Total synthesis of the thiopeptide antibiotic amythiamicin D

Article abstract of DOI:10.1021/ja0547937

The thiopeptide (or thiostrepton) antibiotics are a class of sulfur containing highly modified cyclic peptides with interesting biological properties, including reported activity against MRSA and malaria. Described herein is the total synthesis of the thi

Full text of DOI:10.1021/ja0547937

Published on Web 10/18/2005
Total Synthesis of the Thiopeptide Antibiotic Amythiamicin D
Rachael A. Hughes,† Stewart P. Thompson,† Lilian Alcaraz,‡ and
Christopher J. Moody*,†,§
Contribution from the School of Chemistry, UniVersity of Nottingham, UniVersity Park,
Nottingham NG7 2RD, United Kingdom, Department of Chemistry, UniVersity of Exeter,
Stocker Road, Exeter EX4 4QD, United Kingdom, and Medicinal Chemistry Department,
AstraZeneca R&D Charnwood, Bakewell Road, Loughborough LE11 5RH, United Kingdom
Received July 18, 2005; E-mail: c.j.moody@nottingham.ac.uk
Abstract: The thiopeptide (or thiostrepton) antibiotics are a class of sulfur containing highly modified cyclic
peptides with interesting biological properties, including reported activity against MRSA and malaria.
Described herein is the total synthesis of the thiopeptide natural product amythiamicin D, which utilizes a
biosynthesis-inspired hetero-Diels-Alder route to the pyridine core of the antibiotic as a key step. Preliminary
studies using a range of serine-derived 1-ethoxy-2-azadienes established that hetero-Diels-Alder reaction
with N-acetylenamines proceeded efficiently under microwave irradiation to give 2,3,6-trisubstituted pyridines.
The thiazole building blocks of the antibiotic were obtained by either classical Hantzsch reactions or by
dirhodium(II)-catalyzed chemoselective carbene N-H insertion followed by thionation, and were combined
to give the bis-thiazole that forms the left-hand fragment of the antibiotic. The key Diels-Alder reaction of
a tris-thiazolyl azadiene with benzyl 2-(1-acetylaminoethenyl)thiazole-4-carboxylate gave the core tetrathia-
zolyl pyridine, which was elaborated into the natural product by successive incorporation of glycine and
bis-thiazole fragments followed by macrocyclization.
tors.6,7 Alternatively, other thiopeptides such as GE2270A act
directly on the elongation factor proteins, inhibiting their action.8
Introduction
The thiopeptide (or thiostrepton) antibiotics are a class of
sulfur containing highly modified cyclic peptides characterized
by several common structural features: the presence of thiazole
and, in some cases, oxazole rings, unusual and dehydro amino
acids, and a heterocyclic centerpiece consisting of a tri- or tetra-
substituted pyridine all in a macrocyclic array.1 The five-
membered heterocycles are derived from amino acids by
nonribosomal peptide synthesis,2-5 followed by cyclization of
serine, threonine, or cysteine side chains. Many of the com-
pounds such as the micrococcins (e.g., micrococcin P1 1) and
thiostrepton 2 itself (Figure 1) have been known for over 50
years, although in many cases, the structures have only been
fully elucidated more recently, and even now some structural
uncertainties remain (see below).
Despite the fascinating biological activity of the thiopeptide
antibiotics, relatively few synthetic studies have been carried
out to date, and only two members of this series of natural
products, which numbers over 75 in total, have succumbed to
total synthesis to date, although, the syntheses of various
fragments of other thiopeptides have been reported; for example
the pyridine clusters of dimethyl sulfomycinamate,9,10 nosihep-
tide,11 A10255,12 and GE2270A.13,14
The structure of micrococcin P1 was originally assigned by
Walker et al. in 1977,15 but was corrected by Bycroft and
Gowland the following year.16 Initially, synthesis served to
confuse rather than clarify the structure of micrococcin P1, since
the first reported success actually described a compound that is
epimeric with the Bycroft-Gowland structure 1 in the isoalaninol
Most of the thiopeptide antibiotics inhibit protein synthesis
in bacteria, and share common modes of action. They act by
binding to the complex of 23SrRNA with ribosomal protein
L11, inhibiting the action of GTP-dependent elongation fac-
(6) Draper, D. E. In The Many Faces of RNA; Eggleston, D. S., Prescott, C.
D., Pearson, N. D., Eds.; Academic Press: San Diego, 1998; p 113-125.
(7) Porse, B. T.; Leviev, L.; Mankin, A. S.; Garrett, R. A. J. Mol. Biol. 1998,
276, 391-404.
(8) Heffron, S. E.; Jurnak, F. Biochemistry 2000, 39, 37-45.
