Elucidation and Total Synthesis of the Correct Structures of Tridecapeptides Yaku'amides a and B. Synthesis-Driven Stereochemical Reassignment of Four Amino Acid Residues

Article abstract of DOI:10.1021/jacs.5b05550

Yaku'amides A (1) and B (2) possess four α,β-dehydroamino acid residues in their linear tridecapeptide sequence and differ in their residue-3 (Gly for 1 and Ala for 2). The highly unsaturated peptide structure, characteristic cytotoxicity profile, and ext

Full text of DOI:10.1021/jacs.5b05550

Article
pubs.acs.org/JACS
Elucidation and Total Synthesis of the Correct Structures of
Tridecapeptides Yaku’amides A and B. Synthesis-Driven
Stereochemical Reassignment of Four Amino Acid Residues
Takefumi Kuranaga,† Hiroyuki Mutoh,† Yusuke Sesoko,† Tomomi Goto,†,‡ Shigeki Matsunaga,§
and Masayuki Inoue*,†
†Graduate School of Pharmaceutical Sciences, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
‡Research Foundation ITSUU Laboratory, 2-28-10 Tamagawa, Setagaya-ku, Tokyo 158-0094, Japan
§Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan
S
* Supporting Information
ABSTRACT: Yaku’amides A (1) and B (2) possess four α,β-
dehydroamino acid residues in their linear tridecapeptide
sequence and differ in their residue-3 (Gly for 1 and Ala for
2). The highly unsaturated peptide structure, characteristic
cytotoxicity profile, and extreme scarcity from natural sources
motivated us to launch synthetic studies of 1 and 2. Here, we
report the total synthesis of the originally proposed structure of
yaku’amide B (2a) by applying the route to 1a, which was
previously established in our group. However, this accomplish-
ment only proved that 2a and natural 2 were structurally different
and prompted investigations directed toward determining the
true structure of 2. Extensive Marfey’s analyses of minute
amounts of natural 2 and its degradation products presented us the possible stereoisomers, all of which were synthetically
prepared for chromatographic comparison with the authentic fragments of 2. Based on this detective work, we proposed a
corrected structure for yaku’amide B (2c), in which the orders of residues-7 and -8 and residues-11 and -12 are reversed. Finally,
the total synthesis of 2c led to confirmation of its structural identity. Moreover, the revised structure of yaku’amide A (1c) was
constructed by switching Ala-3 to Gly-3 and was found to be chromatographically matched with the re-isolated natural 1. The
present work demonstrated the high reliability and sensitivity of the MS- and LC-based structural analyses and the indispensable
role of chemical synthesis in structural elucidation of scarce natural products.
Natural peptides 1 and 2 were reported to exhibit
exceptionally potent cytotoxicity toward P388 murine leukemia
cells (IC50 = 8.5 and 2.4 nM, respectively). Moreover,
examination of the growth-inhibitory activity of 1 against a
panel of 39 human cancer cell lines (JFCR39)3 revealed its
distinct profile from those of 38 anticancer drugs. Therefore, 1
is likely to exert its cytotoxicity through a potentially novel
mode of action.4 Further evaluation of the biological activities
of these potential anticancer agents 1 and 2 has been impeded
because minute amounts of the isolated yaku’amides existed
after 2010. As compounds 1 and 2 were utilized for the
structural and biological studies, none of 1 and approximately
0.1 mg of 2 remained.
INTRODUCTION
■
Yaku’amides A (1, Figure 1) and B (2) were isolated from a
deep-sea sponge Ceratopsion sp. collected at Yakushinsone in
the East China Sea as minute components (1, 1.3 mg; 2, 0.3
mg).1,2 Matsunaga and co-workers disclosed their unprece-
dented tridecapeptide structures in 2010. The linearly
assembled 13 amino acid residues contain one α,β-dehydrova-
line (ΔVal, residue-13), three α,β-dehydroisoleucine (ΔIle,
residues-2, -4, and -9), and seven other nonproteinogenic
amino acids (residues-1, -5, -6, -7, -8, -10, and -12) and are
capped with an N-terminal acyl group (NTA) and C-terminal
amine (CTA). The structural difference between 1 and 2
resides in their residues-3 (Gly for 1 and Ala for 2). The
geometries of the double bonds of the three ΔIle-2, -4, and -9
were determined on the basis of NOESY data, and the absolute
configurations of the component amino acids and CTA were
established by Marfey’s analysis. Taking these data together,
Matsunaga proposed the stereostructures of 1 and 2 to be 1a/
1b and 2a/2b, respectively. Only the C4-stereochemistry of
NTA was left undefined.
