Macrolactamization versus macrolactonization: Total synthesis of FK228, the depsipeptide histone deacetylase inhibitor

Article abstract of DOI:10.1021/jo801866z

(Chemical Equation Presented) The cyclic depsipeptide FK228 is the only natural product histone deacetylase (HDAC) inhibitor that has advanced to clinical trials as an anticancer agent. While currently obtained by fermentation, total synthesis is an attractive alternative that will facilitate the preparation of unnatural analogues. The previous total syntheses of FK228 featured macrocylization by ester bond formation from a seco-hydroxy acid. Such routes are operationally jeopardized by the steric hindrance of the carboxylic acid and the sensitivity of the allylic alcohol toward elimination. We report a strategically different approach whereby the ester bond is formed intermolecularly at an early stage and macrocyclization is efficiently achieved by amide bond formation.

Full text of DOI:10.1021/jo801866z

Macrolactamization versus Macrolactonization: Total Synthesis of
FK228, the Depsipeptide Histone Deacetylase Inhibitor†
Shijun Wen,‡ Graham Packham,§ and A. Ganesan*,‡
School of Chemistry, UniVersity of Southampton, Southampton SO17 1BJ, United Kingdom, and Cancer
Research UK Clinical Centre, Cancer Sciences DiVision, School of Medicine, UniVersity of Southampton,
Southampton SO16 6YD, United Kingdom
ganesan@soton.ac.uk
ReceiVed August 21, 2008
The cyclic depsipeptide FK228 is the only natural product histone deacetylase (HDAC) inhibitor that
has advanced to clinical trials as an anticancer agent. While currently obtained by fermentation, total
synthesis is an attractive alternative that will facilitate the preparation of unnatural analogues. The previous
total syntheses of FK228 featured macrocylization by ester bond formation from a seco-hydroxy acid.
Such routes are operationally jeopardized by the steric hindrance of the carboxylic acid and the sensitivity
of the allylic alcohol toward elimination. We report a strategically different approach whereby the ester
bond is formed intermolecularly at an early stage and macrocyclization is efficiently achieved by amide
bond formation.
Introduction
the acetylation catalyzed by histone acetyltransferases (HATs)
or the deacetylation catalyzed by histone deacetylases (HDACs)
are thus of value as chemical probes of biological function, and
as therapeutic agents. To date, drug discovery efforts have
focused on the zinc-dependent HDACs,2 of which there are 11
(HDAC1-11) in the human genome.
Zinc-dependent HDACs are metallohydrolases sharing a
highly conserved catalytic domain containing an active site zinc
ion. HDAC inhibitors typically contain a “warhead” that
reversibly binds to the catalytic zinc atom, linked by a spacer
to a “cap” that extends beyond the substrate-binding channel
into the enzyme’s surface exposed “rim”. One can envision two
extremes of HDAC inhibitors depending on the relative
importance of warhead and cap binding to the enzyme. Merck’s
vorinostat, the first HDAC inhibitor to reach the market, is a
The reversible acetylation1 of lysine residues represents an
important mode of protein posttranslational modification. As it
converts a basic primary amine protonated at physiological pH
to a neutral acetamide, it has major consequences for protein
structure and function. It is of particular significance in
eukaryotic histone proteins, where the pattern of posttranslational
modification is part of the “epigenetic code” that regulates gene
transcription or repression. Small molecule inhibitors of either
* To whom correspondence should be addressed. School of Chemistry,
University of Southampton, Southampton SO17 1BJ, United Kingdom. Tel:+44
2380593897. Fax: +44 2380596805.
† Abbreviations used: Dhb, dehydrobutyrine; HAT, histone acetyltransferase;
HATU, 2-(1H-7-azabenzotriazol-1-yl)--1,1,3,3-tetramethyluronium hexafluoro-
phosphate; HDAC, histone deacetylase; MSNT: 1-(2-mesitylenesulfonyl)-3-nitro-
1H-1,2,4-triazole; PyBOP: benzotriazol-1-yl-oxytripyrrolidinophosphonium hexaflu-
orophosphate; TMSE: (trimethylsilyl)ethyl.
(2) (a) Bolden, J. E.; Peart, M. J.; Johnstone, R. W. Nature ReV. Drug
DiscoVery 2006, 5, 769–784. (b) Elaut, G.; Rogiers, V.; Vanhaecke, T. Curr.
Pharm. Design 2007, 13, 2584–2620. (c) Yang, X.-J.; Seto, E. Nature ReV. Mol.
Cell Biol. 2008, 9, 206–218. (d) Paris, M.; Porcelloni, M.; Binaschi, M.; Fattori,
D. J. Med. Chem. 2008, 51, 1505–1529. (e) Nolan, L.; Johnson, P. W. M.;
Ganesan, A.; Packham, G.; Crabb, S. J. Br. J. Cancer 2008, 99, 689–694.
‡ School of Chemistry.
§ School of Medicine.
(1) (a) Biel, M.; Wascholowski, V.; Giannis, A. Angew. Chem., Int. Ed. 2004,
44, 3186–3216. (b) Hodawadekar, S. C.; Marmorstein, R. Oncogene 2007, 26,
5528–5540.
10.1021/jo801866z CCC: $40.75
Published on Web 11/08/2008
2008 American Chemical Society
J. Org. Chem. 2008, 73, 9353–9361 9353

Wen et al.
SCHEME 1. Synthesis of the Carboxylic Acid Fragment
Nature has thus far revealed only six examples of the bicyclic
depsipeptides and five structurally related azumamides9 which
contain a carboxylic acid or carboxamide warhead in place of
the thiol. It is a moot point if these natural products are best-
in-class, and they offer limited opportunities for semisynthetic
modification. Thus, the discovery of additional analogues that
may be superior has to rely on de novo total synthesis. Despite
FK228’s progression to human clinical trials, synthetic analogues
wereunknownuntilwerecentlyreportedthefirststructure-activity
relationships of the natural product.10
FIGURE 1. Structures of vorinostat and the bicyclic depsipeptide
HDAC inhibitors.
classic example3 with a potent active site zinc-binding hydrox-
amic acid and a minimal cap. At the other end are compounds
with relatively weak zinc-binding groups that compensate by
having large caps engaged in significant binding interactions
with the enzyme rim. These are primarily natural products, a
historically well established source of biologically active
compounds.4
Natural product HDAC inhibitors with large caps are cyclic
peptides and depsipeptides that can be further subdivided into
two families. The first, exemplified by trapoxin, HC-toxin and
apicidin, are cyclic tetrapeptides containing a mixture of L- and
D-amino acids and cyclic amino acids such as proline or
pipecolic acid. Both these features alleviate the ring strain
associated with cyclic tetrapeptides containing all L or all D
amino acids. The second family comprises bicyclic depsipep-
tides5 (Figure 1) isolated from bacterial fermentation extracts.
These bicyclic depsipeptide HDAC inhibitors are larger mac-
rocycles compared to the cyclic tetrapeptides and are hence less
stereochemically constrained. Typically, the amino acids are of
D-stereochemistry, and the backbone of these natural products
can be mapped in a “retro-inverso”6 manner onto that of the
enzyme substrate. Unlike the tetrapeptides, the bicyclic dep-
sipeptides are prodrugs, with the disulfide bridge undergoing
intracellular reduction and the resulting butenylthiol side chain
then serving as the active site binding group. Although this is
a weaker warhead compared to a hydroxamic acid, the affinity
of the macrocyclic scaffold for the enzyme leads to significantly
higher overall potency than vorinostat. Furthermore, as the
enzyme rim is less conserved than the active site, there is the
potential for depsipeptides to discriminate between the eleven
HDAC isoforms. This is indeed the case, and the depsipeptides
have pronounced selectivity7 for the so-called Class I HDACs
(HDAC1, HDAC2, HDAC3 and HDAC8), which fortuitously
happen to be the isoforms most heavily implicated8 in cancer
and other proliferative disease conditions.
The earliest total synthesis of a bicyclic depsipeptide HDAC
inhibitor was of FK228, reported in 1996 by Simon and co-
workers.11 After nearly a decade of inactivity, there followed
in rapid succession the synthesis of FR901,375 by the Went-
worth-Janda group,12 spiruchostatins by us13 and others,14 and
a second15 FK228 synthesis by Williams. A key challenge in
these syntheses is the preparation of the common ꢀ-hydroxy-
γ,δ-unsaturated-ꢁ-thio-carboxylic acid, containing four func-
tional groups and a stereogenic center within a seven carbon
atom framework. All the early routes converged on the
asymmetric acetate aldol reaction of aldehyde 1 (Scheme 1),
readily obtained12 in two steps from acrolein by the Wentworth-
(8) (a) Kozikowski, A. P.; Butler, K. V. Curr. Pharm. Design 2008, 14, 505–
528. (b) Suzuki, T.; Itoh, Y.; Miyata, N. Curr. Pharm. Design 2008, 14, 529–
544. (c) Bieliauskas, A. V.; Pflum, M. K. H. Chem. Soc. ReV. 2008, 37, 1402–
1413.
