Total synthesis of the large non-ribosomal peptide polytheonamide B

Article abstract of DOI:10.1038/nchem.554

Polytheonamide B is by far the largest non-ribosomal peptide known at present, and displays extraordinary cytotoxicity (EC₅₀ =68 pg ml ^-1 , mouse leukaemia P388 cells). Its 48 amino-acid residues include a variety of non-proteinogenic d- and l-amino acids, and the absolute stereochemistry of these amino acids alternate in sequence. These structural features induce the formation of a stable β-strand-type structure, giving rise to an overall tubular structure over 30A? in length. In a biological setting, this fold is believed to transport cations across the lipid bilayer through a pore, thereby acting as an ion channel. Here, we report the first chemical construction of polytheonamide B. Our synthesis relies on the combination of four key stages: syntheses of non-proteinogenic amino acids, a solid-phase assembly of four fragments of polytheonamide B, silver-mediated connection of the fragments and, finally, global deprotection. The synthetic material now available will allow studies of the relationships between its conformational properties, channel functions and cytotoxicity.

Full text of DOI:10.1038/nchem.554

ARTICLES
PUBLISHED ONLINE: 21 FEBRUARY 2010 | DOI: 10.1038/NCHEM.554
Total synthesis of the large non-ribosomal peptide
polytheonamide B
Masayuki Inoue , Naoki Shinohara, Shintaro Tanabe, Tomoaki Takahashi, Ken Okura, Hiroaki Itoh,
*
Yuki Mizoguchi, Maiko Iida, Nayoung Lee and Shigeru Matsuoka
Polytheonamide B is by far the largest non-ribosomal peptide known at present, and displays extraordinary cytotoxicity
(EC50 5 68 pg ml21, mouse leukaemia P388 cells). Its 48 amino-acid residues include a variety of non-proteinogenic
D- and L-amino acids, and the absolute stereochemistry of these amino acids alternate in sequence. These structural
features induce the formation of a stable b-strand-type structure, giving rise to an overall tubular structure over 30 Å in
length. In a biological setting, this fold is believed to transport cations across the lipid bilayer through a pore, thereby
acting as an ion channel. Here, we report the first chemical construction of polytheonamide B. Our synthesis relies on the
combination of four key stages: syntheses of non-proteinogenic amino acids, a solid-phase assembly of four fragments of
polytheonamide B, silver-mediated connection of the fragments and, finally, global deprotection. The synthetic material
now available will allow studies of the relationships between its conformational properties, channel functions
and cytotoxicity.
olytheonamides A and B (epi-1 and 1, respectively; Fig. 1a) are not been observed for any other known secondary metabolites,
exceptionally cytotoxic linear peptides (EC50 ¼ 68 pg ml21
,
and it has therefore been proposed that the particular conformation
P
mouse leukaemia P388 cells)1. These compounds were isolated of 1 is responsible for them. Because the molecular weights of
from the marine sponge Theonella swinhoei, and are believed to be channel proteins are typically over 100,000 Da, polytheonamide
produced by symbiotic microorganisms that have not yet been can be regarded as a minimalist transmembrane channel. At
identified. The structural elucidation of this peptide by Fusetani, present, little is known about how the proteinogenic and non-pro-
Matsunaga and colleagues revealed that it has a molecular weight teinogenic building blocks serve to exert the specific channel func-
of over 5,000 Da, consists of 48 amino-acid residues, and is tion and the potent cytotoxicity of 1. Understanding the structural
capped at the N-terminus with 5,5-dimethyl-2-oxohexanoate. requirements for its biological function would necessitate detailed
Remarkably, 13 of the 19 different component amino acids of 1 structure–activity relationship (SAR) studies.
are non-proteinogenic (Fig. 1b), strongly indicating that these
The unique function and structure of 1 motivated us to under-
natural products are biosynthesized by non-ribosomal peptide take its total chemical construction. The total synthesis of large
synthases2. Accordingly, polytheonamides represent the largest polypeptides, including proteins, continues to pose significant chal-
non-ribosomal peptides known at present.
lenges in chemistry9. This is particularly the case if the peptide mol-
The unique amino-acid components of polytheonamides include ecules of interest are composed of unusual amino acids and are not
eight g-N-methyl asparagine derivatives (17 and 19, Fig. 1b) and easily amenable to automated solid-phase synthesis or biosynthesis.
fourteen b-methylated derivatives of proteinogenic amino acids Polytheonamide B represents an ideal target molecule for expanding
such as b-methyl valine (10 and 11), b-methyl threonine (14 and our knowledge regarding the reactivity and structural and functional
15), b-methyl isoleucine (16), b-methyl glutamine (18) and b,b- properties of large polypeptides.
dimethyl methionine oxide (20). Sulfoxide-bearing amino acid 20
We report the first total synthesis of the largest non-ribosomal
is unprecedented in both natural and synthetic products. peptide, polytheonamide B (1), as well as the structural elucidation
Interestingly, polytheonamide A (epi-1) and B (1) are epimers at of the R-stereochemistry of its sulfoxide. This attempt to gain
the sulfoxide moieties of this particular amino acid, although their precise atom-by-atom control of the structure of 1 will provide
absolute stereochemistries have not been determined (Fig. 1a).
