Resolution of the Confusion in the Assignments of Configuration for the Ciliatamides, Acylated Dipeptides from Marine Sponges

Article abstract of DOI:10.1021/acs.jnatprod.7b00684

Direct comparison of authentic ciliatamide A with four synthetic isomers (1-4) by means of NMR and chiral-phase HPLC revealed that ciliatamide A possesses the 12R (d-N-MePhe residue) and 22S (l-Lys residue) configurations, which were not identical with either our previous assignment or those proposed by others through total synthesis. The absolute configuration of the methionine sulfoxide residue in ciliatamide D was also revised to be d.

Full text of DOI:10.1021/acs.jnatprod.7b00684

Note
pubs.acs.org/jnp
Resolution of the Confusion in the Assignments of Configuration for
the Ciliatamides, Acylated Dipeptides from Marine Sponges
Kentaro Takada,* Raku Irie, Rei Suo, and Shigeki Matsunaga*
Laboratory of Aquatic Natural Products Chemistry, Graduate School of Agricultural and Life Sciences, The University of Tokyo,
Bunkyo-ku, Tokyo 113-8657, Japan
S
* Supporting Information
ABSTRACT: Direct comparison of authentic ciliatamide A with four synthetic
isomers (1−4) by means of NMR and chiral-phase HPLC revealed that ciliatamide A
possesses the 12R (D-N-MePhe residue) and 22S (L-Lys residue) configurations, which
were not identical with either our previous assignment or those proposed by others
through total synthesis. The absolute configuration of the methionine sulfoxide residue
in ciliatamide D was also revised to be D.
was closest to those of the (12R,22R)-isomer as described by
iliatamide A was originally reported as an antileishmanial
Ccompound from the deep-sea sponge Aaptos ciliata. Its Lindsley et al..3
Next, we rigorously assigned the 1H and 13C NMR signals of
structure was proposed as 1 (12S,22S), being composed of N-
methyl-L-phenylalanyl-L-lysine with the N-terminus 9-decenoy-
lated and the C-terminus cyclized to form an ε-caprolactam.1
Later, Lindsley and co-workers reported that the synthetic
compound with the proposed (12S,22S)-configuration ex-
hibited NMR data identical with the natural product, but the
sign of the specific rotation was opposite.2 They synthesized all
four possible stereoisomers (1−4), measured their NMR
spectra and optical rotations, and concluded that ciliatamide
A was 4 with the (12R,22R)-configuration.2 Later, we isolated
ciliatamide A together with a new analogue, ciliatamide D (5,
assigned as 12S, 18S), from a Stelletta sp. marine sponge.3 At
this time we reinvestigated the configuration of ciliatamide A
and again proposed the (12S,22S)-configuration for ciliatamide
A. Mohapatra et al. synthesized the (12S,22S)-isomer that had a
specific rotation of −33, the same as observed by Lindsley et al.
for the (12S,22S)-isomer.4,5 However, the true configuration of
ciliatamide A has not been fully resolved due to the lack of
direct comparison between the natural product and the
synthetic derivatives.3 Therefore, we synthesized the four
possible stereoisomers of ciliatamide A and compared each of
the (12S,22S)- and (12R,22S)-isomers (1 and 2, respectively)
including those of the minor conformers which arose from the
cis−trans isomerism of the amide bond between the 9-decenoyl
residue and the N-MePhe residue.6 As expected, the NMR
spectra of the enantiomeric pairs [ (12S,22S)- and (12R,22R)-
isomers as well as (12R,22S)- and (12S,22R)-isomers] were
1
superimposable. The H NMR spectra of the diastereomeric
pairs were different, as briefly noted by Lindsley et al.2 In the
major isomer, chemical shifts of H-12 (1: δ 5.38 vs 2: δ 5.43),
H-13a (1: δ 3.04 vs 2: δ 2.96), and H-23b (1: δ 1.93 vs 2: δ
1.89) signals were slightly different between the diastereomers,
whereas noticeable differences were observed for the H-2a (1: δ
1.67 vs 2: δ 1.77), H-4a (1: δ 1.06 vs 2: δ 1.10), H3-20 (1: δ
2.89 vs 2: δ 2.92), H-23a, (1: δ 1.59 vs 2: δ 1.54), and H-23b
(1: δ 1.92 vs 2: δ 1.89) signals in the minor isomers (Table S1).
