Chemical synthesis of ComX pheromone and related peptides containing isoprenoidal tryptophan residues

Article abstract of DOI:10.1016/j.tet.2006.06.074

The ComX pheromone is a post-translationally modified oligopeptide that stimulates natural genetic competence controlled by quorum sensing in Bacillus subtilis. Recently, the structure of the ComX_RO-E-2 pheromone produced by strain RO-E-2 was d

Full text of DOI:10.1016/j.tet.2006.06.074

Tetrahedron 62 (2006) 8907–8918
Chemical synthesis of ComX pheromone and related peptides
containing isoprenoidal tryptophan residues
Masahiro Okada,a, ,y Isao Sato,a Soo Jeong Cho,b David Dubnaub and Youji Sakagamia,
*
*
aGraduate School of Bioagricultural Sciences, Nagoya University, Chikusa, Nagoya 464-8601, Japan
bPublic Health Research Institute, 225 Warren Street, Newark, NJ 07103, USA
Received 18 May 2006; revised 17 June 2006; accepted 21 June 2006
Available online 28 July 2006
Abstract—The ComX pheromone is a post-translationally modified oligopeptide that stimulates natural genetic competence controlled by
quorum sensing in Bacillus subtilis. Recently, the structure of the ComXRO-E-2 pheromone produced by strain RO-E-2 was determined. Based
on the NMR analysis, a geranyl group is bound to the tryptophan residue, which results in the formation of a tricyclic ring structure. It was
proposed that one of the four possible stereochemical isomers was based on a conformational search for model compounds and the assumption
that amino acid residues in the natural pheromone have the L-configuration. All possible modified tryptophan residues and the corresponding
ComXRO-E-2 peptides were synthesized to confirm the precise stereochemistry. Here, the synthesis of the modified tryptophan derivatives was
reported in detail. It was succeeded in synthesizing four optically active modified tryptophan methyl esters from which the four diastereomeric
ComXRO-E-2 peptides were prepared. Since only one of the four diastereomers was spectroscopically identical to the natural pheromone and
exhibited biological activity, the absolute structure of the ComXRO-E-2 pheromone was able to be established unambiguously. Furthermore, it
was noticed that two other bioactive pheromones were present in the culture broth that were co-purified with ComXRO-E-2 pheromone. These
pheromones were presumed to be the N-terminal truncated peptides of ComXRO-E-2 pheromone, i.e., [2-6]ComXRO-E-2 and [3-6]ComXRO-E-2
by LC-MS and NMR analyses. Using Fmoc solid-phase peptide synthesis, ComXRO-E-2 pheromone and the [2-6]ComXRO-E-2 and
,
[3-6]ComXRO-E-2 peptides were prepared. The synthetic peptides were identical to the natural pheromones and also showed significant
biological activity.
Ó 2006 Elsevier Ltd. All rights reserved.
1. Introduction
regulation that depends on cell density is called ‘quorum
sensing’.1–4 The quorum sensing pheromones are oligo-
peptides in Gram-positive bacteria, while Gram-negative
pheromones are mainly N-acylhomoserine lactones. Bacillus
subtilis and related bacilli produce a unique pheromone that
stimulates natural genetic competence controlled by the
quorum sensing.5 This B. subtilis competent factor is a
post-translationally modified oligopeptide known as ComX
pheromone. The ComX pheromone is thought to activate
the signal transduction cascade for natural genetic compe-
tence by binding to the membrane-embedded histidine
autokinase ComP.6,7
Bacteria constitutively secrete specific pheromones with the
pheromone concentration increasing with cell density. When
these concentrations reach threshold levels, bacteria respond
to the pheromone by altering their gene expression. The
Keywords: Bacillus subtilis; ComX; Post-translational modification;
Quorum sensing; Tryptophan.
Abbreviations: Clt, 2-chlorotrityl; [Da]ComXRO-E-2, ComXRO-E-2 peptide
containing the modified D-tryptophan residue with an a-geranyl group;
[Db]ComXRO-E-2, ComXRO-E-2 peptide containing the modified D-trypto-
phan residue with a b-geranyl group; DIPEA, N,N0-diisopropylethylamine;
HATU, O-(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluoro-
phosphate; HBTU, O-benzotriazol-1-yl-N,N,N0,N0-tetramethyluronium
hexafluorophosphate; [La]ComXRO-E-2, ComXRO-E-2 peptide containing the
The amino acid sequence of the ComX pheromone varies
according to the B. subtilis strain, but each possesses a mod-
ified tryptophan residue.8–11 This modification increases the
hydrophobicity and molecular weight of the peptide encoded
by comX.5 ComX is biosynthesized as a pre-protein, which is
then processed and modified by ComQ to become the active
ComX pheromone.12 ComQ contains an isoprenyl transfer-
ase domain, and this putative isoprenoid-binding site in
ComQ is required for the expression of pheromone activity.13
Furthermore, [3H]-5-mevalonate, a precursor of isoprenoid,
is incorporated into ComX pheromone.10 These results
modified L-tryptophan residue with an a-geranyl group; [Lb]ComXRO-E-2
,
ComXRO-E-2 peptide containing the modified L-tryptophan residue with
a b-geranyl group; Pp, 2-phenyl-2-propyl; Su, succinimidyl; TAS-F, tris(di-
methylamino)sulfonium difluorotrimethylsilicate; Teoc, 2-(trimethylsilyl)-
ethoxycarbonyl.
* Corresponding authors. Tel.: +81 22 795 6555; fax: +81 22 795 6557;
e-mail addresses: okada-org@mail.tains.tohoku.ac.jp; ysaka@agr.
nagoya-u.ac.jp
y
Present address: Graduate School of Science, Tohoku University, Aoba,
Sendai 980-8578, Japan.
0040–4020/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2006.06.074

8908
M. Okada et al. / Tetrahedron 62 (2006) 8907–8918
indicate that ComX pheromone is activated by isoprenylation
from an inactive precursor. Isoprenylation of a tryptophan
residue during post-translational modification has not been
previously reported, although the modification of cysteine
residues in proteins by farnesyl or geranylgeranyl groups is
well known.14
We recently isolated the ComXRO-E-2 pheromone using an
Escherichia coli expression system. This required only
comQXP0RO-E-2 from B. subtilis strain RO-E-2, which was
driven by a T7 phage promoter to produce biologically ac-
tive ComXRO-E-2 pheromone.15 We determined the planar
structure of the tryptophan residue by NMR analysis, which
showed a tricyclic ring structure with a geranyl group at-
tached to the 3-position of tryptophan. A conformational
search was then carried out on the L-tryptophan derivative
model compound using a Monte Carlo protocol to elucidate
the stereochemistry of ComXRO-E-2 pheromone. This search
was based on the assumption that the pheromone was
composed only of L-amino acids. We found that only one
stereostructure could explain the NMR data. We thus pro-
posed the absolute structure of the ComXRO-E-2 pheromone
1 and its modified tryptophan residue to be [2S,3aR,8aS]-
3a-geranyl-1,2,3,3a,8,8a-hexahydropyrrolo[2,3-b]indole-
2-carboxylic acid [L-Trp*(a-Ger)] (Fig. 1). To confirm the
structure of the ComXRO-E-2 pheromone, synthesis of all
possible ComXRO-E-2 peptides and investigation of their
biological activities and physiological data were necessary.
There are four possible structures 2–5 for the modified
tryptophan residue (Fig. 2), i.e., [La]ComXRO-E-2 2 has the
modified L-tryptophan residue with an a-geranyl group
[L-Trp*(a-Ger)], [Lb]ComXRO-E-2 3 has the modified L-tryp-
tophan residue with a b-geranyl group [L-Trp*(b-Ger)],
[Da]ComXRO-E-2 4 has the modified D-tryptophan residue
with an a-geranyl group [D-Trp*(a-Ger)], and
[Db]ComXRO-E-2 5 has the modified D-tryptophan residue
with a b-geranyl group [D-Trp*(b-Ger)]. We therefore
synthesized four optically active geranyl tryptophan methyl
esters, and prepared four diastereomeric ComXRO-E-2 pep-
tides to confirm our proposed structure for the ComXRO-E-2
pheromone.
Figure 2. Four possible structures of the modified tryptophan residue in the
ComXRO-E-2 pheromone based on the NMR analysis.
2. Results
2.1. Synthesis of modified tryptophan and stereoisomers
For efficient synthesis of the four stereoisomeric geranyl-
modified tryptophan residues and the corresponding pep-
tides composed of the ComXRO-E-2 sequence, we selected
L- or D-tryptophan methyl esters as chiral starting materials
and geranyl bromide as the geranylating reagent. The opti-
cally active tricyclic tryptophan derivatives were synthe-
sized by cyclization of tryptophan methyl ester derivatives,
followed by geranylation at the C3 position. The resulting
C3 diastereomers were separated, and each isomer was con-
verted to the protected tryptophan derivatives 6–9 prior to
peptide synthesis (Scheme 1).
The cyclization of N-acetyl L-tryptophan methyl ester was
carried out with t-BuOCl (Table 1, entry 1).17,18 The geranyl
group was then introduced at the C3 position of the tricyclic
compound 10a with NaH and geranyl bromide in DMF
(Table 2, entry 1).18 Unfortunately, the acetyl deprotection
of the C3-geranylated diastereomers did not proceed under
the various conditions tried (data not shown). Thus, we
screened for removable protecting groups and conditions
for C3-geranylation. Using N-protected tryptophan deriva-
tives, we synthesized nine tricyclic compounds 10a–i that
were purified by recrystallization (Table 1).19 Using 10a–i,
we optimized the conditions for C3-geranylation (Table
2).20 The usual quenching method with methanol or water
resulted in product decomposition, and quenching with 5%
aqueous KHSO4 gave the C3-b-geranylated compound in
poor yield in addition to the N1-geranylated compound.
We found that immediate neutralization with a phosphate
buffer prevented the decomposition of the C3-geranylated
products. When carbamates were selected as the N-protect-
ing group, C-geranylation of 10b–e did not proceed at all
(Table 2). We also found that the ester was important for
the reaction because the geranylation of 2-chlorotrityl (Clt)
Two additional peptides were co-isolated with ComXRO-E-2
pheromone from the culture broth of E. coli ED413. These
were determined to be N-terminal truncated peptides of
ComXRO-E-2 pheromone. Using solid-phase peptide synthe-
sis as described previously,16 we synthesized these two pep-
tides in addition to ComXRO-E-2 pheromone and investigated
their biological activity.
Figure 1. The proposed structure of ComXRO-E-2 pheromone.