(9) (a) Nakao, Y.; Yoshida, S.; Matsunaga, S.; Shindoh, N.; Terada, Y.; Nagai,
K.; Yamashita, J. K.; Ganesan, A.; van Soest, R. W. M.; Fusetani, N. Angew.
Chem., Int. Ed. 2006, 45, 7553–7557. (b) Izzo, I.; Maulucci, N.; Bifulco, G.; De
Riccardis, F. Angew. Chem., Int. Ed. 2006, 45, 7557–7760. (c) Wen, S.; Carey,
K. L.; Nakao, Y.; Fusetani, N.; Packham, G.; Ganesan, A. Org. Lett. 2007, 9,
1105–1108. (d) Maulucci, N.; Chini, M. G.; Di Micco, S.; Izzo, I.; Cafaro, E.;
Russo, A.; Gallinari, P.; Paolini, C.; Nardi, M. C.; Casapullo, A.; Riccio, R.;
Bifulco, G.; De Riccardis, F. J. Am. Chem. Soc. 2007, 129, 3007–3012.
(10) Yurek-George, A.; Cecil, A.; Mo, A. H. K.; Wen, S.; Rogers, H.; Habens,
F.; Maeda, S.; Yoshida, M.; Packham, G.; Ganesan, A. J. Med. Chem. 2007,
50, 5720–5726.
(11) Li, K. W.; Xing, W.; Simon, J. A. J. Am. Chem. Soc. 1996, 118, 7237–
7238.
(12) Chen, Y.; Gambs, C.; Abe, Y.; Wentworth, P., Jr.; Janda, K. D. J. Org.
Chem. 2003, 68, 8902–8905.
(13) (a) Yurek-George, A.; Habens, F.; Brimmell, M.; Packham, G.; Ganesan,
A. J. Am. Chem. Soc. 2004, 126, 1030–1031. (b) Davidson, S. M.; Townsend,
P. A.; Carroll, C.; Yurek-George, A.; Balasubramanyam, K.; Kundu, T. K.;
Stephanou, A.; Packham, G.; Ganesan, A.; Latchman, D. S. ChemBiochem 2005,
6, 162–170.
(3) Marks, P. A.; Breslow, R. Nat. Biotechnol. 2007, 25, 84–90.
(4) Ganesan, A. Curr. Opin. Chem. Biol. 2008, 12, 306–317.
(5) Towensend, P. A.; Crabb, S. J.; Davidson, S. M.; Johnson, P. W. M.;
Packham, G.; Ganesan, A. In Chemical Biology; Schreiber, S. L., Kapoor, T. M.,
Wess, G., Eds.; Wiley-VCH: Weinheim, 2007; Vol 2, pp 693-720.
(6) Chorev, M.; Goodman, M. Acc. Chem. Res. 1993, 26, 266–273.
(7) Furumai, R.; Matsuyama, A.; Kobashi, N.; Lee, K.-H.; Nishiyama, N.;
Nakajima, H.; Tanaka, A.; Komatsu, Y.; Nishino, N.; Yoshida, M.; Horinouchi,
S. Cancer Res. 2002, 62, 4916–4921.
(14) (a) Doi, T.; Iijima, Y.; Shin-ya, K.; Ganesan, A.; Takahashi, T.
Tetrahedron Lett. 2006, 47, 1177–1180. (b) Takizawa, T.; Watanabe, K.; Narita,
K.; Oguchi, T.; Abe, H.; Katoh, T. Chem. Commun. 2008, 1677–1679. (c)
Takizawa, T.; Watanabe, K.; Narita, K.; Kudo, K.; Oguchi, T.; Abe, H.; Katoh,
T. Heterocycles 2008, 76, 275–290.
(15) Greshock, T. J.; Johns, D. M.; Noguchi, Y.; Williams, R. M. Org. Lett.
2008, 10, 613–616.
9354 J. Org. Chem. Vol. 73, No. 23, 2008

Macrolactamization Versus Macrolactonization
Janda synthesis. The desired aldol was first achieved by Simon11
using Carreira’s reagent-based conditions in 99% yield and
>98% enantiomeric excess. However, neither Wentworth-Janda
nor ourselves could reliably reproduce this result, and we opted
for substrate-based control with oxazolidinone16 auxiliaries. As
the Evans auxiliary is not diastereoeselective in acetate aldols,17
Wentworth-Janda used the 2-chloroacetyl enolate followed by
reductive removal of the chloride after the aldol reaction. In
our group, we used the 4-alkylthiazolidinethione Fujita-Nagao
auxiliary which does give high stereoselectivity in acetate
aldols.18 In their later synthesis,14a Doi-Takahashi used See-
bach’s quaternary oxazolidinone as an alternative solution.
Meanwhile, Katoh14b and Williams15 devised nonaldol but
lengthier routes respectively relying on the chiral pool and
Noyori asymmetric reduction.
FIGURE 2. Compounds tested in macrolactonization reactions.
suggested21 by analysis of the FK228 producing gene cluster.
Reversing the order of macrocyclizations appears to be coun-
terproductive in the laboratory. Thus, Doi-Takahashi were
unsuccessful14a in macrolactonization of a monocycle with the
disulfide bridge in place, although modeling suggests the
existence of low energy conformers suitably oriented for
cyclization.
From a practical point of view, the thiazolidinedione aldol
sequence is concise, providing 2 in four steps from acrolein
and we have routinely performed the synthesis on multigram
scale. As it is an auxiliary-based diastereoselective approach,
the aldol impurity is a diastereomer easily removed by con-
ventional chromatography. In contrast, a reagent-based enan-
tioselective approach would need a chiral separation from the
undesired minor enantiomer. Furthermore, in our aldol route
the auxiliary can be recovered after hydrolysis and advanta-
geously recycled.
Results and Discussion
We have prepared over a hundred FK228 and spiruchostatin
analogues by the Shiina macrolactonization, with every case
being successful. Given this background, the total synthesis of
FK228 itself by the same methodology seemed trivial but proved
to be surprisingly insurmountable. The attempted Shiina mac-
rolactonization of 3 or 4 (Figure 2) did not yield cyclized product
in detectable quantities. With hindsight, we believe this failure
is due to high steric hindrance at the carboxylic acid undergoing
macrocyclization. In these compounds, there is an adjacent
isopropyl group whereas the corresponding intermediate in our
spiruchostatin A total synthesis is unencumbered at the R-posi-
tion. Although we had made analogues with substitution at this
position such as methyl and benzyl, none had involved
R-branching like the isopropyl group in FK228. Apparently, this
difference in steric demand is sufficient to prevent the cyclization.
Needing a supply of FK228 for our biological studies, we
then expended considerable effort in repeating Simon’s total
synthesis.11 Since his Mitsunobu macrolactonization involves
inversion of the alcohol, this meant that we had to go back and
prepare epi-4, the epimer of 4 at the alcohol chiral center. The
Mitsunobu reaction of epi-4, when carried out under Simon’s
conditions, was productive and resulted in the desired cyclodep-
sipeptide. Nevertheless, due to the large excess of Ph3P (25
equiv), diisopropyl azodicarboxylate (20 equiv) and TsOH (5
equiv), the product is a very minor component of the reaction
mixture. Chromatographic purification eventually provided
10-20% of the depsipeptide, a significant decrease over Simon’s
reported yield. Overall, we decided the Simon synthesis was
not practical for scale-up, as a large amount of the precious
late stage intermediate epi-4 is sacrificed in the macrocyclization.
Similar findings were reported by Williams,15 who also repeated
the Simon macrolactonization in their FK228 synthesis. Despite
numerous trials, the cyclization yield was only 24%.
A second key operation is the macrocylization to form the
cyclodepsipeptide and disulfide rings. Simon selected the ester
bond for macrolide cyclization19 from the linear seco-hydroxy
acid, but a variety of carboxyl activating conditions gave only
recovered starting material or side-products. They state that
Keck’s modification of Steglich’s carbodiimide esterification did
afford the cyclic depsipeptide, but in a “disappointingly low”
yield. Faced with this obstacle, they switched to an “umpolung”
macrocyclization by activation of the alcohol under Mitsunobu
conditions and displacement by the carboxylate. As the allylic
alcohol is susceptible to elimination, extensive optimization of
this reaction was needed and ultimately furnished the product
in 62% yield. A similar Mitsunobu macrolactonization was
employed in the Wentworth-Janda FR901,375 synthesis, oc-
curring in 59% yield. In our spiruchostatin A synthesis, we
returned to carboxyl activation of the linear hydroxy acid, and
the popular Yamaguchi conditions were found to give the cyclic
depsipeptide in a respectable 53% yield. In their second
generation approach,14a Doi-Takahashi improved this further
by using the more reactive Shiina reagent.20 The Shiina reagent
enables the macrolactonization to proceed under milder room
temperature conditions thus minimizing the formation of side-
products.
Once the cyclodepsipeptide is formed, the second disulfide
macrocycle is readily installed by oxidative conditions. Interest-
ingly, this sequence of macrolactonization followed by disulfide
bond formation mirrors the final stages of the biosynthesis, as
These results prompted us to explore a new route wherein
the difficult ester bond is installed at an earlier stage, and
macrocyclization performed by lactamization rather than lac-
tonization. Indeed, the literature on cyclic depsipeptides suggests
that amide bond formation is often easier than ester bond
formation for the macrocyclization step.22 Nevertheless, in the
total syntheses of depsipeptide HDAC inhibitors, all previous
(16) (a) Vela´zquez, F.; Olivo, H. F. Curr. Org. Chem. 2002, 6, 1–38. (b)
Evans, D. A.; Shaw, J. T. Actual. Chim. 2003, 263-264, 35–38.
(17) Evans, D. A.; Bartroli, J.; Shih, T. L. J. Am. Chem. Soc. 1981, 103,
2127–2129.
(18) (a) Nagao, Y.; Hagiwara, Y.; Kumagai, T.; Ochiai, M.; Inoue, T.;
Hashimoto, K.; Fujita, E. J. Org. Chem. 1986, 51, 2391–2393. (b) Aiguade´, J.;
Gonza´lez, A.; Urp´ı, F.; Vilarrasa, J. Tetrahedron Lett. 1996, 37, 8949–8952.
(19) Parenty, A.; Moreau, X.; Campagne, J.-M. Chem. ReV. 2006, 106, 911–
939.
(21) Cheng, Y.-C.; Yang, M.; Matter, A. M. Appl. EnViron. Microbiol. 2007,
73, 3460–3469.
(20) Shiina, I.; Kubota, M.; Ibuka, R. Tetrahedron Lett. 2002, 42, 7535–
7539.
(22) (a) Davies, J. S. J. Peptide Sci. 2003, 9, 471–501. (b) Stawikowski, M.;
Cudic, P. Methods Mol. Biol. 2007, 386, 321–339.
J. Org. Chem. Vol. 73, No. 23, 2008 9355

Wen et al.
SCHEME 2. Synthesis of a Linear Depsipeptide
routes11-15 have selected macrolactonization instead of mac-
rolactamization. One reason is the initial success reported by
Simon, suggesting that macrolactonization is a viable strategy.
Second, macrolactamization requires a more complex set of
protecting groups, as will be seen below.
rest of the peptide sequence. The solution would then be to carry
out this ester bond formation at the very beginning. The new
route began with the carbodiimide-mediated coupling between
Fmoc-L-valine and 6. This early stage esterification proceeded
in high yield to furnish diester 14 (Scheme 3), confirming that
valine by itself is significantly more reactive than the homolo-
gated peptides. Meanwhile, starting from D-cysteine, the trip-
eptide Fmoc-D-Val-D-Cys(Tr)-Dhb-OH 18 was prepared by
stepwise peptide coupling and introduction of the Dhb side chain
by elimination from threonine.
The Fmoc group of 14 was removed, and the amine coupled
with tripeptide 18 to furnish linear depsipeptide 13 (Scheme
4). This new route proceeded in high yields throughout and was
a significant improvement over the late stage esterification of
12. With practical quantities of 13 in hand, the orthogonal Fmoc
and TMSE protecting groups were then removed to unmask
the linear seco-amino acid. The crucial macrolactamization using
HATU (2-(1H-7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluro-
nium hexafluorophosphate) as condensing agent afforded dep-
sipeptide 19 at the first attempt, and reliably occurred in >70%
yield in subsequent experiments. In these reactions, we did not
observe significant byproducts resulting from guanidinylation,
a known issue with HATU.23 Although not investigated, the
macrocylization yields may be further improved by switching
to coupling reagents such as 7-azabenzotriazol-1-yloxy-tris-
(pyrrolidino)phosphonium hexafluorophosphate (PyAOP) that
avoid this side reaction.24
Our synthesis began with the aldol product 5 (Scheme 2).
Hydrolysis of the auxiliary releases ꢀ-hydroxy acid 2, which
was protected as the 2-(trimethylsilyl)ethyl (TMSE) ester 6. The
choice of ester was dictated by the need for later deprotection
to be accomplished under neutral conditions and orthogonality
to other functionality in the molecule. For the peptide fragment,
Fmoc-L-valine was similarly protected as TMSE ester 7. Since
the deprotection mechanism involves reaction at silicon rather
than the carbonyl, this was expected to minimize issues with
steric hindrance from the valine side chain. Following Fmoc
deprotection, sequential coupling with Fmoc-L-threonine and
Fmoc-D-cysteine gave dipeptide 8 and tripeptide 9 respectively.
Peptide 9 was subjected to mesylation to give 10, followed by
elimination and coupling with Fmoc-D-valine to provide tet-
rapeptide 11. The TMSE ester was next deprotected by fluoride.
As this was accompanied by cleavage of the Fmoc protecting
group, the crude reaction mixture was treated with Fmoc-Cl,
affording 12 in 70% overall yield. The key intermolecular
esterification with the ꢀ-hydroxy ester 6 was then investigated,
and proved to be as challenging as the intramolecular macro-
lactonization of 3 or 4. Best results were obtained with 1-(2-
mesitylenesulfonyl)-3-nitro-1H-1,2,4-triazole (MSNT) which
furnished 13 in a modest 34% yield, accompanied by the
revovery of 42% unreacted alcohol 6. Prolonged reaction times
did not help, instead leading to epimerization.
The successful cyclization of 13 confirms that the amide-
bond forming strategy toward the FK228 macrocycle is far
superior to ester-bond formation. It avoids the difficulties
associated with macrolactonization whether by carboxylic acid
or alcohol activation, due to the sterically hindered nature of
Our observations indicate that esterification of FK228 precur-
sors by carboxyl activation is a difficult process, whether it is
intermolecular as in 12 or an intramolecular macrocylization
as in 3 and 4. Since valine itself is not known to be so
recalcitrant, this suggests that the poor reactivity is not only
due to the isopropyl group but is further exacerbated by the
(23) Albericio, F.; Bofill, J. M.; El-Faham, A.; Kates, S. A. J. Org. Chem.
1998, 63, 9678–9683.
(24) (a) Albericio, F. Curr. Opin. Chem. Biol. 2004, 8, 211–221. (b)
Montalbetti, C. A. G. N.; Falque, V. Tetrahedron 2005, 61, 10827–10852.
9356 J. Org. Chem. Vol. 73, No. 23, 2008

Macrolactamization Versus Macrolactonization
SCHEME 3. Synthesis of Depsipeptide and Peptide Fragments for Intermolecular Coupling
SCHEME 4. Final Stages of the FK228 Total Synthesis
the former and the sensitivity of the latter toward elimination.
The NMR spectrum of depsipeptide 19 indicated the presence
of the tetrabutylammonium cation (from the TBAF deprotection
of the ester in 13). We believe this may be due to encapsulation
of the cation by the macrocyclic scaffold. The cyclodepsipeptide
was taken forward, and oxidative disulfide bond formation
provided pure FK228 after silica column chromatography.
Synthetic FK228 was identical in physical properties to a sample
of naturally isolated material, and spectroscopic data matched
that reported in the literature. Furthermore, synthetic FK228
showed equal activity to the natural product as an inhibitor of
cell growth and HDAC enzymes in vitro.
In summary, we have completed the first total synthesis of
FK228 using amide bond formation for the key macrocylization
step. Our route is convergent, taking six steps from acrolein to
depsipeptide 14 and seven steps from Fmoc-D-Cys(Tr)-OH to
peptide 18, followed by their union and a further four steps to
complete the total synthesis of FK228. Importantly, this bypasses
the difficulties encountered in macrolactonization for less
reactive substrates, as is certainly true for FK228 and likely to
be the case for unnatural analogues with similarly bulky
substituents proximal to the carboxylic acid. The successful route
hinged upon early formation of the difficult ester bond, and the
choice of orthogonal protecting groups to facilitate selective
manipulation of advanced intermediates.
molecule total synthesis and the difficulty of reproducing these
in a new environment. We believe that our second-generation
FK228 synthesis circumvents these obstacles, and the route
described here was successfully transferred to a contract research
organization for a gram scale synthesis of FK228.
Experimental Section
(4E,3S)-3-Hydroxy-7-mercapto-[S-triphenylmethyl]-4-hepteno-
ic acid (2). To a stirred solution of 5 (3.80 g, 6.76 mmol) in THF
(120 mL) cooled in an ice bath was added a solution of LiOH (324
mg, 13.53 mmol) in water (40 mL). After stirring for 1 h, the
solution was acidified to pH 2-3 with 1 M aq HCl, and diluted
with EtOAc (300 mL). The phases were separated, and the aqueous
phase extracted with EtOAc (150 mL × 3). The combined organic
phases were dried (Na2SO4), concentrated, and purified by flash
chromatography (eluent: EtOAc/Hexane ) 1/3-3/2) to provide 2
1
(2.71 g, 96%) as a white solid: [R]25 -4.4 (c 1.0, CH2Cl2); H
D
NMR (400 MHz, CDCl3) δ 7.14-7.49 (m, 15H), 5.51-5.64 (m,
1H), 5.42 (dd, 1H, J ) 15.4, 6.4 Hz), 4.39-4.50 (m, 1H), 2.46-2.59
(m, 2H), 2.14-2.27 (m, 2H), 2.01-2.13 (m, 2H); 13C NMR (100
MHz, CDCl3) 176.7, 145.0, 131.8, 130.9, 129.7, 128.0, 126.8, 68.6,
66.8, 41.3, 31.5, 31.4; MS m/z 417 (M-H+).
(E)-(S)-3-Hydroxy-7-tritylsulfanyl-hept-4-enoic Acid 2-Tri-
methylsilanyl-ethyl Ester (6). To a stirred solution of 2 (300 mg,
0.71 mmol) and trimethlysilylethanol (0.41 mL, 2.84 mmol) cooled
in an ice bath was added DMAP (5 mg) and DCC (161 mg, 0.78
mmol). After stirring for 2 h, the solution was filtered and
concentrated. The residue was purified by flash chromatography
(eluent: EtOAc/Hexane ) 1/10-1/5) to give 6 (296 mg, 80%) as
a colorless liquid: IR 3400, 1729, 1249 cm-1; 1H NMR (400 MHz,
CDCl3) δ 7.35-7.38 (m, 6 H), 7.15-7.26 (m, 9 H), 5.53 (m, 1 H),
5.37 (dd, 1H, J ) 6.0, 15.2 Hz), 4.38 (m, 1 H), 4.15 (t, 2 H, J )
8.9 Hz), 2.42 (m, 2 H), 2.17 (m, 2 H), 2.04 (m, 2 H), 0.94 (m, 2
The total synthesis of natural products is considered a major
touchstone of organic chemistry. At the same time, it is an
activity that is rarely put to the test of verifiability common to
many scientific endeavors. The difficulties we and others have
encountered in repeating either Simon’s enantioselective aldol
for the preparation of 2, or the Mitsunobu macrolactonization
of epi-4, well illustrate the subjective elements of complex
J. Org. Chem. Vol. 73, No. 23, 2008 9357

Wen et al.
H), 0.00 (s, 9 H); 13C NMR (100 MHz, CDCl3) δ 172.6, 145.1,
132.2, 130.3, 129.7, 128.0, 126.7, 68.7, 66.7, 63.2, 41.7, 31.6, 31.5,
17.5, -1.4. MS m/z 542 (M + Na+); HRMS m/z calcd for [M +
Na]+ C31H38NaO3SSi 541.2203, found 541.2201.
pound 9 (6.98 g, 91%) as a white solid: [R]25 -29.7 (c, 0.83,
D
CHCl3); mp 80-83 °C; IR 3307 (brs), 1647 cm-1
;
1H NMR
(300 MHz, CDCl3) δ 7.72 (m, 2 H), 7.54 (m, 2 H), 7.28-7.42
(m, 7 H), 7.14-7.26 (m, 12 H), 7.00 (d, 1 H, J ) 8.7 Hz), 6.80
(d, 1 H, J ) 7.7 Hz), 5.14 (d, 1 H, J ) 7.2 Hz), 4.25-4.40 (m,
5 H), 4.14-4.24 (m, 3 H), 3.80 (q, 1 H, J ) 6.9 Hz), 2.62 (d,
2 H, J ) 6.9 Hz), 2.10 (m, 1 H), 1.04 (d, 3 H, J ) 6.2 Hz), 0.92
(m, 2 H), 0.79 (t, 6 H, J ) 6.9 Hz), 0.0 (s, 9 H); 13C NMR (75
MHz, CDCl3) δ 171.6, 171.2, 170.7, 144.4, 143.9, 143.7, 141.4,
129.6, 128.2, 127.8, 127.2, 127.1, 125.1, 120.0, 67.5, 67.3, 66.2,
63.8, 57.5, 56.8, 54.3, 47.2, 34.1, 30.8, 19.1, 18.1, 17.7, 17.6,
-1.5; MS m/z 909 (M + Na+); HRMS m/z calcd for [M + Na]+
C51H59N3NaO7SSi 908.3735, found 908.3726.
(S)-2-(9H-Fluoren-9-ylmethoxycarbonylamino)-3-methyl-butyr-
ic Acid 2-Trimethylsilanyl-ethyl Ester (7). To a stirred solution
of Fmoc-L-Val-OH (3.4 g, 10 mmol) and trimethylsilylethanol
(2.0 mL, 14 mmol) in anhydrous dichloromethane (20 mL)
cooled in an ice bath was added a solution of DCC (3.1 g, 15
mmol) and DMAP (122 mg, 1 mmol) in anhydrous dichlo-
romethane (30 mL). After stirring for 2 h, the solution was
filtered, concentrated and purified by flash column chromatog-
raphy (eluent: EtOAc/Hexane ) 1/10) to give 7 (4.8 g, 100%)
as a colorless oil: [R]25 -2.3 (c, 0.76, CHCl3); IR 2957, 1715
(S)-2-{(2S,3R)-2-[(S)-2-(9H-Fluoren-9-ylmethoxycarbonylamino)-
3-tritylsulfanyl-propionylamino]-3-methanesulfonyloxy-butyry-
lamino}-3-methyl-butyric Acid 2-Trimethylsilanyl-ethyl Ester (10).
To a stirred solution of compound 9 (5.0 g, 5.64 mmol) in
anhydrous dichloromethane (50 mL) cooled in an ice bath was
added DMAP (40 mg), Et3N (1.5 mL, 10.75 mmol) and MsCl
(0.64 mL, 7.72 mmol). After stirring for 30 min, sat aq NaHCO3
(15 mL) was added followed by dichloromethane (20 mL). The
phases were separated, and the aqueous phase extracted with
dichloromethane (30 mL × 3). The combined organic phases
were washed with sat aq NH4Cl (25 mL) and brine (25 mL)
separately, dried (Na2SO4), and concentrated to give crude 10
(5.48 g, quantitative) as a light yellow solid: [R]25 -12.3 (c,
D
cm-1;1H NMR (300 MHz, CDCl3) δ 7.78 (d, 2 H, J ) 7.50
Hz), 7.62 (d, 2H, J ) 7.14 Hz,), 7.29-7.47 (m, 4H), 5.33 (d,
1H, J ) 8.78 Hz,) 4.35-4.49 (m, 2H), 4.21-4.35 (m, 3H),
2.11-2.28 (m, 1H), 1.02-1.09 (m, 2H), 1.06 (d, 1H), 1.00 (d,
3H, J ) 7.41 Hz), 0.94 (d, 3H, J ) 6.9 Hz,), 0.07 (s, 9H); 13C
NMR (75 MHz, CDCl3) 172.3, 156.3, 144.0, 141.5, 127.8, 127.2,
125.2, 120.1, 67.2, 63.8, 59.3, 47.4, 31.5, 19.1, 17.7, 17.7, -1.4;
MS m/z 463 (M + Na+), 903 (2 M + Na+).
(S)-2-[(2S,3R)-2-(9H-Fluoren-9-ylmethoxycarbonylamino)-3-hy-
droxy-butyrylamino]-3-methyl-butyric Acid 2-Trimethylsilanyl-
ethyl Ester (8). At room temperature, to a stirred solution of 7
(4.4 g, 10 mmol) in dichloromethane (50 mL) was added
diethylamine (10 mL). After stirring for 8 h, the solution was
concentrated, and the residue purified by flash chromatography
0.58, CHCl3); mp 75-80 °C; IR 1651, 1175 cm-D1
;
1H NMR
(300 MHz, CDCl3) δ 7.35 (dd, 2 H, J ) 4.7, 7.3 Hz), 7.55 (m,
2 H), 7.33-7.46 (m, 8 H), 7.16-7.32 (m, 11 H), 6.98 (d, 1 H,
J ) 8.0 Hz), 6.54 (d, 1 H, J ) 8.4 Hz), 5.28 (m, 1 H), 5.03 (d,
1 H, J ) 6.9 Hz), 4.62 (dd, 1 H, J ) 3.3, 8.4 Hz), 4.30-4.42
(m, 3 H), 4.08-4.20 (m, 3 H), 3.48 (m, 1 H), 3.00 (s, 3 H),
2.68 (m, 2 H), 2.14 (m, 1 H), 1.37 (d, 3 H, J ) 6.2 Hz), 0.95
(m, 2 H), 0.90 (d, 3 H, J ) 3.6 Hz), 0.88 (d, 3 H, J ) 3.6 Hz),
0.00 (s, 9 H); 13C NMR (75 MHz, CDCl3) δ 171.2, 171.0, 167.8,
144.3, 144.0, 143.7, 141.5, 141.4, 129.7, 128.3, 127.9, 127.2,
125.2, 125.1, 120.1, 67.7, 67.5, 63.7, 58.1, 56.6, 54.7, 47.2, 38.2,
33.3, 31.0, 19.2, 18.1, 18.0, 17.6, -1.4; MS m/z 965 (M + H+),
982 (M + NH4+), 987 (M + Na+); HRMS m/z calcd for [M +
Na]+ C52H61N3NaO9S2Si 986.3511, found 986.3482.
(eluent: EtOAc/Hexane ) 1/10-1/2, then MeOH/CH2Cl2
)
1/20-1/10) to give the free amine (2.19 g) as a colorless oil.
To a stirred solution of Fmoc-L-Thr-OH (3.4 g, 10 mmol) and
the above amine (2.19 g) in anhydrous dichloromethane (50 mL)
cooled in an ice-bath was added PyBOP (5.7 g, 11 mmol) and
N,N-diisopropylethylamine (2.1 mL, 12 mmol). After stirring
for another 0.5 h, the solution was warmed to room temperature,
stirred for 1.5 h, followed by addition of sat aq NH4Cl (20 mL).
The phases were separated, and the aqueous phase extracted with
dichloromethane (25 mL × 3). The combined organic phases
were washed with brine (20 mL), dried (Na2SO4), filtered and
concentrated. The residue was purified by flash chromatography
(eluent: EtOAc/Hexane ) 1/5-1/2) to give 8 (4.8 g, 88%) as a
(S)-2-((Z)-2-{(S)-2-[(R)-2-(9H-Fluoren-9-ylmethoxycarbonylamino)-
3-methyl-butyrylamino]-3-tritylsulfanyl-propionylamino}-but-2-
enoylamino)-3-methyl-butyric Acid 2-Trimethylsilanyl-ethyl Ester
(11). To a stirred solution of compound 10 (5.40 g, 5.6 mmol)
in anhydrous dichloromethane (50 mL) cooled in an ice bath
was added DABCO (6.28 g, 56 mmol). After stirring for 30 min,
the solution was warmed to room temperature, and was stirred
for 1.5 h before diethylamine (2.9 mL) was added. After stirring
for another 3.5 h, the solution was concentrated, and purified
by flash chromatography (eluent: EtOAc/Hexane ) 1/10-1/2,
then MeOH/dichloromethane ) 1/40-1/20) to give the amine
colorless oil: [R]25 -35 (c, 0.56, CHCl3); IR 3372 (br), 1731
D
cm-1; H NMR (300 MHz, CDCl3) δ 7.71 (d, 2 H, J ) 7.3 Hz),
1
7.54 (d, 2 H, J ) 7.6 Hz), 7.35 (t, 2 H, J ) 7.2 Hz), 7.26 (t, 2
H, J ) 7.2 Hz), 6.93 (d, 1 H, J ) 8.1 Hz), 5.81 (d, 1 H, J ) 7.7
Hz), 4.32-4.48 (m, 4 H), 4.15-4.25 (m, 4 H), 2.15 (m, 1 H),
1.14 (d, 3 H, J ) 6.2 Hz), 0.96 (t, 2 H, J ) 8.4 Hz), 0.88 (d, 3
H, J ) 6.9 Hz), 0.84 (d, 3 H, J ) 6.9 Hz), 0.00 (s, 9 H); 13C
NMR (75 MHz, CDCl3) δ 171.8, 171.1, 156.9, 143.8, 141.4,
127.8, 127.2, 125.1, 120.1, 67.4, 67.0, 63.9, 58.3, 57.5, 47.2,
30.9, 191.1, 18.0, 17.7, 17.6, -1.5; MS m/z 542 (M + H+), 564
(M + Na+).
(3.24 g, 90%) as a colorless oil: [R]25 ) 1.8 (c, 0.70, CHCl3).
D
(S)-2-{(2S,3R)-2-[(S)-2-(9H-Fluoren-9-ylmethoxycarbonylamino)-
3-tritylsulfanyl-propionylamino]-3-hydroxy-butyrylamino}-3-meth-
yl-butyric Acid 2-Trimethylsilanyl-ethyl Ester (9). At room tem-
perature, to a stirred solution of compound 8 (4.69 g, 8.70 mmol)
in anhydrous acetonitrile (45 mL) was added diethylamine (4.5
mL). After stirring for 4 h, the solution was concentrated, and
the residue dissolved in dichloromethane and concentrated again.
The procedure was repeated once, and the residue dried by high
vacuum to give the crude amine (4.04 g) as an oil. To a
suspension of Fmoc-D-Cys(Tr)-OH (5.35 g, 9.13 mmol) and the
above amine in anhydrous dichloromethane (50 mL) and
acetonitrile (5 mL) cooled in an ice-bath was added PyBOP (4.98
g, 9.57 mmol) and N,N-diisopropylethylamine (1.91 mL, 10.96
mmol). The solution was warmed to room temperature slowly
and stirred overnight. The reaction mixture was worked up
following the general procedure, and purified by flash chroma-
tography (eluent: EtOAc/Hexane ) 1/10-1/2.5) to give com-
To a stirred solution of Fmoc-D-Val-OH (1.7 g, 5.0 mmol) and
the above amine (2.95 g, 4.56 mmol) cooled in an ice bath was
added PyBOP (2.85 g, 5.48 mmol) and N,N-diisopropylethy-
lamine (1.20 mL, 6.89 mmol). After stirring for 70 min, the
reaction mixture was worked up following the general procedure.
The residue was purified by flash chromatography (eluent:
EtOAc/Hexane ) 1/4-1/1.5) to give 11 (4.02 g, 83%) as a white
25
solid: [R]D 7.0 (c, 0.78, CHCl3); mp 97-100 °C; IR 3289,
1
1643, 1495 cm-1; H NMR (300 MHz, CDCl3) δ 7.73 (d, 2 H,
J ) 7.3 Hz), 7.50 (d, 2 H, J ) 7.7 Hz), 7.31-7.42 (m, 9 H),
7.10-7.26 (m, 11 H), 6.60-6.71 (m, 2 H), 6.38 (brs, 1 H), 5.35
(brs, 1 H), 4.46 (dd, 1 H, J1 ) 5.5, 8.4 Hz), 4.34 (dd, 1 H, J )
7.2, 10.5 Hz), 4.18-4.32 (m, 1 H), 4.06-4.17 (m, 3 H),
3.90-4.05 (m, 2 H), 2.82 (m, 1 H), 2.62 (m, 1 H), 2.08 (m, 2
H), 1.66 (d, 3 H, J ) 6.9 Hz), 0.82-0.98 (m, 14 H), 0.00 (s, 9
H); 13C NMR (75 MHz, CDCl3) δ 172.1, 171.8, 168.7, 164.1,
156.9, 144.3, 143.9, 143.7, 141.5, 131.5, 129.6, 129.0, 128.2,
9358 J. Org. Chem. Vol. 73, No. 23, 2008

Macrolactamization Versus Macrolactonization
127.9, 127.2, 127.1, 125.1, 120.1, 67.6, 67.3, 63.4, 60.6, 58.0,
53.2, 47.3, 32.9, 31.3, 30.8, 19.4, 19.2, 18.3, 17.8, 17.6, 13.4,
-1.4; MS ESI m/z 968 (M + H+), 990 (M + Na+).
(eluent: EtOAc/Hexane ) 1/10-1/5) to give 14 (522 mg, 85%)
as a colorless oil: [R]25 15 (c, 0.69, CHCl3); IR 1712, 1219
D
cm-1; H NMR (400 MHz, CDCl3) δ 7.74 (d, 2 H, J ) 7.5 Hz),
1
(S)-2-((Z)-2-{(S)-2-[(R)-2-(9H-Fluoren-9-ylmethoxycarbonylamino)-
3-methyl-butyrylamino]-3-tritylsulfanyl-propionylamino}-but-2-
enoylamino)-3-methyl-butyric Acid (12). At room temperature, to
a stirred solution of 11 (1.5 g, 1.55 mmol) in anhydrous THF (8
mL) was added 1 M TBAF in THF (4.65 mL). After stirring for
7.5 h, the solution was cooled in an ice bath, followed by the
addition of sat aq NaHCO3 (2 mL), Na2CO3 (164 mg, 1.55 mmol)
and Fmoc-Cl (481 mg, 1.86 mmol). After stirring for 2 h, the
solution was acidified to pH 4-5 with 1 M aq HCl. The phases
were separated, and the aqueous phase extracted with EtOAc
(20 mL × 5). The combined organic phases were dried (Na2SO4),
filtered and concentrated. The residue was purified by flash
chromatography (EtOAc/Hexane ) 1/5-1/1.5, then MeOH/
dichloromethane ) 1/20-1/10) to give 12 (0.94 g, 70%) as a
7.57 (m, 2 H), 7.20-7.38 (m, 19 H), 5.58-5.70 (m, 2 H), 5.34
(dd, 1 H, J ) 7.5, 15.5 Hz), 5.25 (d, 1 H, J ) 9.0 Hz), 4.30-4.42
(m, 2 H), 4.16 - 4.29 (m, 2 H), 4.05-4.15 (m, 2 H), 2.64 (dd,
1 H, J ) 8.0, 16.0 Hz), 2.53 (dd, 1 H, J ) 5.5, 12.0 Hz), 2.08
-2.20 (m, 3 H), 2.01 (m, 2 H), 0.90-0.98 (m, 2 H), 0.89 (d, 3
H, J ) 6.5 Hz), 0.78 (d, 3 H, J ) 7.0 Hz), 0.00 (s, 9 H); 13C
NMR (100 MHz, CDCl3) δ 171.0, 169.8, 156.3, 145.0, 144.8,
144.1, 144.0, 141.5, 134.2, 129.7, 128.0, 127.9, 127.2, 126.8,
125.3, 120.1, 72.0, 67.2, 66.8, 63.3, 58.9, 47.4, 39.8, 31.6, 31.5,
31.3, 19.1, 17.5, 17.5, -1.4; MS m/z 858 (M + NH+), 863 (M
+ Na+); HRMS m/z calcd for [M + Na]+ C51H57NNaO6SSi
862.3568, found 862.3546.
(2S,3R)-2-[(S)-2-(9H-Fluoren-9-ylmethoxycarbonylamino)-3-tri-
tylsulfanyl-propionylamino]-3-hydroxy-butyric Acid Methyl Ester
(15). To a stirred solution of Fmoc-D-Cys(Tr)-OH (4.00 g, 6.83
mmol) in anhydrous dichloromethane (50 mL) cooled in an ice
bath was added PyBOP (4.00 g, 7.68 mmol), N,N-diisopropyl-
ethylamine (3.00 mL, 17.22 mmol), and then L-Thr-OMe.HCl
(1.27 g, 7.49 mmol). After stirring for 1.5 h, the reaction mixture
was worked up following the general procedure. The residue
was purified by flash chromatography (eluent: EtOAc/Hexane
white solid: [R]25 8.0 (c, 0.63, CHCl3); mp 118-120 °C; IR:
D
3277, 1651, 1504 cm-1
;
1H NMR (300 MHz, CDCl3) δ 7.90
(brs, 1 H), 7.70 (d, 2 H, J ) 7.7 Hz), 7.45 (d, 2 H, J ) 7.7 Hz),
7.05-7.36 (m, 20 H), 6.80-6.90 (m, 2 H), 6.67 (m, 1 H), 5.43
(d, 1 H, J ) 6.6 Hz), 4.28-4.40 (m, 2 H, 4.20 (m, 1 H),
4.00-4.12 (m, 2 H), 3.89 (brs, 1 H), 2.75 (m, 1 H), 2.60 (m, 1
H), 2.00-2.12 (m, 2 H), 1.58 (d, 3 H, J ) 6.9 Hz), 0.75-0.92
(m, 12 H); 13C NMR (75 MHz, CDCl3) δ 173.7, 172.0, 169.3,
164.7, 156.9, 144.2, 143.8, 143.5, 141.3, 133.1, 129.4, 128.1,
127.8, 127.1, 127.0, 125.0, 120.0, 67.4, 60.6, 58.9, 53.1, 47.1,
32.8, 30.8, 30.3, 19.1, 18.3, 13.5; MS m/z 866
(M-H+); HRMS m/z calcd for [M + Na]+ C51H54N4NaO7S
889.3605, found 889.3613.
) 1/2-2/1) to give 15 (4.62 g, 97%) as a white solid: [R]25
D
0.6 (c, 0.75, CHCl3); mp 90-95 °C; IR 3356 (brs), 1727, 1667,
1216 cm-1; 1H NMR (400 MHz, CDCl3) δ 7.74 (t, 2 H, J ) 7.2
Hz), 7.56 (d, 2 H, J ) 7.5 Hz), 7.36-7.48 (m, 8 H), 7.16-7.30
(m, 11 H), 6.56 (d, 1 H, J ) 9.0 Hz), 5.08 (d, 1 H, J ) 7.5 Hz),
4.49 (dd, 1 H, J ) 2.5, 8.5 Hz), 4.37 (d, 2 H, J ) 7.0 Hz),
4.18-4.27 (m, 2 H), 3.88 (q, 1 H, J ) 6.5 Hz), 3.69 (s, 3 H),
2.67 (d, 2 H, J ) 6.5 Hz), 1.14 (d, 3 H, J ) 6.5 Hz); 13C NMR
(100 MHz, CDCl3) δ 170.9, 170.5, 144.4, 143.8, 143.7, 141.4,
129.6, 128.2, 127.8, 127.2, 127.0, 125.1, 120.1, 68.2, 67.5, 67.2,
57.3, 54.2, 52.7, 47.2, 34.0, 20.1; MS m/z 723 (M + Na+);
HRMS m/z calcd for [M + Na]+ C42H40N2NaO6S 723.2499,
found 723.2487.
(E)-(S)-3-((S)-2-{(Z)-2-[(S)-2-((R)-2-(9H-Fluoren-9-ylmethoxycar-
bonylamino)-3-methyl-butyrylamino)-3-tritylsulfanyl-propiony-
lamino]-but-2-enoylamino}-3-methyl-butyryloxy)-7-[(Z)-2-eth-(E)-
ylidene-1,1-diphenyl-hexa-3,5-dienylsulfanyl]-hept-4-enoic Acid
2-Trimethylsilanyl-ethyl Ester (13). To a stirred solution of 12
(100 mg, 0.12 mmol) and 6 (74 mg, 0.14 mmol) in anhydrous
dichloromethane (4 mL) cooled in an ice bath was added MSNT
(90 mg, 0.30 mmol) and 1-methylimidazole (28 uL, 35 mmol).
After stirring for 5 h, the solution was warmed to room
temperature and stirred for 100 min, followed by addition of
sat aq NH4Cl (5 mL). The phases were separated, and the
aqueous phase extracted with dichloromethane (10 mL × 3).
The combined organic phases were washed with brine (5 mL),
dried (Na2SO4), filtered, and concentrated. The residue was
purified by flash chromatography (eluent: EtOAc/Hexane ) 1/5
- 1/1.5) to give recovered 6 (30.6 mg, 42%) and 13 (53.6 mg,
(2S,3R)-2-{(S)-2-[(R)-2-(9H-Fluoren-9-ylmethoxycarbonylamino)-
3-methyl-butyrylamino]-3-tritylsulfanyl-propionylamino}-3-hydroxy-
butyric Acid Methyl Ester (16). At room temperature, to a stirred
solution of 15 (4.62 g, 6.59 mmol) in anhydrous dichloromethane
(32 mL) was added piperidine (3.2 mL, 32.4 mmol). After
stirring for 70 min, the solution was concentrated, and the residue
purified by flash chromatography (eluent: EtOAc/Hexane )
1/4-1/2, then MeOH/dichloromethane ) 1/10) to give an amine
(2.95 g) which was used in the next step directly. To a stirred
solution of the above amine and Fmoc-D-Val-OH (2.23 g, 6.59
mmol) in anhydrous dichloromethane (50 mL) cooled in an ice
bath was added PyBOP (3.80 g, 7.30 mmol) and N,N-diisopro-
pylethylamine (1.72 mL, 9.87 mmol). After stirring overnight
at room temperature, the reaction mixture was worked up
following the general procedure. The residue was purified by
flash chromatography (eluent: EtOAc/Hexane ) 1/2-2/1) to give
34%) as a white solid: [R]25 -0.97 (c, 0.52, CHCl3); mp
D
95-100 °C; IR 3306, 1731, 1644, 1489 cm-1; H NMR (400
1
MHz, CDCl3) δ 7.74 (d, 2 H, J ) 7.5 Hz), 7.52 (d, 2 H, J ) 7.1
Hz), 7.12-7.40 (m, 35 H), 6.62 (m, 2 H), 6.29 (brs, 1 H),
5.52-5.65 (m, 2 H), 5.32 (dd, 1 H, J ) 7.1 Hz, 15.6 Hz), 5.24
(m, 1 H), 4.48 (dd, 1 H, J ) 5.5 Hz, 8.5 Hz), 4.40 (dd, 1 H, J
) 7.1, 10.8 Hz), 4.23 (m, 1 H), 4.00-4.13 (m, 4 H), 3.92 (m,
1 H), 2.89 (m, 1 H), 2.58-2.70 (m, 2 H), 2.50 (dd, 1 H, J )
6.2, 15.8 Hz), 1.94-2.18 (m, 6 H), 1.68 (d, 3 H, J ) 6.0 Hz),
0.75-0.95 (m, 14 H), 0.00 (s, 9 H); 13C NMR (100 MHz, CDCl3)
δ 171.7, 170.8, 169.8, 168.6, 163.9, 156.9, 145.0, 144.3, 141.4,
133.6, 129.7, 129.6, 128.3, 128.0, 127.9, 127.2, 127.1, 126.7,
125.1, 125.1, 120.1, 71.6, 67.6, 67.4, 66.7, 63.2, 60.6, 57.6, 53.1,
47.3, 39.8, 32.9, 31.5, 31.4, 31.2, 30.7, 19.4, 19.1, 18.0, 17.8,
17.4, 13.7, -1.4; MS m/z 1390 (M + Na+).
16 (4.30 g, 82%) as a white solid: [R]25 5.0 (c, 0.50, CHCl3);
D
mp 148-150 °C; IR 3357 (brs), 1727, 1667, 1216, 1032 cm-1
;
1H NMR (300 MHz, CDCl3:CD3OD ) 9:1) δ 7.76 (d, 2 H, J )
7.3 Hz), 7.60 (dd, 2 H, J ) 7.3, 11.7 Hz), 7.17-7.42 (m, 19 H),
4.22-4.45 (m, 5 H), 4.19 (t, 1 H, J ) 6.3 Hz), 3.90 (d, 1 H, J
) 5.5 Hz), 3.67 (s, 3 H), 2.67 (dd, 1 H, J ) 5.4, 12.6 Hz), 2.57
(dd, 1 H, J ) 8.4, 12.6 Hz), 2.05 (m, 1 H), 1.16 (d, 3 H, J ) 6.6
Hz), 0.95 (d, 3 H, J ) 6.9 Hz), 0.93 (d, 3 H, J ) 6.9 Hz); 13C
NMR (75 MHz, CDCl3:CD3OD ) 9:1) δ 171.9, 171.4, 170.6,
157.4, 144.3, 143.8, 143.5, 141.2, 129.4, 127.9, 127.7, 127.0,
126.7, 124.9, 119.9, 67.5, 67.1, 66.9, 61.0, 58.0, 52.3, 52.3, 47.0,
33.1, 30.5, 19.5, 19.0, 17.7; MS m/z 822 (M + Na+); HRMS
m/z calcd for [M + Na]+ C47H49N3NaO7S 822.3183, found
822.3167.
(E)-(S)-7-[(Z)-2-Eth-(E)-ylidene-1,1-diphenyl-hexa-3,5-dienylsul-
fanyl]-3-[(S)-2-(9H-fluoren-9-ylmethoxycarbonylamino)-3-methyl-
butyryloxy]-hept-4-enoic Acid 2-Trimethylsilanyl-ethyl Ester (14).
To a stirred solution of Fmoc-L-Val-OH (297 mg, 0.875 mmol)
and 2 (378 mg) in anhydrous dichloromethane (8 mL) cooled in
an ice bath was added DAMP (8 mg) and DCC (196 mg, 0.95
mmol). After stirring for 3 h, the solution was filtered and
concentrated. The residue was purified by flash chromatography
J. Org. Chem. Vol. 73, No. 23, 2008 9359

Wen et al.
(Z)-2-[(S)-2-((R)-2-Amino-3-methyl-butyrylamino)-3-tritylsulfa-
nyl-propionylamino]-but-2-enoic Acid Methyl Ester (17). To a
stirred solution of 16 (4.2 g, 5.25 mmol) in anhydrous dichlo-
romethane (45 mL) cooled in an ice bath was added DMAP (32
mg, 0.26 mmol), Et3N (1.2 mL, 8.60 mmol) and methanesulfonyl
chloride (0.53 mL, 6.82 mmol). After stirring for 1 h, the reaction
was quenched by adding sat aq NH4Cl (20 mL). The phases
were separated, and the aqueous phase extracted with dichlo-
romethane (25 mL × 3). The combined organic phases were
washed with sat aq NH4Cl (15 mL) and brine (15 mL), dried
(Na2SO4), filtered, and concentrated. The residue was purified
by flash chromatography (eluent: EtOAc/Hexane ) 1/2-2/1) to
then MeOH/dichloromethane ) 1/20-1/10) to give the free amine
(0.24 g, 89%) as a solid. To a stirred solution of 18 (342 mg, 0.445
mmol) and the above amine (230 mg, 0.372 mmol) in anhydrous
dichloromethane (6 mL) cooled in an ice bath was added PyBOP
(232 mg, 0.445 mmol) and N,N-diisopropylethylamine (0.1 mL,
0.574 mmol). After stirring for 1 h, the solution was warmed to
room temperature and stirred overnight. The reaction mixture was
worked up following the general procedure, and the residue purified
by flash chromatography (eluent: EtOAc/Hexane ) 1/3-1/1) to
give 13 (356 mg, 84%) as a white solid: [R]25 -1.12 (c, 0.4,
D
CHCl3), other data as above.
(3S,9S,12R,16S)-6-Eth-(Z)-ylidene-3,12-diisopropyl-16-((E)-4-tri-
tylsulfanyl-but-1-enyl)-9-tritylsulfanylmethyl-1-oxa-4,7,10,13-tetraaza-
cyclohexadecane-2,5,8,11,14-pentaone (19). At room temperature,
to a stirred solution of 13 (305 mg, 0.223 mmol) in anhydrous
dichloromethane (5 mL) was added piperidine (0.5 mL). After
stirring for 1 h, the solution was concentrated, and the residue
purified by flash chromatography (eluent: EtOAc/Hexane )
1/3-1/1, then MeOH/dichloromethane ) 1/20-1/10) to give the
amine (182 mg, 70%) as an oil. At room temperature, to a stirred
solution of the above amine (174 mg, 0.152 mmol) in anhydrous
THF (4 mL) was added 1 M TBAF (0.22 mL, 0.220 mmol).
After stirring for 1 h 40 min, the solution was concentrated,
and the residue dried by high vacuum. The crude amino acid
was dissolved in anhydrous dichloromethane (35 mL). To a
vigorously stirred solution of HATU (144 mg, 0.378 mmol) and
N,N-diisopropylethylamine (0.10 mL, 0.574 mmol) in anhydrous
dichloromethane (40 mL) was added the preprepared solution
above slowly over 4 h. After stirring overnight at room
temperature, sat aq NH4Cl (20 mL) was added and the phases
separated. The aqueous phase was extracted with dichlo-
romethane (20 mL × 3), and the combined organic phases were
washed with brine (20 mL), dried (Na2SO4), filtered, and
concentrated. The residue was purified by flash chromatography
(eluent: EtOAc/Hexane ) 1/4-1/1-2/1) to give compound 19
25
give the mesylate of 16 (3.81 g, 83%) as a white solid: [R]D
26.8 (c, 0.52, CHCl3); mp 158-160 °C. To a stirred solution of
the mesylate (3.50 g, 3.98 mmol) in anhydrous dichloromethane
(40 mL) cooled in an ice bath was added DABCO (4.47 g, 39.8
mmol). After stirring for 10 min, the solution was warmed to
room temperature and stirred for another 1.5 h. Piperidine (2.0
mL) was then added, and the reaction mixture stirred for 40
min. The solution volume was concentrated to one-third, and
the residue purified by flash chromatography (eluent: EtOAc/
Hexane ) 1/3-1/2, then MeOH/dichloromethane ) 1/20-1/
10) to give 17 as a white glass (1.99 g, 89%): [R]25 28.7 (c,
D
1
0.52, CHCl3); IR 3312 (brs), 1651, 1489, 1273 cm-1; H NMR
(400 MHz, CDCl3) δ 7.70 (d, 1 H, J ) 8.0 Hz), 7.58 (s, 1 H),
7.18-7.46 (m, 15 H), 6.74 (q, 1 H, J ) 7.0 Hz), 4.15 (q, 1 H,
J ) 7.0 Hz), 3.69 (s, 3 H), 3.22 (d, 1 H, J ) 4.0 Hz), 2.67 (d,
2 H, J ) 7.0 Hz), 2.24 (m, 1 H), 1.70 (d, 3 H, J ) 7.0 Hz), 0.96
(d, 3 H, J ) 7.0 Hz), 0.80 (d, 3 H, J ) 7.0 Hz); 13C NMR (100
MHz, CDCl3) δ 175.3, 168.6, 164.6, 144.6, 134.3, 129.7, 128.1,
126.9, 126.0, 67.2, 60.2, 52.4, 52.3, 33.0, 30.9, 19.7, 16.3, 14.7;
MS m/z 560 (M + H+), 582 (M + Na); HRMS m/z calcd for [M
+ Na]+ C32H37N3NaO4S 582.2397, found 582.2404.
(Z)-2-{(S)-2-[(R)-2-(9H-Fluoren-9-ylmethoxycarbonylamino)-3-
methyl-butyrylamino]-3-tritylsulfanyl-propionylamino}-but-2-eno-
ic Acid Methyl Ester (18). To a stirred solution of 17 (1.04 g,
1.85 mmol) in THF (12 mL) cooled in an ice bath was added an
aqueous solution of LiOH (109 mg, 4.55 mmol) in H2O (4 mL).
After stirring for 5 h, the solution was neutralized by adding 1
M aq HCl, followed by addition of NaHCO3 (311 mg, 3.7 mmol)
and Fmoc-Cl (622 mg, 2.40 mmol). After stirring for 1 h, the
solution was acidified to pH 3-4 with 1 M HCl, and diluted
with EtOAc (20 mL). The phases were separated, and the
aqueous phase extracted with EtOAc (15 mL × 3). The combined
organic phases were dried (Na2SO4), filtered, and concentrated.
The residue was purified by flash chromatography (eluent:
EtOAc/Hexane ) 1/1, then MeOH/dichloromethane ) 1/30-1/
1
as a white solid (120 mg, 77%): H NMR (400 MHz, CDCl3:
CD3OD ) 9:1) δ 7.20-7.42 (m, 30 H), 6.37 (q, 1 H, J ) 7.0
Hz), 5.78 (m, 1 H), 5.56 (m, 1 H), 5.36 (dd, 1 H, J ) 7.5, 15.6
Hz), 4.04 (d, 1 H, J ) 7.5 Hz), 3.83 (d, 1 H, J ) 9.0 Hz), 3.60
(m, 1 H), 2.84 (dd, 1 H, J ) 8.5, 13.1 Hz), 2.62-2.70 (m, 2 H),
2.45 (m, 1 H), 2.06-2.30 (m, 4 H), 2.00 (m, 2 H), 1.89 (m, 1
H), 0.95 (d, 6 H, J ) 6.5 Hz), 0.88 (d, 3 H, J ) 7.0 Hz), 0.81
(d, 3 H, J ) 6.5 Hz); 13C NMR (100 MHz, CDCl3:CD3OD )
9:1) δ 177.1, 173.0, 172.2, 170.7, 170.4, 144.8, 144.2, 132.7,
129.5, 129.4, 128.0, 127.9, 127.8, 126.9, 126.6, 118.9, 71.9, 67.4,
66.6, 60.4, 58.9, 53.8, 40.2, 32.0, 31.9, 31.1, 30.3, 29.6, 19.2,
19.0, 18.9, 18.4, 12.8; MS m/z: 1028 (M + H+), 1045 (M +
NH+), 1050 (M + Na+).
15) to give 18 (0.91 g, 64%) as a white solid: [R]25 33.8 (c,
D
1
0.61, CHCl3); IR 3295 (brs), 1678, 1651, 1237 cm-1; H NMR
(E)-(1S,4S,10S,21R)-7-Eth-(Z)-ylidene-4,21-diisopropyl-2-oxa-
12,13-dithia-5,8,20,23-tetraaza-bicyclo[8.7.6]tricos-16-ene-3,6,9,19,22-
pentaone (FK228). At room temperature, to a vigorously stirred
solution of iodine (284 mg, 1.12 mmol) in MeOH/dichlo-
romethane (1/10, 220 mL) was added a solution of 19 (115 mg,
0.112 mmol) in MeOH/dichloromethane (1/10, 110 mL) over
1 h. After stirring for 30 min, the solution was cooled down in
an ice bath, and 0.1 M Na2S2O3 (25 mL) added followed by
brine (25 mL). The phases were separated, and the aqueous phase
extracted with dichloromethane (50 mL × 3). The combined
organic phases were washed with brine (50 mL), dried (Na2SO4),
filtered, and concentrated. The residue was purified by flash
chromatography (eluent: 1/2-/10/1) to give FK228 (30 mg, 50%)
as a white solid: [R]25D 34 (c, 0.17, CHCl3); 1H NMR (400 MHz,
CDCl3:CD3OD)9:1) δ 6.20 (q, 1 H, J ) 7.0 Hz), 5.50-5.60
(m, 3 H), 4.55 (dd, 1 H, J ) 5.0, 11.0 Hz), 4.37 (d, 1 H, J ) 4.0
Hz), 3.83 (m, 1 H), 2.95-3.05 (m, 3 H), 2.80-2.90 (m, 1 H),
2.70 (m, 1 H), 2.45-2.58 (m, 3 H), 2.22 (m, 1 H), 2.05 (m, 1
H), 1.59 (d, 3 H, J ) 7.0 Hz), 0.98 (d, 3 H, J ) 7.0 Hz), 0.96
(d, 3 H, J ) 7.0 Hz), 0.88 (d, 3 H, J ) 7.0 Hz), 0.85 (d, 3 H,
J ) 7.0 Hz); 13C NMR (100 MHz, CDCl3:CD3OD)9:1) δ 172.4,
(400 MHz, CDCl3) δ 7.65 (d, 2 H, J ) 7.5 Hz), 7.45-7.55 (m,
3 H), 6.90-7.36 (m, 20 H), 6.78 (q, 1 H, J ) 7.0 Hz), 5.87 (d,
1 H, J ) 7.5 Hz), 4.32 (m, 1H), 4.20 (m, 1 H), 4.02-4.12 (m,
2 H), 3.91 (brs, 1 H), 2.59 (d, 2 H, J ) 6.0 Hz), 1.87 (m, 1 H),
1.64 (d, 3 H, J ) 7.0 Hz), 0.75 (brs, 6 H); 13C NMR (100 MHz,
CDCl3) δ 168.4, 166.7, 156.7, 144.4, 144.0, 143.9, 141.4, 136.1,
129.6, 128.2, 127.8, 127.2, 127.0, 125.4, 120.1, 67.4, 67.3, 60.3,
52.8, 52.7, 47.3, 32.9, 31.5, 19.2, 18.0, 15.0; MS m/z 791 (M +
Na+), 767 (M + H+); HRMS m/z calcd for [M + Na]+
C46H45N3NaO6S 790.2921, found 790.2918.
(E)-(S)-3-((S)-2-{(Z)-2-[(S)-2-((R)-2-(9H-Fluoren-9-ylmethoxycar-
bonylamino)-3-methyl-butyrylamino)-3-tritylsulfanyl-propiony-
lamino]-but-2-enoylamino}-3-methyl-butyryloxy)-7-[(Z)-2-eth-(E)-
ylidene-1,1-diphenyl-hexa-3,5-dienylsulfanyl]-hept-4-enoic Acid
2-Trimethylsilanyl-ethyl Ester (13) (Second Method). To a stirred
solution of 14 (362 mg, 0.43 mmol) in anhydrous dichloromethane
(5 mL) cooled in an ice bath was added piperidine (0.64 mL, 6.45
mmol). After stirring for 30 min, the solution was warmed to room
temperature, stirred for 1 h, and concentrated. The residue was
purified by flash chromatography (eluent: EtOAc/Hexane ) 1/5,
9360 J. Org. Chem. Vol. 73, No. 23, 2008

Macrolactamization Versus Macrolactonization
171.6, 169.4, 168.5, 165.8, 130.4, 129.8, 129.3, 70.3, 62.5, 58.3,
56.7, 37.9, 37.7, 34.7, 31.7, 30.1, 29.0, 19.2, 19.0, 18.4, 18.2,
13.2; MS m/z 541 (M + H+), 563 (M + Na+).
Shin-ya of the University of Tokyo for a generous gift of
naturally isolated FK228.
Supporting Information Available: General experimental
methods and NMR spectra for all compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. We are grateful to Cancer Research UK
for financial support, Dr. Alexander Cecil of Karus Therapeutics
for macrolactonization studies with 3 and 4 and Professor Kazuo
JO801866Z
J. Org. Chem. Vol. 73, No. 23, 2008 9361