the first chemical basis for systematically correlating its molecular
The most striking structural feature of 1 is that its entire 48- structure and biological function10–13
residue peptide sequence is composed of alternating D- and L-
.
amino acids, only interrupted by eight Gly residues (Fig. 1a). Results and discussion
Extensive NMR studies of 1 in methanol/chloroform solution indi- Our total synthesis of polytheonamide made use of a flexible and
cated that the peptide sequence with alternating chirality folded into general strategy consisting of four independent stages: (1) synthesis
a b-helix of 6.5 residues per turn, stabilized by inter-residue hydro- of the non-proteinogenic amino-acid monomers14; (2) construction
gen bonding3. As a consequence, overall, it forms a tubular structure of peptide segments by stepwise assembly using solid-phase syn-
spanning over 30 Å, with all the side chains oriented to point thesis methods15; (3) convergent couplings of the peptide segments;
outside the helix4–7. Polytheonamide B has been shown to and (4) global deprotection of the protected polytheonamide. For
conduct monovalent cations across a lipid bilayer and to have a the first stage, it was necessary to prepare the eight amino-acid
voltage-dependent gating property similar to that of an ion monomers 13–20 from commercially available materials (Fig. 1b).
channel protein8. Such specific ion channel characteristics have In the second stage, the presence of multiple acid-labile amino-acid
Graduate School of Pharmaceutical Sciences, The University of Tokyo, and PRESTO, Japan Science and Technology Agency, Hongo, Bunkyo-ku, Tokyo
113-0033, Japan. e-mail: inoue@mol.f.u-tokyo.ac.jp
*
280
NATURE CHEMISTRY | VOL 2 | APRIL 2010 | www.nature.com/naturechemistry
© 2010 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.554
ARTICLES
a
OH
O
O
O
O
O
O
O
O
O
H
N
H
N
H
N
H
N
H
N
H
N
H
N
H
N
O
N
H
N
H
N
H
N
H
N
H
N
N
H
N
H
H
O
O
O
O
O
O
O
O
O
O
NH
1
MeHN
HN
O
H2N
O
NH
O
O
O
O
O
O
H
N
H
N
H
N
H
N
H
N
H
N
H
N
MeHN
N
H
N
H
N
H
N
H
N
H
N
H
O
N
H
O
O
O
O
O
O
O
O
O
O
O
NH
OH
O
O
O
OH
OH
O
NHMe
NHMe
NHMe
NH2
R/S
+
S
–
O
HN
NHMe
O
O
NH
OH
OH
O
O
O
O
O
H
H
H
H
N
H
H
N
N
N
N
N
N
H
N
H
N
H
N
H
N
H
N
H
O
O
O
O
O
HO
O
O
O
O
O
HO
O
OH
44
48
NHMe
NHMe
NH2
NH2
b
Proteinogenic amino acids
O
H2N
OH
OH
OH
OH
H2N
OH
H2N
H2N
OH
H2N
O
OH
H2N
O
H2N
O
O
O
O
2: Gly (1, 3, 11, 17, 3: L-Ala (10, 12, 4: L-Val (36, 40, 42)
19, 25, 26, 32) 14, 18, 24, 38)
5: L-Thr (48)
6: L-Ile (28, 34)
7: L-Gln (46)
Non-proteinogenic amino acids
O
OH
OH
H2N
OH
H2N
OH
OH
H2N
OH
H2N
H2N
H2N
O
O
O
O
O
8: D-Ala (7)
9: D-Ser (41)
10: D-Tle (5, 9, 13)
11: L-Tle (4,
6, 8, 20, 30)
12: D-Asn (43, 45)
O–
H2N
O
+
S
O
O
OH
OH
OH
OH
OH
MeHN
MeHN
OH
OH
OH
OH
H2N
OH
H2N
H2N
OH
OH
H2N
H2N
H2N
H2N
H2N
O
O
O
O
O
O
O
O
13 (47)
14 (23, 31)
15 (16)
16 (2)
17 (15, 21, 27,
33, 35, 39)
18 (22)
19 (29, 37)
20 (44)
Figure 1 | Structure of polytheonamide. D-Chirality and L-chirality of the amino acids are indicated by red and blue, respectively. a, Polytheonamide A (epi-1)
and B (1) are the epimers of the sulfoxide of the 44th amino acid. b, Component amino acids of polytheonamide. Polytheonamide contains six proteinogenic
amino acids and thirteen non-proteinogenic amino acids. The residue numbers are indicated in parentheses. Compounds 13–20 required
synthetic preparation.
residues, such as b-hydroxy amino acids, required the use of during the third stage of the process, segment condensation. To
Fmoc (9-fluorenylmethoxycarbonyl) chemistry16 for solid-phase maximize the yields of the elongation processes, the primary
peptide synthesis (SPPS) rather than Boc (t-butoxycarbonyl) chem- amides at residues 43, 45 and 46, and the alcohols at residues 41,
istry17, which would require harsh acidic conditions for global 47 and 48, were planned to be protected with triphenylmethyl
deprotection. Design of the peptide segments by dissection of the (Tr) and t-butyl (t-Bu) groups, respectively. In the third stage,
entire molecule was crucial for the success of the synthesis. Agþ-mediated amide formation between thioesters and
Preliminary studies showed that solid-phase synthesis was only effi- amines19,20 was selected for assembly of the gigantic peptides
cient up to 16 residues. Beyond this length, smooth elongation of the because of the high chemoselectivity and reactivity of this reaction.
peptides was impeded by the sterically hindered b-tetrasubstituted After derivatization of the three segments to thioesters and sub-
amino acids, which have low reactivity, and by asparagine and gluta- sequent condensation from the C-terminal segment, the three Tr
mine derivatives, the side-chain amides of which have a strong ten- and three t-Bu groups of the protected polytheonamide were simul-
dency to form interstrand aggregates in the peptide–resin matrix18. taneously removed under acidic conditions in the fourth stage.