To supply sufficient amounts of 1 and 2, their chemical
constructions are the sole realistic solution. Motivated by their
unique structures and bioactivities, we started the synthetic
studies of yaku’amides, and accomplished the total synthesis of
1a and its C4-epimer 1b in 2013 (Figure 1c).5−7 The α,β-
Received: May 29, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.5b05550
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. (a) The originally proposed structures of yaku’amide A (1a/1b) and B (2a/2b). (b) The revised structures of yaku’amides A (1c) and B
(2c). The corrected structures are highlighted in yellow. (c) Structures of compounds 3−7 used for total synthesis of 1a/1b and 2a/2b. (d) Cu-
mediated enamide formation used in this study. Boc = tert-butoxycarbonyl; ΔIle = α,β-dehydroisoleucine; ΔVal = α,β-dehydrovaline; OHIle =
hydroxyisoleucine; OHVal = hydroxyvaline.
unsaturated amino acid residues of fragments 4, 5, and 7 were
constructed in a geometrically selective fashion by employing
the powerful Cu-mediated cross-coupling reactions (e.g., A + B
→ C, Figure 1d).8 Stepwise assembly of the four synthetic
fragments 3a/3b, 4, 5, and 7 then resulted in the entire
RESULTS AND DISCUSSION
■
Total Syntheses of the Two Possible Proposed
Structures of Yaku’amide B. To validate the unestablished
C4-stereochemisty at NTA, we initially planned to synthesize
the two C4-isomers of the proposed structures of yaku’amide B
(2a and 2b, Figure 1). The fragments that were used in the
total synthesis of 1a and 1b would be directly utilized except for
5.5 Thus, the requisite fragment 6 was first prepared prior to
assembly of 2a/2b (Scheme 1). When Boc-L-Ala-NH2 (8) and
ethyl (E)-2-iodo-3-methylpent-2-enoate (9) were treated with
catalytic CuI, N,N′-dimethylethylenediamine, and Cs2CO3 in
dioxane at 70 °C, the hindered C−N bond was stereoselectively
formed to afford dipeptide 10 with the E-didehydroisoleucine
moiety.9,10 Saponification of 10 afforded carboxylic acid 11,
which was transformed into the requisite fragment 6 in three
steps: (i) allyl ester formation, (ii) Boc introduction at the
enamide nitrogen atom, (iii) Pd-catalyzed removal of the allyl
group at the C-terminus.11,12
Total syntheses of 2a and 2b were accomplished by
elongation from nonapeptide 7 by the stepwise condensation
with fragments 6, 4, and 3a/3b (Scheme 2). Two cycles of
TFA-promoted Boc removal and PyBOP13/HOAt14-mediated
amidation were performed on 7 using dipeptides 6 and 4. As a
result, tridecapeptide 13 was obtained in 51% yield in 4 steps (7
→ 12 → 13). Next, treatment of 13 with TFA liberated the N-
terminal amine, which was conjugated with NTA fragments 3a
and 3b by the action of COMU15 and 2,4,6-collidine to deliver
the two C4-epimers 2a and 2b, respectively. Hence the two
1
structure of 1a/1b. Since the H and 13C NMR spectra of
natural 1 matched those of 1a in comparison to 1b, we
concluded that natural 1 possessed the C4S-stereochemistry of
1a.
Herein, we report the total syntheses of the C4-epimers of
the originally proposed structures of yaku’amides B (2a and 2b,
Figure 1a) by application of the route to 1a and 1b.
Chromatographic comparison between natural 2 and 2a/2b
validated that 2a/2b did not represent the true structure of the
natural substance. This recognition prompted investigations
directed at determining the true structure of natural 2. The
structural information was gathered through degradative studies
of the minute amounts of natural 2 available and synthetic
preparation of the possible structures of the degraded partial
structures. Combination of the degradation and synthesis
allowed us to revise the structure of natural 2 to be 2c (Figure
1b), which was chemically constructed to confirm its structural
authenticity. These studies further led to correction of the
structure of natural yaku’amide A. Thus, the revised structure
1c was synthesized and was shown to be chromatographically
identical with natural 1, which was re-isolated in this study.
Below we present the details of these synthetic and analytical
efforts.
B
DOI: 10.1021/jacs.5b05550
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Scheme 1. Synthesis of Dipeptide 6
Figure 2. UHPLC charts of natural 2 and synthetic 2a/2b. (a)
Synthetic 2a/2b. (b) Authentic 2. (c) Co-injection of 2 and 2a/2b.
