Synthesis and structure-activity correlation of natural-product inspired cyclodepsipeptides stabilizing F-actin

Article abstract of DOI:10.1021/ja9095126

The fundamental role played by actin In the regulation of eukaryotic cell maintenance and motility renders it a primary target for small-molecule intervention. in this arena, a class of potent cytotoxic cyclodepsipeptide natural products has emerged over the last quarter-century to stimulate the fields of biology and chemistry with their unique actin-stabilizing properties and complex peptide-polyketide hybrid structures. Despite considerable research effort, a structural basis for the activity of these secondary metabolites remains elusive, not least for the lack of high-resolution structural data and a reliable synthetic route to diverse compound libraries. in response to this, an efficient solid-phase approach has been developed and successfully applied to the total synthesis of Jasplakinolide and chondramide C and diverse analogues. The key macrocylization step was realized using ruthenium-catalyzed ring-closing metathesis (RCM) that in the course of a library synthesis produced discernible trends in metathesis reactivity and E/Z-selectivity, After optimization, the RCM step could be operated under mild conditions, a result that promises to facilitate the synthesis of more extensive analogue libraries for structure-function studies. The growth inhibitory effects of the synthesized compounds were quantified and structure-activity correlations established which appear to be in good alignment with relevant biological data from natural products. in this way a number of potent unnatural and simplified analogues have been found. Furthermore, potentially important stereochemical and structural components of a common pharmacophore have been identified and rationalized using molecular modeling. These data will guide in-depth mode-of-action studies, especially into the relationship between the cytotoxicity of these compounds and their actin-perturbing properties, and should inform the future design of simplified and functionalized actln stabilizers as well.

Full text of DOI:10.1021/ja9095126

Published on Web 02/11/2010
Synthesis and Structure-Activity Correlation of
Natural-Product Inspired Cyclodepsipeptides Stabilizing
F-Actin
Rene´ Tannert, Lech-Gustav Milroy, Bernhard Ellinger, Tai-Shan Hu,
Hans-Dieter Arndt,* and Herbert Waldmann*
Max-Planck-Institut fu¨r molekulare Physiologie, Otto-Hahn-Strasse 11, 44227 Dortmund,
Germany, and Technische UniVersita¨t Dortmund Fakulta¨t Chemie, Otto-Hahn-Strasse 6,
44221 Dortmund, Germany
Received November 9, 2009; E-mail: hans-dieter.arndt@mpi-dortmund.mpg.de;
herbert.waldmann@mpi-dortmund.mpg.de
Abstract: The fundamental role played by actin in the regulation of eukaryotic cell maintenance and motility
renders it a primary target for small-molecule intervention. In this arena, a class of potent cytotoxic
cyclodepsipeptide natural products has emerged over the last quarter-century to stimulate the fields of
biology and chemistry with their unique actin-stabilizing properties and complex peptide-polyketide hybrid
structures. Despite considerable research effort, a structural basis for the activity of these secondary
metabolites remains elusive, not least for the lack of high-resolution structural data and a reliable synthetic
route to diverse compound libraries. In response to this, an efficient solid-phase approach has been
developed and successfully applied to the total synthesis of jasplakinolide and chondramide C and diverse
analogues. The key macrocylization step was realized using ruthenium-catalyzed ring-closing metathesis
(RCM) that in the course of a library synthesis produced discernible trends in metathesis reactivity and
E/Z-selectivity. After optimization, the RCM step could be operated under mild conditions, a result that
promises to facilitate the synthesis of more extensive analogue libraries for structure-function studies.
The growth inhibitory effects of the synthesized compounds were quantified and structure-activity
correlations established which appear to be in good alignment with relevant biological data from natural
products. In this way a number of potent unnatural and simplified analogues have been found. Furthermore,
potentially important stereochemical and structural components of a common pharmacophore have been
identified and rationalized using molecular modeling. These data will guide in-depth mode-of-action studies,
especially into the relationship between the cytotoxicity of these compounds and their actin-perturbing
properties, and should inform the future design of simplified and functionalized actin stabilizers as well.
transduction events,3 and represent promising leads for the
development of new cancer4 and malaria5 chemotherapies.
Introduction
In all eukaryotic cells, the tightly regulated assembly and
disassembly of actin fibers is fundamental to cellular growth
and dynamics, and crucial for cell motility and shape mainte-
nance, phagocytosis, and cytokinesis.1 As such, the potent and
selective perturbation of actin-mediated processes by means of
small molecules carries considerable therapeutic potential. In
nature, functionally distinct actin-modulator compounds have
been found that act according to their ability to either (a) disrupt
actin filament assembly, (b) induce actin polymerization and
stabilize actin fibers, or (c) inhibit actin-binding proteins that
control processes such as actin polymer nucleation.2 These
natural products have been important tools for the study of actin
structure and function,2c,d as well as actin-mediated signal
Moreover, actin-targeting small molecules serve as privileged
scaffolds that fundamentally address biomacromolecular interac-
tions, as protein-protein interactions dominate the assembly
and regulation of actin.6
The bicyclic heptapeptide phalloidin (Figure 1, 1)7 was the
first natural product shown to stabilize F-actin fibers.8 Fluores-
cent conjugates of 1 are routinely microinjected into cells for
the localization of actin fibers,9 but weak cell permeability limits
its use in the study of live cells and its pursuit as a drug lead.
More recently, however, a class of cyclodepsipeptides has
(3) Kustermans, G.; Piette, J.; Legrand-Poels, S. Biochem. Pharmacol.
2008, 76, 1310.
(4) (a) Jordan, M. A.; Wilson, L. Curr. Opin. Cell Biol. 1998, 10, 123.
(b) Rao, J. Y.; Li, N. Curr. Cancer Drug Targets 2004, 4, 345.
(5) Fotie, J.; Morgan, R. E. Mini ReV. Med. Chem. 2008, 8, 1088.
(6) For an example of parallel small-molecule and RNA interference
screening leading to the discovery of novel cytokinesis inhibitors, see:
Eggert, U. S.; Kiger, A. A.; Richter, C.; Perlman, Z. E.; Perrimon,
N.; Mitchison, T. J.; Field, C. M. PLoS Biol. 2004, 2, e379.
(7) Wieland, T. Science 1968, 159, 946.
(1) (a) Disanza, A.; Steffen, A.; Hertzog, M.; Frittoli, E.; Rottner, K.;
Scita, G. Cell. Mol. Life Sci. 2005, 62, 955. (b) Schmidt, A.; Hall,
M. N. Annu. ReV. Cell DeV. Biol. 1998, 14, 305.
(2) (a) Allingham, J. S.; Klenchin, V. A.; Rayment, I. Cell. Mol. Life Sci.