Therefore, four segments of 7 to 16 residues were designed (Fig. 2):
In the first stage of total synthesis, the eight non-proteinogenic
residues 1–11 (21: Ncap-[1–11]-OH), 12–25 (22: Fmoc-[12–25]- amino acids 13, 14, 15, 16, 17, 18, 19 and 20 were synthesized as
OH), 26–32 (23: Fmoc-[26–32]-OH) and 33–48 (24: H-[33–48 protected forms, together with 5,5-dimethyl-2-oxohexanoic acid
(t-Bu)3(Tr)3]-OH). The first three peptides included glycine as the (Ncap-OH) conjugated at the N-terminus (see Supplementary
C-terminal amino acid to eliminate the risk of epimerization Information for details). To determine the unknown sulfoxide
NATURE CHEMISTRY | VOL 2 | APRIL 2010 | www.nature.com/naturechemistry
© 2010 Macmillan Publishers Limited. All rights reserved.
281

NATURE CHEMISTRY DOI: 10.1038/NCHEM.554
ARTICLES
OH
O
O
O
O
O
O
O
O
O
H
H
N
H
H
H
H
H
N
H
N
N
N
N
N
N
O
N
N
N
N
N
OH
FmocHN
N
H
N
H
H
H
H
H
H
O
O
O
O
O
O
1
11
O
O
O
O
NH
12
Fmoc-[12–25]-OH (22)
Ncap-[1–11]-OH (21)
MeHN
HN
O
H2N
O
32
O
O
O
26
H
N
N
H
O
H
H
HO
N
N
NH
33
N
N
NHFmoc
O
O
O
MeHN
NHH
H
H
N
H
N
H
H
O
O
O
N
OH
HO
N
N
H
N
H
O
O
O
OH
O
H
O
NH
25
O
O
O
NHMe
NHMe
O
O
OH
Fmoc-[26–32]-OH (23)
NHMe
–
O
+
S
HN
O
NHMe
O
NHTr
H-[33–48(t-Bu)3(Tr)3]-OH (24)
O
Ot-Bu
OH
NH
O
O
O
O
O
O
H
H
H
H
H
H
N
N
N
N
N
N
N
N
N
N
H
N
N
H
H
H
H
H
O
O
O
O
O
O
O
O
O
O
O
OH
t-BuO
t-BuO
NHMe
NHMe
NHTr
NHTr
41
43
45
46
47
48
Figure 2 | Four peptide segments for the total synthesis of polytheonamide. Segments Ncap-[1–11]-OH (21), Fmoc-[12–25]-OH (22) and Fmoc-[26–32]-OH
(23) have glycine as the C-terminal amino acid to eliminate the risk of epimerization during segment condensation. The primary amides at residues
43, 45 and 46, and the alcohols at residues 41, 47 and 48 in the C-terminus segment 24 were protected with Tr and t-Bu groups, respectively.
Fmoc ¼ 9-fluorenylmethoxycarbonyl; Ncap ¼ Me3C(CH2)2(CO)CO; t-Bu ¼ t-butyl; Tr ¼ triphenylmethyl.
Tf2O, 2,6-lutidine, 3Å MS,
O
O
OH
a
KN(SiMe3)2, THF, –78 to –40 °C;
MeI, –40 to –20 °C
CH2Cl2, –98 to –60 °C;
DIBAL, CH2Cl2, –78 to –40 °C
99%
NaSMe, DMF, –78 to –10 °C
MeO
PhFlHN
MeO
Ot-Bu
Ot-Bu
Ot-Bu
80%
PhFlHN
R1HN
100%
1. H2, Pd(OH)2/C,
O
O
O
AcOEt/MeOH, rt
2. Boc2O, Et3N,
THF, rt
25
26
27: R1 = PhFl
28: R1 = Boc
95% (2 steps)
R
–
S-31 (0.02 eq)
H2NCONH2·H2O2
MeOH, 0 °C
O
+
SMe
S
O
Ti
O
O
N
N
O
N
Ot-Bu
S-31=
PhFl =
OR3
R2HN
97% (85% de)
O
N
Ti
O
N
N
FmocHN
O
O
OH HO
Ph Ph
29: R2 = Boc
30: R2 = Fmoc
32 (85% de): R3 = t-Bu
32 (96% de): R3 = t-Bu
33: R3 = H
1. HCl, AcOEt, 0 to 40 °C
SiO2 purification
N
N
2. FmocCl, NaHCO3,
H2O/dioxane, rt
69% (2 steps)
=
TFA/CH2Cl2,
0 °C to rt, 85%
OH HO
O
O
MeO
Ph
–
N
MeO
–
O
OH
H
S
2+
Ph
H
1. Et2NH, CH2Cl2, 0 °C
2. p-BrBzCl,
–
–
NH
PyBOP, HOBt, pyridine, rt
47% (2 steps)
O
+
S
–
O
OMe
S
buffer (pH = 10)/toluene, rt