Column: Accucore phenyl hexyl 2.1 × 150 mm; column oven: 45 °C;
flow rate: 0.25 mL/min; eluent: 35% 1-PrOH/H2O containing 1%
AcOH; detection: 254 nm.
Re-elucidation of the Planar Structure of Natural 2.
Since the proposed structure could contain one or more errors
in unknown positions, the planar structure of the natural
yaku’amide B (2) was first re-evaluated. The sequence of the
amino acid residues, NTA, and CTA was verified by
combination of the MS and NMR data. As shown in Figure
3, the in-source MS fragmentation of 2 enabled the highly
sensitive peptide sequencing of 2 by detection of the diagnostic
MS fragments, corroborating the sequence from residue-2 to
residue-10. Although connectivities of NTA/residue-1 and
residue-11/-12/-13/CTA were not established by the MS
experiments, the reported interresidue HMBC correlations (red
arrows) clearly indicated the correct order of these
components.1 On the other hand, the isolation paper
unambiguously assigned the double-bond geometries of the
three ΔIle moieties of natural yaku’amide A (1) based on the
NOESY correlations (blue arrows). 1H and 13C NMR chemical
a
DMF = N,N-dimethylformamide; DMAP = N,N-dimethyl-4-amino-
pyridine; THF = tetrahydrofuran.
possible structures of the originally proposed structures of
yaku’amide B (2a and 2b) were chemically constructed.
However, neither of the final products 2a and 2b matched
the natural yaku’amide B (2) by the chromatographic analysis.
Specifically, injection of 2a/2b and 2 to ultrahigh-performance
liquid chromatography (UHPLC) afforded peaks of signifi-
cantly different retention times (Figure 2a,b). Co-injection re-
inforced the distinct nature of these peaks (Figure 2c),
clarifying that 2a or 2b did not represent the true structure
of natural yaku’amide B. This recognition set in motion a next
wave of investigations directed at determining the true structure
of yaku’amide B.
a
Scheme 2. Total Synthesis of the Proposed Structures of Yaku’amide B (2a/2b)
a
TFA = trifluoroacetic acid; PyBOP = benzotriazol-1-yl-oxytripyrrolidinophosphonium hexafluorophosphate; HOAt = 1-hydroxy-7-azabenzotriazole;
COMU = (1-cyano-2-ethoxy-2-oxoethylidenaminooxy)dimethylamino-morpholino-carbenium hexafluorophosphate.
C
DOI: 10.1021/jacs.5b05550
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 3. Confirmation of the planar structure of natural yaku’amide B (2).
Scheme 3. Degradation of Natural Yaku’amide B (2) and Marfey’s Analyses of Total and Partial Hydrosate
shifts corresponding to ΔIle-2, ΔIle-4, and ΔIle-9 are almost
identical among 1, natural 2, and synthetic 2a/2b, indicating
that the stereochemistry of these tetra-substituted olefins is the
same. Taken together, these data confirmed that the planar
structure of natural 2 contained no mis-assignments.
(COSMOSIL π-NAP) was used instead of the standard ODS
column to attain the highly resolved HPLC peaks. In fact, the
four FDAA-derivatives of L-Ile/L-allo-Ile and D-Ile/D-allo-Ile
were found to be separable only upon use of COSMOSIL π-
NAP, presumably due to its stronger attractive interaction with
the dinitrophenyl moieties (Figure 4). Consequently, the L- and
D-FDAA-derivatives of (2R,3R)-/(2S,3R)-/(2R,3S)-/(2S,3S)-
hydroxyisoleucine (OHIle), L-/D-Ala, L-/D-Val, L-Ile/L-allo-Ile/
D-Ile/D-allo-Ile, L-/D-hydroxyvaline (OHVal), and (R)-/(S)-
CTA were used as the standard samples, and the obtained
retention times of these compounds were compared with the L-
and D-FDAA-treated hydrolysate of natural 2 (Marfey’s analysis
Marfey’s Analysis of Natural 2 and its Degradation
Product 15. Having confirmed the planar structure of 2, we
decided to re-establish the absolute configurations of all the 13
stereocenters within the entire structure of 2. In doing so,
degradative studies of natural 2 were executed to collect the
stereochemical information by Marfey’s analyses.16 This turned
out to be an extremely challenging task, because a maximum of
0.1 mg of 2 could be utilized. First, natural 2 (0.04 mg in total)
was subjected to aqueous 6 M HCl solution at 110 °C for 12 h
to induce complete hydrolysis, leading to a mixture of the
component amino acids (Scheme 3). The resulting amino acids
were treated with 1-fluoro-2,4-dinitrophenyl-5-alanine amide
(FDAA), and then the derivatives were analyzed by the LC-MS
method. To realize highly sensitive analysis using a trace
amount of the hydrolysate, the following experimental
conditions were designed. Namely, the amino acids were
derivatized by both enantiomers of FDAA to cross-check the
information, and the online MS detection of the analyte was
employed to acquire the structural data in real-time.17 In
addition, a naphthylethyl-conjugated silica gel column
Figure 4. UHPLC charts of the L-FDAA derivatives of L-Ile, L-allo-Ile,
D-Ile, and D-allo-Ile. FDAA = 1-fluoro-2,4-dinitrophenyl-5-alanine
amide. Column: COSMOSIL 2.5 π-NAP 2.0 × 100 mm; column oven:
40 °C; flow rate: 0.4 mL/min; eluent: linear gradient from 10% to 50%
MeCN/H2O containing 0.05% TFA over 32 min; detection: 340 nm.