2006, 63, 2119. (b) Fenteany, G.; Zhu, S. T. Curr. Top. Med. Chem.
2003, 3, 593. (c) Peterson, J. R.; Mitchison, T. J. Chem. Biol. 2002,
9, 1275. (d) Spector, I.; Braet, F.; Shochet, N. R.; Bubb, M. R. Microsc.
Res. Tech. 1999, 47, 18.
(8) Lengsfeld, A. M.; Lo¨w, I.; Wieland, T.; Dancker, P.; Hasselbach, W.
Proc. Natl. Acad. Sci. U.S.A. 1974, 71, 2803.
9
10.1021/ja9095126  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 3063–3077 3063

A R T I C L E S
Tannert et al.
Figure 1. Phalloidin (1), jasplakinolide/jaspamide (2), chondramide C (3c), geodiamolide H (4), seragamide A (5), and doliculide (6).
emerged (Figure 1, 2-6) which competes with phalloidin. In
addition to their ability to potently stabilize actin fibers in a
manner comparable with 1, these structurally unique secondary
metabolites are also freely cell permeable to render them exciting
targets for drug development.
The 19-membered cyclodepsipeptide, jasplakinolide (Figure
1, 2),10 isolated from the marine sponge Jaspis splendans in
1986 was the first example and has consequently received the
most attention. Isolated independently as “jaspamide”,11 2 has
found widespread use in cell biology research as a cell-
permeable probe of actin dynamics.3 Jasplakinolide competes
in Vitro with 1 for binding to F-actin and stabilizes actin
filaments in ViVo,12 leading to actin lump formation and
polynucleation.13a Furthermore, 2 has shown potent cytotoxicity
toward various cancer cell lines,13 in addition to its fungicidal,10a,14
insecticidal,14a anthelmintic,10a,15 ichthyotoxic,16 herbicidal,14a
and antimalarial activity.17 More recently, the 18-membered
chondramide C (Figure 1, 3c), isolated from strains of the
myxobacterium Chondromyces crocatus,18 was found to potently
induce actin polymerization similar to 2,19 in addition to its
potent antiproliferative activity toward various mammalian
cancer cell lines.18,20 The geodiamolide (for example, geodi-
amolide H, 4, Figure 1)10c,21,22 and seragamide families (for
example, segeramide A, 5, Figure 1),23 and doliculide (Figure
1, 6),24,25 are further examples of this natural product class. The
structural similarity of 2-6 motivates the search for a unifying
structural basis for their activity,25 which would be of substantial
benefit in the design of new small-molecule probes and more
(16) Braekman, J. C.; Daloze, D.; Moussiaux, B.; Riccio, R. J. Nat. Prod.
1987, 50, 994.
(17) Mizuno, Y.; Makioka, A.; Kawazu, S.; Kano, S.; Kawai, S.; Akaki,
M.; Aikawa, M.; Ohtomo, H. Parasitol. Res. 2002, 88, 844.
(18) Kunze, B.; Jansen, R.; Sasse, F.; Ho¨fle, G.; Reichenbach, H. J. Antibiot.
1995, 48, 1262.
(9) (a) Wulf, E.; Deboben, B.; Bautz, F. A.; Faulstich, H.; Wieland, T.
Proc. Natl. Acad. Sci. U.S.A. 1979, 76, 4498. (b) Cooper, J. A. J. Cell
Biol. 1987, 105, 1473.
(19) (a) Waldmann, H.; Hu, T.-S.; Renner, S.; Menninger, S.; Tannert, R.;
Oda, T.; Arndt, H.-D. Angew. Chem. 2008, 120, 6573. Angew. Chem.,
Int. Ed. 2008, 47, 6473. (b) Eggert, U.; Diestel, R.; Sasse, F.; Jansen,
R.; Kunze, B.; Kalesse, M. Angew. Chem. 2008, 120, 6578. Angew.
Chem., Int. Ed. 2008, 47, 6478.
(10) Isolation of jasplakinolide: (a) Crews, P.; Manes, L. V.; Boehler, M.
Tetrahedron Lett. 1986, 27, 2797 See also. (b) Crews, P.; Farias, J. J.;
Emrich, R.; Keifer, P. A. J. Org. Chem. 1994, 59, 2932. (c) Talpir,
R.; Benayahu, Y.; Kashman, Y.; Pannell, L.; Schleyer, M. Tetrahedron
Lett. 1994, 35, 4453.
(20) Sasse, F.; Kunze, B.; Gronewold, T. M. A.; Reichenbach, H. J. Natl.
Cancer Inst. 1998, 90, 1559.
(21) (a) Coleman, J. E.; Dilip de Silva, E.; Kong, F.; Andersen, R. J.; Allen,
T. M. Tetrahedron 1995, 51, 10653. (b) Coleman, J. E.; Van Soest,
R.; Andersen, R. J. J. Nat. Prod. 1999, 62, 1137. (c) Chan, W. R.;
Tinto, W. F.; Manchard, P. S.; Todaro, L. J. J. Org. Chem. 1987, 52,
3091. (d) Tinto, W. F.; Lough, A. J.; McLean, S.; Reynolds, W. F.;
Yu, M.; Chan, W. R. Tetrahedron 1998, 54, 4451. (e) Dilip de Silva,
E.; Anderson, R. J.; Allen, T. M. Tetrahedron Lett. 1990, 31, 489.
(22) (a) Rangel, M.; Prado, M. P.; Konno, K.; Naoki, H.; Freitas, J. C.;
Machado-Santelli, G. M. Peptides 2006, 27, 2047. (b) Freitas, V. M.;
Rangel, M.; Bisson, L. F.; Jaeger, R. G.; Machado-Santelli, G. M.