(9) Kelly, T. R.; Lang, F. J. Org. Chem. 1996, 61, 4623-4633.
(10) Bagley, M. C.; Dale, J. W.; Xiong, X.; Bower, J. Org. Lett. 2003, 5, 4421-
4424.
† University of Exeter.
‡ AstraZeneca.
(11) Umemura, K.; Noda, H.; Yoshimura, J.; Konn, A.; Yonezawa, Y.; Shin,
C. G. Bull. Chem. Soc. Jpn. 1998, 71, 1391-1396.
§ University of Nottingham.
(1) Bagley, M. C.; Dale, J. W.; Merritt, E. A.; Xiong, X. Chem. ReV. 2005,
(12) Umemura, K.; Ikeda, S.; Yoshimura, J.; Okumura, K.; Saito, H.; Shin, C.
G. Chem. Lett. 1997, 1203-1204.
105, 685-714.
(2) Li, Y. M.; Milne, J. C.; Madison, L. L.; Kolter, R.; Walsh, C. T. Science
1996, 274, 1188-1193.
(13) Okumura, K.; Saito, H.; Shin, C.; Umemura, K.; Yoshimura, J. Bull. Chem.
Soc. Jpn. 1998, 71, 1863-1870.
(3) Roy, R. S.; Gehring, A. M.; Milne, J. C.; Belshaw, P. J.; Walsh, C. T. Nat.
Prod. Rep. 1999, 16, 249-263.
(14) Heckmann, G.; Bach, T. Angew. Chem., Int. Ed. 2005, 44, 1199-1201.
(15) Walker, J.; Olesker, A.; Valente, L.; Rabanal, R.; Lukacs, G. J. Chem.
Soc. Chem. Commun. 1977, 706-708.
(4) Schwarzer, D.; Finking, R.; Marahiel, M. A. Nat. Prod. Rep. 2003, 20,
275-287.
(16) Bycroft, B. W.; Gowland, M. S. J. Chem. Soc. Chem. Commun. 1978, 256-
258.
(5) Sieber, S. A.; Marahiel, M. A. Chem. ReV. 2005, 105, 715-738.
9
15644
J. AM. CHEM. SOC. 2005, 127, 15644-15651
10.1021/ja0547937 CCC: $30.25 © 2005 American Chemical Society

Total Synthesis of Amythiamicin D
A R T I C L E S
Figure 2. Amythiamicins.
tion of thiostrepton 1 itself.23-26 We now describe the details
of the first total synthesis of the thiopeptide antibiotic amythia-
micin D 8 by a route that, like Nicolaou’s synthesis of thio-
strepton, uses a biomimetic strategy to establish the 2,3,6-tri-
substituted pyridine core of the natural product.27
Results and Discussion
The amythiamicins (A-D, 5-8) (Figure 2) are among the
most interesting thiopeptide antibiotics, and were isolated from
a strain of Amycolatopsis sp. MI481-42F4, and their structures
determined by degradative and spectroscopic techniques.28-30
They are among the very few thiopeptides that do not contain
a dehydroalanine residue,1 and they are also some of the most
biologically active. Not only are they reported to inhibit the
growth of Gram-positive bacteria, including methicillin-resistant
Staphylococcus aureus (MRSA),28 they have also been shown
to inhibit the action of elongation factor Tu (EF-Tu), a GTP-
dependent translation factor. Such inhibitors exhibit antimalarial
Figure 1. Structures of the thiopeptide antibiotics micrococcin P1,
thiostrepton, and the promothiocins.
side-chain.17,18 Subsequently, Ciufolini’s elegant synthesis of
structure 1,19 together with his detailed NMR studies,20 clarified
the situation: the Bycroft-Gowland structure 1 for micrococcin
P1 is correct in terms of connectivity of residues, but contains
a stereochemical misassignment at one or more centers. Hence,
in view of the structural uncertainty associated with micrococcin
P1, the first definitive synthesis of a thiopeptide natural product
was the preparation of promothiocin A 3, carried out in our
own laboratory in 1998.21,22 Very recently, in a landmark
synthesis, Nicolaou and co-workers have reported the prepara-
(23) Nicolaou, K. C.; Safina, B. S.; Zak, M.; Estrada, A. A.; Lee, S. H. Angew.
Chem., Int. Ed. 2004, 43, 5087-5092.
(24) Nicolaou, K. C.; Zak, M.; Safina, B. S.; Lee, S. H.; Estrada, A. A. Angew.
Chem., Int. Ed. 2004, 43, 5092-5097.