D
DOI: 10.1021/jacs.5b05550
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
A). The experiments revealed the structures of residues-1 and
-6 to be (2S,3R)-OHIle and D-allo-Ile, respectively, thus
establishing four of the stereocenters of 2. Furthermore, the
L/D ratios of the three component amino acids were
determined as follows: Ala (L/D = 1:1), Val (L/D = 1:2), and
OHVal (L/D = 1:1). On the other hand, the FDAA-derivatives
of CTA were not detected because of its instability toward the
highly acidic conditions.
Scheme 4. Syntheses of Eight Possible Isomers of 16
When another 0.01 mg of 2 was treated with aqueous 1 M
HCl solution at 70 °C for 36 h, the three fragments 14, 15, and
16 were produced (Scheme 3). Interestingly, this controlled
hydrolysis transformed the enamide moieties of ΔIle to the N-
terminal dicarbonyl structures of 14 and 15 (1:1 stereoisomers
at their branched carbons). Therefore, 15 was utilized for the
optimized Marfey’s analysis. After complete hydrolysis of 15
with 6 M HCl, the lysate was functionalized with L-FDAA to
identify the presence of D-Ala at residue-10 (Marfey’s analysis
B). Considering that 2 possesses a 1:1 ratio of L- and D-Ala, the
structure of residue-3 was determined to be L-Ala at this stage.
Therefore, the degradative studies indicated that the originally
assigned absolute stereochemistry of residues-1, -3, -6, and -10
should be correct [(2S, 3R)-OHIle-1, L-Ala-3, D-allo-Ile-6, D-
Ala-10, respectively].
Syntheses of Eight Possible Isomers of 16. The
structural ambiguities in 2 resided in the seven absolute
configurations of NTA, residues-5, -7, -8, -11, and -12, and
CTA. Since fragment 16 generated by the partial hydrolysis
contained residues-11 and -12 and CTA (Scheme 3), we
planned to determine the three unknown stereocenters of 16
by chromatographic comparison with the eight possible isomers
of synthetic 16a−h (Scheme 4). Among them, compound 16a
corresponded to the proposed structure and was already
prepared via 21a.5
A divergent synthesis of other seven isomers 16b−h is
summarized in Scheme 4.18 Four isomers of 19 were prepared
utilizing the Cu-catalyzed coupling reaction of vinyl iodide 17a/
17b and primary amide 18a/18b. TFA-treatment of 19
removed the Boc group, and the resulting amine reacted with
Boc-L- or D-Val-OH in the presence of COMU to generate the
seven isomers of 20. Next, Nα-deprotection of 20 and COMU-
promoted amidation with Boc-D-Ala-OH gave rise to 21b−h. It
is of note that use of COMU in these two amidations was
important to suppress the Cα-epimerization of the sterically
hindered amino acids. Finally, the obtained 21b−h were
converted to the requisite amines 16b−h.
The synthesized eight possible isomers 16a−h were then
chromatographically compared with authentic 16 by LC-MS
experiments. After extensive screening of the HPLC conditions,
16a−h was found to exhibit eight separate peaks as shown in
Figure 5a. The retention time of authentic 16 matched that of
synthetic 16d, but not those of the other seven stereoisomers,
by separate injection (Figure 5b) as well as by coinjection
(Figure 5c). Consequently, the absolute structures of residues-
11, and -12 and CTA were determined to be D-Val, L-Val, and
(S)-CTA, respectively. As Marfey’s analysis of the total
hydrolysate shed light on the existence of one L-Val and two
D-Val in 2 (vide supra), residue-5 was fixed to be D-Val.