1H NMR signals of natural ciliatamide A for H-12 and H-13a in
the major isomer and H-4a and H3-20 in the minor isomer were
identical with those of the (12R,22S)-isomer but different from
those of the (12S,22S)-isomer (Figure 1 and Tables S1−S3).7
Marfey’s analysis of the natural ciliatamide A clearly showed the
presence of D-N-MePhe and L-Lys, demonstrating that the
structure of the natural product was 2 (12R,22S) (Figure S17).
The configuration of the N-MePhe residue was erroneously
assigned in our previous publications.1,3,8
1
them with the natural product by H NMR spectroscopy and
HPLC using a chiral stationary phase.
We employed a synthetic route essentially identical with the
one reported by Lindsley et al. (Scheme 1).2 Either
commercially available L-α-amino-ε-caprolactam hydrochloride
or its D-isomer prepared from N-α-Boc-D-lysine was condensed
with D- or L-N-methyl-phenylalanine (N-MePhe) and 9-
decenoic acid to afford 1 with the (12S,22S)-configuration, 2
with the (12R,22S)-configuration, 3 with the (12S,22R)-
configuration, and 4 with the (12R,22R)-configuration. The
signs and magnitudes of the specific rotations of the four
compounds prepared in this study were in agreement with
those reported by Lindsley et al. (Table 1), and the sign and
magnitude of the specific rotation value of the natural product
Due to the absence of the detailed experimental data for our
first report,1 we speculated the cause of the misassignment
during the Marfey’s analysis of ciliatamide A was as follows.
The retention times of the N-(2,4-dinitrophenyl)-L-alanine
amide (DAA) derivatives of N-MePhe and Lys are as follows: L-
N-MePhe (13.1 min), D-N-MePhe (13.6 min), L-Lys (13.7
min), D-Lys (17.8 min), among which D-N-MePhe and L-Lys
are overlapped (Figure S17). Our hydrolysis experiments
Received: August 9, 2017
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
DOI: 10.1021/acs.jnatprod.7b00684
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Note
Chart 1
Scheme 1. Synthesis of 1
of DAA-Phe, not of DAA-N-MePhe (Figure S10 in ref 2). The
D-N-MePhe and L-N-MePhe used as the standards were
prepared from D-N-Boc-Phe and L-N-Boc-Phe, respectively,
and used without rigorous purification. In the Marfey’s analysis
of the natural ciliatamide A, a peak coeluting with DAA-L-Phe
derived from the unreacted L-Phe during the preparation of L-
N-MePhe was detected. The misleading peak turned out to
derive from an unidentified impurity.
Table 1. Specific Rotation Values (MeOH) for Natural
Ciliatamide A and Synthetic Diastereomers
Lindsley
(c 0.05)
Mohapatra
(c 0.05)
this work
(c 0.2)
natural product
(c 0.1)
a
b
12S,22S
12S,22R
12R,22S
12R,22R
−35
+70
−48
+42
−33
−42
+65
−67
+46
+40 (+52 )
With the four isomers as well as the natural product in hand,
we proceeded to analyze them by HPLC with a chiral stationary
phase. By using a CHIRALPAK IG-3 column, the (12R,22R)-,
(12S,22S)-, and (12R,22S)-isomers each gave a sharp peak with
different retention times. However, the (12S,22R)-isomer (3)
did not give a peak under the same HPLC condition. We
examined a variety of chiral stationary phases,9 among which
CHIRALPAK ID-3 afforded a peak for 3 and baseline separated
the four isomers (Figure 2). Under this condition natural
ciliatamide A coeluted with the (12R,22S)-isomer, in
accordance with the assignment described above.
