Termination of the structural confusion between plipastatin A1 and fengycin IX

Article abstract of DOI:10.1016/j.bmc.2012.04.040

Plipastatin A1 and fengycin IX were experimentally proven to be identical compounds, while these had been considered as diastereomers due to the permutation of the enantiomeric pair of Tyr in most papers. The ¹H NMR spectrum changed to become quite similar to that of plipastatin A1, when the sample which provided resembled spectrum of fengycin IX was treated with KOAc followed by LH-20 gel filtration. Our structural investigations disclosed that the structures of these molecules should be settled into that of plipastatin A1 by Umezawa (l-Tyr4 and d-Tyr10).

Full text of DOI:10.1016/j.bmc.2012.04.040

Bioorganic & Medicinal Chemistry 20 (2012) 3793–3798
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Termination of the structural confusion between plipastatin A1 and fengycin IX
Miho Honma a, Kazuaki Tanaka a, Katsuhiro Konno b, Kenji Tsuge c, Toshikatsu Okuno a,d
,
Masaru Hashimoto a,
⇑
a Department of Agriculture and Bioscience, Hirosaki University, 3-Bunkyo-cho, Hirosaki 036-8561, Japan
b Institute of Natural Medicine, University of Toyama, 2630 Sugitani, Toyama 930-0194, Japan
c Institute of Advanced Bioscience, Keio University, 403-1 Nipponkoku, Daihouji, Tsuruoka, Yamagata 997-0035, Japan
d University of Akita Nursing and Welfare, 2-3-4 Shimizu, Odate 017-0046, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Plipastatin A1 and fengycin IX were experimentally proven to be identical compounds, while these had
been considered as diastereomers due to the permutation of the enantiomeric pair of Tyr in most papers.
The 1H NMR spectrum changed to become quite similar to that of plipastatin A1, when the sample which
provided resembled spectrum of fengycin IX was treated with KOAc followed by LH-20 gel filtration. Our
structural investigations disclosed that the structures of these molecules should be settled into that of
Received 28 March 2012
Revised 19 April 2012
Accepted 20 April 2012
Available online 27 April 2012
plipastatin A1 by Umezawa (L-Tyr4 and D-Tyr10).
Keywords:
Ó 2012 Elsevier Ltd. All rights reserved.
Plipastatin A1
Fengycin IX
Structural confusion
Stereochemistry
Determination
1. Introduction
Budzikiewicz’s structure as fengycin IX, considerable number of re-
cent reports mentioned that these are identical compounds15,20–22
supposedly in order to avoid the contradiction between the struc-
ture and the biogenesis. However, there was no experimental evi-
dence. The confusion has become more serious, because at least
five other structures exist as fengycin IX (some structures did not
match up to their own discussions).8–14,16,23–25 The present studies
experimentally proved that plipastatin A1 is the K+ salt whereas
fengycin IX is the free form or the TFA salts. Although these mole-
cules gave considerably different 1H NMR spectra, the quite similar
1H NMR spectrum to that of fengycin IX was changed to provide ni-
cely accorded spectrum to that of plipastatin A1 when the sample
was converted into the K+ salt. Our structural studies led a conclu-
sion that the structures of these compounds should be settled into
that of plipastatin A1 by Umezawa.
Plipastatin A1 is a macrocyclic depsilipopeptide which was first
isolated by Umezawa et al. in 1986 from Bacillus cereus BMG3O2-
fF67 as a phospholipase A2 inhibitor.1–4 In the same issue of the
journal, Jung et al. independently reported fengycin from Bacillus
subtilis strain F-29-3 as an antibiotic lipopeptide.5 Although the fi-
nal structure was not disclosed in that report, they obtained the
identical amino acids and the fatty acid to those from plipastatin
A1 by the acid hydrolysis. The structure of fengycin was revealed
in 1999 by Budzikiewicz, that it is a diastereomer of plipastatin
A1 (named fengycin IX) possessing
while plipastatin A1 has the reversed permutation,
-Tyr10.8 Due to the promising antibiotic activity of this substance
D-Tyr4 and
L
-Tyr10 residues,6
-Tyr4 and
L
D
and empirical safeness of the producer Bacillus subtilis, fengycin is
expected to be a novel biocontrol7 to produce more than 220
scientific papers and counting. Some of those papers dealt with
the biogenesis of fengycins8–17 to disclose NRP (non-ribosormal-
peptide) synthase clusters. Those involved sequences suggesting
racemases. However, positions of the expected racemases accorded
with plipastatin A1 in spite of fengycin producers.18,19 The enzyme
cluster which rationally explains Budzikiewicz’s fengycin IX has
not so far been reported. These facts have brought structural con-
fusion between these molecules. Although many papers followed
2. Results and discussions
Bacillus subtilis H336B was found to produce an antifungal cyclic
peptide 1 which showed the same molecular weight 1462 as both
fengycin IX and plipastatin A1 by ESIMS [1463.8019, (calcd for
1463.8038, [M+H]+ C72H112N12O20)]. Isolation was performed by a
series of conventional chromatographies (XAD-7, ODS MPLC, and
ODS HPLC). Analyses involving Maefey’s configurational determi-
nation26 after acidic hydrolyses disclosed
-Tyr, -allo-Thr, -Ala, -Pro, -Ile, and 3-hydroxyhexadecanoic
acid (see Figure 1). Although our amino acid analyses were not
L-Glx (Â3),
L-Orn, L-Tyr,
D
D
D
L
L
⇑
Corresponding author. Tel./fax: +81 172 39 3782.