J. Cell Physiol. 2008, 216, 583.
(11) Isolation of jaspamide: (a) Zabriskie, T. M.; Klocke, J. A.; Ireland,
C. M.; Marcus, A. H.; Molinski, T. F.; Faulkner, D. J.; Xu, C.; Clardy,
J. J. Am. Chem. Soc. 1986, 108, 3123 See also. (b) Chevallier, C.;
Richardson, A. D.; Edler, M. C.; Hamel, E.; Harper, M. K.; Ireland,
C. M Org. Lett. 2003, 5, 3737.
(12) Bubb, M. R.; Senderowicz, A. M. J.; Sausville, E. A.; Duncan,
K. L. K.; Korn, E. D. J. Biol. Chem. 1994, 269, 14869.
(13) (a) Senderowicz, A. M. J.; Kaur, G.; Sainz, E.; Laing, C.; Inman,
W. D.; Rodriguez, J.; Crews, P.; Malspeis, L.; Grever, M. R.; Sausville,
E. A.; Duncan, K. L. K. J. Natl. Cancer I. 1995, 87, 46. (b) Stingl, J.;
Andersen, R. J.; Emerman, J. T. Cancer Chemother. Pharmacol. 1992,
30, 401.
(23) Tanaka, C.; Tanaka, J.; Bolland, R. F.; Marriott, G.; Higa, T.
Tetrahedron 2006, 62, 3536.
(14) (a) Peng, J.; Shen, X.; El Sayed, K. A.; Dunbar, D. C.; Perry, T. L.;
Wilkins, S. P.; Hamann, M. T.; Bobzin, S.; Huesing, J.; Camp, R.;
Prinsen, M.; Krupa, D.; Wideman, M. A. J. Agric. Food Chem. 2003,
51, 2246. (b) Scott, V. R.; Boehme, R.; Matthews, T. R. Antimicrob.
Agents Chemother. 1988, 32, 1154.
(24) (a) Ishiwata, H.; Nemoto, T.; Ojika, M.; Yamada, K. J. Org. Chem.
1994, 59, 4710. (b) Ishiwata, H.; Sone, H.; Kigoshi, H.; Yamada, K.
J. Org. Chem. 1994, 59, 4712. (c) Ishiwata, H.; Sone, H.; Kigoshi,
H.; Yamada, K. Tetrahedron 1994, 50, 12853.
(25) Bai, R.; Covell, D. G.; Liu, C.; Ghosh, A. K.; Hamel, E. J. Biol. Chem.
2002, 277, 32165.
(15) Crews, P.; Hunter, L. M. Mar. Biotechnol. 1993, 1, 343.
9
3064 J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010

Structure-Function Studies of Cyclodepsipeptides
A R T I C L E S
Figure 2. Design of chondramide C (7a) and jasplakinolide/jaspamide analogues (7b).
efficacious and synthetically accessible actin-targeting drugs.26
Progress toward this desirable goal, however, is severly
hampered by the lack of high-resolution crystal structure
data19a,27 and, not least, by flexible synthetic routes to access
diverse structural analogues.
Scheme 1. General Retrosynthetic Planning
To facilitate investigations into the common molecular
parameters of these natural product lead scaffolds,28 we sought
to develop a new and efficient synthesis of 2 and 3c that in
extension could be suited to the synthesis of focused sets of
analogues. In particular, we intended on studying the relationship
between structure and biological activity, initially in terms of
macrocycle configuration, through the synthesis and biological
evaluation of stereochemically diverse analogues of 3c (Figure
2, 7a). The polyketide region had been surmised to ‘program’
the peptide domain into the biologically active conformation.29
We19a and Kalesse and co-workers19b have since realized
successful total syntheses of 3c, which permitted firm stereo-
chemical assignment of the macrocycle scaffold. Here, we
sought to scrutinize the crosstalk between polyketide stereo-
chemistry and bioactivity by preparing an extensive collection
of stereochemically diverse polyketide analogues of 3c in search
of other biologically active macrocycle configurations. Interest-
ingly, for stereochemically diverse actin-targeting analogues of
bistramide A, improved target potency could be shown while
our work was ongoing.30
Furthermore, we sought to quantify the significance of
structural features found recurrently in this class of cyclodep-
sipeptide by evaluating a series of strategically simplified
analogues inspired by 2. As we had found jasplakinolide to be
quite resistant to major structural simplification,31 in line with
other research groups,29,32,33 we adopted a systematic approach
(Figure 2, 7b). On the basis of our binding hypothesis for 3c,19a
we focused our attention on modifying the ꢀ-tyrosine and
tryptophan unitssboth believed to make important contributions
to binding to the actin trimer. A desirable outcome of this project
was the identification of potent simplified polyketide analogues
of 2. For this reason and for their potential ease of synthesis
(commercially available building blocks - Vide infra), analogues
lacking polyketide-based C6- and C8-methyl groups were to
be investigated (Figure 2, 7b). These single-site modifications
were expected to increase macrocycle flexibility through the loss
of 1,3-allylic and syn-pentane interactions29 and hence enable us
to test to what extent expected conformational preferences in this
region of the molecule impact on biological activity. Here we
provide a full account of the synthesis of these molecules along
with cytotoxicity and immunostaining studies. We report results
from a structure-function analysis of the analogues that is in line
with literature data previously reported for other natural products
(26) Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2004, 67, 1216.
(27) Recently, the ligand-free F-actin structure was refined to a resolution
of 3.3 Å radial and 5.6 Å equatorial: Oda, T.; Iwasa, M.; Aihara, T.;
Maeda, Y.; Narita, A. Nature 2009, 457, 441.
(28) Kumar, K.; Waldmann, H. Angew. Chem. 2009, 121, 3272. Angew.
Chem., Int. Ed. 2009, 48, 3224.
(29) (a) Marimganti, S.; Yasmeen, S.; Fischer, D.; Maier, M. E.
Chem.sEur. J. 2005, 11, 6687. (b) Marimganti, S.; Wieneke, R.;
Geyer, A.; Maier, M. E. Eur. J. Org. Chem. 2007, 2779.
(30) Wrona, I. E.; Lowe, J. T.; Turbyville, T. J.; Johnson, T. R.; Beignet,
J.; Beutler, J. A.; Panek, J. S. J. Org. Chem. 2009, 74, 1897.
(31) For the synthesis of simplified triazole-derived jasplakinolide ana-
logues, see: Hu, T.-S.; Tannert, R.; Arndt, H.-D.; Waldmann, H. Chem.
Commun. 2007, 3942; in biological tests, these compounds were
inactive.
(32) Kahn, M.; Nakanishi, H.; Su, T.; Lee, J. Y. H.; Johnson, M. E. Int. J.
Pept. Protein Res. 1991, 38, 324.
9
J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010 3065

A R T I C L E S
Tannert et al.