3. Me3SiCHN2,
2+
p-BrBzHN
S-36
O
–
benzene/MeOH, rt
MesSO3NH2, CH2Cl2, rt
33
OMe
OMe
O
p-BrBzHN
41% (3 steps)
p-BrBzHN
MeO
Ph
O
–
34
35
O
N
O
R1
R2
O
H
b
MeO
OH
S
2+
–
–
N
SMe
+0.10
CH3
Ph
H
O
2+
S
MeO
+0.25
CH3
PyBOP, HOBt, pyridine, rt
44% (2 steps)
CH3
OMe
p-BrBzHN
R-36
O
H3C
O
CH
O
–0.02
p-BrBzHN
H
OMe
+0.12
S-36: R1 = Ph, R2 = H
R-36: R1= H, R2 = Ph
4.9
4.8
3.3
1.4
ppm
1.3
1.2
1.1
1.0
Figure 3 | Synthesis and structural determination of the Fmoc-protected amino acid of residue 44. a, Stereoselective synthesis of 33, and derivatization of
33 to S-36 and R-36. b, Determination of the absolute configuration of the sulfoxide. The coloured numbers are differences (Dd) in the 1H chemical shifts
between S-36 and R-36 (Dd ¼ d (S-36)–d (R-36)). Boc ¼ t-butoxycarbonyl; DIBAL ¼ diisobutylaluminium hydride; Fmoc ¼ 9-fluorenylmethoxycarbonyl;
HOBt ¼ 1-hydroxybenzotriazole; 2,6-lutidine ¼ 2,6-dimethylpyridine; Mes ¼ mesityl ¼ 2,4,6-trimethylphenyl; MS ¼ molecular sieves; PyBOP ¼ benzotriazol-
1-yl-oxy-tris-pyrrolidinophosphonium hexafluorophosphate; Tf ¼ trifluoromethanesulfonyl; TFA ¼ trifluoroacetic acid.
282
NATURE CHEMISTRY | VOL 2 | APRIL 2010 | www.nature.com/naturechemistry
© 2010 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.554
ARTICLES
HS(CH2)2CO2Et, HOBt
CyN=C=NCy, THF, rt
Ncap-[1–11]-OH (21)
Fmoc-[12–25]-OH (22)
Ncap-[1–11]-S(CH2)2CO2Et (37)
1. Solid-phase peptide synthesis
O
(i) N -Fmoc-amino acid
FmocHN
(or Ncap-OH)
13% overall yield (24 steps)
O
HS(CH2)2CO2Et, HOBt
HBTU, HOBt, i-Pr2NEt, NMP
i-PrN=C=Ni-Pr, THF, rt
(ii) 22% piperidine/NMP
O
Fmoc-[12–25]-S(CH2)2CO2Et (38)
2. TFA/H2O=95/5, rt
11% overall yield (28 steps)
HS(CH2)2CO2Et, HOBt
CyN=C=NCy, THF, rt
Fmoc-[26–32]-OH (23)
Fmoc-[26–32]-S(CH2)2CO2Et (39)
Fmoc-Gly-Wang resin
9% overall yield (14 steps)
1. Solid-phase peptide synthesis
(i) N -Fmoc-amino acid,
N
N
N
PF6
O
HATU, HOAt, i-Pr2NEt, NMP
N
N
(ii) 22% piperidine/NMP
H2N
N
N
N
X
X
O
H-[33–48(t-Bu)3(Tr)3]-OH (24)
OH
O
2. (CF3)2CHOH/CH2Cl2=1/3, rt
t-BuO
Cl
7% overall yield (31 steps)
HOBt: X = CH
HOAt: X = N
HBTU: X = CH
HATU: X = N
H-Thr(t-Bu)-2-chloro trityl resin
Figure 4 | Solid-phase synthesis of the four peptide segments. Segments Ncap-[1–11]-OH (21), Fmoc-[12–25]-OH (22) and Fmoc-[26–32]-OH (23) were
synthesized from Fmoc-Gly-Wang resin. After cleavage from the resin, the three segments were separately converted to the thioesters. Segment
H-[33–48(t-Bu)3(Tr)3]-OH (24) was synthesized from H-Thr(t-Bu)-2-chloro trityl resin, and was then cleaved from the resin. Cy ¼ cyclohexyl;
HATU ¼ O-(7-azabenzotriazole-1-yl)-N,N,N’,N’-tetramethyluronium hexafluorophosphate; HBTU ¼ O-(benzotriazole-1-yl)-N,N,N’,N’-tetramethyluronium
hexafluorophosphate; HOAt ¼ 1-hydroxy-7-azabenzotriazole; HOBt ¼ 1-hydroxybenzotriazole.