Through these experiments, four of the seven stereocenters
were resolved.
Figure 5. LC-MS charts of authentic 16 and synthetic 16a−h. (a)
Mixture of 16a−h. (b) Authentic 16. (c) Co-injection of 16 and 16a−
h. Column: COSMOSIL 2.5 π-NAP 2.0 × 100 mm; column oven: 30
°C; flow rate: 0.3 mL/min; eluent: linear gradient from 10% to 15%
MeCN/H2O containing 0.05% TFA over 18 min; detection: m/z =
497.
the next issue for the structural determination. Thus, two
possible isomers 14a and 14b with these residues were planned
to be prepared and analyzed by UHPLC along with the
degradatively obtained 14. Compound 21d, which was
prepared according to Scheme 4, was elongated to 14a and
14b in 10 steps (Scheme 5). Treatment of 21d with TFA
liberated the amine, and condensation of the amine with
enantiomeric dipeptides 22a5 and 22b by using PyBOP and
HOAt furnished hexapeptides 23a and 23b, respectively. Three
rounds of TFA-promoted Nα-deprotection and COMU-
mediated amidation transformed 23a and 23b into non-
apeptides 25a and 25b, respectively, when 24a/24b5 (24a for
25a, 24b for 25b), Boc-D-allo-Ile, and Boc-D-Val were
sequentially used. Finally, TFA treatment of 25a and 25b,
Syntheses of Two Possible Isomers of 14. Marfey’s
analysis A specified the presence of one L- and one D-OHVal
within the peptide sequence of 2 (Scheme 3). Accordingly,
correct placement of L- and D-OHVal to residues-7 and -8 was
E
DOI: 10.1021/jacs.5b05550
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 5. Syntheses of Two Possible Isomers of 14
followed by amidation with acid 26 by the action of PyBOP/
HOAt, gave rise to ketones 14a and 14b.
Figure 6a−c depicts the UHPLC charts of the two synthetic
isomers 14a and 14b, authentic 14, and their combination,
2a/2b were corrected to be D-OHVal-7, L-OHVal-8, D-Val-11,
and L-Val-12 of 2c/2d, respectively.
Total Synthesis of the Two Possible Revised
Structures of Yaku’amide B. The possible structures of
natural yaku’amide B (2) were narrowed down to the two C4-
epimers 2c and 2d. Determination of the absolute stereo-
chemistry of the last remaining C4-stereocenter of NTA
necessitated the total synthesis of 2c and 2d (Scheme 6).
Similar to the route to 2a and 2b in Scheme 2, the stepwise Nα-
deprotection/amidation from nonapeptide 25b delivered the
target molecules in six transformations. Nonapeptide 25b was
deprotected and coupled with acid 6 by the action of PyBOP
and HOAt to yield the undecapeptide, which underwent in situ
deprotection by TFA to deliver amine 27.19 Condensation of
27 with 4 in turn delivered tridecapeptide 29. Lastly, TFA
induced the detachment of the Boc group of 29 to afford the
amine, which was separately coupled with enantiomeric
carboxylic acids 3a and 3b using COMU and 2,4,6-collidine
to provide 2c and 2d, respectively.
Figure 6. UHPLC charts of authentic 14 and synthetic 14a/14b. (a)
Synthetic 14a/14b. (b) Authentic 14. (c) Co-injection of 14 and 14a/
14b. Column: Accucore phenyl hexyl 2.1 × 150 mm; column oven: 40
°C; flow rate: 0.5 mL/min; eluent: 40% MeCN/H2O containing
0.05% TFA; detection: 226 nm.
When the UHPLC experiments of natural 2 and the two
synthetic isomers 2c and 2d were conducted (Figure 7), the
retention time of 2c with the C4S-stereochemistry was found to
be identical to natural 2, thus defining the entire stereostructure
of natural yaku’amide B to be that shown in Scheme 6.
respectively. Each compound gave two peaks, because they
were 1:1 diastereomeric mixtures at the methyl group of the N-
terminal acyl portion. Nevertheless, it was clear that the
retention time of 14b was identical to 14. Hence, residues-7
and -8 were established to be D-OHVal and L-OHVal,
respectively.
The series of degradative and synthetic efforts described
herein permitted us to unambiguously determine the entire
stereostructure of natural yaku’amide B except for the NTA
moiety. Most importantly, the revised structure 2c/2d differs
from the originally proposed structure 2a/2b in the orders of
residues-7 and -8 and residues-11 and -12 (Scheme 6).