Scheme 2. Synthesis of Building Blocksa
a Reagents and conditions: (a) TIPSCl or TBSCl, imidazole, DMF, 0 °C f rt, 3 h; (b) (S)-N-(1-phenylethyl)benzylamine, nBuLi, THF, -78 °C, 20-30
min; (c) H2 (10 bar), Pd(OH)2/C, AcOH/EtOH, rt, 23 h; (d) NaHCO3, FmocOSu, 1,4-dioxane/H2O, 0 °C f rt, 3 h; (e) benzaldehyde, MeOH, rt, 2 h, then
NaCNBH3, 0 °C f rt, 15 h; (f) NaCNBH3, MeOH, rt, then (CH2O)n, 0 °C, 24 h; (g) H2 (8 bar), Pd(OH)2/C, AcOH/EtOH, rt, 20 h; (h) NaOH, NaHCO3,
FmocOSu, 1,4-dioxane/H2O, 0 °C f rt, 3 h; (i) MeOH, EDC ·HCl, (iPr)2NEt, DMAP, CH2Cl2, rt, 17 h; (j) NBS, CH2Cl2, rt, 10 min; (k) Me3SnOH (10
equiv), 1,2-dichloroethane, 80 °C, 20 h; (l) LDA, ZnBr2, THF, -78 °C, 30 min, then ꢀ-methallyl bromide, -78 °C f -15 °C, 14 h; (m) NaOH, THF/
MeOH/H2O, rt, 2.5 h.
from the same family and are able to align the observed trends
with a computationally derived binding mode.
only the configurations at the C6 and C7 positions in significant
doubt (Figure 2, 7a). Therefore, on the basis of substantial
literature precedent for the preparation of cyclopeptides using
ring closing metathesis (RCM),36 it was envisaged that the
trisubstiuted olefin (8, Scheme 1) could be formed in a similar
manner. In principle, the linear chain could then be simplified
to alcohol 12 and peptide acid 11, which in turn could be
disconnected into readily available building blocks 13-15 for
solid-phase synthesis.
Preparation of Metathesis Precursors. Silyl-protected Fmoc-
ꢀTyr-OH (Scheme 2, 13a,b) was synthesized by capitalizing
on the efficient procedure developed by Kocienski for the
preparation of H-ꢀTyr(OTBS)-OMe.34e First, silyl-protection of
the commercially available benzyl cumarate (16) was then
followed by stereoselective Michael addition of a chiral benzylic
amide to yield amine 17.37 Hydrogenolysis and subsequent
Fmoc-protection yielded acids 13a and b in a reliable 38-45%
yield over four steps. The synthesis of Fmoc-D-abrines 14a and
b (Scheme 2) was initiated by sequential reductive aminations
of D-tryptophan in the presence of the free carboxylic acid group
to give N-benzyl-abrine (18).38 Hydrogenation and Fmoc-
protection then yielded Fmoc-abrine (14a) in 62% over two
steps. For C2-bromination of the indole ring (14b), intermediate
Results and Discussion
Chemical Synthesis. Retrosynthetic Planning. Jasplakinolide
(2) has previously been prepared by other research groups using
macrolactonization-centered strategies (Scheme 1, 8 f 9 and
10).34 Concurrent to our efforts,19a chondramide C (3c) was
prepared using a similar approach.19b In our synthesis design,
we opted instead for a novel disconnection across C4-C5
(Scheme 1, 8) that was anticipated to benefit analogue studies
by enabling greater divergence in the polyketide region of the
macrocycle at the later stages of the synthesis. This strategy
was especially important considering the initial uncertainty
surrounding the absolute configuration of 3csespecially in the
polyketide region of the moleculesalthough biosynthetic con-
siderations35 and the striking similarity between 3c and 2 left
(33) (a) Terracciano, S.; Bruno, I.; Bifulco, G.; Copper, J. E.; Smith, C. D.;
Gomez-Paloma, L.; Riccio, R. J. Nat. Prod. 2004, 67, 1325. (b)
Terracciano, S.; Bruno, I.; Bifulco, G.; Avallone, E.; Smith, C. D.;
Gomez-Paloma, L.; Riccio, R. Bioorg. Med. Chem. 2005, 13, 5225.
(c) Terracciano, S.; Bruno, I.; D’Amico, E.; Bifulco, G.; Zampella,
A.; Sepe, V.; Smith, C. D.; Riccio, R. Bioorg. Med. Chem. 2008, 16,
6580.
(34) (a) Grieco, P. A.; Hon, Y. S.; Perez-Medrano, A. J. Am. Chem. Soc.
1988, 110, 1630. (b) Chu, K. S.; Negrete, G. R.; Konopelski, J. P. J.
Org. Chem. 1991, 56, 5196. (c) Hirai, Y.; Yokota, K.; Momose, T.
Heterocycles 1994, 39, 603. (d) Imaeda, T.; Hamada, Y.; Shioiri, T.
Tetrahedron Lett. 1994, 35, 591. (e) Ashworth, P.; Broadbelt, B.;
Jankowski, P.; Kocienski, P.; Pimm, A.; Bell, R. Synthesis 1995, 199.
(f) Ghosh, A. K.; Moon, D. K. Org. Lett. 2007, 9, 2425. See also: (g)
Schmidt, U.; Siegel, W.; Mundinger, K. Tetrahedron Lett. 1988, 29,
1269.
(36) (a) Clark, T. D.; Ghadiri, M. R. J. Am. Chem. Soc. 1995, 117, 12364.
(b) Miller, S. J.; Blackwell, H. E.; Grubbs, R. H. J. Am. Chem. Soc.
1996, 118, 9606. (c) Blackwell, H. E.; Grubbs, R. H. Angew. Chem.
1998, 110, 3469. Angew. Chem., Int. Ed. 1998, 37, 3281. (d)
Schafmeister, C. E.; Po, J.; Verdine, G. L. J. Am. Chem. Soc. 2000,
122, 5891. (e) Chapman, R. N.; Dimartino, G.; Arora, P. S. J. Am.
Chem. Soc. 2004, 126, 12252.
(37) (a) Davies, S. G.; Ichihara, O. Tetrahedron Asym. 1991, 2, 183. (b)
Davies, S. G.; Garrido, N. M.; Kruchinin, D.; Ichihara, O.; Kotchie,
L. J.; Price, P. D.; Mortimer, A. J. P.; Russell, A. J.; Smith, A. D.
Tetrahedron: Asymmetry 2006, 17, 1793.
(35) (a) Bai, R. L.; Covell, D. G.; Liu, C. F.; Ghosh, A. K.; Hamel, E.