E-mail address: hmasaru@cc.hirosaki-u.ac.jp (M. Hashimoto).
0968-0896/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2012.04.040

3794
M. Honma et al. / Bioorg. Med. Chem. 20 (2012) 3793–3798
quantitative enough, for example due to partial O-derivatizations
of Tyr residue with DNS [5-(dimethylamino)naphthalene-1-sulfo-
nyl] chromophore27 and DAA [2,4-dinitrophenyl-5-
L-alanine
amide] chiral auxiliary group,26 the three Glx substructures could
be disclosed by 1H and 13C NMR spectra (see Supplementary data).
Suggested molecular formula led us to assign those to be one glu-
tamine and two glutamic acids. These results consisted well with
both fengycin IX and plipastatin A1. The NMR spectral data of 1 ac-
corded well with those of Budzikiewicz’s fengycin IX both in
CD3OH and DMSO-d6,28 when acidic conditions (H2O–CH3CN, in
the presence of TFA) were employed in the final ODS-HPLC purifi-
cation. These conditions give samples as free acid forms or TFA
salts. However, these spectra obviously disaccorded with those of
plipastatin A1.4,29 Since Umezawa isolated plipastatins as the K+
salts, 1 was converted into the K+ salt by treating with aqueous
KOAc followed by LH-20 gel filtration. Interestingly, these treat-
ments dramatically changed the 1H NMR profile in DMSO-d6 as
shown in Figure 2, for example, the
a-proton signal in Ala7 ap-
Figure 1. Structure of 1.
peared at 5.02 ppm in the K+ salt, while it resonated at 4.58 ppm
before the treatments. These signals could be assigned distinctly
by observing COSY correlations with doublet methyl signals (1.11
and 1.25 ppm, respectively). The spectrum after the treatments be-
came to accord well to that of plipastatin A1 in the Umezawa’s re-
port (see Supplementary data). No structural change occurred
during these operations, which was confirmed by the LCMS reten-
tion time and the MS spectral profile. These results clearly indi-
cated that plipastatin A1 and fengycin IX possess identical
framework but they are different forms, the free acid form or the
TFA salt for fengycin IX and the K+ salt form for plipastatinn
A1.30 These findings grossly contradicted the precedent knowledge
that plipastatin A1 and fengycin IX were diastereomers due to the
pair of enantiomeric Tyr residues (Tyr4 and Tyr10). In other words,
the above results suggested that either (in some case, both) struc-
ture of these molecules is wrong. In order to resolve the above con-
fliction, we determined the full structure of 1 independently.
A mono-cyclic structure was suggested based on the degree of
unsaturation as well as the amino acid residues involved. Basic
methanolysis/hydrolysis took place smoothly under mild condi-
tions to give the methanol adduct 2 (m/z 1495.8350) and hydroly-
sate
3 (m/z 1481.8117), respectively, disclosing that 1 is
depsipeptide. Peptide bonding should not be cleaved under these
conditions. Acyclic molecule results only fragmentation by solvol-
ysis. A doublet signal due to the ortho-position of the phenol ring
on one Tyr residue appeared at considerably higher frequency
(7.21 ppm) than that of corresponding protons on the other Tyr
(6.75 ppm) in 1. However, the hydrolysis neutered it to display
both of them at around 6.73 ppm. These observations revealed that
the phenolic hydroxy group in one Tyr is assignable to the lactone
ring. The carboxy part of the lactone was determined to be Ile11 by
Figure 2. 1H NMR spectra of 1 in DMSO-d6 (500 MHz, upper: K+ salt, lower: free acid form).
Figure 3. Structures and observed CID-MS/MS fragment ions of the methanol adduct 2 and those of hydrolysate 3 (in brackets).

M. Honma et al. / Bioorg. Med. Chem. 20 (2012) 3793–3798
3795
mental evidence. Umezawa had concluded the same configuration
by measuring the specific rotation value of methyl 3-hydroxyhex-
adecanoate ([
pastatin A1. However much smaller value ([
a
]
À35) obtained by the acidic methanolysis of pli-
D
a]
À6.0) was also
D
reported for this molecule in the recent synthetic paper.32 This sug-
gested low accuracy for our particular case because only small
quantity of the sample was available from the natural product.
We investigated that with circular dichroism (CD) by taking advan-
tage of its high sensitivity. After acidic methanolysis, the resulting
methyl 3-hydroxyhexadecanoate was reduced with LiAlH4 to give
hexadecane-1,3-diol as shown in Scheme 1. With expecting large
D
e
values,33,34 we introduced 2-naphthoyl group as the chromo-
phore to give bis-1,3-O-(2-naphthoyl) derivative nat-4. In spite of
small amount (870 g based on UV absorption), this sample
l
showed a typical negative split Cotton effect at 230 nm with reli-
able quality. The authentic sample (R)-4 was prepared by conven-
tional transformations from (R)-pentadecen-1-2-oxide which was
readily provided by asymmetric dihydroxylation of 1-pentadecene
with AD-mixb followed by a selective 1-sulfonylation and an epox-
ide formation under basic conditions. The 1H NMR of (R)-4 was
identical to that of nat-4. Comparison of the CD profiles led us to
undoubtedly establish the (R)-configuration for Acyl1 moiety.