a
b
The value reported in ref 1. The value reported in ref 3.
showed that N-MePhe was partially racemized, but Lys was not
racemized (Figure S17). Because DAA2-L-Lys and DAA-D-N-
MePhe overlap, the hydrolysate of 2 (composed of D-N-MePhe
and L-Lys) will give a small peak of DAA-L-N-MePhe, generated
by partial racemization, and a large overlapped peak of DAA2-L-
Lys and DAA-D-N-MePhe. By detecting the HPLC with UV
absorption, it was not possible to distinguish between DAA2-L-
Lys and DAA-D-N-MePhe. After detection of L-N-MePhe, it
was reasonable to consider another peak as derived only from L-
Lys (Figure S17).
The revision of the configuration of ciliatamide A prompted
us to reanalyze the configuration of ciliatamide D (5), a 1:1
epimeric mixture of diastereomers due to asymmetry at the
sulfur atom of the methionine sulfoxide residue.3 We conducted
In our second report3 Marfey’s analysis of N-MePhe was
conducted by inadvertently using the mass chromatogram with
detection at m/z 418, which corresponded to the [M + H]+ ion
B
DOI: 10.1021/acs.jnatprod.7b00684
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Note
consideration was apparently given to the latter evidence. It is
worthy to note that the 1H NMR spectra of synthetic
ciliatamide A derivatives are highly reproducible, but the
magnitude of specific rotation is prone to fluctuate due to the
inconsistent purity and errors in weighing of the sample. It is
not possible for us to study the configuration of ciliatamides B
and C due to the lack of authentic samples. The synthetic
studies showed that the sign and magnitude of specific rotations
of ciliatamide A and its isomers depend largely on the
configuration of the N-MePhe residue; the D-N-MePhe residue
gives a positive value, whereas L-N-MePhe gives a negative
value. Because the Lys residue was correctly assigned by
Marfey’s analysis in the two isolation studies,1,3 we anticipate
that the configurational assignment of the Orn residue, a minor
homologue of Lys, was also correct. From these analyses and
the reported positive specific rotation values of both
ciliatamides B and C, we speculate that the N-MePhe residues
in these compounds are in the D-form and the Lys and Orn
residues in the L-form. This study provides a cautionary note
for those who conduct structure elucidation of natural products
that the direct comparison of the synthetic compounds with the
natural product is crucial.11
EXPERIMENTAL SECTION
■
General Experimental Procedures. Optical rotations were
measured on a Jasco DIP-1000 polarimeter. UV spectra were
measured on a Shimadzu Biospec 1600. NMR spectra were recorded
on a JEOL alpha 600 NMR spectrometer. Chemical shifts were
referenced to a solvent peak: δH 3.30 and δC 49.0 (CD3OD). HRESI
mass spectra were measured on a JEOL JMS-T100LC. LC-MS
experiments were performed on a Shimadzu LC-20AD solvent delivery
system and interfaced to a Bruker amaZon SL mass spectrometer.
Ciliatamides A and D were isolated as described in ref 3.
Figure 1. 1H NMR (600 MHz) spectra of 1, 2, and natural ciliatamide
A.
the Marfey’s analysis of ciliatamide D after converting MetO to
MetO2, which permitted us to detect D-MetO2 (Figure S18)
and L-Lys, indicating that the absolute configuration of
ciliatamide D has to be revised as 6 (12R,18S).9
In this study, we showed that ciliatamide A possesses the
(12R,22S)-configurations, which were not identical with either
our previous assignment or those proposed by others through
total synthesis.1−3 In spite of possessing all isomers for
comparison with the reported data of ciliatamide A, Lindsley
et al. drew an incorrect conclusion.2 Because the differences of
the NMR data between the diastereomers were small and the
value of the specific rotation of the synthetic (12R,22R)-isomer
matched well with that of the natural product, more serious
Syntheses of (12S,22S)- and (12R,22S)-Ciliatamide A. To a
solution of L-α-amino-ε-caprolactam hydrochloride (24.5 mg, 0.15
mmol) and N-Boc-N-methyl-L-phenylalanine (41.6 mg, 0.15 mmol)
and 1-[(1-(cyano-2-ethoxy-2-oxoethylideneaminooxy)-
dimethylaminomorpholino)]uronium hexafluorophosphate (COMU)
(64.2 mg, 0.15 mmol) in DMF (1 mL) was added 2,4,6-collidine
(0.060 mL, 0.45 mmol) at 0 °C. The reaction mixture was allowed to
warm to room temperature (rt) and stirred for 21 h. The reaction
mixture was then cooled to 0 °C, quenched with saturated aqueous
Figure 2. LC-MS analyses of the synthetic ciliatamide A isomers and natural product. (a) Chromatogram of a mixture of the four synthetic isomers.
(b) Chromatogram of the natural ciliatamide A. (c) Co-injection of the mixture of the four synthetic isomers and natural ciliatamide A. LC-MS
conditions are as follows: CHIRALPAK ID-3 (ϕ 4.6 × 250 mm); flow rate, 0.8 mL/min.; solvent, MeCN/H2O containing 20 mM NH4HCO3 (7:3).
C
DOI: 10.1021/acs.jnatprod.7b00684
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Note
NaHCO3, extracted with EtOAc, washed with brine, and dried over
MgSO4. The solution was concentrated and dissolved in CH2Cl2 (5
mL), to which was added trifluoroacetic acid (TFA) (1 mL) at 0 °C.
The reaction mixture was stirred at rt for 4 h and concentrated to
afford a crude dipeptide, which was used in the next reaction without
further purification. To a solution of the crude dipeptide, COMU
(64.2 mg, 0.15 mmol), and 9-decenoic acid (0.028 mL, 0.15 mmol) in
dimethylformamide (DMF) (1 mL) was added 2,4,6-collidine (0.040
mL, 0.30 mmol) at 0 °C. The reaction mixture was allowed to warm to
rt and stirred for 16 h. The reaction mixture was then cooled to 0 °C,
quenched with saturated aqueous NaHCO3, extracted with EtOAc,
washed with brine, and dried over MgSO4. The solution was
concentrated, and the residue was purified by RP-HPLC (50−65%
MeCN containing 0.5% AcOH) to afford (12S,22S)-ciliatamide A (1,
29 mg) as a colorless oil. (12R,22S)-Ciliatamide A (2) was synthesized
by using N-Boc-N-methyl-D-phenylalanine instead of N-Boc-N-methyl-
L-phenylalanine.
MeOH, and eluted with MeOH to afford ciliatamide D sulfone (m/z
458 [M + H]+).
Marfey’s Analysis. Ciliatamide D sulfone was dissolved in 6 N
HCl (100 μL) and heated at 110 °C for 2 h. The solution was
concentrated and redissolved in 100 μL of H2O. L-FDAA (1%) in
acetone (100 μL) and 1 M NaHCO3 (10 μL) were added to the
solution. The mixture was heated at 55 °C for 30 min. After cooling to
rt, the reaction mixture was quenched with 2 N HCl (5 μL),
concentrated, and redissolved in MeOH. D- and L-Methionine sulfone
were treated with L-DAA in the same manner. The L-FDAA derivatives
were analyzed by LC-MS [Cosmosil 2.5C18-MS-II (ϕ 2.0 × 100 mm);
flow rate, 0.5 mL/min; solvent, MeOH/MeCN containing 1% formic
acid/H2O (27:5:68)].11 Retention times (tR) of the amino acids were
as follows: N-methyl-D-methionine sulfone derivative (tR = 22.2 min)
and N-methyl-L-methionine sulfone derivative (tR = 23.7 min). The tR
of the derivatized methionine sulfone from ciliatamide D was 22.2 min.
1
(12S,22S)-Ciliatamide A (1): H NMR (600 MHz) and 13C NMR
ASSOCIATED CONTENT
* Supporting Information
■
(150 MHz), Table S21; HRESIMS m/z 464.2867 [M + Na]+ (calcd
S
for C26H40N3O3Na, 464.2884).
1
(12R,22S)-Ciliatamide A (2): H NMR (600 MHz) and 13C NMR
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jnat-
prod.7b00684.
(150 MHz), see Table S32; HRESIMS m/z 464.2866 [M + Na]+
(calcd for C26H40N3O3Na, 464.2884).
Syntheses of (12S,22R)- and (12R,22R)-Ciliatamide A. To a
solution of N-α-Boc-D-lysine (150.0 mg, 0.60 mmol) and COMU
(771.1 mg, 1.80 mmol) in CH2Cl2/DMF (9:1, 120 mL) was added
2,4,6-collidine (0.480 mL, 3.60 mmol) at 0 °C. The reaction mixture
was allowed to warm to rt and stirred for 13 h. The reaction mixture
was then cooled to 0 °C, quenched with saturated aqueous NaHCO3,
extracted with EtOAc, washed with brine, and dried over MgSO4. The
solution was concentrated, the residue was extracted with Et2O and
filtered, and the filtrate was concentrated. The residue was dissolved in
CH2Cl2 (5 mL), to which was added TFA (1 mL) at 0 °C. The
reaction mixture was stirred at rt for 4 h and concentrated to afford a
crude D-α-amino-ε-caprolactam (78.5 mg), which was used in the next
reaction without further purification. (12S,22R)- and (12R,22R)-
Ciliatamide A were synthesized as described above by using D-α-
amino-ε-caprolactam.
NMR and MS data of 1 and 2 (PDF)
AUTHOR INFORMATION
Corresponding Authors
*E-mail: atakada@mail.ecc.u-tokyo.ac.jp.
*E-mail: assmats@mail.ecc.u-tokyo.ac.jp.
ORCID
■
Shigeki Matsunaga: 0000-0002-8360-2386
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
1
(12S,22R)-Ciliatamide A (3): H NMR (600 MHz) and 13C NMR
This work was partly supported by JSPS KAKENHI Grant
Numbers 25252037, 16H04980, and 17H06403 from the
Ministry of Education, Culture, Sports, Science and Technol-
ogy, Japan. We thank Mr. K. Ashizawa of Daicel Chemical
Industries Ltd., for help in the HPLC column screening and
Prof. J. Kimura of Aoyama Gakuin University for valuable
discussion.
(150 MHz) data were identical with those of 2 (Table S3); HRESIMS
m/z 464.2868 [M + Na]+ (calcd for C26H40N3O3Na, 464.2884).
1
(12R,22R)-Ciliatamide A (4): H NMR (600 MHz) and 13C NMR
(150 MHz) data were identical with those of 1 (Table S2); HRESIMS
m/z 464.2868 [M + Na]+ (calcd for C26H40N3O3Na, 464.2884).
LC-MS Analysis of the Four Isomers with a Chiral Stationary
Phase. The mode of separation of 1−4 was screened by a
combination of a chiral stationary phases (CHIRALPAK IA, IB, IC,
ID, IE, IF, IG, AS-RH, AY-RH, OD-RH, OJ-RH, and OZ-RH) and
various solvent systems (mixtures of H2O, MeOH, and MeCN). The
four isomers were separated by LC-MS [CHIRALPAK ID-3 (ϕ 4.6 ×
250 mm); flow rate, 0.8 mL/min; solvent, MeCN/H2O containing 20
mM NH4HCO3 (7:3)].
Preparation of N-Methyl-D- and N-Methyl-L-methionine
Sulfone. To a solution of N-Boc-L-methionine (50 mg) in MeOH
(2 mL) was added Oxone (246 mg), and the solution was stirred at rt
for 1 h. The mixture was diluted with H2O and extracted with CH2Cl2
to give N-Boc-L-methionine sulfone (m/z 282 [M + H]+). To a
solution of the above N-Boc-L-methionine sulfone in THF (2 mL) was
added NaH (15 mg) and MeI (165 μL), and the mixture was stirred at
rt overnight. The reaction mixture was diluted with H2O and extracted
with EtOAc. The organic layer was concentrated to afford N-Boc-N-
methyl-L-methionine sulfone (m/z 296 [M + H]+), which was
dissolved in 6 N HCl (100 μL) and heated at 110 °C for 2 h. The
solution was concentrated to provide N-methyl-L-methionine sulfone
(m/z 196 [M + H]+). N-Methyl-D-methionine sulfone was synthesized
from N-Boc-D-methionine in the same manner.
REFERENCES
■
(1) Nakao, Y.; Kawatsu, S.; Okamoto, C.; Okamoto, M.; Matsumoto,
Y.; Matsunaga, S.; van Soest, R. W. M.; Fusetani, N. J. Nat. Prod. 2008,
71, 469−472.
(2) Lewis, J. A.; Daniels, R. N.; Lindsley, C. W. Org. Lett. 2008, 10,
4545−4548.
(3) Imae, Y.; Takada, K.; Okada, S.; Ise, Y.; Yoshimura, H.; Morii, Y.;
Matsunaga, S. J. Nat. Prod. 2013, 76, 755−758.
(4) Avula, K.; Mohapatra, D. K. Tetrahedron Lett. 2016, 57, 1715−
1717.
(5) In a symposium paper, it was mentioned that the synthetic 1
exhibited the specific rotation value as reported by Lindsley et al.
Akiyama, S.; Nakao, Y.; Matsumoto, Y.; Goto, Y.; Sanjoba, C.; Osada,
Y.; Umehara, M.; Kimura, J. Symposium Paper of 55th Symposium on
The Chemistry of Natural Products 2013, 55, 591−596.
(6) 1H NMR signal assignments for the ε-caprolactam residue
reported in refs 1 and 3 have been corrected.
(7) This analysis contradicts with the one reported by Lindsley et al.,2
who stated that the NMR spectra of the synthetic 1 were identical with
those of the natural product.
Oxidation of Ciliatamide D. To a solution of ciliatamide D (50
μg) in MeOH (0.5 mL) was added Oxone (1 mg in 100 μL of H2O),
and the mixture was stirred at rt for 2 h. The solution was diluted with
H2O, applied on InertSep PLS-2 (GL Science), washed with 10%
(8) Zhang, T.; Nguyen, D.; Franco, P. J. Chromatogr. A 2008, 1191,
214−222.
D
DOI: 10.1021/acs.jnatprod.7b00684
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Note
(9) In our previous study of ciliatamide D, the Marfey’s analysis was
conducted by using an ODS column (Cosmosil MS-II) as the
stationary phase and a gradient elution of 10−50% MeCN containing
0.5% AcOH as the mobile phase. Under this condition DAA-D-MetO2
and FDAA-L-MetO2 were barely separable. The assignment of the
configuration of MetO2 was conducted by considering the difference
of the retention times observed in experiments carried out on separate
days as the true difference, leading to the misassignment. In the
current study HPLC of the DAA derivatives was conducted with the
same stationary phase but with a different mobile phase [an isocratic
elution with
a mixture of MeOH/MeCN/H2O/HCOOH
(27:5:68:1)],10 which gave two distinct peaks for the two DAA
derivatives and permitted us to unambiguously assign the chirality of
MetO2 as D.
(10) Vijayasarathy, S.; Prasad, P.; Fremlin, L. J.; Ratnayake, R.; Salim,
A. A.; Khalil, Z.; Capon, R. J. J. Nat. Prod. 2016, 79, 421−427.
(11) Kuranaga, T.; Mutoh, H.; Sesoko, Y.; Goto, T.; Matsunaga, S.;
Inoue, M. J. Am. Chem. Soc. 2015, 137, 9443−9451.
E
DOI: 10.1021/acs.jnatprod.7b00684
J. Nat. Prod. XXXX, XXX, XXX−XXX