M. Okada et al. / Tetrahedron 62 (2006) 8907–8918
8909
Scheme 1. Synthetic strategy toward four ComXRO-E-2 peptides.
Table 1. Cyclization of various tryptophan derivatives
Table 2. C- or N-selective geranylation under various conditions
Entry
R1
R2
Yield, %a (product)
1
2
3
4
5
6
7
8
9
Ac
Boc
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CH2OClt
92 (10a)
42 (10b)
49 (10c)
78 (10d)
69 (10e)
85 (10f)
79 (10g)
95 (10h)
62 (10i)
Substrate (R1, R2)
Solvent
Yield, %a
C-Ger (a:b)b
CO2Bn
CO2Me
CO2Et
COCCl3
COCF3
Bz
N-Ger
10a (Ac, CO2Me)
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
NMP
THF
63 (33:30)
0
0
0
0
9
0
0
14
8
0
10b (Boc, CO2Me)
10c (CO2Bn, CO2Me)
10d (CO2Me, CO2Me)
10e (CO2Et, CO2Me)
10f (COCCl3, CO2Me)
10g (COCF3, CO2Me)
10h (Bz, CO2Me)
10i (Bz, CH2OClt)
10h (Bz, CO2Me)
10h (Bz, CO2Me)
Bz
0
0
a
Isolated yield with recrystallization.
0
81 (44:37)
0
77 (32:45)
83 (47:36)
15
72
18
11
ether 10i gave only the N-geranylated compound. The Bz-
protected substrate 10h gave a better yield of C3-geranylated
diastereomers when compared with 10a. The solvent for ger-
anylation also affected the yield and diastereomeric ratio of
the geranylated products. THF increased the yield slightly
and the ratio of the C3-a-geranyl product improved, but
the reverse was true for NMP. The key intermediate 11
was obtained in high yield using the optimized conditions.
Mixtures of the C3-geranyl isomers were acceptable for
the purpose of synthesizing the four diastereomers. Since
the C3-geranylated diastereomers could not be separated
by column chromatography, the diastereomixture 11 was
used directly as the starting material in the next step.
a
Isolated yield.
a/b ratios were determined by 1H NMR, and the stereochemistries were
determined by NOE.
b
determined the a-geranyl configuration of 12 based on
NOE analysis, howeverpurified 13 was not sufficiently stable
for direct determination of its stereochemistry. Because com-
pound 13 was readily prepared from the b-geranyl isomer of
11 by treatment with DIBAL-H in nearly quantitative yield,
and compound 13 was also determined to have the b-geranyl
configuration. It was necessary to protect the unstable sec-
ondary amines 12 and 13 with an Fmoc group before the re-
duction of the imine. After Fmoc protection, the reduction of
the imines 14 and 15 was carried out with catecholborane to
give amines 16 and 17, respectively. Fmoc was found to
be the best protecting group for the reduction. Bz-protected
11 required prolonged reduction with catecholborane and
was difficult to deprotect,20 and the free amine 12 was
Next, deprotection, separation of diastereomers, and chemo-
selective reduction of the imine were attempted (Scheme 2).
Treatment of 11 with DIBAL-H selectively cleaved the Bz
group. The reductive diastereomeric products 12 and 13
were easily separated by column chromatography using a
solvent system of chloroform and methanol to afford the
optically active tryptophan derivatives 12 and 13. We

8910
M. Okada et al. / Tetrahedron 62 (2006) 8907–8918
Scheme 2. Synthesis of L-Trp*(a-Ger)-OMe 6 and L-Trp*(b-Ger)-OMe 7. Reaction conditions: (a) DIBAL-H, THF, –78 ꢀC, 2 h; (b) Fmoc-OSu, 1.0 M aqueous
NaHCO3, dioxane, rt, 1.5 h; (c) catecholborane, THF, 0 ꢀC, 2 h; (d) piperidine, CH3CN, rt, 1.5 h.
decomposed by reduction. Cleavage of the Fmoc group gave
L-Trp*(a-Ger)-OMe 6 for the synthesis of [La]ComXRO-E-2
2, which has our proposed configuration for natural
ComXRO-E-2 pheromone. We confirmed that L-Trp*(a-
Ger)-OMe 6 had a cis configuration based on the NOE
between the H-2 proton and the geranyl group. We also syn-
thesized L-Trp*(b-Ger)-OMe 7 from intermediate 13 using
the same procedure (Scheme 2), and D-Trp*(a-Ger)-OMe
8 and D-Trp*(b-Ger)-OMe 9 from the corresponding
N-benzoyl-D-tryptophan methyl esters.
2.2. Synthesis of ComXRO-E-2 peptides
We synthesized peptides containing the same amino acid se-
quence as the ComXRO-E-2 pheromone using each modified
tryptophan methyl ester (Scheme 3). The ComXRO-E-2 pher-
omone is very acid labile.15,16 Therefore, we used Clt as a
protecting group for the carboxy groups because it can be
cleaved under mild acidic conditions. Methyl ester 6 was
treated with 2-(trimethylsilyl)ethoxycarbonylphenylalanine
(Teoc-Phe) to obtain dipeptide 18. From the analysis of the
HMBC spectrum, we found that L-Trp*(a-Ger)-OMe 6 reacts
with amino acids only at the Na-position under common
amino acid coupling conditions (data not shown). Dipeptide
methyl ester 18 was hydrolyzed under alkaline conditions,
followed by coupling with Glu(Clt)-Gln-OClt. Purification
of the resulting tetrapeptide was carried out by silica gel col-
umn chromatography with 1% triethylamine so as not to
cleave the Clt esters. After cleavage of the Teoc group of the
purified tetrapeptide, the peptide was coupled to Teoc-Gly-
Ile. Finally, the Clt esters were cleaved with 50% aqueous
AcOH, followed by removal of the Teoc group with tris(di-
methylamino)sulfonium difluorotrimethylsilicate (TAS-F).
The resultant compound was purified by HPLC to give the
desired [La]ComXRO-E-2 peptide 2. The other three dia-
stereomers, [Lb]ComXRO-E-2 peptide 3, [Da]ComXRO-E-2
peptide 4, and [Db]ComXRO-E-2 peptide 5, were synthesized
using L-Trp*(b-Ger)-OMe 7, D-Trp*(a-Ger)-OMe 8, and
D-Trp*(b-Ger)-OMe 9, respectively, as shown in Figure 3.
Scheme 3. Synthesis of [La]ComXRO-E-2 2. Reaction conditions: (a) Teoc-
Phe, HATU, HOAt, Et3N, CH2Cl2, 0 ꢀC, 2 h; (b) LiOH, THF, MeOH, H2O,
rt, 1 h; (c) Glu(Clt)-Gln-OClt, HATU, HOAt, Et3N, CH2Cl2, 0 ꢀC, 2 h; (d)
TAS-F, CH2Cl2, rt, 24 h; (e) Teoc-Gly-Ile, HATU, HOAt, Et3N, CH2Cl2,
0 ꢀC, 2 h; (f) 50% aqueous AcOH, 4 ꢀC, 24 h.
2.3. Comparison of synthetic ComXRO-E-2 peptides and
natural pheromone
1
The H NMR spectra of the four synthetic peptides were
compared (Fig. 4). [La]ComXRO-E-2 peptide 2 possessed
the proposed stereostructure and exhibited the same 1H
NMR spectrum as the natural peptide pheromone. The spec-
tra of the three other diastereomers 3–5 showed considerable
differences, particularly in the region of the modified trypto-
phan residues.
We then investigated the biological activities of these
peptides by b-galactosidase assay with o-nitrophenyl-b-D-