These results corroborated both Umezawa’s plipastatin A1 and
Scheme 1. Determination of the configuration for the Acyl1 moiety.
Budzikiewicz’s fengycin IX. The permutation of D-Tyr and L-Tyr
was only the argument which remained to be determined. In other
words, either structure of these molecules is identical to 1, and the
other is incorrect in the order of
D
- and
L-Tyr residues. Umezawa
concluded -Tyr10 by observing O-DNP-
D
D-Tyr (DNP = 2,4-dinitro-
phenyl group) by the acidic hydrolysis after introduction of DNP
group into plipastatin A1. However, they also observed Gln from
the hydrolysate 3 by carboxypeptidase Y digestion. Since this en-
zyme is well established to cleave off
-Tyr10 should terminate the digestion before releasing
Budzikiewicz reached the other conclusion by detecting only
-Tyr by the acidic hydrolysis after CrO3 oxidation of fengycin IX.
Since these experiments involved ambiguity as described, we
investigated that with a certain methodology.
L
-amino acids selectively,
D
L-Gln9.
D
Scheme 2. Determination of the configurations for Tyr4 and tyr10 residues.
As shown in Scheme 2, 10-Tyr was labeled with DNS group. This
condition also furnished DNS on the d-amine group of Orn3, afford-
ing the bis-DNS derivative giving single LCMS signal (m/z 1929.9).
Subsequent acidic hydrolysis librated L-Tyr and D-(O-DNS)-Tyr.
Their configurations were determined by LCMS after Marfey’s
DAA derivatization. Since the phenolic OH in Tyr was also reacted
identifying Ile-OMe substructure in methanolysis product 2 in the
ESI-CID-MS/MS spectrum as shown in Figure 3. The CID of 2 (target
ion: m/z 1495.8) provided a fragment ion at 309.2 which was
assigned as y2 ion ([Tyr-Ile-OMe+H]+). The corresponding ion was
observed at 295.2 when hydrolysate 3 was employed (target ion:
m/z 1481.8). On the other hand 2 and 3 gave the common b9 and
under the conditions, a mixture of DAA-(O-DAA)-
and DAA-(O-DNS)- -Tyr was obtained by these transformations.
Authentic DAA- -Tyr, DAA-(O-DNS)- -Tyr, and DAA-(O-DAA)- -Tyr
as well as the corresponding -Tyr derivatives were prepared from
N-Boc- -Tyr and N-Boc- -Tyr, respectively. The LCMS ion chro-
matograms choosing m/z 434.1, 686.2, and 667.2 detected DAA-
Tyr, DAA-(O-DAA)- -Tyr, and DAA-(O-DNS)- -Tyr, respectively, as
L-Tyr, DAA-L-Tyr,
b10 ions at m/z 1187.1 and 1351.8, respectively. These also indicate
D
Tyr-Ile-OMe substructure in 2.
L
L
L
Amino acid sequence was determined also with ESI-CID-MS/MS
analyses. In the similar manner as above, comparison of the MS/MS
spectra between 2 and 3 undoubtedly distinguished the C-termi-
nus and N-terminus ions. Although b8 and y3 ions were missing
in these spectra, these were assigned to be a dipeptide unit Pro
and Gln (total residue mass = 225.1 u) based on the mass differ-
D
L
D
L-
L
D
shown in Table 1. Standing the reaction mixture without neutral-
ization allowed a gradual isomerization of the amino acid moie-
ties26 to provide the corresponding diastereomers. These were
also helpful in LCMS peak assignments. These analyses distinctly
ences between y9 and y7 as well as between y4 and y2 (both
Dm/
z = 225.1 u). The order of these two amino acid residues was deter-
mined to be Pro8-Gln9 based on the oxazolidine intermediate the-
ory in the ESI MS/MS which suggests that CID cleavages hardly
occur at the C-side of proline residues.31 These analyses revealed
the sequence Acyl1-Glu2-Orn3-Tyr4-allo-Thr5-Glu6-Ala7-Pro8-
Gln9-Tyr10-Ile11. As described above, one of the two phenolic
groups should form the lactone ring. We determined that Tyr4
should be the responsible residue. Lactone ring with Tyr10 might
logically be possible, but it bends the tyrosine aromatic ring to
bring extraordinary strain.
disclosed L- and D-configurations for Tyr4 and Tyr10, respectively.
As described, the full structure of 1 was thus established as shown
in Figure 1 which consists with Umezawa’s plipastatin A1 but not
with Budzikiewicz’s fengycin IX.
3. Conclusion
We succeeded in disclosing that fengycin IX and plipastatin A1
are identical compounds although these had been considered as
diastereomers at the two Tyr residues. Although reported their
NMR spectra showed discordance, it could be explained by their
Configuration of Acyl1 moiety was next determined.