J. Biol. Chem. 2002, 277, 32165. (b) Rachid, S.; Krug, D.; Weissman,
K. J.; Mu¨ller, R. J. Biol. Chem. 2007, 282, 21810.
9
3066 J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010

Structure-Function Studies of Cyclodepsipeptides
Scheme 3. Synthesis of Alkenesa
A R T I C L E S
a Reagents and conditions: (a) KOtBu, nBuLi, THF, -45 °C, 20 min, then (-)-B-methoxydiisopinocamphenylborane, -45 °C, 30 min, BF3 ·Et2O, CH3CHO,
-78 °C, 4.5 h, then aq NaOH/H2O2, reflux, 1 h, 72%; (b) as for (a) using instead (+)-B-methoxydiisopinocamphenylborane, 64%; (c) p-nitrobenzoic acid,
PPh3, DIAD, THF, 0 °C, 19 h; (d) NaOH, THF/MeOH/H2O, 0 °C, 15 h; (e) TBSOTf, 2,6-lutidine, CH2Cl2, -30 f -10 °C, 40 min (68% starting from 12a);
(f) O3, CH2Cl2, -78 °C, 5 min; (g) 5-hexenyltriphenylphosphonium iodide, nBuLi, THF, -78 °C, 1 h, then add aldehyde, -78 °C, 2 h; 34% (starting from
12a) or 49% (starting from 12c), three steps, single isomer (exclusively Z-olefin); (h) CSA, MeOH/THF, 0 °C f rt, 3 h (88%, 12e; 30%, four steps), or
TBAF, THF, 0 °C f rt, 24 h (73%, 12f; 36%, four steps); (i) LDA, LiCl, THF, -78 °C, then (R)-propylene oxide, 0 °C, 3.5 h, 86% d.e., (j) aq H2SO4 in
1,4-dioxane, rt, then reflux, 90 min; (k) DIBAL-H, CH2Cl2, -78 °C, 1 h; (l) methyltriphenylphosphonium iodide, nBuLi, THF or 5-hexenyltriphenylphosphonium
iodide, nBuLi, THF, -78 °C, 1 h, then add 21, 2 h, -78 °C, then 2 h, 0 °C (in the case of 12i, exclusively Z-olefin).
protection of the acid function was necessary for reliable yields
(Scheme 2, conditions i-k).34e The synthesis of the (R)- and
(S)-enantiomers of pentenoic acid 15 was achieved through the
asymmetric R-alkylation of the propanoylated derivative of
Seebach’s oxazolidinone (19)39 with methallyl bromide followed
by basic hydrolysis (Scheme 2). X-ray-crystallography unam-
biguously confirmed the configuration at the C2-position.40
For the synthesis of chondramide C (3c) and its stereochem-
ical analogues, Brown’s crotylboration methodology41 provided
us with convenient one-step access to three of the four C6-C7
homoallylic alcohol diastereomers 12a-c starting from either
cis- or trans-butene (Scheme 3). Isomer 12d was most conve-
niently prepared by Mitsunobu inversion at the C7-position of
12a followed by saponification (Scheme 3). Toward the
synthesis of 2 and simplified analogues, alcohols 12g-j were
synthesized using Myers’ pseudoephedrine methodology (Scheme
3).42 The lithium enolate derived from pseudoephedrine deriva-
tive 20 was alkylated with (R)-propylene oxide in the diaste-
reoselectively “matched” sense in good diastereomeric excess
(86% d.e.), followed by acid-mediated cleavage of the auxiliary
and reduction of the intermediate lactone to deliver lactol 21.
At this juncture, Wittig olefination using methyltriphenylphos-
phonium iodide (12g) and Mitsunobu inversion produced alcohol
12h in the correct configuration for the synthesis of 2. To bolster
our strategic options for RCM studies (Vide infra), the elaborate
dienes 12e,f were synthesized using a sequence of alcohol
protection, ozonolysis, and Wittig olefination steps (Scheme 3),
and 12j by first reacting 5-hexenyltriphenylphosphonium iodide
with lactol 21 to form 12i, followed by Mitsunobu inversion.
In each case, the Z-olefin was exclusively detected. The less
complex alcohols 12k and 12l were commercially available.
With the synthetic building blocks in hand, diverse peptide
acids were assembled using Fmoc-based solid-phase peptide
synthesis. Silyl-protected Fmoc-ꢀTyr-OH 13a,b or Fmoc-ꢀAla-
OH 13c were first attached to 2-chlorotrityl resin under basic
conditions (Scheme 4, 22a-c). Sequential basic deprotection
and carbodiimide-mediated couplings of Fmoc-Trp-OH 14a-c,
Fmoc-Ala-OH, and (S)- or (R)-pentenoic acid (15a or 15b)
followed to yield a variety of peptide acids 11a-g after resin
cleavage (see Supporting Information for structures).43 The
esterification of the acid was found to be non-trivial. Standard
Mitsunobu conditions44 were ineffective. Fortunately, Steglich
esterification proved to be highly effective and quite general
(Scheme 4) as evidenced by the successful coupling of alcohols
12a-f for chondramide C analogues (n ) 0) and 12 h-l for
jasplakinolide analogues (n ) 1) to afford structurally diverse
peptide dienes 23a-q in moderate to good yields.
Macrocycle Formation by RCM: Chondramide C Ana-
logues. The pivotal ring-closing metathesis (RCM) step was
studied next. For the synthesis of macrocycle-embedded
(43) On cleavage of (OTBS)-or (OTIPS)-ꢀ-tyrosine-containing peptidic
acids 11c-g from the polymer support (Scheme 4), varying quantities
of the phenolic OH unprotected form were additionally isolated
(Supporting Information). Further investigations suggested that the
peptide coupling conditions (carbodiimide in combination with ami-
noacid and benzotriazole) were favoring undesired silyl deprotection.
Fortunately, the desilylated peptides could be readily reprotected in
solution using standard conditions (Supporting Information).
(44) For a recent review see: But, T. Y. S.; Toy, P. H. Chem. Asian J.
2007, 2, 1340.
(38) (a) White, K. N.; Konopelski, J. P. Org. Lett. 2005, 7, 4111 See also.
(b) Ohfune, Y.; Kurokawa, N.; Higuchi, N.; Saito, M.; Hashimoto,
M.; Tanaka, T. Chem. Lett. 1984, 441.
(39) Hintermann, T.; Seebach, D. HelV. Chim. Acta 1998, 81, 2093.
(40) Tannert, R.; Schu¨rmann, M.; Preut, H.; Arndt, H.-D.; Waldmann, H.
Acta Crystallogr., Sect. E 2007, 63, O4381.