H-[33–48(t-Bu)3(Tr)3]-OH (24)
stereochemistry of polytheonamides A and B, it was decided to
synthesize the R-sulfoxide diastereomer of 20 of residue 44 in a
stereoselective fashion. Figure 3a summarizes preparation of 33,
AgNO3, HOOBt,
the Fmoc-protected version of 20 with the R-sulfoxide, from the
known aspartate 25. The steric bulk of the 9-phenylfluorenyl
(PhFl) group21 attached to the amine of 25 effectively controlled
the first two chemoselective transformations by preventing abstrac-
tion of the Ca-proton and reduction of the t-butyl ester. Specifically,
treatment of 25 with potassium bis(trimethylsilyl)amide and methyl
iodide enabled the bis-substitution at the Cb of 25 to generate b,b-
dimethylated 26 (ref. 22), the side-chain ester of which was then
selectively reduced using diisobutylaluminium hydride (DIBAL)
to produce primary alcohol 27. Alteration of the protecting group
of 27 from PhFl to Boc was performed by means of hydrogenolysis
and (Boc)2O treatment in the presence of Et3N, leading to 28. The
less nucleophilic nature of the carbamate of 28 compared to the
tertiary amine of 27 was particularly important for preventing unde-
sired intramolecular cyclization in the subsequent step. Alcohol 28
was converted into the corresponding triflate by the action of
trifluoromethanesulfonic anhydride (Tf2O) and 2,6-lutidine, and
the leaving group was in situ displaced by sodium methanethiolate
in an intermolecular fashion to produce methyl sulfide 29. The Boc
group was in turn replaced by an Fmoc group by the standard two-
step procedure (29 ! 30).
Fmoc-[26–32]-S(CH2)2CO2Et (39)
i-Pr2NEt, THF, rt
Fmoc-[26–48(t-Bu)3(Tr)3]-OH (40)
N
O
N
N
Piperidine, THF, rt
83% (2 steps)
OH
HOOBt
H-[26–48(t-Bu)3(Tr)3]-OH (41)
AgNO3, HOOBt,
i-Pr2NEt, DMF/THF=1/2, rt
Fmoc-[12–25]-S(CH2)2CO2Et (38)
Fmoc-[12–48(t-Bu)3(Tr)3]-OH (42)
Piperidine, THF, rt
86% (2 steps)
Stereoselective synthesis of the sulfoxide from sulfide 30 and its
stereochemical assignment were not simple tasks. Many chiral
reagents for sulfide oxidation were found to exhibit no diastereo-
selectivity. Gratifyingly, it was found that the Katsuki conditions
promoted highly diastereoselective oxidation of sulfide 30 to sulfox-
ide 32 (ref. 23). Treatment of 30 with a catalytic amount of S-31 in
the presence of urea hydrogen peroxide in methanol led to 32 in
85% de (96% de after the second SiO2 chromatography purifi-
cation). Finally, removal of the t-Bu groups from 32 using
trifluoroacetic acid (TFA) in CH2Cl2 gave rise to the requisite
Fmoc-protected amino acid 33. The stereochemistry of the sulfoxide
of 33 was then unambiguously established using the NMR-based
methodology developed by Kusumi24. After the three-step reactions
H-[12–48(t-Bu)3(Tr)3]-OH (43)
AgNO3, HOOBt,
i-Pr2NEt, DMF/THF=1/2, 50 °C
91%
Ncap-[1–11]-S(CH2)2CO2Et (37)
Ncap-[1–48(t-Bu)3(Tr)3]-OH (44)
Figure 5 | Convergent assembly of the four peptide segments. The
protected polytheonamide 44 was synthesized by the three Agþ-mediated
couplings between the amines and the thioesters. HOOBt ¼ 3,4-dihydro-3-
hydroxy-4-oxo-1,2,3-benzotriazine.
NATURE CHEMISTRY | VOL 2 | APRIL 2010 | www.nature.com/naturechemistry
283
© 2010 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.554
ARTICLES
OH
O
O
O
O
O
O
O
O
O
H
H
N
H
N
H
N
H
N
H
N
H
N
H
N
N
O
N
H
N
H
N
H
N
H
N
H
N
N
H
N
H
H
O
O
O
O
O
O
O
O
O
O
NH
1
MeHN
HN
O
H2N
O
NH
O
O
O
O
O
O
O
H
N
H
N
H
N
H
N
H
N
H
N
H
N
MeHN
N
H
N
H
N
H
N
H
N
H
N
H
N
H
O
O
O
O
O
O
O
O
O
O
O
NH
OH
O
O
O
OH
OH
O
NHMe
NHMe
NHMe
–
R
+
S
NHR2
O
HN
NHMe
O
O
NH
OR1
OH
O
O
O
O
O
H
N
H
H
H
H
H
N
N
N
N
N
N
H
N
H
N
H
N
H
N
H
N
H
O
O
O
O
O
R1O
O
O
O
O
O
R1O
O
OH
NHMe
NHMe
NHR2
NHR2
41
43
45
46
47
48
TFA/H2O = 20/1, 0 °C (1h) 30 °C (1h)
99%
Ncap-[1–48(t-Bu)3(Tr)3]-OH
(44: R1 = t-Bu, R2 = Tr)
Ncap-[1–48]-OH
(1: R1 = H, R2 = H)
Polytheonamide(R-sulfoxide) = Polytheonamide B
Figure 6 | Completion of the total synthesis of polytheonamide B. The three t-Bu and three Tr groups were simultaneously removed under acidic conditions.
from 33, O-mesitylsulfonylhydroxyamine transformed sulfoxide and CH2Cl2)30 without deprotection of the peptide side chains. HPLC
34 to sulfoximine 35 with the retention of chirality at the sulfur purification of the crude mixture resulted in H-[33–48(t-Bu)3(Tr)3]-OH
atom25. Imine 35 was conjugated with the chiral anisotropic (24) with an overall yield of 7%.