(25) Nicolaou, K. C.; Safina, B. S.; Zak, M.; Lee, S. H.; Nevalainen, M.; Bella,
M.; Estrada, A. A.; Funke, C.; Zecri, F. J.; Bulat, S. J. Am. Chem. Soc.
2005, 127, 11159-11175.
(17) Okumura, K.; Ito, A.; Yoshioka, D.; Shin, C.-G. Heterocycles 1998, 48,
1319-1324.
(26) Nicolaou, K. C.; Zak, M.; Safina, B. S.; Estrada, A. A.; Lee, S. H.;
Nevalainen, M. J. Am. Chem. Soc. 2005, 127, 11176-11183.
(27) Hughes, R. A.; Thompson, S. P.; Alcaraz, L.; Moody, C. J. Chem. Commun.
2004, 946-948.
(18) Okumura, K.; Suzuki, T.; Nakamura, Y.; Shin, C.-G. Bull. Chem. Soc. Jpn.
1999, 72, 2483-2490.
(19) Ciufolini, M. A.; Shen, Y.-C. Org. Lett. 1999, 1, 1843-1846.
(20) Fenet, B.; Pierre, F.; Cundliffe, E.; Ciufolini, M. A. Tetrahedron Lett. 2002,
43, 2367-2370.
(28) Shimanaka, K.; Kinoshita, N.; Iinuma, H.; Hamada, M.; Takeuchi, T. J.
Antibiotics 1994, 47, 668-674.
(21) Moody, C. J.; Bagley, M. C. J. Chem. Soc. Chem. Commun. 1998, 2049-
(29) Shimanaka, K.; Takahashi, Y.; Iinuma, H.; Naganawa, H.; Takeuchi, T. J.
Antibiotics 1994, 47, 1145-1152.
2050.
(22) Bagley, M. C.; Bashford, K. E.; Hesketh, C. L.; Moody, C. J. J. Am. Chem.
Soc. 2000, 122, 3301-3313.
(30) Shimanaka, K.; Takahashi, Y.; Iinuma, H.; Naganawa, H.; Takeuchi, T. J.
Antibiotics 1994, 47, 1153-1159.
9
J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005 15645

A R T I C L E S
Hughes et al.
Scheme 1. Retrosynthetic Analysis of Amythiamicin D 8.
Scheme 2. Bycroft-Floss Hypothesis for the Biosynthesis of the
Pyridine Core of the Thiopeptide Antibiotics.
A 3,21,22 we also used a ring synthesis approach, employing the
little known Bohlmann-Rahtz reaction35 as a route to the pyri-
dine core of the natural product.36 On this occasion, however,
we elected to draw our inspiration from Nature, and use a
biosynthesis-mimicking Diels-Alder route to the pyridine core
of the antibiotic as a key step.
In 1978, Bycroft and Gowland, in addition to reporting the
structure of micrococcin P1,16 suggested that its pyridine ring
(as well as the tetrahydropyridine in thiostrepton) could result
biogenetically “from the interaction of two dehydroalanine
units,” themselves derived from serine residues. This interesting
proposal for the biosynthesis of the pyridine ring in thiopeptides
was subsequently supported by isotopic labeling experiments
by Floss and co-workers, who showed that serine, the precursor
of dehydroalanine, was incorporated into the pyridine as
predicted by such a proposal.37,38 Floss also viewed the Bycroft
proposal for the biosynthesis of the pyridine ring 15 as a formal
Diels-Alder cycloaddition (i.e. 13, but not necessarily con-
certed) followed by aromatization as shown in Scheme 2, the
intermediate tetrahydropyridine 14 being clearly related to the
corresponding structural unit in thiostrepton.
activity against Plasmodium falciparum,31 the parasite that
causes the majority of malarial infections in humans, the most
potent being amythiamicin A (IC50 0.01 µM). It was shown that
the thiopeptide binds to P. falciparum EF-Tu, thereby blocking
protein synthesis in the parasite, and suggesting that drugs which
target this mechanism might be useful in the treatment of this
disease.
Although the detailed structural assignment of amythiamicin
D 8 was reported,29 the stereochemistry of the three chiral cen-
ters was not disclosed. Therefore, we assumed that they derive
from natural L-amino acids, an assumption supported by the
structure of the very closely related antibiotic GE2270A.8,32 Our
synthetic strategy is indicated in Scheme 1, and “obvious”
disconnections at the amide bonds reveal the building blocks
as the core pyridine 9, a simple glycine derivative 10, and
two thiazoles 11 and 12. However, it is the synthesis of the
core pyridine 9 that lies at the center of the problem, and in
view of the difficulties associated with the synthesis of poly-
substituted pyridines, we decided at the outset that we would
construct the pyridine fragment by a ring synthesis rather than
by an approach involving sequential modification and substitu-
tion of a preformed pyridine ring. In the thiopeptide arena, the
latter approach is by far the most common, although there are
some exceptions.33,34 Indeed in our synthesis of promothiocin
(32) Tavecchia, P.; Gentili, P.; Kurz, M.; Sottani, C.; Bonfichi, R.; Selva, E.;
Lociuro, S.; Restelli, E.; Ciabatti, R. Tetrahedron 1995, 51, 4867-4890.