M. Okada et al. / Tetrahedron 62 (2006) 8907–8918
8911
A
40
20
a
bc
0
0
10
20
30
Retention time (min)
15
b
c
B
a
0.6
0.4
0.2
0
10
5
0
19
20
21
22
23
24
25
Retention time (min)
Figure 5. (a) HPLC of a ComXRO-E-2 solution partially purified. (b) The
effect of each fraction in bioassay and the corresponding HPLC. Gray
bars represent optical density at 420 nm.
Figure 3. Other three ComXRO-E-2 peptides.
galactopyranoside using a B. subtilis tester strain (BD3020)
as described previously.9,10 In addition to the natural
ComXRO-E-2 pheromone, only one of the four synthetic
peptides, [La]ComXRO-E-2 2, showed obvious biological
activity at 1.0 nM, while diastereomers 3–5 showed no
biological activity up to 300 nM (Fig. 5).
2.4. Identification and solid-phase synthesis of N-termi-
nal truncated natural pheromones
We analyzed the natural ComXRO-E-2 pheromone solution
from the culture broth using an E. coli expression system,
Natural ComXRO-E-2 pheromone 1
[Lα]ComXRO-E-2
[Lβ]ComXRO-E-2
2
3
[Dα]ComXRO-E-2
[Dβ]ComXRO-E-2
4
5
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5 ppm
Figure 4. 1H NMR spectra of the natural ComXRO-E-2 pheromone and the synthetic peptides. Each DHO signal of [La]ComXRO-E-2 2 or [Db]ComXRO-E-2 5 was
irradiated for suppression.

8912
M. Okada et al. / Tetrahedron 62 (2006) 8907–8918
Table 3. HRMS of the ComXRO-E-2 pheromone and its truncated peptides
Peak
HRMS [M+H]+
Molecular composition (D ppm)
Amino acid sequence
Pheromone
a
b
c
745.39152
915.49830
858.47591
C40H53N6O8 (ꢁ0.56)
C48H67N8O10 (+0.91)
C46H64N7O9 (ꢁ0.11)
FW*EQ
GIFW*EQ
IFW*EQ
[3-6]ComXRO-E-2
ComXRO-E-2
[2-6]ComXRO-E-2
W* represents the modified tryptophan residue with a geranyl group.
as reported previously.9,10,15 Following the final purification
step, we identified three biologically active fractions, includ-
ing the ComXRO-E-2 pheromone, as shown in Figure 5. Be-
cause the other two peptides in the active fractions showed
[M+H]+ molecular ions of 858.47591 and 745.39152 on
HRMS analysis, and these peptides were assigned molecular
formulas of C46H64N7O9 and C40H53N6O8, respectively
(Table 3). These results suggest that the two peptides were
N-terminal truncated peptides of ComXRO-E-2 pheromone,
ComXRO-E-2 pheromone and two N-terminal truncated
ComXRO-E-2 peptides, [3-6]ComXRO-E-2 peptide 20 and [2-
6]ComXRO-E-2 peptide 21, using solid-phase synthesis. We
first synthesized the desired Fmoc-protected modified trypto-
phan for Fmoc solid-phase peptide synthesis. L-Trp*(a-Ger)-
OMe 6 was hydrolyzed under alkaline conditions, followed
by Fmoc protection to give Fmoc-L-Trp*(a-Ger) 19 (Scheme
4). Peptide bond formation was accomplished with a peptide
synthesizer, although a previously reported manual synthesis
procedure was also used (Scheme 5).16 Both resin and pro-
tecting groups were smoothly cleaved under mild acidic con-
ditions by treatment with 5% TFA at 4 ꢀC for 20 h. The
desired peptides were obtained after HPLC purification and
i.e., [2-6]ComXRO-E-2 and [3-6]ComXRO-E-2
.
In order to confirm that these active compounds are N-termi-
nal truncated pheromones, we attempted to synthesize
1
analyzed by H NMR spectroscopy and HRMS. The two
synthetic N-terminal truncated peptides, [3-6]ComXRO-E-2
peptides 20 and [2-6]ComXRO-E-2 peptides 21, were found
to be identical to the natural truncated pheromones.
The biological activity of these peptides was investigated by
b-galactosidase assay (Fig. 6). The two peptides showed
strong activities, although these were approximately 10-
fold weaker than that of the ComXRO-E-2 (Fig. 7). The
dose–response curve of [2-6]ComXRO-E-2 (peptide 21) was
saturated at 100 nM with an EC50 value of 7 nM. [3-
6]ComXRO-E-2 (peptide 20) showed an EC50 value of 8 nM,
but its dose–response curve did not reach 100%, even at
1 mM.
Scheme 4. Synthesis of Fmoc-L-Trp*(a-Ger) 19. Reaction conditions: (a)
LiOH, THF, CH3OH, H2O, rt, 1.5 h; (b) Fmoc-OSu, dioxane, 1.0 M aqueous
NaHCO3, rt, 1.5 h.
Scheme 5. Solid-phase peptide synthesis of ComXRO-E-2 peptides. Reaction conditions: (a) Fmoc-L-Trp*(a-Ger) 19, HATU, HOAt, Et3N, CH2Cl2, 0 ꢀC, 2 h; (b)
Stepwise solid-phase peptide synthesis with peptide synthesizer using HBTU/HOBt/DIPEA chemistry; (c) 5% TFA, 5% trifluoroethanol, CH3CN, 4 ꢀC, 20 h.

M. Okada et al. / Tetrahedron 62 (2006) 8907–8918
8913
diastereomers for this synthesis, we did not investigate the
a/b selectivity in the C3-geranylation in this work, although
we observed that selectivity was affected by the choice of
solvent. At present, we require only the a-diastereomer to
synthesize the natural peptides. We are currently optimizing
a-selectivity at the C3 position by investigating alternative
solvents and various ester protecting groups for the effective
synthesis of ComX pheromones and related peptides.
100
80
60
40
20
0
[Lα]ComXRO-E-2 2
[Lβ]ComXRO-E-2 3
[Dα]ComXRO-E-2 4
[Dβ]ComXRO-E-2 5
The second key step was chemoselective reduction to pro-
duce modified tryptophan methyl esters 6 and 7 (or 8 and
9). Compounds 11, 14, and 15 contain various reactive func-
tional groups such as olefin, methyl ester, imine, and benz-
amide (or carbamate). Reduction of 11 using an equivalent
amount of DIBAL-H mediated the chemoselective cleavage
of benzamide, rather than the reduction of other functional
groups. Employment of NaBH3CN, NaBH4, excess DI-
BAL-H or BH3$SMe2 gave various reductive compounds
non-chemoselectively. Employment of pinacholborane
gave only L-Trp*(b-Ger)-OMe 7 in 51% yield (data not
shown). Reduction of 14 (or 15) with Lindlar’s reagent22 or
Et3SiH under acidic conditions did not facilitate imine reduc-
tion. By contrast, catecholborane was found to be the best
reagent for the chemoselective reduction of the imine.23
0.1
1
10
Concentration (nM)
100
Figure 6. Dose–response curves obtained using the ComXRO-E-2 peptide
and its diastereomers. Error bars are SD of the means of triplicate samples.
100
90
80
70
60
50
40
30
Additionally, it was very important to obtain the modified
tryptophans in optically pure form. We were able to separate
diastereomers 12 and 13 by column chromatography easily,
although we were not able to separate the diastereomers in
other steps at preparative scale. However, a single isomer is
afforded after cleavage of the Bz group, which allowed us
to synthesize all four tryptophan residues. The 1H NMR spec-
trum of Fmoc-Trp*(Ger)-OMe 16 indicated two different
compounds in a 1:1 ratio, although both have the cis config-
uration as indicated by an NOE between the H-2 proton and
the geranyl group. We believe that the two compounds are
conformational isomers, since NMR and other analyses
showed that cleavage of the Fmoc group gives only optically
active esters 6. We also observed similar phenomena on com-
pounds 17 and 18, although the ratio of conformers is differ-
ent on 18 (10:1). Furthermore, the conformational search
previously reported for the model study showed two stable
conformers.15 One of the conformers should have a similar
three-dimensional structure to the modified tryptophan resi-
due in ComXRO-E-2 pheromone. The coupling constants
between a and one of the b protons are nearly zero,
indicating that the dihedral angle between a and one of the
b protons is about 90ꢀ.24 The conformational search clearly
suggested this stereostructure, which is important for the
expression of biological function. Subsequent investigation
into the structure–activity relationships of ComXRO-E-2
analogs will clarify the importance of this stereostructure
for pheromone activity.
ComX
[2-6]ComX
[3-6]ComX
2
RO-E-2
RO-E-2
RO-E-2
20
10
0
21
20
0.1
1
10
Concentration (nM)
100
Figure 7. Dose–response curves obtained using the N-terminal truncated
ComXRO-E-2 peptides. Error bars are SD of the means of triplicate samples.
3. Discussion
The first key step in the synthesis of the modified tryptophan
residue was C-geranylation. We screened a variety of condi-
tions for C3-geranylation of the tricyclic N-protected trypto-
phan methyl esters. We found that Bz-Trp-OMe was the
most suitable reactant for C-geranylation. Immediate neu-
tralization with buffer was particularly important to prevent
degradation of the C-geranylated products in the subsequent
workup. The conditions established are also useful for syn-
thesizing related compounds using other functional groups,
such as prenyl, farnesyl, and geranylgeranyl.21 Based on
the hard–soft acid–base theory, we expected C3-geranyl-
ation to occur predominantly in THF, a relatively nonpolar
solvent. However, THF gave little enhancement of C-gera-
nylation when compared with relatively polar solvents
such as DMF or NMP. C/N selectivity in the dimethylallyl-
ation of 10a using alkylating methods involving quaternary
ammonium salts was reported to be ineffective by Cardoso
et al.18 On the other hand, geranylation of ether 10i gave
only the N1-geranylated compound in good yield and we
were thus able to preferentially synthesize either the N1-
or C3-geranylated compound. The synthetic conditions
were optimized to yield diastereomeric mixtures of C3-
geranyl-modified tryptophans. Because we required both
The four peptides were synthesized using a liquid-phase
method instead of a solid-phase method for two reasons.
First, we wanted to confirm the amino acid coupling reaction
with the modified tryptophans at the Na-position rather than
the N1-position. L-Trp*(a-Ger)-OMe 6 was reacted with
Teoc-Phe under typical coupling conditions, and HMBC
analysis showed that the reaction occurred only at the Na-
position. Second, we were not able to obtain modified
Fmoc-protected D-tryptophan with a free carboxyl group,