Therefore, L-OHVal-7, D-OHVal-8, L-Val-11, and D-Val-12 of
1
Additionally, the H and 13C NMR chemical shifts of synthetic
2c were assigned based on HMBC and HSQC correlations and
were in accordance with those of natural 2. Considering the
NMR spectra of 2c changed depending on the conditions in the
NMR tube (e.g., concentration, temperature, and purity), the
1H NMR spectra of the mixture of natural 2 and synthesized 2c
were also obtained for further confirmation of their structural
identity.
F
DOI: 10.1021/jacs.5b05550
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 6. Total Synthesis of the Revised Structures of Yaku’amide B (2c), A (1c), and their C4-Epimers (2d, 1d)
After optimizing the purification protocol, <0.1 mg of natural 1
was isolated from the side fraction, which was then used for the
UHPLC experiments. As clearly shown in Figure 8, retention
times of re-isolated 1 and previously synthesized 1a/1b were
completely different, suggesting assignment errors in the
proposed structure of yaku’amide A.
Figure 7. UHPLC charts of natural 2 and synthetic 2c/2d. (a)
Synthetic 2c/2d. (b) Authentic 2. (c) Co-injection of 2 and 2c/2d.
Column: Accucore C18 2.1 × 150 mm; column oven: 40 °C; flow
rate: 0.5 mL/min; eluent: 70% MeCN/H2O containing 0.05% TFA;
detection: 226 nm.
Re-isolation, Structural Revision, and Total Synthesis
of Yaku’amide A. Our previous study reported the total
synthesis of the originally determined structure of yaku’amide A
C4S-1a and C4R-1b and permitted the assignment of the C4S-
stereochemistry of NTA based on the similarities of the NMR
spectra of 1a with those of natural 1.5 However, the
aforementioned structural revision of yaku’amide B (2)
jeopardized the credibility of the original structure of natural
1 and thus forced us into defining the correct stereostructure of
1. Since variable NMR spectra of peptide natural products often
complicate structural determination,20 our NMR-based com-
parison of 1 and 1a could have led to the incorrect conclusion.
Chromatographic comparison has been demonstrated to be a
powerful method for the structural assignment of large
peptides. No sample of authentic 1 remained, and thus re-
isolation of authentic 1 was required to apply the chromato-
graphic method. In this situation, we focused our attention on
the remaining side-fraction in the original isolation procedures.1
Figure 8. UHPLC charts of re-isolated 1 and synthetic 1a/1b. (a)
Synthetic 1a/1b. (b) Authentic 1. (c) Co-injection of 1 and 1a/1b.
Column: Accucore phenyl hexyl 2.1 × 150 mm; column oven: 45 °C;
flow rate: 0.25 mL/min; eluent: 30% 1-PrOH/H2O containing 1%
AcOH; detection: 254 nm.
The above results led us to synthesize 1c, which shares all the
structure of the revised structure of yaku’amide B except
residue-3 (Gly for 1c and L-Ala for 2c). Scheme 6 shows the
total synthesis of 1c along with its C4-epimer 1d. The common
intermediate 25b underwent chain elongation through
repeating the Nα-deprotection and amidation by using 5, 4,
and 3a/3b to yield 1c and 1d (25b → 28 → 30 → 1c/1d).
The retention time of C4S-epimer 1c was indeed identical to
natural 1 (Figure 9), while 1 and 1d exhibited distinct
chromatographic behavior. Therefore, the absolute structure of
yaku’amide A should be revised from 1a to 1c.
G
DOI: 10.1021/jacs.5b05550
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was financially supported by the Funding
Program for a Grant-in-Aid for Scientific Research (A) to M.I.
REFERENCES
Figure 9. UHPLC charts of re-isolated 1 and synthetic 1c/1d. (a)
Synthetic 1c/1d. (b) Authentic 1. (c) Co-injection of 1 and 1c/1d.
Column: Accucore C18 2.1 × 150 mm; column oven: 40 °C; flow
rate: 0.5 mL/min; eluent: 65% MeCN/H2O containing 0.05% TFA;
detection: 226 nm.
■
(1) Ueoka, R.; Ise, Y.; Ohtsuka, S.; Okada, S.; Yamori, T.; Matsunaga,
S. J. Am. Chem. Soc. 2010, 132, 17692.
(2) For selected reviews of marine natural products, see:
(a) Yasumoto, T.; Murata, M. Chem. Rev. 1993, 93, 1897. (b) Fusetani,
N.; Matsunaga, S. Chem. Rev. 1993, 93, 1793. (c) Bewley, C. A.;
Faulkner, D. J. Angew. Chem., Int. Ed. 1998, 37, 2162. (d) Blunt, J. W.;
Copp, B. R.; Keyzers, R. A.; Munro, M. H. G.; Prinsep, M. R. Nat.