Budzikiewicz determined R-configuration for this part by analogy
with other bacterial lipopeptides, there was however no experi-

3796
M. Honma et al. / Bioorg. Med. Chem. 20 (2012) 3793–3798
Table 1
Comparison of LCMS retention times of DAA-Tyr derivatives
DAA-amino acids
[M+H]+ (m/z)
Retention time (min)
Judged configuration
Authentic
L
-isomer
Authentic
D-isomer
Signals from 1
DAA-Tyr
DAA-(O-DAA)-Tyr
DAA-(O-DNS)-Tyr
434.1
686.2
667.2
20.1
38.9
47.9
22.3
44.1
50.3
20.4
39.7
50.3
L
L
D
Conditions: Inertsil ODS 2.1 mm  100 mm, 20–100% CH3CN/H2O for 80 min (containing 0.1% HCOOH) 0.20 mL/min flow.
forms; plipastatin is the K+ salt, while fengycin is the free form or
the TFA salts. The present studies disclosed that structures of these
molecules should be settled into that of plipastatin A1 by Umeza-
wa. This structure showed more advantageous to rationally explain
fengycin biosynthesis than the precedent structure by Bud-
zikiewicz. The confusion was probably caused by the preoccupa-
tion that these two must be different molecules because of
disaccording NMR profiles. Many groups utilized MS/MS spectra
in fengycin/plipastatin identifications. Of course mass spectrome-
try has brought incredible progress in the last decade to enable
us determine molecules within a short period. However, instant
structural conclusions only by LC–MS/MS magnified the confusion
in this particular case. In other words, the present studies redis-
cover the importance of detailed structural elucidation based on
chemical degradations and derivatizations. Our results experimen-
tally terminate the structural ambiguity between these com-
pounds, and would somehow accelerates the fengycin/plipastatin
researches such as detailed lactonization mechanism in the biosyn-
thesis,12 application as a functional molecule,35 and detailed mech-
anistic studies of the biological activity.36
ESIMS (rel. int.%) m/z 1485.7915 (4, calcd for 1485.7857, [M+Na]+
C72H110N12O20Na), 1463.8019, (25, calcd for 1463.8038, [M+H]+
C72H112N12O20), 751.3769 (9.0, calcd for 751.3837 [M+H+K]2+
,
,
,
C
72H111N12O20K), 743.3934 (11, calcd for 739.3967 [M+H+Na]2+
C72H111N12O20Na), 732.4006 (100, calcd for 738.4005, [M+2H]2+
C72H111N12O20). The detailed NMR spectral data are shown in Sup-
plementary data.
4.2. Transformation into the K+ salt
Pure 1 (4.4 mg) was diluted with aqueous KOAc solution
(200 mg in 10 mL) and the resulting solution was lyophilized. After
the obtained powder was dissolved in a minimum amount of H2O
(ca. 300 lL), gel filtration was performed with LH-20 (25 mm I.
D. Â 300 mm) to give the K+ salt (ca 2.2 mg) after lyophilization.
No reaction occurred during these operations, which was verified
by LCMS.
4.3. Amino acid analysis
Budzikiewicz’s fengycin IX does not exist. The NMR spectrum of
this virtual diastereomer would show different profile from the
natural product, and it will be the final proof for this argument.
Synthesis of this isomer is undergoing in our laboratories.
Pure 1 (500 lg) was treated with 6.0 M HCl (500 lL) in a sealed
tube at 110 °C for 12 h. After water (3.0 mL) was added, the result-
ing solution was lyophilized. The residue was stirred with 5-
(dimethylamino)naphthalene-1-sulfonyl chloride (DNSCl, 600
lg,
2.22
(100
l
l
mol) in a mixture of 0.5 M NaHCO3 (100
L) at room temperature for 14 h. The resulting solution
l
L) and acetone
4. Experimental
was directly analyzed with LCMS [InertsilÒ ODS (2.1 mm
4.1. Fermentation and isolation
I.D. Â 100 mm, 30–100% H2O/CH3CN for 15 min containing 0.1%
HCOOH, 0.20 mL/min flow) equipped with
a UV detector
Bacillus subtilis H336B was isolated as a contaminant from a
fungal culture and was identified through BLAST search of Genbank
based on 16S rRNA gene sequence. It was deposited at the Japan
Collection of Microorganisms of Riken Bioresource Center as JCM
18293. The bacterium was cultured in a medium prepared from
glycerol (100 g), meat extract (20 g), polypeptone (20 g), yeast ex-
tract (40 g), NaCl (8.0 g), MgSO4Á7H2O (2.0 g), K2HPO4 (2.0 g),
CaCO3 (12.8 g), and H2O (4.0 L) under shaking conditions
(110 rpm). After 3 days, the culture broth was filtered in suction
through a Celite pad and the filtrate was concentrated under re-
duced pressure until the whole volume became 1.0 L. The resulting
aqueous suspension was loaded on an Amberlite XAD-7HP (70 mm
I.D. Â 700 mm) column and eluted with 30%, 50%, 90%, and 100%
aqueous methanol solution. The activity was recovered in 50%
and 90% aqueous methanol fractions to give a residue (231 mg)
after lyophilization. Then, the residue was charged on Waters
ODS Sep-PakÒ Vac 35 cc (10 g). After washing with 50% aqueous
methanol, the active substance was eluted with 90% aqueous
methanol. After concentration, the residue was further subjected
to medium pressured liquid chromatography with UltraPack
(ODS, Size B, Yamazen Co.) with 85% aqueous methanol. The active
fractions were combined and concentrated under reduced pressure
to give crude 1 (27 mg). Pure 1 (12 mg) was obtained by prepara-
tive ODS HPLC (Sunfire™ Prep C18 OBD™ 19 mm I.D. Â 150 mm,
45–75% CH3CN/H2O containing 0.1% TFA for 15 min, 10.0 mL/min
flow) followed by lyophilization. tR = 8.5 min (above conditions).