8914
M. Okada et al. / Tetrahedron 62 (2006) 8907–8918
which is a requirement for solid-phase synthesis, nor was the
compound sufficiently stable for this purpose.
isoprenylation is widely known, but we have not yet been
able to confirm the consensus sequence for tryptophan
isoprenylation among the ComX pheromones. We have no
evidence that this novel modification occurs in any other
peptide. However, the present tryptophan isoprenylation in
a pheromone that governs genetic exchange in B. subtilis
reminds us that cysteine isoprenylation was first discovered
in the sex pheromones of basidiomycetous yeasts.32–36
Synthetic peptide 2 and natural ComXRO-E-2 pheromone
exhibited identical 1H NMR spectra and showed very simi-
lar biological activity. Thus, the absolute structure of
ComXRO-E-2 pheromone was confirmed to be the same as
our proposed structure [La]ComXRO-E-2 2, and the structure
of the modified tryptophan residue was determined to be
[2S,3aR,8aS]-3a-geranyl-1,2,3,3a,8,8a-hexahydropyrrolo-
[2,3-b]indole-2-carboxylic acid.
4. Conclusion
We had previously synthesized putative pheromones with 1-,
2-, 4-, 5-, 6- or 7-geranyl substituted tryptophan residues.16
All of these peptides, including the three newly synthesized
diastereomers reported here, had the same molecular
formula and showed similar hydrophobicity to the natural
ComXRO-E-2 pheromone on LC analysis, but none exhibited
its biological activity. These results suggest that the structure
of the tryptophan residue, including its stereostructure, is
essential both for the interaction of ComXRO-E-2 pheromone
with receptor ComP and for the signal transduction that
results in the expression of genetic competence.
We synthesized four diastereomers of unique tricyclic trypto-
phan residues modified with a geranyl group and their four
corresponding peptides. We established the precise structure
of ComXRO-E-2 pheromone, including the modified tryp-
tophan residue, as [2S,3aR,8aS]-3a-geranyl-1,2,3,3a,8,
8a-hexahydropyrrolo[2,3-b]indole-2-carboxylic acid by
comparing NMR spectra and biological activities of the
synthetic peptides with those of the natural pheromone.
We also identified other natural ComXRO-E-2 pheromones.
We found that these pheromones were N-terminal truncated
peptides of ComXRO-E-2 pheromone, i.e., [2-6]ComXRO-E-2
and [3-6]ComXRO-E-2. These pheromones were synthesized
using the solid-phase method and were identical to the
natural pheromones analytically and biologically.
Furthermore, we identified two additional bioactive peptides
in the E. coli ED413 culture broth. Their structures were pre-
dicted to be [2-6]ComXRO-E-2 and [3-6]ComXRO-E-2 based
on their molecular weights. The Fmoc-modified L-Trp hav-
ing a free carboxyl group was sufficiently stable to permit
the synthesis of these two peptides and ComXRO-E-2 phero-
mone itself by solid-phase methods. This method will be
useful for future studies on structure–activity relationships
and to synthesize probes for investigating ligand–receptor
interactions.
5. Experimental
5.1. General methods
High-performance liquid chromatography was performed on
a HPLC system equipped with Jasco LC-980 series. HRMS
(ESI-TOF, positive) was recorded on a Mariner system (Ap-
plied Biosystems) using either an angiotensin/bradykinin/
neurotensin mixture or a polypropylene glycol solution as
a calibration standard. Optical densities were measured
with an AE-15F photoelectric colorimeter (Erma). NMR
spectra were recorded on a Bruker ARX-400 or a Bruker
AMX-600 spectrometer. Optical rotations were measured
on a Jasco DIP-370 digital polarimeter. Organic solvents
were purchased as anhydrous grade except for the following
solvents, which were freshly distilled prior to use: THF and
diethyl ether were dried by distillation from Na and benzo-
phenone ketyl. Moisture-sensitive reactions were carried
out under a dry nitrogen atmosphere in well-dried equipment
with a tightly fitted rubber septum. Open column chromato-
graphy was performed using silica gel BW-300 (Fuji silysia)
or ODS Cosmosil 140C18-OPN (Nacalai Tesque). Solid-
phase peptide synthesis was performed with a model 433A
peptide synthesizer (Applied Biosystems). Fmoc-protected
amino acid derivatives except modified tryptophan residues
and Clt resin were purchased from commercial sources
(Watanabe chemical, Nova biochem).
The two synthetic N-terminal truncated peptides were iden-
tical to the natural peptides. Even the [3-6]ComXRO-E-2,
which has only four amino acid residues, showed low but ap-
parent biological activity. This suggests that while the mod-
ified tryptophan residue is essential for biological activity,
the two N-terminal amino acid residues are not absolutely
required.
Several other ComX pheromones have been reported, and
some of these are thought to possess farnesyl groups. The
structure of the modified tryptophan residues in these pher-
omones may be similar to that of ComXRO-E-2 pheromone.
We are now able to synthesize these pheromones using
a method similar to that described here. It will be interesting
to determinewhether thesepheromoneshavethesamestereo-
structures as ComXRO-E-2 pheromone.
Numerous secondary metabolites of isoprenylated trypto-
phan derivatives are known;25–27 however, only four exam-
ples of post-translational modification on tryptophan
residues have been reported to date.28–31 The structure of
the geranyl-modified tryptophan residue described here is
the fifth example of the post-translational modification of
tryptophan. Post-translational isoprenylation, farnesyl or
geranylgeranyl modification on the C-terminal cysteine
residue is known to be quite common.14 However, isoprenyl-
ation on a tryptophan residue is unprecedented, and geranyl-
ation is also unknown. The consensus sequence of cysteine
5.2. Bacterial strains, pheromone production, and
biological activity
The E. coli ComXRO-E-2 producer strain (ED413) and the
B. subtilis tester strain (BD3020) were grown4,5 and phero-
mone production was carried out as described previously.4

