enteritidis, Proteus sp., Serratia marcescens, Pseu-
domonas aeruginosa, Klebsiella pneumoniae and Candida albicans, and
by using Mueller–Hinton agar supplemented with 5% defibrinated sheep
blood (Nippon Biotest Laboratories Inc., Tokyo, Japan) under 5% CO2
for Streptococcus pneumoniae and Mueller–Hinton agar enrichment sup-
plemented with 5% Fildes enrichment peptic digest of sheep blood
(Becton, Dickinson and Company) under 5% CO2 for Haemophilus in-
fluenzae. The MIC was observed against acid-fasted bacteria after incu-
bation for 42 h at 378C, against Candida albicans 3147 after incubation
for 18 h at 278C, and against other bacteria after incubation for 18 h at
378C.
Isolation of the producing organism and taxonomy of the strains: The
amycolamicin-producing organism, strain MK575-fF5, was isolated from
a soil sample collected at Sendai, Miyagi prefecture, Japan. The strain
MK575-fF5 formed staggered and fragmented substrate mycelia and
straight to flexuous aerial mycelia. The aerial mycelia and substrate my-
celia were white to light yellow orange and colorless to pale orange, re-
spectively. Whole-cell hydrolysates of the strain contained meso-diamino-
pimelic acid, galactose, and arabinose. The strain had type PII phospholi-
pid and MK-9(H4) as the major components of menaquinone. The N-
acyl-type of muramic acid in the cell wall was the acetyl type. Mycolic
acid was absent. The partial 16S ribosomal RNA gene sequence
(1455 bp) of the strain showed high identity with those of genus Amyco-
latopsis such as A. balhimycina DSM44591T (98.8%), A. mediterranei
JCM4789T (98.6%) and A. kentuckyensis JCM12670T (99.1%). The mor-
phological characteristics and genetic analysis of the strain MK575-fF5
suggested that it belongs to the genus Amycolatopsis.[28] Therefore, the
strain was designated as Amycolatopsis sp. MK575-fF5 (FERMP-21465).
Inhibition of macromolecular synthesis: Inhibition of macromolecular
synthesis were assayed by measurement of the incorporation of radioac-
tive precursors of peptidoglycan, fatty acid, DNA, RNA, and protein syn-
thesis into the precipitate of the cell extract with 10% trichloroacetic
acid (TCA) as reported previously.[29] Briefly, samples taken from the
early-exponential phase (optical density at 600 nm [OD600]=0.3) of S.
aureus Smith grown in nutrient broth were plated into 96-well plates at
90 mL per well, and then preincubated at 378C for 5 min with antibiotics.
Nutrient broth consisted of 1% polypeptone (Wako, Osaka, Japan), 1%
nutrient (Kyokuto, Tokyo, Japan), and 0.2% NaCl (Wako, Osaka, Japan)
in deionized water (pH7.0 before sterilization). The following radiola-
beled compounds were added to the cells for the indicative assays: pepti-
doglycan assay: 1 mCi of N-acetyl-d-[1-3H] glucosamine; fatty acid assay:
1 mCi of [1-14C] acetate; DNA assay: 1 mCi of [methyl-3H] thymidine;
RNA assay: 1 mCi of [5,6-3H] uridine; protein assay: 5 mCi of l-[4,5-3H]
leucine. All radioactive chemicals were purchased from GE Healthcare
Bioscience. Mixtures were incubated for 10 min at 378C, and then the re-
actions were stopped by adding an equal volume of 10% TCA. After
10 min incubation at room temperature, the solutions were transferred
into 96-well filter plates (Millipore, MultiScreenHTS, Billerica, MA, USA)
and filter-washed five times with 5% TCA. After drying, the radioactivi-
ty remaining on the filter plate was counted by using a liquid-scintillation
counter (Tri-Carb 2800TR, PerkinElmer, Waltham, MA, USA).