(41) Brown, H. C.; Bhat, K. S. J. Am. Chem. Soc. 1986, 108, 293.
(42) Myers, A. G.; McKinstry, L. J. Org. Chem. 1996, 61, 2428.
9
J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010 3067

A R T I C L E S
Tannert et al.
Scheme 4. Solid-Phase-Based Synthesis of Metathesis Precursorsa
a Reagents and conditions: (a) 13a-c (see Scheme 2 conditions a-d), 2-chlorotrityl chloride resin, (iPr)2NEt, CH2Cl2, rt, 1 h; (b) DBU, piperidine, NMP
(5 ×), rt; (c) 14a-c (see Scheme 2, conditions e-k), DIC, HOBt, NMP or DMF, rt, 2-3 h; (d) Fmoc-R-Ala-OH, HATU, HOAt, (iPr)2NEt, NMP, rt, 2-3
h; (e) S-15 or R-15 (see Scheme 2, conditions l, m), DIC or HATU, HOBt, (iPr)2NEt, NMP, rt, 2-3 h; (f) HFIP/CH2Cl2 or AcOH/TFE/CH2Cl2 (1 × 20 min,
1 × 10 min); (g) 12a-f for chondramide C analogues (n ) 0) or 12h-l for jasplakinolide analogues (n ) 1) - see Scheme 3, EDC ·HCl, (iPr)2NEt, DMAP,
CH2Cl2/DMF, rt, 14 h.
Scheme 5. Development of General Metathesis Conditions and Formation of Chondramide C (3c) and Analogues 28 and 29 (see Table 1)a
a Reagents and conditions: (a) conditions: see Table 1 conditions; (b) TBAF, THF, 30 min.
R-branched trisubstituted olefins by RCM, Grubbs’ second-
generation catalyst (Scheme 5, 25)45 has found frequent use,
typically in refluxing CH2Cl2,46 though the rates of conversion
and the E/Z-selectivities were largely substrate dependent. For
our purposes, treating diene 23a with 10 mol % of 25 in
degassed refluxing CH2Cl2 produced no conversion, even after
further catalyst additions and extended heating (Table 1, entry
1). Disappointingly, more forcing conditions (Table 1, entry 2),
and a switch to the Hoveyda-Grubbs second-generation or
Grela-type catalysts (Scheme 5, 26 and 27),47 did not improve
the outcome (Table 1, entries 3 and 4).48
To account for the steric hindrance of the C6-R-olefin methyl
group, we next attempted ring closure using relay metathesis
(RRCM).49 Additionally, it was speculated that purging the
reaction of ethylene with argon may improve catalyst reactiv-
ity.50 We were pleased to find that treating precursor 23b with
20 mol % of 25 in refluxing toluene with purging argon
(47) (a) Garber, S. B.; Kingsbury, J. S.; Gray, B. L.; Hoveyda, A. H. J. Am.
Chem. Soc. 2000, 122, 8168. (b) Grela, K.; Harutyunyan, S.;
Michrowska, A. Angew. Chem. 2002, 114, 4210. Angew. Chem., Int.
Ed. 2002, 41, 4038. (c) Review: Samojlowicz, C.; Bieniek, M.; Grela,
K. Chem. ReV. 2009, 109, 3708.
(45) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1,
953.
(46) (a) Alexander, M. D.; Fontaine, S. D.; La Claire, J. J.; DiPasquale,
A. G.; Rheingold, A. L.; Burkhart, M. D. Chem. Commun. 2006, 4602.
(b) Smith, A. B., III; Mesaros, E. F.; Meyer, E. A. J. Am. Chem. Soc.
2006, 128, 5292. (c) Feyen, F.; Jantsch, A.; Altmann, K.-H. Synlett
2007, 415. (d) Jin, J.; Chen, Y.; Li, Y.; Wu, J.; Dai, W.-M. Org. Lett.
2007, 9, 2585. For a recent review see: (e) Gradillas, A.; Pe´rez-Castells,
J. Angew. Chem. 2006, 118, 8264. Angew.Chem. Int. Ed., 2006, 45,
6086.
(48) Conditions for generating sterically hindered olefins were tested as
well: Rost, D.; Porta, M.; Gessler, S.; Blechert, S. Tetrahedron Lett.
2008, 49, 5968.
(49) For a recent review see: Wallace, D. Angew. Chem. 2005, 117, 1946.
Angew. Chem., Int. Ed., 2005, 44, 1912.
(50) Nosse, B.; Schall, A.; Jeong, W. B.; Reiser, O. AdV. Synth. Catal.
2005, 347, 1869.
9
3068 J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010

Structure-Function Studies of Cyclodepsipeptides
A R T I C L E S
Table 1. Synthesis of Chondramide C 3c, Silyl-Protected Macrocycles 28, and Chondramide C Analogues 29 (see Scheme 5)
a All solvents degassed prior to use. b 3 mM dilution. c 1 mM dilution and continuous argon purging. d No change after a further 20 mol % addition
and 5 h. e No change after a further 30 mol % addition and overnight heating. f No change after a further 50 mol % addition and overnight heating.
g Yield calculated over two steps including silyl-deprotection step (b). h 1:1.5 E/Z mixture after cyclization. i Yields determined after purification by
j
column chromatography. Inseparable 1:1.4 E/Z mixture; 1:1.1 after desilylation (R1 ) TIPS f H).
produced our first chondramide C analogues as a separable 1:1.5
mixture of E/Z isomers (E-29a and Z-29a) in yields of 14 and
23%, respectively, after deprotection of the silyl group (Table
1, entry 5). The same conditions were then successfully applied
to diene 23c, leading this time to a single isomer (Z-29b) in a
promising 33% yield over two steps (Table 1, entry 6). Despite
initial concerns raised by the non-reactivity of 23a, we were
fortunate to find that the newly developed metathesis conditions
were generally applicable to the other dienes in the series 23d-j
(Table 1, entries 7-13) with the exception of 23e (Table 1,
entry 8). In each successful case, the geometric outcome of ring
closure could conveniently be determined by NOESY or
ROESY analysis (Scheme 5). The negligible turnover of dienes
23a and 23e nonetheless serves as an important reminder of
the substrate dependency of the RCM methodology.51 Pleas-
ingly, among the macrocycle products, the (2S,4E,6R,7R)-
configured compound (Table 1, entry 9) was found to be
indentical to chondramide C (3c) by comparing compound
characterization and cell screening data with those measured
for material isolated from natural sources.19a
On closer inspection of the metathesis results (Table 1), the
E/Z-outcome appeared to be strongly influenced by the config-
uration of the polyketide, especially at the C2-position. First,
while (2R,6S,7R)- and (2S,6S,7R)-configured dienes (23b and
23g, respectively) both yielded 2:3 E/Z mixtures (Table 1, entries
5 and 10), the diene pair 23d and 23h instead produced opposite
olefin geometries (Table 1, compare entries 7 and 11). A similar
switch in selectivity was also observed in the case of dienes
23c and 23i (Table 1, compare entries 6 and 12). Finally,
whereas 23f reacted to form E-olefin 3c (after silyl-deprotection
of E-28d), diene 23e was unreactiVe (Table 1, compare entries
8 and 9). In each case, the stereochemistries of the diene
precursors under comparison differ only by inversion at the C2-
position. The observed selectivities are difficult to rationalize.