M. Okada et al. / Tetrahedron 62 (2006) 8907–8918
8915
Biological activity was measured by a b-galactosidase assay
using the B. subtilis tester strain (BD3020) employed the ex-
pression of a srfA-lacZ fusion, under control of the natural
srfA promoter, which responds to add ComXRO-E-2 phero-
mone. BD3020 was grown overnight to stationary phase
and was then diluted 100-fold. The diluted culture (0.5 ml)
was added to a sample solution (3 ml of 50% aqueous
CH3CN solution), and incubated at 37 ꢀC for 5 h. After
toluene (5 ml) was added to the mixture, biological activity
was measured at 420 nm (%) with a standard method using
o-nitrophenyl-b-D-galactopyranoside at 30 ꢀC.
(m, 4H), 2.37 (dd, 1H, J¼10.9, 12.9 Hz), 2.56 (dd, 1H,
J¼7.8, 14.2 Hz), 2.57 (d, 1H, J¼12.9 Hz), 2.77 (dd, 1H,
J¼6.7, 14.2 Hz), 3.86 (s, 3H), 4.86 (m, 1H), 5.05 (m, 1H),
5.60 (d, 1H, J¼10.9 Hz), 7.03 (dt, 1H, J¼1.8, 7.2 Hz),
7.18–7.26 (m, 3H), 7.43 (dd, 1H, J¼7.5, 7.8 Hz), 7.56 (dt,
1H, J¼1.1, 7.5 Hz), 7.61 (dd, 1H, J¼1.1, 7.8 Hz); 13C
NMR (150 MHz, CDCl3) d 16.4, 17.6, 25.6, 26.4, 30.8,
34.8, 39.8, 52.8, 60.5, 65.0, 117.2, 119.9, 122.6, 123.7,
123.9, 128.1, 128.5, 129.5, 131.6, 132.5, 133.2, 139.5,
139.7, 158.2, 168.5, 171.4, 181.5; Anal. Calcd for
C29H32N2O3: C, 76.29; H, 7.06; N, 6.14. Found: C, 76.31;
H, 7.07; N, 6.24.
5.3. Purification of ComXRO-E-2 derivatives
5.4.2.2. Secondary amines 12 and 13. To a solution of 11
(2.23 g, 4.89 mmol) in THF (49 ml) was slowly added DI-
BAL-H (1.01 M solution in hexane, 9.70 ml, 9.80 mmol)
at –78 ꢀC. After stirring for 2 h at –78 ꢀC, the reaction mix-
ture was poured into 0.1 M phosphate buffer (pH 7.0, 50 ml).
It was extracted with EtOAc (4ꢂ50 ml). The organic layer
was washed with saturated aqueous NaCl, dried over
Na2SO4, and evaporated. The residue was purified by silica
gel column chromatography (CHCl3/CH3OH 100/1/80/1)
to give a-isomer 12 (871 mg, 2.47 mmol, 51%) as a colorless
oil and b-isomer 13 (751 mg, 2.13 mmol, 44%) as a colorless
oil.
The pre-purification was as reported previously. The two ac-
tive fractions were purified by HPLC on an ODS column
(4.6ꢂ250 mm ID, Develosil ODS-HG-5, Nomura Chemical)
at a flow rate of 1.0 ml/min, with CH3CN and 0.1% aqueous
ammonium acetate to give pure ComXRO-E-2 derivatives. The
solution was concentrated to remove CH3CN, and freeze-
dried several times to give pure ComXRO-E-2 derivatives.
5.4. Synthesis of ComXRO-E-2 pheromone
5.4.1. Typical cyclization. To a solution of N-protected tryp-
tophan methyl ester (1.00 mmol) and Et3N (4.0 equiv) in
CH2Cl2 (10 ml) at 0 ꢀC under nitrogen was slowly added
t-BuOCl (1.00 mmol). The reaction mixture was stirred at
0 ꢀC for 1 h, and warmed to room temperature for 12 h.
Et2O (200 ml) and H2O (200 ml) were added to the mixture.
The two layers were separated, and the organic layer was
washed with saturated aqueous NaCl, dried over Na2SO4,
and evaporated. The residue was recrystallized from
CH2Cl2/Et2O (or appropriate solvents) to give the tricyclic
compound as a powder.
5.4.2.3. a-Geranyl secondary amine 12. [a]2D6¼ +288 (c
1.0, CHCl3); 1H NMR (400 MHz, CDCl3) d 1.45 (s, 3H), 1.62
(s, 3H), 1.70 (s, 3H), 2.00–2.09 (m, 4H), 2.09 (dd, 1H, J¼9.8,
11.9 Hz), 2.18 (dd, 1H, J¼7.4, 13.9 Hz), 2.44 (dd, 1H, J¼8.0,
13.9 Hz), 2.67 (dd, 1H, J¼5.1, 11.9 Hz), 3.77 (s, 3H), 4.95
(dd, 1H, J¼5.1, 9.8 Hz), 5.09 (m, 1H), 5.22 (m, 1H), 6.83
(ddd, 1H, J¼1.0, 7.4, 7.5 Hz), 7.01 (dd, 1H, J¼1.0,
7.7 Hz), 7.09 (dd, 1H, J¼0.8, 7.4 Hz), 7.15 (ddd, 1H,
J¼0.8, 7.5, 7.7 Hz); 13C NMR (100 MHz, CDCl3) d 16.0,
17.7, 25.7, 26.5, 34.4, 39.4, 39.9, 52.1, 60.2, 71.0, 111.8,
118.6, 120.4, 123.9, 124.2, 128.1, 131.5, 133.2, 139.3,
149.5, 173.4, 184.6; Anal. Calcd for C22H28N2O2: C,
74.97; H, 8.01; N, 7.95. Found: C, 74.96; H, 7.87; N, 7.50.
5.4.1.1. Tricyclic compound from Bz-L-Trp-OMe 10h.
[a]2D3 ꢁ104 (c 1.0, CHCl3); 1H NMR (600 MHz, DMSO-d6)
d 11.50 (br, 1H), 7.60 (d, 2H, J¼7.1 Hz), 7.53–7.36 (m, 4H),
7.26 (dd, 1H, J¼4.9, 8.6 Hz), 6.98–6.95 (m, 2H), 5.58 (d,
1H, J¼8.2 Hz), 3.60 (dd, 1H, J¼8.2, 9.2 Hz), 3.44 (s, 3H),
3.06 (d, 1H, J¼9.2 Hz); 13C NMR (150 MHz, DMSO-d6)
d 171.5, 166.8, 143.1, 138.0, 135.2, 131.0, 128.8, 127.5,
122.9, 120.3, 119.5, 117.2, 112.9, 99.0, 67.1, 52.7, 29.4;
Anal. Calcd for C19H16N2O3: C, 71.24; H, 5.03; N, 8.74.
Found: C, 71.24; H, 4.92; N, 8.87.
5.4.2.4. D-Enantiomer of 12. [a]2D8¼ꢁ279 (c 1.0,
CHCl3).
5.4.2.5. b-Geranyl secondary amine 13. 1H NMR
(400 MHz, CDCl3) d 1.35 (s, 3H), 1.60 (s, 3H), 1.69 (s,
3H), 1.93–2.04 (m, 4H), 2.16 (dd, 1H, J¼7.1, 13.9 Hz),
2.38–2.49 (m, 2H), 2.64 (d, 1H, J¼12.8 Hz), 3.81 (s, 3H),
4.93 (d, 1H, J¼9.6 Hz), 5.04–5.11 (m, 2H), 6.86 (dt, 1H,
J¼0.7, 7.6 Hz), 6.93 (dd, 1H, J¼0.7, 7.6 Hz), 7.09 (dd,
1H, J¼0.9, 7.6 Hz), 7.16 (dt, 1H, J¼0.9, 7.6 Hz); 13C
NMR (100 MHz, CDCl3) d 16.2, 17.6, 25.6, 26.6, 37.4,
38.1, 39.9, 51.9, 53.2, 55.9, 117.2, 117.5, 121.1, 121.7,
124.2, 131.1, 137.5, 138.3, 139.1, 155.8, 172.5, 174.7;
HRMS (ESI+) m/z: calcd for C22H29N2O2 ([M+H]+)
353.2224. Found 353.2228.
5.4.1.2. D-Enantiomer of 10h. [a]D28 +106 (c 1.0, CHCl3).
5.4.2. Typical geranylation. To a solution of 2 (1.00 mmol)
in THF (10 ml) at 0 ꢀC under nitrogen was added NaH
(1.3 equiv). After stirring at 0 ꢀC for 1 h, geranyl bromide
(1.1 eq) was added to the mixture. After stirring for 5 h at
0 ꢀC, the reaction mixture was quenched and neutralized
with 0.1 M phosphate buffer (pH 7). The reaction mixture
was extracted with Et2O (4ꢂ20 ml), washed with saturated
aqueous NaCl, dried over Na2SO4, and evaporated. The
residue was purified by silica gel column chromatography
using appropriate solvents to give the C-geranylated dia-
stereoisomeric mixtures and N-geranylated compound.
5.4.2.6. Fmoc-protected compound 14. To a solution of
12 (100 mg, 0.284 mmol) in dioxane (3.0 ml) and 1.0 M
aqueous NaHCO3 (3.0 ml) was added Fmoc-OSu (120 mg,
0.356 mmol) at room temperature. After stirring for 2 h,
water (10 ml) was added to the reaction mixture and the
mixture was extracted with EtOAc (4ꢂ10 ml). The organic
layer was washed with saturated aqueous NaCl, dried over
5.4.2.1. b-Geranyl compound of 11. 1H NMR(600 MHz,
CDCl3) d 1.57 (s, 3H), 1.60 (s, 3H), 1.68 (s, 3H), 1.98–2.05