Fermentation and purification of AMM: A slant culture of the AMM-
producing organism was inoculated into a 500 mL baffled Erlenmeyer
flask containing 110 mL of a seed medium consisting of 2.0% galactose,
2.0% dextrin, 1.0% Bacto soytone (Becton, Dickinson and Company,
Franklin Lakes, NJ, USA), 0.5% corn steep liquor, 0.2% (NH4)2SO4, and
0.2% CaCO3 in deionized water (pH7.4 before sterilization). The culture
was incubated on a rotary shaker (200 rpm) at 308C for three days. The
seed culture (2 L) was transferred into a 200 L jar fermenter, containing
100 L of a producing medium consisting of the half strength of the seed
medium. The fermentation was carried out at 278C for four days with ag-
itation at 200 rpm and aeration of 100 NLminÀ1.The fermentation broth
(80 L) was separated into the mycelial cake and the supernatant by cen-
trifugation. The 2.5 kg of the mycelia were extracted with 12 L of MeOH.
The MeOH solutions were concentrated in vacuum to 2 L. The solution
containing active compound was extracted with 6 L of EtOAc and con-
centrated in vacuum to yield brown oil (18 g). The brown oil was washed
with 200 mL of n-hexane to yield a brown cake (3.3 g). The brown cake
containing the active substance was purified on a SephadexLH-20 (GE
Healthcare Bio-Science, Fairfield, CT, USA, column: 28 mmꢂ450 mm)
eluted with MeOH. The active fractions were collected and concentrated
in vacuo to give an orange solid (1584 mg), which was further purified by
using a silica gel column with stepwise development of CHCl3/MeOH/
water=100:0:0 (200 mL), 95:5:0 (450 mL), 95:5:0.25 (450 mL), and
90:10:0.5 (450 mL). The active components were eluted with CHCl3/
MeOH/water=95:5:0.25 and 90:10:0.5, collected, and concentrated in
vacuo to give a pale orange solid (827 mg). A small amount of the solid
(54 mg) was further purified on a reversed-phase HPLC by elution with a
MeOH/water=1:1 to 1:0 linear gradient for 75 min (CAPCELL PAK
C18 UG-120 5 mm, 30 mmꢂ250 mm, Shiseido Co., Ltd., Tokyo, Japan,
flow rate=15 mLminÀ1). Active fractions were eluted at 55 to 60 min
and collected to give a white solid (35 mg), which was suspended in
EtOAc, washed with phosphate buffer (0.5m, pH7.0), and 0.001m aque-
ous HCl solution. The EtOAc extract was dried over MgSO4 and filtered.
The filtrate was concentrated to give 34 mg of AMM. The physical data
of AMM were given in the following order: pale yellow amorphous
solid; [a]2D3 =À36.58 (c=1 in MeOH); 1H and 13C NMR data see Table 1;
IR (KBr): n˜ =3460, 2930, 1730, 1610, 1540, 1460, 1380, 1310, 1240,
In vivo plasmid DNA topology assay: The topology of the plasmid DNA
in E. coli was evaluated as reported previously[30] with modifications. E.
coli tolC (efflux-deficient) strain CAG12184[31] harboring pUC18 was cul-
tivated in Mueller–Hinton broth containing 100 mgmLÀ1 ampicillin at
378C until mid-log phase (OD600 =0.5). The cell culture was divided into
2 mL aliquots, and compound or inhibitor was added to create two-fold
dilutions that spanned the MICs of each compound. The tubes were fur-
ther incubated at 378C with shaking for one hour. Plasmids were purified
from the above-described cells by using a QuickLyse Miniprep Kit
(QIAGEN, Valencia, CA, USA), according to the instructions of the
manufacturer. The 100 ng of purified DNA was analyzed by electropho-
resis along with supercoiled DNA size markers (Promega, Madison, WI,
USA) in a 0.8% agarose gel. After electrophoresis, the gel was stained
with ethidium bromide and photographed. Unusual DNA topology
caused by inappropriate topoisomerase reactions was detected as an un-
usual mobility shift.