(33) Ciufolini, M. A.; Shen, Y. C. J. Org. Chem. 1997, 62, 3804-3805.
(34) Yonezawa, Y.; Konn, A.; Shin, C. G. Heterocycles 2004, 63, 2735-2746.
(35) Bohlmann, F.; Rahtz, D. Chem. Ber. 1957, 90, 2265-2272.
(36) Bagley, M. C.; Dale, J. W.; Jenkins, R. L.; Bower, J. Chem. Commun.
2004, 102-103.
(37) Mocek, U.; Knaggs, A. R.; Tsuchiya, R.; Nguyen, T.; Beale, J. M.; Floss,
H. G. J. Am. Chem. Soc. 1993, 115, 7557-7568.
(38) Mocek, U.; Zeng, Z.; O’Hagan, D.; Zhou, P.; Fan, L.-D. G.; Beale, J. M.;
Floss, H. G. J. Am. Chem. Soc. 1993, 115, 7992-8001.
(31) Clough, B.; Rangachari, K.; Strath, M.; Preiser, P. R.; Wilson, R. J. M.
Protist 1999, 150, 189-195.
9
15646 J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005

Total Synthesis of Amythiamicin D
A R T I C L E S
Table 1. Preparation of 1-alkoxy-2-azabutadienes 19
Scheme 3. LR ) Lawesson’s Reagent.
R2
17 yield/%
R
19 yield/%
a
b
c
d
e
f
Ph
a
b
b
b
b
a
b
Me
Me
Me
Me
48
4-Cl-C6H4
2-thienyl
2-pyridyl
2-Me-4-thiazolyl
Ph
Ae
Bf
55c
65c
38c
72c
64
Me
PNBd
Me
g
h
31c
50
∼100
PNBd
a Commercial ethyl benzimidate hydrochloride used. b Imidate used
directly in next step. c Yield is over the two steps. d PNB ) 4-nitrobenzyl.
e A ) (S)-2-(2-tert-butoxycarbonylamino-3-methyl)propylthiazol-4-yl. f B
) (S)-2-[(2-tert-butoxycarbonylamino-3-methyl)propylthiazol-4-yl]thiazol-
4-yl.
steps) using the modified Hantzsch procedure to avoid racem-
ization.49 Conversion into the amide 16g was quantitative, and
thionation and a second Hantzsch reaction, again under Holza-
pfel’s modified conditions, gave the bis-thiazole 21 in excellent
yield. To check the stereochemical integrity of the bis-thiazole
21, it was converted into R-methoxy-R-trifluoromethyl pheny-
lacetic acid (Mosher) amides. Removal of the N-Boc-group with
TFA was followed by separate coupling reactions to (R)- and
(S)-Mosher’s acid to give the corresponding amides. Analysis
by 19F NMR spectroscopy established the purity of each dia-
stereomeric amide, confirming that no racemization had occurred
in the synthesis of the thiazole rings. The ester group in 21 was
finally converted into the corresponding amide 16h by treatment
with ammonia (Scheme 3).
In keeping with the desire to mimic Nature, the first dieno-
philes investigated were the serine derived N-acetyldehydro-
alanine esters 24a and 24b, the NHAc group mimicking the
N-terminus peptide chain in the proposed biosynthetic route (cf.