8916
M. Okada et al. / Tetrahedron 62 (2006) 8907–8918
Na2SO4, and evaporated. The residue was purified by silica
gel column chromatography (hexane/EtOAc 150/2/125/2)
to give 14 (143 mg, 0.249 mmol, 88%) as a colorless oil.
7.5 Hz), 6.38 (d, 1H, J¼7.7 Hz), 5.19 (m, 1H), 5.13 (m,
1H), 4.59 (s, 1H), 3.65 (dd, 1H, J¼5.8, 10.7 Hz), 3.24 (s,
3H), 2.35 (dd, 1H, J¼7.8, 14.5 Hz), 2.31 (dd, 1H, J¼7.1,
14.5 Hz), 2.27 (dd, 1H, J¼5.8, 12.0 Hz), 2.09–2.05 (m,
2H), 2.01–1.96 (m, 2H), 1.89 (dd, 1H, J¼10.7, 12.0 Hz),
1.68 (s, 3H), 1.53 (s, 3H), 1.44 (s, 3H); 13C NMR
(100 MHz, CD3COCD3) d 173.6, 150.9, 136.8, 132.9,
130.6, 127.5, 124.1, 123.1, 120.3, 117.2, 107.8, 82.2, 59.1,
58.1, 50.9, 44.0, 39.5, 36.4, 26.2, 24.8, 16.7, 15.4; Anal.
Calcd for C22H30N2O2: C, 74.54; H, 8.53; N, 7.90. Found:
C, 74.52; H, 8.57; N, 7.99.
[a]2D9 +317 (c 1.0, CHCl3); 1H NMR (600 MHz, CD3COCD3)
d 1.42 (s, 3H), 1.56 (s, 3H), 1.64 (s, 3H), 1.90–2.04 (m, 4H),
2.27 (dd, 1H, J¼9.9, 12.0 Hz), 2.46 (d, 2H, J¼7.9 Hz), 2.73
(dd, 1H, J¼5.2, 12.0 Hz), 3.78 (s, 3H), 4.38 (dd, 1H, J¼6.9,
8.1 Hz), 4.55 (dd, 1H, J¼8.1, 10.4 Hz), 4.59 (dd, 1H, J¼6.9,
10.4 Hz), 5.06 (m, 1H), 5.09 (m, 1H), 5.21 (dd, 1H, J¼5.2,
9.9 Hz), 7.13 (dt, 1H, J¼0.9, 7.5 Hz), 7.29 (ddd, 1H,
J¼1.2, 7.5, 7.9 Hz), 7.31–7.36 (m, 3H), 7.39–7.44 (m, 2H),
7.82–7.96 (m, 3H), 7.96 (dd, 1H, J¼0.6, 7.5 Hz), 8.26 (dd,
1H, J¼0.6, 7.5 Hz); 13C NMR (100 MHz, CDCl3) d 16.1,
17.8, 25.8, 27.3, 35.8, 39.9, 40.6, 47.4, 52.3, 60.8, 69.4,
74.1, 115.9, 118.3, 118.6, 120.6, 120.7, 124.8, 124.9,
125.0, 126.6, 127.4, 128.0, 128.2, 128.6, 128.7, 129.1,
131.9, 133.1, 140.6, 141.9, 142.2, 143.3, 143.4, 143.8,
151.6, 173.8, 176.4; HRMS (ESI+) m/z: calcd for
C37H39N2O4 ([M+H]+) 575.2904. Found 575.2922.
5.4.2.11. D-Trp*(a-Ger)-OMe 8 (D-enantiomer of 7).
[a]2D8 +75.4 (c 1.0, CHCl3).
5.4.3. Synthesis of the ComXRO-E-2 peptides. We chose
Teoc as the N-terminal protecting group, and Clt as the car-
boxy-protecting group. The Teoc group was cleaved under
neutral condition using TAS-F, and the Clt group was
cleaved under mild acidic conditions using 50% aqueous
AcOH. All reactions were monitored with LC/MS and
TLC. Peptides were purified by silica gel column chromato-
graphy containing 1% Et3N so as not to cleave the Clt group.
All Clt esters were unstable, so they were carried on to the
next step immediately without further purification.
5.4.2.7. D-Enantiomer of 14. [a]D28 –314 (c 0.50, CHCl3).
5.4.2.8. L-Trp*(a-Ger)-OMe 6. To a solution of 14
(458 mg, 0.797 mmol) in THF (31 ml) was added catechol-
borane (1.01 M solution in THF, 1.0 ml, 1.00 mmol) at 0 ꢀC.
After stirring for 2 h, water (30 ml) was added to the reaction
mixture, followed by extraction with EtOAc (4ꢂ30 ml). The
organic layer was washed with saturated aqueous NaCl,
dried over Na2SO4, and evaporated. The residue was purified
by silica gel column chromatography (hexane/acetone
10/1/8/1) to give a reductive compound. To a solution
of the reductive compound in CH3CN (2.7 ml) at room
temperature was added piperidine (0.3 ml). After stirring
for 1 h, the reaction mixture was quenched and neutralized
with 0.1 M phosphate buffer (pH 7.0, 3.0 ml). The reaction
mixture was extracted with EtOAc (4ꢂ5 ml). The organic
layer was washed with saturated aqueous NaCl, dried over
Na2SO4, and evaporated. The residue was purified by silica
gel column chromatography (hexane/acetone 6/1/5/1) to
give 6 (202 mg, 0.570 mmol, 72%, in two steps) as a color-
less oil.
5.4.3.1. Teoc-Phe-L-Trp*(a-Ger)-OMe 18. To a solution
of L-Trp*(a-Ger)-OMe 6 (80.5 mg, 0.227 mmol), Teoc-Phe
(150 mg, 0.485 mmol), and Et3N (94.6 mg, 0.935 mmol)
in CH2Cl2 (5.0 ml) under nitrogen were added HOAt
(66.0 mg, 0.485 mmol) and HATU (185 mg, 0.487 mmol)
at 0 ꢀC. After stirring for 2 h, the reaction mixture was
quenched and neutralized with 0.1 M phosphate buffer
(10 ml). The reaction mixture was extracted with EtOAc
(4ꢂ10 ml). The organic layer was washed with saturated
aqueous NaCl, dried over Na2SO4, and evaporated. The
residue was purified by silica gel column chromatography
(hexane/EtOAc 6/1/4/1) to give 18 (141 mg, 0.217 mmol,
96%) as a colorless oil.
1H NMR (600 MHz, CDCl3) d 7.34–7.27 (m, 3H), 7.23 (d,
2H, J¼7.1 Hz), 6.96 (dd, 1H, J¼7.3, 7.7 Hz), 6.86 (d, 1H,
J¼7.4 Hz), 6.59 (dd, 1H, J¼7.3, 7.4 Hz), 6.11 (d, 1H,
J¼7.7 Hz), 5.59 (d, 1H, J¼8.2 Hz), 5.10–5.04 (m, 2H),
5.04 (s, 1H), 4.39 (m, 1H), 4.16–4.05 (m, 2H), 3.57 (d, 1H,
J¼8.3 Hz), 3.14 (dd, 1H, J¼4.7, 12.7 Hz), 3.10 (s, 3H),
2.85 (dd, 1H, J¼10.6, 12.7 Hz), 2.34 (d, 1H, J¼12.8 Hz),
2.24 (dd, 1H, J¼7.5, 14.4 Hz), 2.19 (dd, 1H, J¼7.6,
14.4 Hz), 2.08–2.00 (m, 4H), 1.70 (s, 3H), 1.68 (s, 3H),
1.52–1.47 (m, 4H), 0.96 (m, 2H), 0.02 (s, 9H); 13C NMR
(150 MHz, CDCl3) d 171.7, 170.4, 155.4, 149.5, 138.8,
136.2, 131.6, 130.0, 129.4, 128.7, 127.2, 124.1, 123.9,
118.5, 118.4, 109.0, 80.9, 63.1, 59.7, 55.1, 53.7, 52.3, 41.3,
40.0, 37.9, 35.3, 26.6, 25.7, 17.8, 17.7, 16.2, –1.53; HRMS
(ESI+) m/z: calcd for C37H52N3O5Si ([M+H]+) 646.3671.
Found 646.3684.
1
[a]2D6¼+52.2 (c 1.0, CHCl3); H NMR (600 MHz, C6D6)
d 6.99 (t, 1H, J¼7.6 Hz), 6.96 (d, 1H, J¼7.6 Hz), 6.69 (t,
1H, J¼7.6 Hz), 6.40 (d, 1H, J¼7.6 Hz), 5.24 (m, 1H), 5.14
(m, 1H), 4.63 (s, 1H), 3.71 (dd, 1H, J¼3.3, 7.8 Hz), 3.07 (s,
3H), 2.46 (dd, 1H, J¼3.3, 12.6 Hz), 2.35 (dd, 1H, J¼7.9,
14.5 Hz), 2.31 (dd, 1H, J¼7.6, 14.5 Hz), 2.18 (dd, 1H,
J¼7.8, 12.6 Hz), 2.10–2.06 (m, 2H), 2.01–1.97 (m, 2H),
1.67 (s, 3H), 1.53 (s, 3H), 1.46 (s, 3H); 13C NMR
(100 MHz, CD3COCD3) d 174.8, 151.2, 137.8, 131.7,
128.4, 125.1, 124.4, 121.4, 120.3, 118.3, 109.6, 83.5, 60.2,
58.1, 51.7, 41.9, 40.6, 37.2, 27.2, 25.8, 17.7, 16.4; Anal.
Calcd for C22H30N2O2: C, 74.54; H, 8.53; N, 7.90. Found:
C, 74.52; H, 8.49; N, 7.83.
5.4.2.9. D-Trp*(b-Ger)-OMe 9 (D-enantiomer of 6).
[a]2D8 –52.1 (c 1.0, CHCl3).
5.4.3.2. [La]ComXRO-E-2 2 (ComXRO-E-2 pheromone).
1H NMR (600 MHz, CD3CN/D2O 6/4) d 7.33–7.39 (m, 3H),
7.27 (d, 2H, J¼7.2 Hz), 7.01 (dd, 1H, J¼7.4, 7.8 Hz), 6.85
(d, 1H, J¼7.3 Hz), 6.63 (d, 1H, J¼7.8 Hz), 6.60 (dd, 1H,
J¼7.3, 7.4 Hz), 5.08 (m, 1H), 5.06 (s, 1H), 5.00 (m, 1H),
4.47 (dd, 1H, J¼4.9, 11.8 Hz), 4.21 (d, 1H, J¼8.3 Hz),
5.4.2.10. L-Trp*(b-Ger)-OMe 7. [a]2D9 –76.1 (c 1.0,
CHCl3); 1H NMR (600 MHz, C6D6) d 7.01 (dd, 1H, J¼7.5,
7.7 Hz), 6.94 (d, 1H, J¼7.4 Hz), 6.73 (dd, 1H, J¼7.4,