agents (S)- and (R)-methoxyphenylacetic acids to generate S-36
The third stage involved the assembly of the four structurally
and R-36, respectively. The differences in the 1H-chemical shifts complex segments. The molecular weight of each segment exceeded
(Dd) between S-36 and R-36 were calculated to determine the indicated 1,000 Da (Fig. 5). After many unsuccessful attempts, Aimoto con-
stereochemistry of the sulfoximine (Fig. 3b). This structural assignment ditions20 (AgNO3 and 3,4-dihydro-3-hydroxy-4-oxo-1,2,3-benzo-
indicated that the catalyst S-31 installed the R-stereocentre at the sulfur triazine (HOOBt), and N,N-diisopropylethylamine in THF/DMF)
atom in the reaction from 30 to 32. Accordingly, 33 was used for the were found to be exceptionally effective for coupling Fmoc-[26–
total synthesis of polytheonamide with the R-sulfoxide.
32]-S(CH2)2CO2Et (39) and H-[33–48(t-Bu)3(Tr)3]-OH (24). The
Having synthesized 33 and the other non-proteinogenic amino thioester was transformed in situ by the action of silver salt to the
acids as protected forms, the second stage of the total synthesis corresponding HOOBt ester, which was attacked by the amine,
was to prepare the four requisite peptide segments using an auto- resulting in the adduct Fmoc-[26–48(t-Bu)3(Tr)3]-OH (40).
mated solid-phase peptide synthesizer. Segments Ncap-[1–11]- Piperidine treatment of the product, followed by HPLC purification,
OH (21), Fmoc-[12–25]-OH (22) and Fmoc-[26–32]-OH (23) gave rise to the half structure of the protected polytheonamide,
were synthesized from commercially available Fmoc-Gly-Wang H-[26–48(t-Bu)3(Tr)3]-OH (41), with a yield of 83% for two
resin26 using standard Fmoc-based SPPS coupling conditions with steps. The obtained amine 41 was next coupled with Fmoc-
O-(benzotriazole-1-yl)-N,N,N0,N0-tetramethyluronium hexafluoro- [12–25]-S(CH2)2CO2Et (38) under similar reaction conditions, and
phosphate (HBTU)/1-hydroxybenzotriazole (HOBt) activation subsequently deprotected to produce H-[12–48(t-Bu)3(Tr)3]-OH
(Fig. 4). After cleavage from the Wang resin by treatment with (43) with a yield of 80% for two steps and the subsequent
95% aqueous TFA, the three segments were separately converted purification. Finally, the protected polytheonamide Ncap-[1–48
to the corresponding thioesters using the reagent combination of (t-Bu)3(Tr)3]-OH (44) was synthesized by Agþ-mediated coupling
HS(CH2)2CO2Et, HOBt and N,N0-dialkylcarbodiimide27. The crude of 43 with Ncap-[1–11]-S(CH2)2CO2Et (37), and purified
peptides were then purified using high-performance liquid chrom- (91% yield). These three high-yield couplings clearly demonstrated
atography (HPLC) to obtain Ncap-[1–11]-S(CH2)2CO2Et (37), the broad applicability of the present coupling protocol for
Fmoc-[12–25]-S(CH2)2CO2Et (38) and Fmoc-[26–32]-S(CH2)2 the generation of highly complex molecules with numerous
CO2Et (39) in overall yields of 13%, 11% and 9%, respectively, reactive functionalities.
from the Fmoc-Gly-Wang resin. The synthesis of the C-terminus
The last stage of the total synthesis was the removal of the six pro-
segment, H-[33–48(t-Bu)3(Tr)3]-OH (24), began with H-Thr tective groups (Fig. 6). The protected peptide was subjected to carefully
(t-Bu)-2-chloro trityl resin using O-t-Bu-protected Fmoc amino acids controlled acidic conditions. Exposure of Ncap-[1–48(t-Bu)3(Tr)3]-
for residues 41 and 47, and N-Tr-protected Fmoc amino acids OH (44) to TFA/H2O (20:1) mixture for 1 h at 0 8C and an additional
for residues 43, 45 and 46 (ref. 28). A more reactive reagent system 1 h at 30 8C, and subsequent purification by reversed-phase HPLC,
(O-(7-azabenzotriazole-1-yl)-N,N,N0,N0-tetramethyluronium hexa- delivered polytheonamide with the R-sulfoxide in a yield of 99%.
fluorophosphate (HATU)/1-hydroxy-7-azabenzotriazole (HOAt)29) Longer reaction times or higher temperatures resulted in lower
was applied for the SPPS coupling, because HBTU/HOBt conditions yields, presumably owing to acid-sensitive functionalities such as the
resulted in a significantly lower yield of the peptide. The highly acid- b-tertiary alcohols at residues 16, 23 and 31. Thus, by applying the
labile 2-chloro trityl resin linker allowed orthogonal cleavage from the four described synthetic stages, the largest non-ribosomal peptide
resin under mild acidic conditions (1,1,1,3,3,3-hexafluoro-2-propanol was chemically constructed for the first time.
284
NATURE CHEMISTRY | VOL 2 | APRIL 2010 | www.nature.com/naturechemistry
© 2010 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.554
ARTICLES
11. Bong, D. T., Clark, T. D., Granja, J. R. & Ghadiri, M. R. Self-assembling organic
HPLC analyses of the synthetic and natural polytheonamide
indicated that the polytheonamide with R-sulfoxide is polytheona-
mide B (1). The retention time of synthetic 1 (tR ¼ 29.15 min)
matched that of natural polytheonamide B (1: tR ¼ 29.03 min), but
not that of its sulfoxide epimer polytheonamide A (epi-1: tR ¼
24.75 min) (Inertsil C8, 4.6 ꢀ 150 mm, UV 210 nm, n-PrOH/H2O
35:65, 0.5 ml min21, 45 8C). Detailed NMR data for synthetic 1
were collected (800 MHz), and the assigned signals clarified that syn-
thetic 1 is identical to natural 1 in all respects (1H NMR, DQF-COSY,
nanotubes. Angew. Chem. Int. Ed. 40, 988–1011 (2001).
´
12. Matile, S., Som, A. & Sorde, N. Recent synthetic ion channels and pores.
Tetrahedron 60, 6405–6435 (2004).
13. Koert, U., Al-Momani, L. & Pfeifer, J. R. Synthetic ion channels. Synthesis
1129–1146 (2004).
14. Humphrey, J. M. & Chamberlin, A. R. Chemical synthesis of natural product
peptides: coupling methods for the incorporation of noncoded amino acids into
peptides. Chem. Rev. 97, 2243–2266 (1997).
15. Chan, W. C. & White, P. D. Fmoc Solid Phase Peptide Synthesis—A Practical
Approach (Oxford Univ. Press, 2000).
TOCSY and NOESY: see Supplementary Information for details). 16. Carpino, L. A. & Han, G. Y. The 9-fluorenylmethoxycarbonyl function, a new
base-sensitive amino-protecting group. J. Am. Chem. Soc. 92, 5748–5749 (1970).
17. Merrifield, R. B. Solid phase peptide synthesis. I. The synthesis of a tetrapeptide.
J. Am. Chem. Soc. 85, 2149–2154 (1963).
18. Kent, S. B. H. Chemical synthesis of peptides and proteins. Ann. Rev. Biochem.
Overall, these data unambiguously show that polytheonamides A
and B have the S-sulfoxide and R-sulfoxide, respectively.
A preliminary toxicity study of the natural product and the synthetic
compound was carried out using mouse leukaemia P388 cells. Synthetic
1 displayed an EC50 value (98 pM) comparable to that of the natural
form (79 pM). However, the protected analogue 44 surprisingly did
not exhibit detectable toxicity (.1 mM), in spite of it having the
same molecular length as parent 1. Presumably, the bulky and hydro-
phobic protective groups impede the efficient formation of the mem-
brane-spanning ion channel by masking important polar residues,
although the structural basis of this hypothesis awaits clarification.
In summary, the first total synthesis and structural elucidation of
polytheonamide B, the largest known non-ribosomal peptide, was
achieved through the convergent assembly of four complex peptides.
Our versatile and modular strategy should be useful for synthesizing
various analogues to obtain structural insights into the conformation-
al behaviour of this peptide, as well as its channel function and potent
toxicity. Future investigations will include more detailed studies
towards the elucidation of the molecular mode of action of polytheo-
namide and the rational design of new tailor-made channel peptides.
57, 957–989 (1988).
19. Blake, J. & Li, C. H. New segment-coupling method for peptide synthesis in
aqueous solution: application to synthesis of human [Gly17]-b-endorphin.
Proc. Natl Acad. Sci. USA 78, 4055–4058 (1981).
20. Aimoto, S. Polypeptide synthesis by the thioester method. Biopolymers 51,
247–265 (1999).
21. Gmeiner, P., Feldman, P. L., Chu-Moyer, M. Y. & Rapoport, H. Efficient and
practical total synthesis of (þ)-vincamine from L-aspartic acid. J. Org. Chem. 55,
3068–3074 (1990).
22. Kawahara, N., Weisberg, M. & Goodman, M. Synthesis of b,b-dimethylated
amino-acid building blocks utilizing the 9-phenylfluorenyl protecting group.
J. Org. Chem. 64, 4362–4369 (1999).
23. Saito, B. & Katsuki, T. Ti(salen)-catalyzed enantioselective sulfoxidation using
hydrogen peroxide as a terminal oxidant. Tetrahedron Lett. 42, 3873–3876 (2001).
24. Yabuuchi, T. & Kusumi, T. NMR spectroscopic determination of the absolute
configuration of chiral sulfoxides via N-(methoxylphenylacetyl)sulfoximines.
J. Am. Chem. Soc. 121, 10646–10647 (1999).
25. Johnson, C. R., Kirchhoff, R. A. & Corkins, H. G. Synthesis of optically active
sulfoximines from optically active sulfoxides. J. Org. Chem. 39, 2458–2459 (1974).
26. Wang, S. p-Alkoxybenzyl alcohol resin and p-alkoxybenzyloxycarbonylhydrazide
resin for solid phase synthesis of protected peptide fragments. J. Am. Chem. Soc.
95, 1328–1333 (1973).
27. Futaki, S., Sogawa, K., Maruyama, J., Asahara, T., Niwa, M. & Hojo, H.
Preparation of peptide thioesters using Fmoc-solid-phase peptide synthesis and
its application to the construction of a template-assembled synthetic protein
(TASP). Tetrahedron Lett. 38, 6237–6240 (1997).
28. Barlos, K., Chatzi, O., Gatos, D. & Stavropoulos, G. 2-chlorotrityl chloride resin.
Int. J. Peptide Protein Res. 37, 513–520 (1991).
Received 9 September 2009; accepted 13 January 2010;
published online 21 February 2010; corrected online 23 February 2010
References
1. Hamada, T., Matsunaga, S., Yano, G. & Fusetani, N. Polytheonamide A and B,
highly cytotoxic, linear polypeptides with unprecedented structural features, from
the marine sponge, Theonella swinhoei. J. Am. Chem. Soc. 127, 110–118 (2005).
2. Schwarzer, D., Finking, R. & Marahiel, M. A. Nonribosomal peptides: from
genes to products. Nat. Prod. Rep. 20, 275–287 (2003).
29. Carpino, L. A. 1-hydroxy-7-azabenzotriazole. An efficient peptide coupling
additive. J. Am. Chem. Soc. 115, 4397–4398 (1993).
30. Bollhagen, R., Schmiedberger, M., Barlos, K. & Grell, E. A new reagent for the
cleavage of fully protected peptides synthesized on 2-chlorotrityl chloride resin.
Chem. Commun. 2559–2560 (1994).
3. Hamada, T., Matsunaga, S., Fusetani, N., Fujiwara, M. & Fujita, K. Structure
elucidation of polytheonamide B, a highly cytotoxic polypeptide from the
marine sponge Theonella swinhoei, by NMR spectroscopy. Tennen Yuki
Kagobutsu Toronkai Koen Yoshishu 37, 695–700 (1995).
Acknowledgements
This work was supported financially by the Takeda Science Foundation and the Naito
Foundation. Fellowships for N.S. and S.T. from the Japan Society for the Promotion of
Science are gratefully acknowledged. We thank S. Matsunaga for providing the natural
polytheonamides A and B, T. Hamada for valuable information, T. Katsuki for providing
the catalyst and M. Hirama for valuable suggestions. The 800 MHz 1H NMR spectra were
recorded at RIKEN SSBC, Yokohama, Japan.
4. Urry, D. W. The gramicidin A transmembrane channel: a proposed p(L,D) helix.
Proc. Natl Acad. Sci. USA 68, 672–676 (1971).
5. Ketchem, R. R., Hu, W. & Cross, T. A. High-resolution conformation of
gramicidin A in a lipid bilayer by solid-state NMR. Science 261, 1457–1460 (1993).
6. Stankovic, C. J., Heinemann, S. H., Delfino, J. M., Sigworth, F. J. &
Schreiber, S. L. Transmembrane channels based on tartaric acid–gramicidin A
hybrids. Science 244, 812–817 (1989).
Author contributions
7. Navarro, E., Fenude, E. & Celda, B. Solution structure of a D,L-alternating
oligonorleucine as a model of double-stranded antiparallel b-helix. Biopolymers
64, 198–209 (2002).
8. Oiki, S., Muramatsu, I., Matsunaga, S. & Fusetani, N. A channel-forming peptide
toxin: polytheonamide from marine sponge (Theonella swinhoei). Folia
Pharmacol. Jpn 110 (Suppl. 1), 195P–198P (1997).
M. Inoue conceived and designed the study. N.S. and S.T. performed the total synthesis.
N.S. and S.M. contributed the structural analyses. T.T., K.O., H.I., Y.M. and M.Iida
performed the fragment syntheses. M.Iida and N.L. performed the bioassay. M.Inoue, N.S.
and S.M. co-wrote the paper.
9. Kimmerlin, T. & Seebach, D. ‘100 years of peptide synthesis’: ligation methods Additional information
for peptide and protein synthesis with applications to beta-peptide assemblies.
J. Peptide Res. 65, 229–260 (2005).
10. Koeppe, R. E. II & Andersen, O. S. Engineering the gramicidin channel. Annu.
Rev. Biophys. Biomol. Struct. 25, 231–258 (1996).
The authors declare no competing financial interests. Supplementary information and
chemical compound information accompany this paper at www.nature.com/
naturechemistry. Reprints and permission information is available online at http://npg.nature.com/
reprintsandpermissions/. Correspondence and requests for materials should be addressed to M.I.
NATURE CHEMISTRY | VOL 2 | APRIL 2010 | www.nature.com/naturechemistry
285
© 2010 Macmillan Publishers Limited. All rights reserved.

Nature chemistry doi: 10.1038/nchem.613
articles
erratUM
total synthesis of the large non-ribosomal peptide polytheonamide B
Masayuki Inoue, Naoki Shinohara, Shintaro Tanabe, Tomoaki Takahashi, Ken Okura, Hiroaki Itoh, Yuki Mizoguchi,
Maiko Iida, Nayoung Lee and Shigeru Matsuoka
Nature Chemistry doi:10.1038/nchem.554 (2010); published online: 21 February 2010; corrected online 23 February 2010.
In the version of this Article originally published online, an in-house error led to the incorrect representation of stereochemistry in
Figs 1, 2 and 6. ese have now been corrected in all versions of the Article.
nature chemistry | VOL 2 | APRIL 2010 | www.nature.com/naturechemistry
© 2010 Macmillan Publishers Limited. All rights reserved
335