Scheme 2). Although not commonly used as dienophiles be-
cause of their poor reactivity, dehydroalanine derivatives do par-
ticipate in Diels-Alder reactions.50-57 On the other hand, there
is no precedent for simple enamides such as the phenyl sub-
stituted enamide 24c acting as a dienophile, and therefore this
compound, along with the more relevant 2-thiazolyl sub-
stituted derivatives 24d-24f, formed part of our early fea-
sibility studies. Methyl 2-acetamidoacrylate 24a is com-
mercially available, and its ethyl analogue 24b was prepared
from the corresponding acid. Dienophiles 24c-24f were
prepared by reduction of the corresponding oximes 23 using
iron with acetic anhydride-acetic acid in toluene.58 The two
thiazolyl ketones 22e and 22f were prepared as shown in Scheme
Although the case for a Diels-Alderase enzyme remains the
subject of debate,39-42 it does not invalidate biosynthesis-
inspired Diels-Alder approaches to target molecules. Our first
task was, therefore, to realize Bycroft’s original biosynthesis
proposal in a biomimetic cycloaddition route to 2,3,6-trisubsti-
tuted pyridines, involving the Diels-Alder reaction of serine-
derived 1-alkoxy-2-azadienes with dehydroalanine derivatives
(cf. 13).43 The dienes chosen for study were the 1-alkoxy-2-
azadienes 19, which mimic the dehydroalanine dipeptide diene
proposed by Bycroft and Floss, by fixing it in the required ‘enol’
form (cf. Scheme 2).44,45 The dienes were prepared from
O-acetylserine esters 18 by reaction with the imidates 17, which
are commercially available (R2 ) Ph) or obtained by reaction
of the corresponding carboxamide 16 with triethyloxonium
hexafluorophosphate,46 using a procedure based on a literature
route to 2-azadiene 19a.47 As summarized in Table 1, a range
of 2-azadienes 19 incorporating aryl (19a, 19b, 19f) and
heteroaryl (19c - e) groups was prepared. The 1-ethoxy-1-
phenyl-2-azadiene system was prepared with both methyl (19a)
and 4-nitrobenzyl (PNB) (19f) esters groups, the latter being
chosen to offer the possibility of selective ester deprotection at
a subsequent stage.
Also, two model thiopeptide dienes (19g, 19h), incorporating
a more complex thiazole and a bis-thiazole at the 1-position,
were prepared from the (S)-thiazole-4-carboxamides 16g and
16h, obtained from the known thiazole-4-carboxylate 2048 as
shown in Scheme 3. The chiral thiazole 20 was readily obtained
from N-Boc-valine on a 5-20 g-scale (∼65% yield over three
(39) Laschat, S. Angew. Chem., Int. Ed. Engl. 1996, 35, 289-291.
(40) Stocking, E. M.; Williams, R. M. Angew. Chem., Int. Ed. 2003, 42, 3078-
3115.
(49) Bredenkamp, M. W.; Holzapfel, C. W.; vanZyl, W. J. Synth. Commun.
1990, 20, 2235-2249.
(50) Wulff, G.; Lindner, H. J.; Bohnke, H.; Steigel, A.; Klinken, H. T. Liebigs
Ann. 1989, 527-531.
(41) Ose, T.; Watanabe, K.; Mie, T.; Honma, M.; Watanabe, H.; Yao, M.;
Oikawa, H.; Tanaka, I. Nature 2003, 422, 185-189.
(42) Guimaraes, C. R. W.; Udier-Blagovic, M.; Jorgensen, W. L. J. Am. Chem.
Soc. 2005, 127, 3577-3588.
(51) Cativiela, C.; Lopez, P.; Mayoral, J. A. Tetrahedron-Asymmetry 1990, 1,
379-388.
(52) Crossley, M. J.; Stamford, A. W. Aust. J. Chem. 1994, 47, 1695-1711.
(53) Burkett, B. A.; Chai, C. L. L.; Hockless, D. C. R. Aust. J. Chem. 1998, 51,
993-997.
(54) Burkett, B. A.; Chai, C. L. L. Tetrahedron Lett. 1999, 40, 7035-7038.
(55) Burkett, B. A.; Chai, C. L. L. Tetrahedron Lett. 2000, 41, 6661-6664.
(56) Avenoza, A.; Cativiela, C.; Busto, J. H.; Fernandez-Recio, M. A.; Peregrina,
J. M.; Rodriguez, F. Tetrahedron 2001, 57, 545-548.
(57) Avenoza, A.; Barriobero, J. I.; Cativiela, C.; Fernandez-Recio, M. A.;
Peregrina, J. M.; Rodriguez, F. Tetrahedron 2001, 57, 2745-2755.
(58) Burk, M. J.; Casy, G.; Johnson, N. B. J. Org. Chem. 1998, 63, 6084-
6085.
(43) Moody, C. J.; Hughes, R. A.; Thompson, S. P.; Alcaraz, L. Chem. Commun.
2002, 1760-1761.
(44) Buonora, P.; Olsen, J. C.; Oh, T. Tetrahedron 2001, 57, 6099-6138.
(45) Jayakumar, S.; Ishar, M. P. S.; Mahajan, M. P. Tetrahedron 2002, 58, 379-
471.
(46) Pattenden, G.; Thom, S. J. Chem. Soc., Perkin Trans. 1 1993, 1629-1636.