M. Okada et al. / Tetrahedron 62 (2006) 8907–8918
8917
3.92 (dd, 1H, J¼4.8, 8.3 Hz), 3.46 (d, 1H, J¼8.7 Hz), 3.26
(d, 1H, J¼16.7 Hz), 3.23–3.20 (m, 2H), 3.17 (dd, 1H,
J¼6.4, 8.5 Hz), 2.87 (dd, 1H, J¼11.8, 12.2 Hz), 2.28 (d,
1H, J¼12.7 Hz), 2.14–1.98 (m, 10H), 1.94–1.84 (m, 2H),
1.80–1.75 (m, 2H), 1.68 (m, 1H), 1.66 (s, 3H), 1.58 (s,
3H), 1.49 (m, 1H), 1.48 (s, 3H), 1.10 (m, 1H) 1.00 (dd,
1H, J¼8.7, 12.8 Hz), 0.96 (d, 3H, J¼6.8 Hz), 0.83 (dd,
3H, J¼7.3, 7.4 Hz); HRMS (ESI+) m/z: calcd for
C48H68N8O10 ([M+2H]2+) 458.2524. Found 458.2525.
was added dropwise at room temperature. After the reaction
mixture had been stirred for 30 min, it was neutralized with
0.1 M phosphate buffer (10 ml). The reaction mixture was
extracted with EtOAc (5ꢂ10 ml). The organic solvent was
removed by evaporation. The residue was dissolved in diox-
ane (8.8 ml) and 1 M aqueous Na2CO3 (4.4 ml). To the solu-
tion was added Fmoc-OSu (386 mg, 1.16 mmol) at room
temperature. After the reaction mixture had been stirred for
2 h, the reaction was neutralized with 0.1 M phosphate buffer
(10 ml). The reaction mixture was extracted with EtOAc
(4ꢂ30 ml). The organic layer was washed with saturated
aqueous NaCl, dried over Na2SO4, and evaporated. The res-
idue was purified by silica gel column chromatography
(CHCl3/CH3OH 100/1/50/1) to give Fmoc-L-Trp*(a-
Ger) 19 (482 mg, 0.857 mmol, 97% in two steps) as a color-
less oil.
5.4.3.3. [Lb]ComXRO-E-2 3. 1H NMR (600 MHz,
CD3CN/D2O, 6/4) d 7.34–7.28 (m, 4H), 7.25 (dd, 1H,
J¼7.2, 7.4 Hz), 7.15 (d, 1H, J¼7.4 Hz), 7.07 (dd, 1H,
J¼7.3, 7.8 Hz), 6.78 (dd, 1H, J¼7.3, 7.4 Hz), 6.59 (d, 1H,
J¼7.8 Hz), 5.25 (s, 1H), 5.00 (m, 1H), 4.91 (m, 1H), 4.87
(dd, 1H, J¼4.8, 9.5 Hz), 4.18 (dd, 1H, J¼5.3, 8.9 Hz), 4.12
(d, 1H, J¼7.4 Hz), 4.09 (dd, 1H, J¼4.9, 8.1 Hz), 4.05 (dd,
1H, J¼7.5, 8.9 Hz), 3.25 (s, 2H), 3.12 (dd, 1H, J¼4.8,
14.1 Hz), 2.89 (dd, 1H, J¼9.5, 14.1 Hz), 2.53 (dd, 1H,
J¼7.5, 12.8 Hz), 2.40 (m, 2H), 2.25–2.11 (m, 5H,), 2.08
(dd, 1H, J¼8.9, 12.8 Hz), 2.05–1.93 (m, 3H), 1.90–1.81
(m, 4H), 1.69 (m, 1H), 1.60 (s, 3H), 1.51 (s, 3H), 1.46 (s,
3H), 1.20 (m, 1H), 0.98 (m, 1H), 0.74 (dd, 3H, J¼7.3,
7.4 Hz), 0.68 (d, 3H, J¼6.8 Hz); HRMS (ESI+) m/z: calcd
for C48H68N8O10 ([M+2H]2+) 458.2524. Found 458.2530.
Anal. Calcd for C36H38N2O4: C, 76.84; H, 6.81; N, 4.98.
Found: C, 76.95; H, 7.12; N, 5.08.
5.4.5. Syntheses of ComXRO-E-2 pheromones with solid-
phase peptide synthesis. Peptide bond formation was
carried out according to the procedure for putative
ComXRO-E-2 peptides16 except for the final cleavage and
the deprotection procedures. Cleavage and deprotection of
the resin was carried out as follows. To a suspension of the
attached resin in CH3CN was added 5% TFA and 5% tri-
fluoroethanol at 0 ꢀC. The reaction mixture was shaken in
a rotary shaker for 20 h at 4 ꢀC.
5.4.3.4. [Da]ComXRO-E-2 4. 1H NMR (600 MHz,
CD3CN/D2O, 6/4) d 7.26–7.23 (m, 4H) 7.21 (dd, 1H,
J¼7.2, 7.4 Hz), 7.14 (d, 1H, J¼6.8 Hz), 7.09 (dd, 1H,
J¼7.4, 7.9 Hz), 6.82 (dd, 1H, J¼6.8, 7.4 Hz), 6.64 (d, 1H,
J¼7.9 Hz), 5.43 (s, 1H), 5.42 (dd, 1H, J¼6.4, 8.8 Hz),
5.03 (m, 1H), 4.99 (m, 1H), 4.23 (dd, 1H, J¼3.7, 9.8 Hz),
4.13–4.09 (m, 2H) 4.00 (dd, 1H, J¼7.3, 9.4 Hz), 3.80 (d,
1H, J¼15.8 Hz), 3.63 (d, 1H, J¼15.8 Hz), 3.09 (dd, 2H,
J¼6.4, 13.9 Hz), 2.81 (dd, 1H, J¼8.8, 13.9 Hz), 2.60 (dd,
1H, J¼7.3, 13.0 Hz), 2.45 (dd, 1H, J¼7.6, 14.0 Hz), 2.39
(dd, 1H, J¼7.9, 14.0 Hz), 2.26 (dd, 2H, J¼7.2, 8.5 Hz),
2.19–2.06 (m, 5H), 2.04–1.89 (m, 5H), 1.81 (m, 1H), 1.74
(m, 1H), 1.65 (s, 3H), 1.56 (s, 3H), 1.37 (s, 3H), 1.18 (m,
1H), 1.03 (m, 1H), 0.77 (dd, 3H, J¼7.3, 7.4 Hz), 0.65 (d,
3H, J¼6.8 Hz); HRMS (ESI+) m/z: calcd for C48H68N8O10
([M+2H]2+) 458.2524. Found 458.2533.
5.4.5.1. [3-6]ComXRO-E-2 20. 1H NMR (600 MHz,
CD3CN/D2O, 6/4) d 7.42–7.34 (m, 3H), 7.28 (d, 2H,
J¼6.9 Hz), 7.02 (dd, 1H, J¼7.5, 7.7 Hz), 6.88 (d, 1H,
J¼7.2 Hz), 6.63–6.60 (m, 2H), 5.09 (m, 1H), 5.07 (s, 1H),
5.02 (m, 1H), 4.28 (dd, 1H, J¼4.8, 11.8 Hz), 3.96 (dd, 1H,
J¼4.8, 7.8 Hz), 3.38 (d, 1H, J¼8.7 Hz), 3.27 (dd, 1H,
J¼4.8, 12.2 Hz), 3.21 (dd, 1H, J¼5.2, 6.1 Hz), 2.99 (dd,
1H, J¼11.8, 12.2 Hz), 2.28 (d, 1H, J¼12.9 Hz), 2.19–2.12
(m, 2H), 2.11 (d, 1H, J¼6.8 Hz), 2.17–1.90 (m, 12H), 1.76
(m, 1H), 1.65 (s, 3H), 1.58 (s, 3H), 1.52 (m, 1H), 1.48 (s,
3H), 1.17 (dd, 1H, J¼8.7, 12.9 Hz); HRMS (ESI+) m/z: calcd
for C40H53N6O8 ([M+H]+) 745.3919. Found 745.3939.
5.4.3.5. [Db]ComXRO-E-2 5. 1H NMR (600 MHz,
CD3CN/D2O, 6/4) d 7.37–7.27 (m, 5H), 6.94 (dd, 1H,
J¼7.5, 7.7 Hz), 6.90 (d, 1H, J¼7.4 Hz), 6.60 (dd, 1H,
J¼7.4, 7.5 Hz), 6.49 (d, 1H, J¼7.7 Hz), 5.01 (m, 1H),
4.84–4.79 (m, 2H), 4.53 (d, 1H, J¼9.0 Hz), 4.49 (s, 1H),
4.28 (d, 1H, J¼7.9 Hz), 3.98 (dd, 1H, J¼6.4, 6.5 Hz), 3.76
(dd, 1H, J¼5.7, 7.5 Hz), 3.46 (d, 1H, J¼17.1 Hz), 3.42 (d,
1H, J¼17.1 Hz), 3.11 (dd, 1H, J¼5.6, 12.6 Hz), 3.05 (dd,
1H, J¼12.6, 12.6 Hz), 2.39 (d, 1H, J¼12.7 Hz), 2.18–2.09
(m, 2H), 1.98–1.71 (m, 13H), 1.64 (s, 3H), 1.55 (s, 3H),
1.54 (m, 1H), 1.41 (s, 3H), 1.37 (m, 1H), 1.08 (m, 1H),
0.84 (d, 3H, J¼6.8 Hz), 0.79 (d, 3H, J¼7.3, 7.4 Hz);
HRMS (ESI+) m/z: calcd for C48H68N8O10 ([M+2H]2+)
458.2524. Found 458.2521.
5.4.5.2. [2-6]ComXRO-E-2 21. 1H NMR (600 MHz,
CD3CN/D2O, 6/4) d 7.39–7.34 (m, 3H), 7.28 (d, 2H,
J¼7.2 Hz), 7.01 (dd, 1H, J¼7.6, 7.7 Hz), 6.87 (d, 1H,
J¼7.5 Hz), 6.63–6.59 (m, 2H), 5.08 (s, 1H), 5.07 (m, 1H),
5.00 (m, 1H), 4.51 (dd, 1H, J¼5.1, 11.2 Hz), 3.94 (dd, 1H,
J¼4.8, 7.9 Hz), 3.55 (d, 1H, J¼8.8 Hz), 3.23–3.17 (m, 3H),
2.87 (dd, 1H, J¼11.2, 12.2 Hz), 2.29 (d, 1H, J¼12.8 Hz),
2.09 (d, 2H, J¼7.2 Hz), 2.08–1.92 (m, 9H), 1.89–1.68 (m,
3H), 1.66 (s, 3H), 1.61 (m, 1H), 1.58 (s, 3H), 1.49 (s, 3H),
1.41 (m, 1H), 1.12–1.05 (m, 2H), 0.91 (d, 3H, J¼6.8 Hz),
0.82 (dd, 3H, J¼7.4, 7.5 Hz); HRMS (ESI+) m/z: calcd for
C46H64N7O9 ([M+H]+) 858.4760. Found 858.4763.
Acknowledgements
5.4.4. Synthesis of Fmoc-protected tryptophan residue
for solid-phase peptide synthesis.
5.4.4.1. Fmoc-L-Trp*(a-Ger) 19. To a solution of
L-Trp*(a-Ger)-OMe 6 (312 mg, 0.880 mmol) in THF
(3 ml) and CH3OH (3.0 ml), 1 M aqueous LiOH (3.0 ml)
The work performed in Nagoya was supported by Grant-
in-Aid for COE (14COEA02) and for Scientific Research
(No. 18101009). The work in Newark was supported by
NIH grant GM57720.

8918
M. Okada et al. / Tetrahedron 62 (2006) 8907–8918
19. Other protected compounds (Boc, Teoc, Trt, TBS, and Fmoc)
References and notes
could not be recrystallized or decomposed by the cyclization.
20. Teoc group was mediated to only b-geranylated compound 13
with the treatment with Teoc-O-nitrophenyl.
21. We have actually confirmed to react the material with farnesyl
bromide in good yield (data not shown).
22. Lindlar, H. Helv. Chim. Acta 1952, 35, 446.
23. Chemoselective imine reduction with diacid–borane complex
has been reported (see: Liu, Z.-H.; Bhongle, N.; Su, X.; Ribe,
S.; Senanayake, C. H. Tetrahedron Lett. 2002, 43, 8617), butthe
reagent was very low reactivity for 14 (or 15) (data not shown).
24. Karplus, M. J. Phys. Chem. 1959, 30, 11.
25. Cui, C. B.; Kakeya, H.; Okada, G.; Onose, R.; Ubukata, M.;
Takahashi, I.; Isono, K.; Osada, H. J. Antibiot. 1995, 48, 1382.
26. Robbers, J. E.; Floss, H. G. Tetrahedron Lett. 1969, 23, 1857.
27. Ishida, K.; Matsuda, H.; Murakami, M.; Yamaguchi, K.
Tetrahedron 1996, 52, 9025.
1. Williams, P.; Baldwin, T. J.; Downie, J. A. Microbial Signaling
and Communication; England, R. R., Hobbs, G., Bainton, N. J.,
Roberts, D. McL., Eds.; Cambridge University Press:
Cambridge, 1999; pp 1–32.
2. Lyon, G. J.; Muir, T. W. Chem. Biol. 2003, 10, 1007.
3. Taga, M. E.; Bassler, B. L. Proc. Natl. Acad. Sci. U.S.A. 2003,
100, 4549.
4. Hamoen, L. W.; Venema, G.; Kuipers, O. P. Microbiology 2003,
149, 9.
5. Magnuson, R.; Solomon, J.; Grossman, A. D. Cell 1994, 77,
207.
6. Weinrauch, Y.; Penchev, R.; Dubnau, E.; Smith, I.; Dubnau, D.
Genes Dev. 1990, 4, 860.
7. Piazza, F.; Tortosa, P.; Dubnau, D. J. Bacteriol. 2001, 181,
4540.
28. McIntire, W. S.; Wemmer, D. E.; Chistoserdov, A.; Lindstrom,
M. E. Science 1991, 252, 817.
8. Tran, L.-S. P.; Nagai, T.; Itoh, Y. Mol. Microbiol. 2000, 37,
1159.
€
€
29. Hofsteenge, J.; Muller, D. R.; de Beer, T.; Loffler, A.; Richter,
W. J.; Vliegenthart, J. F. G. Biochemistry 1994, 33, 13524.
30. Jimenez, E. C.; Craig, A. G.; Watkins, M.; Hillyard, D. R.;
Gray, W. R.; Gulyas, J.; Rivier, J. E.; Cruz, L. J.; Olivera,
B. M. Biochemistry 1997, 36, 989.
9. Tortosa, P.; Logsdon, L.; Kraigher, B.; Itoh, Y.; Mandic-Mulec,
I.; Dubnau, D. J. Bacteriol. 2001, 183, 451.
10. Ansaldi, M.; Dubnau, D. Mol. Microbiol. 2002, 44, 1561.
11. Ansaldi, M.; Marolt, D.; Stebe, T.; Mandic-Mulec, I.; Dubnau,
D. J. Bacteriol. 2004, 186, 15.
31. Anderson, L. B.; Maderia, M. M.; Ouellette, A. J. A.; Putnam-
Evans, C.; Higgins, L.; Krick, T.; MacCoss, M. J.; Lim, H.;
Yates, J. R., III; Barry, B. A. Proc. Natl. Acad. Sci. U.S.A.
2002, 99, 14676.
12. Weinrauch, Y.; Msadek, T.; Kunst, F.; Dubnau, D. J. Bacteriol.
1991, 173, 5685.
13. Bacon Schneider, K.; Palmer, T. M.; Grossman, A. D.
J. Bacteriol. 2002, 184, 410.
32. Kamiya, Y.; Sakurai, A.; Takahashi, N.; Tsuchiya, E.; Abe, K.;
Fukui, S. Agric. Biol. Chem. 1979, 43, 363.
33. Sakagami, Y.; Isogai, M.; Suzuki, A.; Kitada, C.; Fujino, M.
Agric. Biol. Chem. 1979, 43, 2643.
34. Sakagami, Y.; Yoshida, M.; Isogai, A.; Suzuki, A. Science
1981, 212, 1525.
14. Clarke, S. Annu. Rev. Biochem. 1992, 61, 355.
15. Okada, M.; Sato, I.; Cho, S. J.; Iwata, H.; Nishio, T.; Dubnau,
D.; Sakagami, Y. Nat. Chem. Biol. 2005, 1, 23.
16. Okada, M.; Sato, I.; Cho, S. J.; Suzuki, Y.; Ojika, M.; Dubnau,
D.; Sakagami, Y. Biosci. Biotechnol. Biochem. 2004, 68,
2374.
35. Rilling, H. C.; Breunger, E.; Epstein, W. W.; Crain, P. F.
Science 1990, 247, 318.
17. Ohno, M.; Spande, T. F.; Witkop, B. J. Am. Chem. Soc. 1980,
92, 343.
36. Farnsworth, C. C.; Gelb, M. H.; Glomset, J. A. Science 1990,
247, 320.
18. Cardoso, A. S.; Srinivasan, N.; Lobo, A. M.; Prabhakar, S.
Tetrahedron Lett. 2001, 42, 6663.