Enzyme inhibitory assay of DNA gyrase, topoisomerase IV, and human
topoisomerase II: DNA gyrase and topoisomerase IV assays were per-
formed by using the Gyrase Supercoiling assay kit (Wako, John Innes En-
terprise LTD., Osaka, Japan) and the topoisomerase IV Relaxing kit
(Wako, John Innes Enterprise LTD.), respectively. The 50% inhibitory
concentrations (IC50s) of each compound for topoisomerases were deter-
mined by measuring the densities of the bands of the digitalized data by
using ImageJ free software (http://rsbweb.nih.gov/ij/). A total of 30 mL
of the gyrase supercoiling reaction mixture, which contained gyrase (E.
coli, GyrA2B2) (1 unit), 7 mm Tris-HCl (pH 7.5), 4.8 mm KCl, 0.4 mm
MgCl2, 0.4 mm dithiothreitol (DTT), 0.36 mm spermidine, 0.2 mm ATP,
20 mgmLÀ1 albumin, 1.3% (v/v) glycerol, and 400 ng relaxed pBR322
DNA was incubated at 378C for 30 min and then terminated by addition
of 5 mL of stop buffer (5% sarkosyl, 0.0025% bromophenol blue, 25%
glycerol), followed by electrophoretic analysis in 0.8% agarose gels. Top-
oisomerase IV relaxing assays were performed as follows. A total of
30 mL of the reaction mixture, consisting of topoisomerase IV (E. coli,
ParC2E2) (1 unit), 40 mm HEPES/KOH (pH 7.6), 100 mm potassium glu-
1080 cmÀ1
; UV/Vis (0.01m HCl in MeOH): lmax (loge)=280 nm
(4.45 molÀ1 dm3 cmÀ1); (0.01m NaOH in MeOH): lmax (loge)= 248 (4.30),
277 nm (4.45 molÀ1 dm3 cmÀ1); HRESI-MS (positive): m/z calcd for
[C44H60Cl2N4O14+H]+: 939.3556 [M+H]+; found: 939.3571.
Antibacterial activity: The minimum inhibitory concentrations (MICs) of
AMM were examined by a serial agar dilution method by using Mueller–
Hinton agar (Becton, Dickinson and Company) for Enterococci, Strepto-
cocci, Staphylococcus aureus, Mycobacterium smegmatis, Escherichia coli,
15780
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 15772 – 15781

Amycolamicin: A Broad-Spectrum Antibiotic
FULL PAPER
tamate, 10 mm magnesium acetate, 10 mm DTT, 4 mgmLÀ1 tRNA, 2 mm
ATP, 50 mgmLÀ1albumin, and 400 ng supercoiled pBR322 DNA was incu-
bated at 378C for 30 min and then terminated by addition of 5 mL of stop
buffer (5% sarkosyl, 0.0025% bromophenol blue, 25% glycerol), fol-
lowed by electrophoretic analysis in 0.8% agarose gels. The human top-
oisomerase II decatenation assay was performed by using the topoisomer-
ase II assay kit (TopoGEN, Inc. Port Orange, FL, USA) using human
DNA topoisomerase II alpha of the p170 form (TopoGEN, Inc.). An ali-
quot of 30 mL of the decatenation assay mixture, which contained 50 mm
Tris-HCl (pH 8.0), 150 mm NaCl, 10 mm MgCl2, 5 mm ATP, 0.5 mm DTT,
30 mgmLÀ1 bovine serum albumin, and 0.2 mg kinetoplast DNA (kDNA)
was incubated at 378C for 30 min and then terminated by addition of
5 mL of stop buffer (5% sarkosyl, 0.0025% bromophenol blue, 25% glyc-
erol), followed by electrophoretic analysis in 0.8% agarose gels with
linear kDNA markers.
[6] N. Bruinsma, K. G. Kristinsson, S. Brozwaer, P. Schrijnemakers, J.
Degener, E. Tiemersma, W. Hryniewics, J. Monen, H. Grundmann,
J. Antimicrob. Chemother. 2004, 54, 1045–1050.
[7] P. Littauer, M. Sangvik, D. A. Caugant, E. A. Høiby, G. S. Simonsen,
A. Sundsfjord, J. Clin. Microb. 2005, 43, 2125–2132.
[8] S. Bae, J. Lee, E. Kim, S. Lee, J. Yu, Y. Kang, Antimicrob. Agents
Chemother. 2010, 54, 65–71.
[9] R. J. Fass, Antimicrob. Agents Chemother. 1983, 24, 568–574.
[10] I. Hayakawa, S. Atarashi, S. Yokohama, M. Imamura, K. Sakano,
M. Furukawa, Antimicrob. Agents Chemother. 1986, 29, 163–164.
[11] I. Momose, S. Hirosawa, H. Nakamura, H. Naganawa, H. Iinuma, D.
Ikeda, T. Takeuchi, J. Antibiot. 1999, 52, 787–796.
[12] D. R. Schroeder, K. L. Colson, S. E. Klohr, M. S. Lee, J. A. Matson,
J. Antibiot. 1996, 49, 865–872.
[13] J. W. Phillips, M. A. Goetz, S. K. Smith, D. L. Zink, J. Polishook, R.
Onishi, S. Salowe, J. Wiltsie, J. Allocco, J. Sigmund, K. Dorso, S.
Lee, S. Skwish, M. de la Cruz, J. Martin, F. Vicente, O. Genilloud, J.
Lu, R. E. Painter, K. Young, K. Overbye, R. G. Donald, S. B. Singh,
Chem. Biol. 2011, 18, 955–965.
[14] B. A. Sherer, K. Hull, O. Green, G. Basarab, S. Hauck, P. Hill, J. T.
Loch III, G. Mullen, S. Bist, J. Bryant, A. Boriack-Sjodin, J. Read,
N. DeGrace, M. Uria-Nickelsen, R. N. Illingworth, A. E. Eakin,
Bioorg. Med. Chem. Lett. 2011, 21, 7416–7420.
Construction of S. aureus Smith mutants resistant to DNA gyrase inhibi-
tors: An AMM-resistant strain of S. aureus Smith was selected by using
continuous subculture and both the DNA gyrase (GyrA and B subunit)
and topoisomerase IV (ParC and E subunit) genes were sequenced to
obtain the mutations responsible for the resistance. The gyrA, gyrB, or
parC structural gene of S. aureus Smith was amplified by PCR and
cloned into the E. coli–S. aureus shuttle vector pHY300PKL (TaKaRa
Bio Inc., Shiga, Japan) containing the tetracycline-resistant gene. Subse-
quently, the obtained or known missense mutations were induced in the
[15] S. Tohyama, Y. Takahashi, Y. Akamatsu, J. Antibiot. 2010, 63, 147–
149.
corresponding gene of this plasmid by using
a recombination PCR
method[32] with PrimeSTAR GXL DNA polymerase (TaKaRa Bio Inc.).
The resulting plasmids were introduced into S. aureus Smith by electro-
poration as previously reported,[33] with slight modifications and selected
by tetracycline. The obtained transformants were challenged with the cor-
responding antibiotic and resistant mutants were selected. These mutants
were then confirmed by their drug resistance, and the sequences of the
gyrB gene on their genome. Because the mutated gyrB gene on the plas-
mid did not have a promoter, the transformants could survive under the
presence of the drug only when the gyrB gene on the genome was re-
placed with the mutated gene on the plasmid. The MICs of AMM, NOV,
COM, and LVFX against the above-constructed strains were determined
by a standard microbroth dilution method by using nutrient broth.
[16] P. S. Steyn, P. L. Wessels, Tetrahedron Lett. 1978, 19, 4707–4710.
[17] B. J. L. Royles, Chem. Rev. 1995, 95, 1981–2001.
[18] I. Ohtani, T. Kusumi, Y. Kashman, H. Kakisawa, J. Am. Chem. Soc.
1991, 113, 4092–4096.
[19] Ok. Kwang-dae, Y. Takagi, T. Tsuchiya, S. Umezawa, H. Umezawa,
Carbohydr. Res. 1987, 169, 69–81.
[20] Y. Takagi, N. Kobayashi, T. Tsuchiya, S. Umezawa, T. Takeuchi, K.
Komuro, C. Nosaka, J. Antibiot. 1989, 42, 1318–1320.
[21] H. Ikeda, N. Matsumori, M. Ono, A. Suzuki, A. Isogai, H. Nagasa-
wa, S. Sakuda, J. Org. Chem. 2000, 65, 438–444.
[22] K. Fujii, Y. Ikai, H. Oka, M. Suzuki, K. Harada, Anal. Chem. 1997,
69, 5146–5151.
[23] H. Hashizume, H. Adachi, M. Igarashi, Y. Nishimura, Y. Akamatsu,
J. Antibiot. 2010, 63, 279–283.
[24] M. Fujimoto-Nakamura, H. Ito, Y. Oyamada, T. Nishino, J. Yama-
gishi, Antimicrob. Agents Chemother. 2005, 49, 3810–3815.
[25] H. S. Isbell, H. L. Frush, J. Org. Chem. 1958, 23, 1309–1319.
[26] C.-H. Liao, W.-C. Ko, J.-J. Lu, P.-R. Hsueh, Antimicrob. Agents Che-
mother. 2012, 56, 3943–3949.
[27] L. A. Jevitt, A. J. Smith, P. P. Williams, P. M. Raney, J. E. McGo-
wan, Jr., F. C. Tenover, Microb. Drug Resist. 2003, 9, 389–393.
[28] M. P. Lechevalier, H. Prauser, D. P. Labeda, J. S. Ruan, Int. J. Syst.
Bacteriol. 1986, 36, 29–37.
Acknowledgements
The authors are grateful to the National bioresource project: E. coli, the
National institute of genetics, Japan for providing the E. coli tolC (efflux-
deficient) strain CAG12184. We are indebted to Dr. Hiroyasu Sato at
Rigaku Corporation for the X-ray crystallographic analysis of 3b. We ex-
press thanks to Dr. Tomoyuki Kimura for assistance with the X-ray anal-
ysis. We thank Yumiko Kubota and Yoshiko Koyama for valuable discus-
sion about the structure determination. We also thank Yuichi Kasahara
for technical assistance with the fermentation of AMM.
[29] H. Hashizume, R. Sawa, S. Harada, M. Igarashi, H. Adachi, Y. Nish-
imura, A. Nomoto, Antimicrob. Agents Chemother. 2011, 55, 3821–
3828.
[30] T. H. Grossman, D. J. Bartels, S. Mullin, C. H. Gross, J. D. Parsons,
Y. Liao, A. L. Grillot, D. Stamos, E. R. Olson, P. S. Charifson, N.
Mani, Antimicrob. Agents Chemother. 2007, 51, 657–666.
[31] M. Singer, T. A. Baker, G. Schnitzler, S. M. Deischel, M. Goel, W.
Dove, K. J. Jaacks, A. D. Grossman, J. W. Erickson, C. A. Gross, Mi-
crobiol. Rev. 1989, 53, 1–24.
[32] D. H. Jones, Genome Res. 1994, 3, S141–S148.
[33] S. Schenk, R. A. Laddaga, FEMS Microbiol. Lett. 1992, 94, 133–
138.
[1] L. D. Saravolatz, J. Pawlak, L. B. Johnson, Int. J. Antimicrob. Agents
2010, 36, 69–72.
[2] R. Leclercq, E. Derlot, M. Weber, J. Duval, P. Courvalin, Antimi-
crob. Agents Chemother. 1989, 33, 10–15.
[3] S. Chang, D. M. Sievert, J. C. Hageman, M. L. Boulton, F. C. Ten-
over, F. P. Downes, S. Shah, J. T. Rudrik, G. R. Pupp, W. J. Brown,
D. Cardo, S. K. Fridkin, N. Engl. J. Med. 2003, 348, 1342–1347.
[4] S. C. A. Meijvis, J. C. Grutters, S. F. Thijsen, G. T. Rijkers, D. H.
Biesma, H. Endeman, Neth. J. Med. 2011, 69, 21–26.
[5] N. Chiba, R. Kobayashi, K. Hasegawa, M. Morozumi, E. Nakayama,
T. Tajima, S. Iwata, K. Ubukata, J. Antimicrob. Chemother. 2005, 56,
756–760.
Received: July 25, 2012
Published online: November 5, 2012
Chem. Eur. J. 2012, 18, 15772 – 15781
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15781