However, we found that the E/Z-ratios did not evidently change
during the course of the reaction, which suggests that double-
bond formation takes place under kinetic control and that the
geometric outcomes described in Table 1 result from distinct
ruthenocyclobutane intermediate conformations.52 The E/Z-
outcome is likely, therefore, to be independent of the metathesis
strategy employed (consider 23c and 23i, RRCM vs RCM). In
addition, minor structural changes to the phenolic silyl protecting
group (Scheme 5, R2) are not believed to influence E/Z-product
ratios (consider 23e and 23f, for example), although the effect
of remote functional groups on ring formation can be strong,53
while proximal groups such as R-olefinic substituents are known
to significantly influence the reactivity and selectivity of RCM
processes.54 Indeed, kinetically controlled RCM of densely
functionalized dienes in the context of natural product synthesis
has been previously reported.55
D-Amino acids are known to induce type-II′ ꢀ-turns,56 and
N-methylated amino acids lower the energy difference between
the cis- and trans-conformation of peptide bonds, leading to
(52) Grubbs, R. H., Ed. Handbook of Metathesis; Wiley-VCH: Weinheim,
2003; Vols. 1-3.
(53) Fu¨rstner, A.; Thiel, O. R.; Blanda, G. Org. Lett. 2000, 2, 3731.
(54) Ram´ırez-Ferna´ndez, J.; Collado, I. G.; Herna´ndez-Gala´n, R. Synlett
2008, 339.
(55) For a related example see: (a) Castoldi, D.; Caggiano, L.; Panigada,
L.; Sharon, O.; Costa, A. M.; Gennari, C. Angew. Chem. 2005, 117,
594. Angew. Chem., Int. Ed. 2005, 44, 588. (b) Castoldi, D.; Caggiano,
L.; Panigada, L.; Sharon, O.; Costa, A. M.; Gennari, C. Chem.sEur.
J. 2006, 12, 51.
(51) Diene 23f was treated once with catalyst 25 in refluxing CH2Cl2
without reaction (data not shown) to indicate that the non-reactivity
of diene 23a is indeed representative of all diene substrates.
(56) Heller, M.; Sukopp, M.; Tsomaia, N.; John, M.; Mierke, D. F.; Reif,
B.; Kessler, H. J. Am. Chem. Soc. 2006, 128, 13806.
9
J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010 3069

A R T I C L E S
Tannert et al.
Table 2. Synthesis of Jasplakinolide/Jaspamide 2, Silyl-Protected Macrocycles 30 and Jasplakinolide Analogues 31 (see Scheme 6)
a Calculated over two steps. b Yield determined after column chromatography and preparative HPLC purification. c Yield determined after column
chromatography purification. d 5 mol %, then heating for a further 5 min (10 min in total). e 6 × 5 mol % (every 5 min), then heating for a further 30
min.
Scheme 6. Formation of Jasplakinolide/Jaspamide 2 and Analogues 30 and 31 by Ring-Closing Metathesis (see Tables 2 and 3)a
a Reagents and conditions: (a) 25 (25-30 mol %, single-portion), PhMe (degassed, 1 mM), reflux, argon purging, 2 h; (b) TBAF, THF, 30 min see Tables
2 and 3.
Macrocycle Formation by RCM: Jasplakinolide Ana-
logues. The synthesis of jasplakinolide/jaspamide (2) and the
desbromo-analogue E-31b by RCM proved to be equally
challenging (Table 2). Nonetheless, using the previously
established metathesis conditions, cyclization was successful,
in each case yielding approximately 1:1 E/Z mixtures, which
were separable by chromatography (Table 2, entries 1 and 2).
The relay RCM approach was also successful (Table 2, entry
3) and slightly higher yielding, but did not produce any
significant change in E/Z-selectivity. Indeed, HPLC analysis of
the crude RCM and RRCM mixtures revealed a nearly identical
formation of E/Z isomers to suggest that the macrocycle-forming
cycloaddition-cycloreversion steps are rate-determining. In
stark contrast, analogues lacking a methyl group at the C6
position of the polyketide (Scheme 6, R3 ) H) were significantly
more straightforward to prepare. Under similar conditions,
dienes 23n-q were fast to react and high yielding and led almost
exclusively to the E-isomer (Table 2, entries 4-7). A more
rigorous control of the rate of conversion of substrates 23n and
23p showed the transformation to be complete within 10 min
using only 5 mol % of catalyst.
the loss of a potentially important hydrogen donor in the
backbone of peptide-based structures.57 In this context, the
N-Me-Trp-ꢀTyr- sequence of 2 has been surmised to adopt a
type-II ꢀ-turn conformation,32 a proposal that has led a number
of research groups to prepare macrocyclic analogues bearing
this sequence alone in the peptide region.32,58 In order to study
the influence of N-methylation on the macrocycle conformation,
and quantify its impact on cytotoxicity, we synthesized E-29h,
norchondramide C (Table 1, entry 13) by replacing D-abrine
with D-tryptophan in the course of the synthesis (Schemes 4
and 5). Compared with the formation of 3c (Table 1, entry 9),
the lack of N-methylation in diene 23j (Table 1, entry 13) did
not evidently alter the outcome of the metathesis reaction
(predominantly E). The same preference for the trans-amide
bond rotamer was observed according to NOESY experiments
(Scheme 5), although the reaction’s yield was somewhat lower.
(57) (a) Chatterjee, J.; Mierke, D.; Kessler, H. J. Am. Chem. Soc. 2006,
128, 15164. (b) Dechantsreiter, M. A.; Planker, E.; Matha¨, B.; Lohof,
E.; Ho¨lzemann, G.; Jonczyk, A.; Goodman, S. L.; Kessler, H. J. Med.