(47) Balsamini, C.; Bedini, A.; Galarini, R.; Spadoni, G.; Tarzia, G.; Hamdan,
M. Tetrahedron 1994, 50, 12375-12394.
(48) Downing, S. V.; Aguilar, E.; Meyers, A. I. J. Org. Chem. 1999, 64, 826-
831.
9
J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005 15647

A R T I C L E S
Scheme 4
Hughes et al.
Most of the Diels-Alder reactions described in Table 2 suffer
from competing ‘dimerization’ of the 2-azadiene to give
pyridines 26. For completeness, a series of separate ‘dimeriza-
tion’ experiments were carried out as summarized in Table 3
(Supporting Information). With the exception of diene 19h, all
azadienes investigated gave pyridines 26, presumably by way
of a Diels-Alder dimerization-aromatization sequence with
the loss of ethanol and the imidate. Similar ‘dimerizations’ of
2-azadienes to give pyridines have been noted previously.60
The yields of the desired 2,3,6-trisubstituted pyridines are
somewhat variable, although no attempts were made to optimize
the reactions, and probably reflect the fact that the diene-
dienophile components are not ideally matched in terms of their
electronic properties for an efficient Diels-Alder process.
Additionally, although initial Diels-Alder adducts were never
observed, the subsequent aromatization sequence may not be
trivial given the relatively poor natures of the acetamide anion
and ethoxide as potential leaving groups. Nevertheless, that is
what the ‘biomimetic’ Diels-Alder-aromatization sequence
requires, albeit under ‘biological’ conditions, rather than the
simple thermal conditions described here.
Scheme 5. 24a, R3 ) CO2Me; 24b, R3 ) CO2Et; 24c, R3 ) Ph;
24d, R3 ) 2-thiazolyl; 24e, R3 ) 4-EtO2C-2-thiazolyl; 24f, R3
)
4-BnO2C-2-thiazolyl.
Scheme 6. 25a, R ) Me, R2 ) Ph, R3 ) CO2Me; 25b, R ) Me,
R2 ) Ph, R3 ) CO2Et.
The above feasibility study has established for the first time
the viability of the long-standing proposal for the biosynthesis
of the pyridine ring in the thiopeptide antibiotics involving a
cycloaddition of two serine-derived dehydroalanine type frag-
ments. Thus, it has not only resulted in the realization of
Bycroft’s original idea, but also allowed us to adopt such a
cycloaddition approach to the core pyridine 9 of the amythia-
micins as the keystone of our overall strategy.
4, and the preparation of the enamide dienophiles 24 is
summarized in Scheme 5.
Initial experiments on the Diels-Alder reactions of the
2-azadienes were carried out with the 1-phenyl diene 19a and
methyl acetamidoacrylate 24a as dienophile. Prolonged heating
of the two components in xylene resulted in a 42% yield of
the pyridine 25a after chromatography (Scheme 6), thereby
establishing for the first time the viability of the ‘biomimetic’
Diels-Alder-aromatization sequence, albeit under thermal rather
than ‘biological’ conditions. Interestingly, the pyridine 25a
was also formed slowly in 79% yield on heating the diene 19a
alone. That the dehydroalanine dienophile ester 24a was ac-
tually involved in the cycloaddition was shown by the use of
the corresponding ethyl ester 24b that gave pyridine 25b,
together with 25a, formed by competing ‘dimerization’ of the
2-azadiene. The reaction was also investigated under micro-
wave irradiation,59 and these conditions were applied to a range
of other 1-ethoxy-2-azadienes 19 and enamide dienophiles
24 to give the corresponding pyridines 25 in modest yield
(Table 2, Supporting Information). Crucially, in terms of our
planned synthesis of a thiopeptide pyridine core, thiazole rings
could be incorporated into both diene and dienophile compo-
nents, resulting in the synthesis of a series of 2,3-dithiazolyl-
pyridines (Table 2, Supporting Information). In particular,
pyridine 25o is fully functionalized, with orthogonal protecting
groups, for further elaboration although in the event, it was
not used in the final synthesis of amythiamicin D, since the
selective deprotection could not be carried out satisfactorily.
The Diels-Alder reactions were regiospecific and gave the
2,3,6-trisubstituted pyridines 25 as evidenced by their 1H
NMR spectra, which showed two doublets (J ≈ 8 Hz) for
the two adjacent pyridine ring protons. There was no evidence
for the formation of the alternative 2,4,6-substituted regioiso-
mers.
With the key methodology established for the synthesis of
2,3,6-trisubstituted pyridines relevant to the core pyridine 9 of
the amythiamicins, attention turned to the preparation of the
thiazoles 11 and 12 that were also required by the retrosynthetic
analysis (Scheme 1). Although the Hantzsch reaction discovered
in 1889 remains one of the most reliable routes to thiazoles,
and has been used by others for the synthesis of the thiazole
building block 12 common to GE2270A and the amythia-
micins,61 we elected to use the versatile rhodium carbene N-H
insertion method developed in our own laboratory.62 The sub-
strate for the rhodium reaction was the aspartate-derived amide
28, readily obtained from commercially available (S)-N-Boc-
aspartic acid benzyl ester 27 using the mixed anhydride method.
The amide 28 underwent chemoselective N-H insertion with
the rhodium carbene derived from methyl 2-diazo-3-oxobutano-
ate by dirhodium tetraoctanoate catalyzed reaction. The resulting
1,4-dicarbonyl compound, ketoamide 29, formed as a mixture
of diastereoisomers, was readily converted into the thiazole 30
upon treatment with two equivalents of Lawesson’s reagent in
boiling THF.63 Use of less reagent or lower temperatures re-
sulted in poorer yields. It thus remained to install the correct
side-chain and this was achieved by hydrogenolysis of the
benzyl ester followed by amide formation to give 31, although
the first step was not without difficulty: fairly forcing condi-
(59) MicrowaVe Assisted Organic Synthesis; Tierney, J. P., Lidstro¨m, P., Eds.;
Blackwell Publishing Ltd: Oxford, 2005.
(60) Palacios, F.; Alonso, C.; Rubiales, G.; Ezpeleta, J. M. Eur. J. Org. Chem.
2001, 2115-2122.
(61) Suzuki, T.; Nagasaki, A.; Okumura, K.; Shin, C. G. Heterocycles 2001,
55, 835-840.
(62) Davies, J. R.; Kane, P. D.; Moody, C. J. Tetrahedron 2004, 60, 3967-
3977.
(63) Gordon, T. D.; Singh, J.; Hansen, P. E.; Morgan, B. A. Tetrahedron Lett.
1993, 34, 1901-1904.
9
15648 J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005

Total Synthesis of Amythiamicin D
A R T I C L E S
Scheme 7. Oct ) Octanoate; LR ) Lawesson’s Reagent.
Scheme 8. BTEAC ) Benzyltriethylammonium Chloride: EDCI )
1-(3-dimethylaminopropyl)-3-ethylcarbodiimide Hydrochloride;
HOBt ) 1-hydroxybenzotriazole.
Scheme 9. LR ) Lawesson’s Reagent.
tions and high catalyst loadings were required to achieve good
yields. Finally, lithium hydroxide hydrolysis revealed the free
thiazole-4-carboxylic acid 12 for a subsequent coupling reaction
(Scheme 7).
The synthesis of the second thiazole 11 was readily achieved
from the known (S)-ethyl thiazole-4-carboxylate 20 that, to
facilitate subsequent deprotection and coupling reactions, was
converted into the corresponding tert-butyl ester 33. Hydrolysis
of the ethyl ester to the acid 32 was followed by re-esterification
using phase-transfer catalyzed alkylation with tert-butyl bro-
mide.64 Selective removal of the N-Boc-group65 by treatment
with HCl in dioxane gave the required amine 11 isolated as its
hydrochloride salt. This was coupled to thiazolecarboxylic acid
12 using carbodiimide methodology to give the bis-thiazole 34
in modest yield. Bis-thiazole 34 was readied for a further
coupling reaction by a second selective removal of an N-Boc-
group in the presence of a tert-butyl ester to give the bis-thiazole
amine hydrochloride 35 (Scheme 8).
With the thiazole building blocks prepared and already co-
joined, the stage was set to exploit the biosynthesis inspired
azadiene Diels-Alder reaction. Both diene and dienophile com-
ponents are required to contain a thiazole-4-carboxylate, and
therefore the ester groups need to be differentiated. Since amy-
thiamicin D bears a thiazole methyl ester at the pyridine-6-
position, this fixes the corresponding 3-substituent of the pro-
posed 2-azadiene component, thereby defining the thiazole 24f
containing an orthogonally protected carboxyl, as the dienophile.
The 2-azadiene component 42 that contains the remaining three
thiazole rings was constructed from the valine-derived bis-
thiazole imidate 17h described earlier and the serine-derived
thiazole 41. N-Boc-Serine methyl ester was converted into the
known silyl derivative 36; subsequent reaction with ammonia
gave the amide 37 and hence the thioamide 38 (Scheme 9).
Hantzsch reaction with methyl bromopyruvate established the
(64) Chevallet, P.; Garrouste, P.; Malawska, B.; Martinez, J. Tetrahedron Lett.
1993, 34, 7409-7412.
(65) Gibson, F. S.; Bergmeier, S. C.; Rapoport, H. J. Org. Chem. 1994, 59,
3216-3218.