Prod. Rep. 2015, 32, 116.
(3) (a) Yamori, T. Cancer Chemother. Pharmacol. 2003, 52, S74.
(b) Yaguchi, S.; Fukui, Y.; Koshimizu, I.; Yoshimi, H.; Matsuno, T.;
Gouda, H.; Hirono, S.; Yamazaki, K.; Yamori, T. J. Natl. Cancer Inst.
2006, 98, 545. (c) Yamori, T.; Matsunaga, A.; Sato, S.; Yamazaki, K.;
Komi, A.; Ishizu, K.; Mita, I.; Edatsugi, H.; Matsuba, Y.; Takezawa, K.;
Nakanishi, O.; Kohno, H.; Nakajima, Y.; Komatsu, H.; Andoh, T.;
Tsuruo, T. Cancer Res. 1999, 59, 4042.
(4) For recent reviews on application of natural products to drug
discovery, see: (a) Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2007, 70,
461. (b) Cragg, G. M.; Grothaus, P. G.; Newman, D. J. Chem. Rev.
2009, 109, 3012. (c) Mishra, B. B.; Tiwari, V. K. Eur. J. Med. Chem.
2011, 46, 4769. (d) Butler, M. S.; Robertson, A. A. B.; Cooper, M. A.
Nat. Prod. Rep. 2014, 31, 1612.
(5) Kuranaga, T.; Sesoko, Y.; Sakata, K.; Maeda, N.; Hayata, A.;
Inoue, M. J. Am. Chem. Soc. 2013, 135, 5467.
(6) For synthetic studies on yaku’amides from other groups, see:
(a) Ma, Z.; Naylor, B. C.; Loertscher, B. M.; Hafen, D. D.; Li, J. M.;
Castle, S. L. J. Org. Chem. 2012, 77, 1208. (b) Saavedra, C. J.; Boto, A.;
́
Hernandez, R. Org. Lett. 2012, 14, 3788. (c) Ma, Z.; Jiang, J.; Luo, S.;
Cai, Y.; Cardon, J. M.; Kay, B. M.; Ess, D. H.; Castle, S. L. Org. Lett.
2014, 16, 4044.
(7) For recent accounts on total synthesis and biological evaluation
of peptidic natural products from our laboratory, see: polytheonamide
B (a) Itoh, H.; Inoue, M. Acc. Chem. Res. 2013, 46, 1567. antillatoxin
(b) Inoue, M. Proc. Jpn. Acad., Ser. B 2014, 90, 56.
(8) For reviews on Cu-mediated enamide formation, see: (a) Evano,
G.; Blanchard, N.; Toumi, M. Chem. Rev. 2008, 108, 3054. (b) Ma, D.;
Cai, Q. Acc. Chem. Res. 2008, 41, 1450. (c) Surry, D. S.; Buchwald, S.
L. Chem. Sci. 2010, 1, 13. (d) Evano, G.; Theunissen, C.; Pradal, A.
Nat. Prod. Rep. 2013, 30, 1467. (e) Kuranaga, T.; Sesoko, Y.; Inoue, M.
Nat. Prod. Rep. 2014, 31, 514.
(9) (a) Klapars, A.; Huang, X.; Buchwald, S. L. J. Am. Chem. Soc.
2002, 124, 7421. (b) Jiang, L.; Job, G. E.; Klapars, A.; Buchwald, S. L.
Org. Lett. 2003, 5, 3667.
(10) Yamashita, T.; Matoba, H.; Kuranaga, T.; Inoue, M. Tetrahedron
2014, 70, 7746.
Cytotoxicity of the Revised Structures of Yaku’amides
A and B. A cytotoxicity assay of the revised structures 1c and
2c against P388 mouse leukemia cells was performed using the
XTT method.21 Yaku’amides A (1c) and B (2c) both displayed
extremely potent cytotoxicity (1c: 0.88 nM and 2c: 0.51 nM),
corroborating the authenticity of our structural revisions.