(350 nm). The following amino acid derivatives were found;
DNS-Glu (tR = 5.8 min, m/z 381.11), DNS-allo-Thr (tR = 5.8 min, m/
z
353.12), DNS-Tyr (tR = 7.5 min, m/z 415.13), DNS-Ala
(tR = 9.3 min, m/z 323.10), DNS-Pro (tR = 11.5 min, m/z 349.12),
DNS-Ile (tR = 13.4 min, m/z 345.15), (DNS)2-Orn (tR = 15.3 min, m/
z 599.20), and (DNS)2-Tyr (tR = 18.6 min, m/z 648.19).
4.4. Determinations of the configurations for constituting
amino acids
Acid degradation of 1 (500
manner as described above. The residue was stirred with (5-flu-
oro-2,4-dinitrophenyl)- -alanine amide ( -FDAA, 300 g) and NaH-
CO3 (21 mg) in a mixture of H2O (500 L) and acetone (200 L) at
lg) was performed in the similar
L
L
l
l
l
40 °C for 90 min. After cooling, the reaction mixture was directly
analyzed by LCMS with ODS column. The retention times and
MS-profiles were checked by authentic amino acids. The retention
times and configurational judgments were summarized in Supple-
mentary data.
4.5. Determination of configuration of 3-hydroxyhecadecanoate
acid (Acyl1) moiety
Natural 1 (14.1 mg) was heated in 10% HCl/MeOH (1.5 mL) at
60 °C for 12 h. The mixture was poured in H2O (15 mL) and

M. Honma et al. / Bioorg. Med. Chem. 20 (2012) 3793–3798
3797
extracted with AcOEt (15 mL). The organic solution was washed
with brine (15 mL), dried over MgSO4 and then concentrated under
reduced pressure. Methyl 3-hydroxyhexadecanoate (1.3 mg) was
isolated by preparative TLC (AcOEt/hexane = 20:80, Rf = 0.6). 1H
NMR (in CDCl3) d 0.88 (3H, t, 6.9 Hz), 1.25 (20H), 1.50–1.57 (2H,
m), 2.41 (1H, dd, J = 9.1, 16.4 Hz), 2.52 (1H, dd, J = 3.0, 16.4 Hz),
3.71 (3H, s), 4.00 (1H, m). After the methyl ester thus obtained
was dissolved in THF (1.0 mL), LiAlH4 (2.0 mg) was added at room
temperature and the mixture was stirred for 30 min. Methanol (2
drops) was added to destroy excess reagent and the resulting mix-
ture was further stirred for additional 30 min. The mixture was
poured in H2O (15 mL) and extracted with AcOEt (15 mL). The or-
ganic solution was washed with brine (15 mL), dried over MgSO4
and then concentrated under reduced pressure. After the residue
was dissolved in pyridine without purification, 2-naphthoyl chlo-
ride (5.0 mg) was added at 60 °C and the mixture was stirred for
2 h at the same temperature. The mixture was concentrated and
then diluted with AcOEt (10 mL). The resulting solution was
washed with H2O (10 mL) and brine (15 mL) successively, and then
dried over MgSO4. After concentration under reduced pressure,
(3H, t, J = 6.7 Hz), 1.23–1.35 (21H), 1.45 (3H, m) 3.43 (1H, dd,
J = 7.7, 11.0 Hz), 3.66 (1H, dd, J = 3.1, 11.0 Hz), 3.72 (1H, m), 13C
NMR (CDCl3) d 14.09, 22.67, 25.52, 29.34, 29.53, 29.57, 29.63,
29.64 (for three carbons), 29.67, 31.91, 33.20, 66.84, 72.32. The
optical purity was determined by converting its bisMTPA ester to
confirm that it was more than 95% ee. The diol (150 mg, 614
was stirred with mesitylenesulfonyl chloride (171 mg, 730
l
l
mol)
mol)
and pyridine (150 lL, 1.83 lmol) at room temperature for 12 h.
The mixture was poured in H2O and extracted with ether (Â3).
The ethereal solution was washed with brine, dried over MgSO4,
and then concentrated under reduced pressure. Silica gel column
chromatography of the residue gave (S)-2-hydroxytetradecyl
mesitylenesulfonate as an oil (135 mg, 51%), (R)-1-hydroxypentad-
ecan-2-yl mesitylenesulfonate (21.0 mg, 8%), and the recovered
diol (35.0 mg, 23%). 1H NMR for 1-sulfonate (CDCl3) d 0.88 (3H, t,
J = 7.0 Hz), 1.25 (20H) 1.28 (1H, m), 1.42 (3H, m), 2.06 (1H, br d,
J = 4.2 Hz), 2.32 (3H, s), 2.04 (6H, s), 3.82 (1H, dd, J = 7.2, 9.5 Hz),
3.85 (1H, m), 3.99 (1H, dd, J = 2.2, 9.5 Hz), 6.98 (2H, s). The obtained
1-sulfonate (135 mg, 316 lmol) was stirred with K2CO3 (50 mg) in
methanol (2.0 mL) at room temperature for 2 h. The suspension
was poured in H2O and extracted with ether (Â3). The ethereal
solution was washed with brine, dried over MgSO4, and then con-
centrated under reduced pressure. Silica gel column chromatogra-
phy of the residue gave (S)-2-pentadecene oxide (70 mg, 98%). 1H
NMR (CDCl3) d 0.88 (3H, t, J = 7.0 Hz), 1.26 (20H), 1.44 (2H, m),
1.53 (2H, m), 2.46 (1H, dd, J = 2.8, 5.0 Hz), 2.74 (1H, dd, J = 4.0,
HPLC purification [Capcell Pak C8 UG120 (5 lm), 4.6 mm
I.D. Â 250 mm, CH3CN/H2O, 50:50–100:0 for 15 min, then keeping
100:0 for 15 min (containing 0.1% TFA), 1.0 mL/min flow] afforded
4 at tR = 21.7 min. Co-injection with synthetic authentic sample
confirm the HPLC peak. The yield was estimated to be ca. 870
lg
based on the UV absorbance at 262 nm (
e
124500). 1H NMR (CDCl3)
d 0.80 (3H, t, J = 6.9 Hz), 1.15 (16H, m), 1.28, 1.39 (each 2H, m),
1.69, 1.77 (each 1H, m), 2.22 (2H, br q, J = 6.5 Hz), 4.48 (2H, m),
5.39 (1H, br quint, J = 6.5 Hz), 7.36–7.52 (4H, m), 7.66–7.83 (6H,
m), 7.92, 7.97 (each br dd. J = 1.6, 8.6 Hz), 8.45, 8.50 (each br s).
The 1H NMR spectrum was identical to that of synthetic authentic
sample. ESIMS m/z 567.36 [M+H]+.
5.0 Hz), 2.90 (1H, m). A solution of the epoxide (70 mg, 310
lmol)
was stirred with KCNe (70 mg, 310 mol) in MeOH (2.0 mL) at
l
40 °C for 16 h. The mixture was poured in H2O (100 mL) and ex-
tracted with ether (80 mL Â 3). The ethereal solution was washed
with brine, dried over MgSO4, and then concentrated under re-
duced pressure. Silica gel column chromatography of the residue
gave (S)-3-hydroxyhexadecanenitrile as plates (65 mg, 83%). Mp
ca. 30 °C (from cold hexane), 1H NMR (CDCl3) d 0.88 (3H, t,
J = 6.8 Hz), 1.24–1.34 (21H), 1.44 (1H, m), 1.59 (2H, m), 1.94 (1H,
br d, J = 5.0 Hz), 2.49 (1H, dd, J = 6.4, 16.7 Hz), 2.57 (1H, dd,
J = 4.6, 16.7 Hz), 3.95 (1H, m). The obtained nitrile (65 mg,
4.6. Determination of the configurations for Tyr4 and Tyr10
A solution of 1 (1.0 mg) was stirred with dansyl chloride
(1.8 mg) and triethylamine (20 lL) in acetonitrile (500 lL) at room
temperature for 3 h. The reaction was monitored with LCMS. After
3 h, the starting 1 (m/z 1463) totally disappeared and a signal (m/z
1929.91, suggesting bis-DNS product) was newly observed in the
LCMS. After concentration, the residue was dissolved in 6 M aque-
237 lmol) was stirred in 12 M HCl (5.0 mL) at 75 °C for 30 min.
After cooling, the mixture was diluted with H2O (10 mL) and ex-
tracted with AcOEt (20 mL Â 3). The combined organic solution
was washed with brine, dried over MgSO4 and the concentrated
under reduced pressure to give the crude carboxylic acid
(65 mg). 1H NMR (CDCl3) d 0.88 (3H, t, J = 6.6 Hz), 1.2–1.6 (24H),
2.48 (1H, dd, J = 8.7, 16.5 Hz), 2.58 (1H, dd, J = 3.0, 16.5 Hz), 4.03
ous HCl solution (500
(3.0 mL) was added to the mixture and the resulting solution was
lyophilized. The residue was stirred with -FDAA (10 mg) and NaH-
CO3 (21 mg) in a mixture of H2O (500 L) and acetone (200 L) at
lL) and heated at 110 °C for 6 h. Water
L
(1H, m). After the crude carboxylic acid (24 mg, 88.1 lmol) was di-
l
l
luted with THF (2.0 mL), LiAlH4 (8.0 mg) was added and the mix-
ture was stirred at room temperature for 2 h. After the mixture
was cooled in an ice bath, methanol (three drops) was added to
decompose the excess reagent. The mixture was poured in 2 M
aqueous HCl solution and extracted with AcOEt (10 mL Â 3). The
combined organic solution was washed with brine (15 mL), dried
over MgSO4 and then concentrated under reduced pressure. Silica
gel column chromatography of the residue with AcOEt/hexane
(40:60) gave (R)-hexadecane-1,3-diol (21 mg, 92%). 1H NMR
(CDCl3) d 0.88 (3H, t, J = 6.9 Hz), 1.23–1.35 (22H), 1.48 (2H, m),
1.70 (2H, m), 2.26, 2.31 (each 1H, br), 3.86 (3H, m). The 1,3-diol
40 °C for 90 min. The reaction mixture was directly analyzed by
LCMS [Inertsil ODS 2.1 mm I.D. Â 100 mm, 20–100% CH3CN/H2O
for 80 min (containing 0.1% HCOOH) 0.20 mL/min flow]. The
DAA-
(m/z 343.13) and 50.3 min (m/z 667.18), respectively. The
corresponding diastereomers DAA- -Tyr-OH and DAA- -(O-DNS)-
L-Tyr-OH and DAA-D-(O-DNS)-Tyr-OH were detected at 20.4
D
L
Tyr-OH appeared at 22.3 and 47.9 min, those were established
employing authentic samples.
4.7. Preparation of syn-4
thus obtained (15 mg, 58
lmol) was stirred with 2-naphtoylchlo-
Pentadecene (231 mg, 1.10 lmol) was stirred with AD-mixb
ride (25 mg, 131 mol) in pyridine (1.0 mL) at 65 °C for 2 h. After
l
(2.0 g) in a mixture of H2O (5.0 mL) and 2-methyl-2-propanol
(5.0 mL) for 3 h at 0 °C. The mixture was poured in H2O (50 mL)
and extracted with ether (30 mL Â 3). The ethereal solution was
washed with brine, dried over MgSO4, and then concentrated un-
der reduced pressure. The residue was diluted with hexane
(3.0 mL) and stood at room temperature for 12 h to give (R)-pen-
concentration, silica gel column chromatography of the residue
(hexane/AcOEt = 7:93) gave the synthetic 4 (22.0 mg, 67%). UV
230.5 nm (
e
= 124,500, c = 1.02 Â 10À5 mol/L, CH3CN), 1H NMR
(CDCl3) d 0.87 (3H, t, J = 6.9 Hz), 1.20–1.50 (22H), 1.82 (2H, m),
2.28 (2H, m), 4.52 (1H, dt, J = 6.6, 11.2 Hz), 4.58 (1H, ddd, J = 4.1,
6.2, 11.2 Hz), 7.52 (4H), 7.76 (1H, br d, J = 8.7 Hz), 7.81 (4H), 7.88
(1H, br d, J = 8.7 Hz), 7.99 (1H, dd, J = 1.7, 8.7 Hz), 8.04 (1H, dd,
J = 1.7, 8.7 Hz), 8.52, 8.58 (each 1H, br s). APCIMS m/z 567.36.
tadecan-1,2-diol as plates (232 mg, 950
l
mol, 86%). Mp 63 °C
24
(from hexane), [a]
+6.3 (c 0.72, MeOH), 1H NMR (CDCl3) d 0.88
D

3798
M. Honma et al. / Bioorg. Med. Chem. 20 (2012) 3793–3798
7. Bie, X.; Lu, Z.; Lu, F. J. Microbiol. Methods 2009, 79, 272.
Acknowledgements
8. Wu, C.-Y.; Chen, C.-L.; Lee, Y.-H.; Cheng, Y.-C.; Wu, Y.-C.; Shu, H.-Y.; Götz, F.;
Liu, S.-T. J. Biol. Chem. 2007, 282, 5608.
The present work was supported in part by the Grants-in-Aid
for Scientific Research from the Japan Society for the Promotion
of Science (JSPS). The authors appreciated Emeritus Professor
Haruo Seto of University of Tokyo for his kind suggestions.
9. Samel, S. A.; Wagner, B.; Marahiel, M. A.; Essen, L.-O. J. Mol. Biol. 2006, 359, 876.
10. Lin, T.-P.; Chen, C.-L.; Fu, H.-C.; Wu, C.-Y.; Lin, G.-H.; Huang, S.-H.; Chang, L.-K.;
Liu, S.-T. Biochim. Biophys. Acta 2005, 1730, 159.
11. Walsh, C. T. Science 2004, 303, 1805.
12. Sieber, S. A.; Tao, J.; Walsh, C. T.; Marahiel, M. A. Angew. Chem. 2004, 43, 493.
13. Sieber, S. A.; Walsh, C. T.; Marahiel, M. A. J. Am. Chem. Soc. 2003, 125, 10862.
14. Steller, S.; Vater, J. J. Chromatogr., B 2000, 737, 267.
15. Steller, S.; Vollenbroich, D.; Leenders, F.; Stein, T.; Conrad, B.; Hofemeister, J.;
Jacques, P.; Thonar, P.; Vater, J. Chem. Biol. 1999, 6, 31.
Supplementary data
16. Lin, T.-P.; Chen, C.-L.; Chang, L.-K.; Tschen, J. S.-M.; Liu, S.-T. J. Biotechnol. 1999,
181, 5060.
17. Tosato, V.; Albertini, A. M.; Zotti, M.; Sonda, S.; Bruschi, C. V. Microbiology 1997,
143, 3443.
18. Tsuge, K.; Ano, T.; Hirai, M.; Nakamura, Y.; Shoda, M. Antimicrob. Agents
Chemother. 1999, 43, 2183.
19. Tsuge, K.; Matsui, K.; Itaya, M. J. Biotechnol. 2007, 129, 592.
20. Wang, J.; Liu, J.; Wang, X.; Yao, J.; Yu, Z. Lett. Appl. Microbiol. 2004, 39, 98.
21. Deleu, M.; Razafindralambo, H.; Popineau, Y.; Jacques, P.; Thonart, P.; Paquot,
M. Colloids Surf., A: Physicochem. Eng. Aspects 1999, 152, 3.
22. Hashizume, H.; Nishimura, Y. Stud. Nat. Prod. Chem. 2008, 35, 693.
23. Vater, J.; Kablitz, B.; Wilde, C.; Franke, P.; Mehta, N.; Cameotra, S. S. Appl.
Environ. Microbiol. 2002, 68, 6210.
Supplementary data (summary for structural confusion of
fengycin in papers, comparison of TIC and mass spectra between
1 and 3, NMR spectra (1H, 13C, DEPT-135, COSY, HMQC, HMBC
and HSQC-TOXSY) of 1, comparison of 1H and 13C NMR signal
assignments of 1 (free form) in CD3OH with those reported by
Budzikiewicz, comparison of the 1H NMR between 1 and 3 in
CD3OD, 1H NMR spectral comparison between 1 and plipastatin
by Umezawa in DMSO-d6, COSY spectrum of 1 (K+ salt, free acid)
in DMSO-d6, CID-MS/MS spectra and assignments of 2 and 3. 1H
NMR spectrum of 4, tables for the HPLC retention times in config-
urational determination of the amino acids, experimental detail for
the preparation of the syn-4) associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.bmc.2012.04.040.
24. Sieber, S. A.; Marrahiel, M. A. J. Biotechnol. 2003, 185, 7036.
25. Stein, T. Mol. Microbiol. 2005, 56, 845.
26. Bhushan, R.; Brückner, H. Amino Acids 2004, 27, 231.
27. Kinoshita, T.; Iinuma, F.; Tsuji, A. Chem. Pharm. Bull. 1974, 22, 2413.
28. Some 13C NMR resonances for
crowdings.
a-carbonyls could not assigned due to signal
29. Volpon, L.; Besson, F.; Lancelin, J.-M. FEBS Lett. 2000, 485, 76.
30. The present studies could not distinguish whether fengycin is the free form or
the TFA salt.
References and notes
31. Kinter, M.; Sharman, N. E. Protein Sequencing and Identification Using Tandem
Mass Spectrometry; John Wiley & Sons: New York, 2000.
32. Blanc, D.; Madec, J.; Popowyck, F.; Ayad, T.; Phansavath, P.; Virginie, R.-V.;
Genet, J.-P. Adv. Synth. Catal. 2007, 349, 943.
33. Harada, N.; Nakanishi, K. Acc. Chem. Res. 1972, 5, 257.
34. Akritopoulou-Zanze, I.; Nakanishi, K.; Stepowska, H.; Grrzeszczyk, B.; Zamojski,
A.; Berova, N. Chirality 1997, 9, 699.
35. Eeman, M.; Pegado, L.; Dufrêne, Y. F.; Paquot, M.; Deleu, M. J. Colloid Interface
Sci. 2009, 329, 253.
36. Ramarathnam, R.; Shen, B.; Yu, C.; Fernando, W. G. D.; Gao, X.; de Kievit, T. Can.
J. Microbiol. 2007, 53, 901.
1. Umezawa, H.; Aoyagi, T.; Nishikiori, T.; Okuyama, A.; Yamagishi, Y.; Hamada,
M.; Takeuchi, T. J. Antibiot. 1986, 39, 737.
2. Nishikiori, T.; Nagasawa, H.; Maruoka, Y.; Aoyagi, T.; Umezawa, H. J. Antibiot.
1986, 39, 755.
3. Nishikiori, T.; Nagasawa, H.; Maruoka, Y.; Aoyagi, T.; Umezawa, H. J. Antibiot.
1986, 39, 745.
4. Nishikiori, T.; Nagasawa, H.; Maruoka, Y.; Aoyagi, T.; Umezawa, H. J. Antibiot.
1986, 6, 860.
5. Vannittanakom, N.; Loeffler, W.; Koch, U.; Jung, G. J. Antibiot. 1986, 39, 888.
6. Schneider, J.; Taraz, K.; Budzikiewicz, H.; Deleu, M.; Thonart, P.; Jacques, P. Z.
Naturforsch. C 1999, 54, 859.