9
J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005 15649

A R T I C L E S
Hughes et al.
Scheme 10. PyBOP ) Benzotriazol-1-yloxy-tris(pyrrolidino)Phosphonium Hexafluorophosphate; DPPA ) Diphenylphosphoryl Azide.
thiazole 39 that was prepared for azadiene formation by de-
silylation and acetylation to give the thiazole 40. Building on
the earlier examples, the key azadiene 42 was accessed by
reaction of amine hydrochloride 41, obtained by simple HCl-
mediated deprotection of 40, with imidate 17h, followed by
elimination of the acetate using DBU as base (Scheme 9).
The synthesis had now reached the critical Diels-Alder
reaction, and heating the azadiene 42 with enamide dienophile
24f under microwave irradiation in toluene at 120 °C for 12 h
gave the required 2,3,6-tris(thiazolyl)pyridine 9, albeit in a
modest 33% yield. Given that pyridine 9 forms the core of the
natural product, it was essential to establish its stereochemical
integrity, and this was again achieved by formation of Mosher
amides. Removal of the N-Boc-group with TFA was followed
by separate coupling reactions to (R)- and (S)-Mosher’s acid to
give the corresponding amides. Analysis by 19F NMR spec-
troscopy established the purity of each diastereomeric amide,
confirming that no racemization had occurred during the
synthesis of diene 42 and hence pyridine 9.
With all the components now in hand, the assembly of the
natural product could be addressed. Deprotection of the ter-
minal N-Boc-group on the bis-thiazole moiety of pyridine 9 was
followed by a PyBOP-mediated coupling to N-Boc-glycine 10
to give the complete right-hand fragment of amythiamicin D
43 (Scheme 10). After much experimentation, it was found
that the benzyl ester in compound 43 could be removed by
hydrogenolysis over palladium black to give the corresponding
carboxylic acid. The use of other palladium catalysts re-
sulted in little or no debenzylation, presumably as a result
of catalyst poisoning by the poly-sulfur system; under more
forcing conditions, reduction of the pyridine ring occurred. A
second PyBOP-mediated reaction successfully coupled the
left-hand bis(thiazole) 35 to provide the cyclization precur-
sor 44. The terminal N-Boc and tert-butyl ester groups in 44
were simultaneously cleaved using TFA, and the resulting amino
acid treated with diphenylphosphoryl azide (DPPA)66-69 and
Hu¨nig’s base in DMF. This resulted in macrolactamization in
a yield of 73% (from 44) to give amythiamicin D 8 (Scheme
10).
Our synthetic material had properties consistent with those
reported for the natural product,29 suggesting that our original
assumption about the stereochemistry of the three chiral centers
was correct. Subsequent correspondence with the original
authors revealed that unpublished X-ray crystallographic data
substantiates the (10S,19S,29S)-stereochemistry thereby provid-
ing final confirmation that we had completed the first synthesis
of the natural product amythiamicin D, and paving the way for
syntheses of other thiopeptide natural products.
Acknowledgment. This paper is dedicated to the former staff
and students of the Department of Chemistry in the University
of Exeter (1901-2005). We thank the EPSRC and AstraZeneca
for financial support, Dr I. Prokes for NMR spectroscopy, the
EPSRC Mass Spectrometry Center at Swansea for mass spectra,
(66) Qian, L.; Sun, Z.; Deffo, T.; Mertes, K. B. Tetrahedron Lett. 1990, 31,
6469-6472.
(67) Shioiri, T.; Hamada, Y.; Matsuura, F. Tetrahedron 1995, 51, 3939-
3958.
(68) Wagner, B.; Beugelmans, R.; Zhu, J. P. Tetrahedron Lett. 1996, 37, 6557-
6560.
(69) Jayaprakash, S.; Pattenden, G.; Viljoen, M. S.; Wilson, C. Tetrahedron
2003, 59, 6637-6646.
9
15650 J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005

Total Synthesis of Amythiamicin D
A R T I C L E S
and Drs. Yoshikazu Takahashi and Yuzuru Akamatsu (Microbial
Chemistry Research Foundation) for helpful exchange of
information.
23f, 24e, 24f, 25i-l, 25n, 25o, 26b, 26c, 26e, 29, 30, 11, 34,
37-39, 42-44, and 8; 19F NMR analyses of Mosher amides of
compounds 21 and 9. This material is available free of charge
via the Internet at http://pubs.acs.org.
Supporting Information Available: Experimental details and
characterization data for all compounds described herein; copies
1
of H and 13C NMR spectra of compounds 16h, 18b, 19b-h,
JA0547937
9
J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005 15651