CONCLUSION
■
In summary, the structural revision and total synthesis of
yaku’amides A and B were achieved by judicious combination
of natural product degradation, MS and chromatographic
analyses, and chemical synthesis.22 The total synthesis of the
proposed structure of yaku’amide B (2a) proved its mistaken
structural identity. The structural information on 2 was then
acquired by applying the highly sensitive Marfey’s analysis of
the minute amounts of natural 2 and its fragments. The
potential substructures suggested by these degradative studies
were chemically synthesized and chromatographically evaluated
to obtain the two possible revised structures 2c and 2d. Finally,
the total synthesis defined the correct structure of yaku’amide B
to be 2c. Moreover, re-isolation of natural 1, total synthesis of
1c, and their chromatographic comparison indicated 1c as the
revised structure of yaku’amide A. Both of the revised structures
possess the inverse configurations at the Cα-positions of the
four amino acid residues (residues-7, -8, -11, and -12) of the
original structures and S-configuration at C4 of NTA. These
studies demonstrated the efficiency and robustness of our
convergent strategy for assembly of the various yaku’amide
stereoisomers and the high reliability and sensitivity of the MS-
and LC-based structural analyses. Moreover, the present work
showed once again the indispensable role of chemical synthesis
in structural elucidation of scarce, but highly valuable, natural
products. Establishment of the true structures of yaku’amides
provides a starting point for systematic synthesis of analogues,
detailed structure−activity relationship studies, and elucidation
of the unknown mode of action of these potent cytotoxins.
Such investigations are currently underway in our laboratory.
(11) Friedrich-Bochnitschek, S.; Waldmann, H.; Kunz, H. J. Org.
Chem. 1989, 54, 751.
(12) To prevent the palladium hydride-promoted isomerization of
the double bond, 2-methyl-2-butene was utilized as its trapping agent.
(13) Coste, J.; Le-Nguyen, D.; Castro, B. Tetrahedron Lett. 1990, 31,
205.
(14) Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397.
(15) El-Faham, A.; Funosas, R. S.; Prohens, R.; Albericio, F. Chem. -
Eur. J. 2009, 15, 9404.
(16) (a) Marfey, P. Carlsberg Res. Commun. 1984, 49, 591.
(b) Bhushan, R.; Bruckner, H. J. Chromatogr. B: Anal. Technol.
Biomed. Life Sci. 2011, 879, 3148.
ASSOCIATED CONTENT
* Supporting Information
■
S
Characterization data for all new compounds and experimental
procedures. The Supporting Information is available free of
charge on the ACS Publications website at DOI: 10.1021/
jacs.5b05550.
̈
AUTHOR INFORMATION
Corresponding Author
*inoue@mol.f.u-tokyo.ac.jp
■
(17) Takada, K.; Ninomiya, A.; Naruse, M.; Sun, Y.; Miyazaki, M.;
Nogi, Y.; Okada, S.; Matsunaga, S. J. Org. Chem. 2013, 78, 6746.
(18) See Supporting Information for details.
H
DOI: 10.1021/jacs.5b05550
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(19) One-pot deprotection procedure was adopted after the coupling
with 6, because the aqueous work-up appeared to cleave the Boc-
protected amide bond between residues-3 and -4. For facile cleavage of
the carbamate protected amide, see: Grehn, L.; Gunnarsson, K.;
Ragnarsson, U. J. Chem. Soc., Chem. Commun. 1985, 1317.
(20) Peptide natural products give variable NMR spectra depending
on concentration, temperature, and purity. Thus, structural identity
between natural and synthetic compounds is often difficult to be
validated by the NMR spectral comparison. For recent examples, see:
(a) Zou, B.; Long, K.; Ma, D. Org. Lett. 2005, 7, 4237. (b) Sugiyama,
H.; Watanabe, A.; Teruya, T.; Suenaga, K. Tetrahedron Lett. 2009, 50,
7343. (c) Ma, B.; Litvinov, D. N.; He, L.; Banerjee, B.; Castle, S. L.
Angew. Chem., Int. Ed. 2009, 48, 6104. (d) Ma, B.; Banerjee, B.;
Litvinov, D. N.; He, L.; Castle, S. L. J. Am. Chem. Soc. 2010, 132, 1159.
(21) (a) Scudiero, D. A.; Shoemaker, R. H.; Paull, K. D.; Monks, A.;
Tierney, S.; Nofziger, T. H.; Currens, M. J.; Seniff, D.; Boyd, M. R.
Cancer Res. 1988, 48, 4827. (b) Roehm, N. W.; Rodgers, G. H.;
Hatfield, S. M.; Glasebrook, A. L. J. Immunol. Methods 1991, 142, 257.
(22) See Supporting Information for possible errors in the original
structural determination.
I
DOI: 10.1021/jacs.5b05550
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX