Chemistry of platensimycin

Article abstract of DOI:10.1016/j.tetlet.2007.06.009

Platensimycin is a novel natural product antibiotic that inhibits bacterial growth by inhibiting fatty acid synthesis specifically inhibiting the elongation condensing enzyme FabF. Reaction with diazomethane at controlled temperatures led to selective met

Full text of DOI:10.1016/j.tetlet.2007.06.009

Tetrahedron Letters 48 (2007) 5429–5433
Chemistry of platensimycin
*
Sheo B. Singh, Kithsiri B. Herath, Jun Wang, Nancy Tsou and Richard G. Ball
Merck Research Laboratories, Rahway, NJ 07065, USA
Received 25 April 2007; revised 4 June 2007; accepted 4 June 2007
Available online 9 June 2007
Abstract—Platensimycin is a novel natural product antibiotic that inhibits bacterial growth by inhibiting fatty acid synthesis spe-
cifically inhibiting the elongation condensing enzyme FabF. Reaction with diazomethane at controlled temperatures led to selective
methylation of the phenolic groups. Methylation, halogenation, reduction, epoxidation, Bayer–Villiger oxidation and details of the
conversion of dihydroplatensimycin to the cyclic enamino-amido forms have been described.
Ó 2007 Elsevier Ltd. All rights reserved.
Platensimycin (1) is a novel antibiotic isolated from
six-center hydrogen bond with the carbonyl oxygen of
the benzoic acid. The location of the methyl group of
2b was deduced by HMBC correlations of the methyl
1
,2
various strains of Streptomyces platensis.
It was
discovered by a novel antisense differential sensitivity
screening strategy in which FabH/FabF was sensi-
0
hydrogens and both of the aromatic hydrogens to C-5
3
–6
0
0
tized. Platensimycin selectively inhibits the elongation
condensing enzyme FabF of the bacterial fatty acid syn-
thesis pathway by interacting with the malonyl binding
site of the catalytic triad of the FabF acyl-enzyme inter-
mediate. It exhibits potent in vitro activity against both
cell-free and whole-cell systems. The in vitro activity
could not be directly translated to an in vivo mouse
model when the drug was administered by conventional
routes. However, when administered by continuous
infusion the drug was highly efficacious. The poor
in vivo activity under conventional administration is
attributed to its pharmacokinetic properties which could
potentially be improved by chemical modification. The
present study describes some chemical modifications of
platensimycin.
and confirmed by NOEs between H-6 and C-5 -OMe.
The methyl ester was readily hydrolyzed by treatment
with LiOH in THF–water overnight affording 2d (from
2b) and 2e (from 2c).
OH
OR
1
7
'
3'
O
O
5'
O5
O
1
'
HO C
RO
2
C
N
H
2
N
H
1
OH
OR
2
7
O
O
1 (Platensimycin)
2a:R = Me, R
1
= R
= H
2
2
2
2
b:R = Me, R = Me, R = H
1
1
2
c:R = Me, R = R = Me
2
d:R = H, R = Me, R2 = H
1
1
2e:R = H, R
= R = Me
2
R
Br
Treating platensimycin (1) with an excess of ethereal
diazomethane at ꢀ78 °C selectively methylated the carb-
R1
OH
O
OH
O
O
O
7
oxyl group and produced the methyl ester (2a) in quan-
HO C
2
N
Br
N
OH
H
H
titative yield. A similar reaction at 0 °C produced a
mixture of mono- and dimethyl ethers 2b and 2c in a
ratio of 4:1 in 30 min. These compounds were readily
separated by reversed phase HPLC affording a combined
yield of 95%. Compound 2b was completely converted
OH
O
O
3
3
a:R = Cl, R
1
= H
4
b:R = H, R = Cl
1
3c:R = Br, R
= H
1
to 2c when the reaction was allowed to continue for
0
4
8 h. The rate of methylation of the C-3 phenolic group
Reaction of 1 with a 10-fold excess of N-chlorosuccini-
mide in a 5:1 mixture of acetone and acetic acid
produced a 7:9 mixture of the two mono-chloro deriva-
tives, 3a and 3b, in a combined yield of 97%; they were
readily separated by reversed phase HPLC. No reaction
was slower due to steric factors or the formation of a
*
Corresponding author. Fax: +1 732 594 6880; e-mail: sheo_singh@
merck.com
0040-4039/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.06.009

5430
S. B. Singh et al. / Tetrahedron Letters 48 (2007) 5429–5433
was observed in the absence of acetic acid. Adding 1.1
equiv of N-bromosuccinimide (NBS) to platensimycin
in a 4:1 mixture of acetone and THF produced exclu-
sively 6 -bromoplatensimycin (3c) in 2 h with an isolated
yield of 85%. When the reaction was performed with
selection of a crystal with the shown atropisomeric form
was fortuitous.
0
Methylation of 8 with diazomethane at 0 °C for 48 h
produced a trimethyl derivative 11. Similar methylation
of the atropisomeric mixture of 9 produced a mixture of
three products 12/13, and 14 (listed in the order of
RPHPLC elution) that were separated by HPLC.
5
equiv of NBS it produced a 5:6 mixture of 3c and the
dibromo-des-carboxyl derivative 4 in less than 1 h. The
structure of 4 was confirmed by 2D NMR experiments.
2
Epoxidation of 1 with m-CPBA in a 2:1 mixture of
methylene chloride–THF failed to produce any product
and left the unreacted starting material intact; a
common feature of most enones. However, epoxidation
under basic conditions using triton B and tert-butyl-
hydroperoxide (TBHP) in a 1:1 mixture of MeOH–
Following our initial report we decided to investigate
the details of the interesting conversion reactions and
stability of compounds 8 and 9. Both compounds were
found to be completely stable in pyridine for a period
of two months. Compounds 8 and 9 do interconvert
and reach a 60/40 (8:9) equilibrium mixture after stand-
ing for 4–5 days in an acetonitrile–water mixture con-
taining 0.1% TFA. This inter-conversion can be
accelerated so equilibrium is reached in 1 h if the solu-
tion is changed to aqueous MeOH with 1% PTSA or
TFA.
THF gave an oxidized product identified as lactone acid
7
5
(Scheme 1). The structure and configuration of com-
pound 5 was elucidated by HMBC (6) and NOEDS (7)
experiments (Fig. 1). For example, H-6 (dH 4.59, s; dC
8
4.3) exhibited HMBC correlations to C-18 (dC 19.5),
acid C-5 (dH 170.4) and lactone carbonyl C-7 (dC
74.6). H -18 (d 1.31, s) exhibited HMBC correlation
1
Larger scale hydrogenation of 1 (1 g in 40 mL MeOH)
was sluggish and required 48 h for completion. During
this reaction 31% of 9 was produced in the absence of
any added acid prompting further investigation of the
conversion. The keto compound 8 was completely con-
verted to enamino-amido product 9 in MeOH. How-
ever, the rate of the conversion was scale dependent.
For example, on a small analytical scale (mg quantities)
and at a concentration of 5 mg/mL the conversion was
relatively fast and 98% of the cyclic product (9) was
formed at room temperature (Fig. 3) in about 17 h.
However, on a larger scale (grams) the conversion was
slower and required heating at 50 °C for 24 h for com-
plete conversion (Fig. 4). Once the enamino-amido cyc-
lic product was produced it was completely stable in
MeOH. Addition of hydrated PTSA to MeOH increased
the rate of the formation of the cyclic product so 97%
conversion occurred in less than 4 h. However, after
about 15 h the cyclic product slowly hydrolyzed
3
H
to C-6; irradiation of H -18 showed strong NOE to H-
3
6
and H-13 while irradiation of H-6 showed reciprocal
NOE to H -18 thus suggesting the equatorial orienta-
3
7
tion of H-6.
2
We recently reported that the catalytic hydrogenation
of platensimycin produced dihydroplatensimycin (8),
which converted to an inseparable 2:1 mixture of
atropisomers of the enamino-amido cyclic product (9),
particularly in the presence of acid (e.g., TFA), via
intermediate 10. This intermediate could not be detected
by LCMS probably because of spontaneous dehydra-
tion. Recently, we crystallized 9 from a nitromethane–
CH Cl –MeOH mixture and confirmed the structure
2
2
of one of the atropisomers by single-crystal X-ray crys-
8
tallography (Fig. 2). Examination of the remaining
1
crystals by H NMR in C D N indicated that these crys-
5
5
tals were still a mixture of the two atropisomers; the
(
Fig. 3) producing about 23% of the keto product. Com-
plete conversion of 8 to 9 was achieved in less than 4 h in
the presence of PPTS and no conversion back to the
keto product was observed indicating that the hydration
of PTSA provided just enough water content for the
hydrolysis to take place. This result prompted us to
screen for other solvents which affect the stability of
these compounds. At 5 mg/mL, the rate of conversion
of 8 was significantly slower in acetonitrile, THF and
DMSO producing 30%, 10%, and 30%, respectively, of
OH
O
CO H
TBHP
2
2
HO C
1
N
H
O
OH
TritonB
H O
O
O
2
5
Scheme 1. Epoxidation of platensimycin (1).
9
in 4 h without any increase over the next 20 h. These
observations, together with the observed stability of
compounds 8 and 9 in pyridine, led us to the assumption
that the acidity of the carboxylic acid was sufficient for
protonation and formation of the enamino-amido prod-
uct 9. We tested this assumption by converting 8 quan-
titatively to a sodium salt by treatment with NaOH.
This salt was completely stable in MeOH at room tem-
perature for one week indicating that the acidity of the
original carboxyl group was sufficient to catalyze the
conversion whereas the acidities of the phenolic groups
were insufficient for such catalysis. Likewise the tri-
methyl keto derivative (11), was completely stable in
O
O
1
8
H
5
CO
H
H
2
CO H
2
6
4
O
O
9
7
O
O
O
O
6
(HMBC)
7 (NOEDS)
Figure 1. Key HMBC and NOEDS correlations of compound 5 in 3:1
CDCl –CD OD.
3
3

S. B. Singh et al. / Tetrahedron Letters 48 (2007) 5429–5433
5431
OH
O
HO
HO
HO
O
O
O
OH
+
OH
+
OH
NH
H O
3
H O
3
N
N
OH
OH
O
OH O
or
O
MeOH
MeOH
O
O
O
8
10
9
OMe
O
MeO
MeO
MeO
O
O
O
OMe
OMe
OMe
NH
N
N
OMe O
OH O
O
O
O
O
11
12/13 (atropisomers)
14
Figure 2. X-ray crystal structure of an atropisomer of 9.
Conversion of Keto to Cyclic Products
Conversion of Keto to Cyclic Product
Preparative Scale)
1
00
(
1
00
8
6
4
2
0
0
0
0
0
8
0
%
%
Keto (MeOH)
60
40
Keto
Cyclic (MeOH)
Cyclic
% Keto (MeOH-PTSA)
%
Cyclic (MeOH-PTSA)
2
0
0
0
5
10
15
20
25
4
8
56
64
72
80
88
96 104
Time (h)
Time
Figure 3. Conversion of keto (8) to cyclic (9) compounds on an
Figure 4. Conversion of keto (8) to cyclic (9) compounds at a
analytical scale.
preparative scale.

5432
S. B. Singh et al. / Tetrahedron Letters 48 (2007) 5429–5433
MeOH but the cyclization reaction could be catalyzed
by addition of either aqueous 1% PTSA or 1% TFA
leading to a 65/35 ratio of the keto and two HPLC-sep-
arable atropisomers of enamino-amido products 12 and
References and notes
1. Wang, J.; Soisson, S. M.; Young, K.; Shoop, W.; Kodali,
S.; Galgoci, A.; Painter, R.; Parthasarathy, G.; Tang, Y.;
Cummings, R.; Ha, S.; Dorso, K.; Motyl, M.; Jayasuriya,
H.; Ondeyka, J.; Herath, K.; Zhang, C.; Hernandez, L.;
1
3 (ratio 20/15, respectively). Each of the purified atrop-
isomers 12 and 13 were highly stable in MeOH but
hydrolyzed cleanly to the keto form, 11, in MeOH–
water (2:1) with 1% PTSA reaching an identical equilib-
rium ratio as that produced from 8. While heating of
atropisomers 9, 12/13, in DMSO (NMR) at 120 °C did
not show any inter-conversion, 12/13 could be converted
via acidic hydrolysis to the keto form. The same ꢁ20/15
ratio observed for 11 was seen when starting with 12/13.
The ratio of the two products was identical whether the
hydroxy and carboxyl groups were protonated or
methylated.
´
Alloco, J.; Basilio, A.; Tormo, J. R.; Genilloud, O.; Vicente,
F.; Pelaez, F.; Colwell, L.; Lee, S. H.; Michael, B.; Felcetto,
T.; Gill, C.; Silver, L. L.; Hermes, J.; Bartizal, K.; Barrett,
J.; Schmatz, D.; Becker, J. W.; Cully, D.; Singh, S. B.
Nature 2006, 441, 358–361.
2. Singh, S. B.; Jayasuriya, H.; Ondeyka, J. G.; Herath, K. B.;
Zhang, C.; Zink, D. L.; Tsou, N. N.; Ball, R. G.; Basilio,
A.; Genilloud, O.; Diez, M. T.; Vicente, F.; Pelaez, F.;
Young, K.; Wang, J. J. Am. Chem. Soc. 2006, 128, 11916–
1
1920, and 15547.
3
. Young, K.; Jayasuriya, H.; Ondeyka, J. G.; Herath, K.;
Zhang, C.; Kodali, S.; Galgoci, A.; Painter, R.; Brown
Driver, V.; Yamamoto, R.; Silver, L. L.; Zheng, Y.;
Ventura, J. I.; Sigmund, J.; Ha, S.; Basilio, A.; Vicente,
F.; Tormo, J. R.; Pelaez, F.; Youngman, P.; Cully, D.;
Barrett, J. F.; Schmatz, D.; Singh, S. B.; Wang, J.
Antimicrob. Agents Chemother. 2006, 50, 519–526.
OH
OH
O
R
O
R
HO C
HO C
2
N
H
2
N
H
4
5
. Singh, S. B.; Phillips, J. W.; Wang, J. Curr. Opin. Drug
Disc. Dev. 2007, 10, 160–166.
OH
OH
O
O
. Wang, J.; Kodali, S.; Lee, S. H.; Galgoci, A.; Painter, R.;
Dorso, K.; Racine, F.; Motyl, M.; Hernandez, L.; Tinney,
E.; Colletti, S.; Herath, K.; Cummings, R.; Salazar, O.;
Gonzalez, I.; Basilio, A.; Vicente, F.; Genilloud, O.; Pelaez,
F.; Jayasuriya, H.; Young, K.; Cully, D.; Singh, S. B. Proc.
Natl. Acad. Sci. U.S.A. 2007, 104, 7612–7616.
. Jayasuriya, H.; Herath, K. B.; Zhang, C.; Zink, D. L.;
Basilio, A.; Genilloud, O.; Diez, M. T.; Vicente, F.;
Gonzalez, I.; Salazar, O.; Pelaez, F.; Cummings, R.; Ha,
S.; Wang, J.; Singh, S. B. Angew. Chem., Int. Ed. 2007, 46,
1
5a:R = β-OH
16a:R = β-OH
16b:R = α-OH
1
5b:R = α-OH
O
O
O
O
O
O
OH
O
O
6
7
HO C
2
NH2
OH
O
O
O
O
1
7
18
19
20
21
4
684–4688.
1
. All compounds were characterized by H (500 MHz) and
1
3
C (125 MHz) NMR and ESIMS analyses. Key assign-
ments of selected compounds are listed except for com-
pound 5 where full assignments are listed. Compound 4b:
Reaction of 1 with 15 equiv of sodium borohydride in
MeOH yielded a mixture of di- and tetra-hydro deriva-
tives 15a, 15b, 16a, and 16b in a ratio of 50:5.8:21:22
0
(
(
4
C
5
D
5
N) d
H
7.81 (1H, s, new H at original C-2 ); d
C
150.4
0
0
0
0
0
C-3 ,5 ), 133.1 (C-2 ), 127.6 (C-2 ), 120.1 (C-6 ), 103.5 (C-
). Compound 5: (CDCl –CD OD, 3:1) d 7.54 (1H, d,
J = 8.5 Hz, H-7 ), 6.41 (1H, d, J = 8.5 Hz, H-6 ), 4.59 (1H,
s, H-6), 4.35 (1H, br s, H-10), 2.76 (1H, ddd, J = 14, 11.5,
4 Hz, H-2), 2.53 (1H, ddd, J = 14, 11.5, 5.5 Hz, H-2), 2.34
(1H, t, J = 6.5 Hz, H-12), 2.28 (1H, br s, H-9), 2.12 (1H, dd,
J = 12, 3.5 Hz, H-13), 2.08 (1H, d, J = 12 Hz, H-14), 2.04
(
HPLC), respectively, in over 90% combined yield be-
0
3 3 H
fore HPLC purification from each other. Acid workup
followed by concentration of the reaction mixture to
dryness lead to rapid conversion to 3-amino-2,4-dihydr-
oxybenzoic acid 17 and corresponding lactones 18–21.
The lactone formation was instantaneous when methan-
olic solution of amides was treated with 1 N HCl. In fact
compounds 15b, 16a, and 16b could not be isolated until
the reaction mixture was quenched with sodium phos-
phate buffer (pH 7) and purified as a sodium salt by
RP HPLC at pH 7 using the phosphate buffer. While
the sodium salt of 15a was stable the sodium salts of
0
0
(
(
1H, ddd, J = 14, 11.5, 5 Hz, H-3), 1.96 (1H, m, H-11), 1.94
1H, m, H-13), 1.92 (1H, m, H-11), 1.86 (1H, d, J = 12 Hz,
H-14), 1.68 (1H, ddd, J = 14, 11.5, 5 Hz, H-3), 1.37 (3H, s,
H-17), 1.31 (3H, s, H-18) d 173.5 (C-1), 30.1 (C-2), 33.2
C-3), 36.6 (C-4), 170.4 (C-5), 84.3 (C-6), 174.6 (C-7), 48.3
C
(
(
(
C-8), 46.7 (C-9), 75.7 (C-10), 39.6 (C-11), 43.9 (C-12), 45.5
C-13), 51.2 (C-14), 86.8 (C-15), 22.2 (C-17), 19.5 (C-18),
0
0
0
0
1
5b, 16a, and 16b were not as stable and the purified
172.0 (C-1 ), 104.8 (C-2 ), 155.0 (C-3 ), 113.4 (C-4 ), 155.0
0
0
0
samples were always contaminated with the correspond-
ing lactones after concentration with a ratio reaching as
high as 1:1.
(C-5 ), 109.8 (C-6 ), 128.4 (C-7 ); ESIMS (m/z) 488 [MꢀH],
512 [M+Na], 490 [M+H]. Compound 11: (C
D
N) d
7.95
5
5
H
0
0
(
(
1H, d, J = 9 Hz, H-7 ), 6.83 (1H, d, J = 9 Hz, H-6 ), 4.45
1H, br s, H-10), 4.10, 3.81, 3.76 (3H each, s). Compound
0
1
2: (C D N) d 8.13 (1H, d, J = 9 Hz, H-7 ), 6.95 (1H, d,
5
5
H
In summary, we have described herein selected func-
tional-group chemistries of platensimycin. It is notewor-
thy that the reduction of the enone olefin or ketone leads
to such a facile formation of the cyclic products 9 and
0
J = 9 Hz, H-6 ), 4.59 (1H, dd, J = 4.5, 3 Hz, H-6), 4.54 (1H,
br t, J = 3.5 Hz, H-10), 3.96, 3.82, 3.74 (3H each, s).
0
Compound 13: (C
6
D
5 5
N) d
H
8.09 (1H, d, J = 9 Hz, H-7 ),
0
.87 (1H, d, J = 9 Hz, H-6 ), 4.69 (1H, t, J = 4 Hz, H-6),
1
8–21. Remarkably nature has protected this antibiotic
4
.56 (1H, br s, H-10), 4.03, 3.84, 3.70 (3H each, s).
1
by introduction of this enone olefin. None of these com-
pounds showed better activity than platensimycin. Bio-
logical activity and SAR will be reported elsewhere.
Compound 14: H (C D N) d 7.92 (1H, d, J = 9 Hz, H-
7 ), 6.72 (1H, d, J = 9 Hz, H-6 ), 4.81 (1H, t, J = 4 Hz, H-
5
5
H
0
0
6), 4.54 (1H, br s, H-10), 3.77, 3.74 (3H each, s).

S. B. Singh et al. / Tetrahedron Letters 48 (2007) 5429–5433
5433
8
. Compound 9 C24
a = 16.4900(8), c = 13.4962(13) A, V = 3178.2(4) A , Z =
H
27NO
6
, M
r
= 425.470, hexagonal, P6
1
,
Germany). The final model was refined using 285 para-
meters and all 6438 data. All non-hydrogen atoms were
refined with anisotropic thermal displacements. The final
agreement statistics are: R = 0.043 (based on 5308 reflec-
tions with I > 2r(I)), wR = 0.095, S = 1.02 with (d/
r)max < 0.01. The maximum peak height in a final difference
˚
˚
3
ꢀ
3
6
,
D = 1.334 g cm
x
,
monochromatized
radiation
˚
ꢀ1
k(Mo) = 0.71073 A, l = 0.10 mm , F(000) = 1356, T =
1
00 K. Data were collected on a Bruker CCD diffractom-
eter to a h limit of 30.48°, which yielded 51,312 reflections.
There are 6438 unique reflections with 5308 observed at the
˚
ꢀ3
Fourier map is 0.305 e A
and this peak is without
2
r level. The structure was solved by direct methods
chemical significance. CCDC 650857 contains the supple-
mentary crystallographic data for this Letter. These data
can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
(
SHELXS-97, Sheldrick, G. M. Acta Crystallogr. 1990, A46,
2
4
(
67–473) and refined using full-matrix least-squares on F
SHELXL-97, Sheldrick, G. M. SHELXL-97. Program for the
Refinement of Crystal Structures. Univ. of G o¨ ttingen,