Chem. 1999, 42, 3033.
(58) Celanire, S.; Descamps-Francois, C.; Lesur, B.; Guillaumet, G.; Joseph,
B. Lett. Org. Chem. 2005, 2, 528.
9
3070 J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010

Structure-Function Studies of Cyclodepsipeptides
A R T I C L E S
Table 3. Optimization of Metathesis Conditions for
non-R-methyl-substituted olefins, milder conditions proved
effective: 5-10 mol % catalyst 25 in dry CH2Cl2 (1 mM)
heating at 40 °C for 3-15 h.
Non-R-branched Dienes 23n-q (see Table 2 and Scheme 6)
prod
(yield/%)b
entry
diene
cat.
mol %a
T/°C
t/h
conv./%
Biological Evaluation. Chondramide C Analogues. Recently,
a study on a series of novel natural jasplakinolide/jaspamide
analogues reported compound isolation, rigorous structure
elucidation by NMR analysis, and the determination of com-
pound cytotoxicity toward MCF-7 and HT-29 cancer cells.60
To allow comparison, we tested our synthetic analogues against
the same cell lines (Table 4). In general the (2R)-configured
analogues were significantly less potent than their corresponding
(2S)-isomers. The two (2R,4Z)-isomers showed some remaining
activity (Table 4, entries 6 and 7), while the (2R,4E)-configured
analogues were completely inactive (Table 4, entries 5 and 8).
By contrast, the (2S)-isomers (Table 4, entries 9-11) were only
10-100-fold less potent than synthetic 3c (Table 4, entry 2),
which itself was found to be equipotent with the natural product
(Table 4, entry 1) as final confirmation to a successful total
synthesis.19a The fact that Z-29f was also active within the same
range is an indication that changes in olefin geometry modulate
but do not extinguish biological activity. During work toward
their own total synthesis of 3c, Kalesse and co-workers observed
similar levels of cytotoxicity for a number of (2S,4E)-configured
isomers against different cancer cell lineages (Figure 3, E-29e,
E-29f, and E-29g), including one that was equally potent as
3c.19b Indeed, the (2S,4E)-stereochemical arrangement is a
structural hallmark of jaspamide/jasplakinolide, geodiamolide,
and seragamide families (Figure 1). Furthermore, compared with
2, synthetic (2R)-2 was reported to display different insect anti-
feedant properties61 while the natural analogue jaspamide J
(Figure 3, 32a)60 was 100-fold less cytotoxic. Methylation of
the tryptophan R-amino group could also be considered essential
for optimal potency on the basis of our result for norchondra-
mide C, E-29h (Table 4, entry 12), where an approximate 10-
fold loss in activity was observed. In support of this, the natural
analogue jaspamide M (Figure 3, 32b), also lacking the
N-methylation, was found to be one-order-of magnitude less
1
2
3
4
5
6
7
23n
23n
23n
23n
23o
23p
23q
25
25
24
26
25
25
25
5
10
10
10
5
23
40
40
40
40
40
40
24
3
24
24
15
3
30c
89d
<5c
65c
57d
91d
75d
E-30c (n.d.)
E-30c (65)
n.d.
E-30c (n.d.)
E-30d (30)
E-30e (75)
E-30c (59)
5
5
15
a Single portion addition. Reaction run in degassed CH2Cl2 at 1 mM
substrate concentration. b Preparative yield, n.d. ) not determined. c By
integration of crude HPLC trace. d By integration of crude 1H NMR
signals.
This significant increase in metathesis reactivity for substrates
lacking the methyl group at the C6 position of the polyketide
allowed us to realize milder conditions. We were gratified to
find, therefore, that treating 23n with 5 mol % of 25 in dry
CH2Cl2 at room temperature resulted in a 30% conversion to
E-30c within 24 h (Table 3, entry 1). This promising result could
be further optimized by using 10 mol % of catalyst and heating
at 40 °C for 3 h, leading to an 89% conversion and isolated
yield of 65% (Table 3, entry 2). When the results for 30b and
31b (Table 2, entries 2 and 3) and 30c (Table 3, entry 2) are
compared, it becomes apparent how the C6-methyl group
influences the reactivity and E/Z-selectivity of ring-closure.
Catalyst 2459 failed to catalyze ring closure (Table 3, entry 3),
while catalyst 26 managed a useful conversion (Table 3, entry
4) to add further scope to this reaction. Indeed, for our purposes,
these conditions were found to be quite general for analogous
substrates (Table 3, entries 5-7) to provide further complex
examples of macrocycle-embedded trisubstituted olefin forma-
tion using RCM under mild reaction conditions.46
In summary, for the formation of macrocycle-embedded
R-methyl olefins, forcing conditions were required: 25-30 mol
% of catalyst 25 in dry, degassed toluene (1 mM) heating under
reflux for 2 h with continuous argon purging. In the case of
Table 4. Cytotoxicity Data for Chondramide C 3c and Analogues 28 and 29
a Natural source. b Synthetic material. c Inseparable 1:1.1 E/Z mixture.
9
J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010 3071

A R T I C L E S
Tannert et al.
analogue, E-31b (Table 5, entry 2),62 recently isolated from
nature as jaspamide Q,63 was essentially equipotent with 2 to
suggest that the indole-based bromine atom is not required for
potent activity. This result might be considered general, as
halogenated and non-halogenated members of the chondramide
family share the same potent cytotoxic properties (A-D, Figure
4, 3a-d)18,20sdespite the recurrence of aromatic halogenation
in nearly all known natural jaspamide,60 geodiamolide,10c,21,22
seragamide23 family members as well as doliculide.24,25 In this
context, an interesting fine-tuning of compound potency and
selectivity has been demonstrated through variation of the
halogen group.22,24c
Two natural analogues of jaspamide lacking methyl substit-
uents in the polyketide region of the molecule (Figure 4, 32c
and 32d) were recently shown to be significantly inactive toward
MCF-7 cells (IC50 > 30 µM).60 This strong drop in activity was
attributed to a loss in actin-binding affinity and related to an
increase in ring flexibility through the simultaneous loss of 1,3-
allylic and/or syn-pentane interactions.29a On the contrary,
however, we were surprised to find the newly synthesized
analogue E-31c to be only 2-fold less potent and E-31d no
greater than 10-fold less active (entries 3 and 4, Table 5). Indeed,
early NMR and molecular mechanics studies into the solution-
phase conformation of 2 identified conformational promiscuity
in the C6-C8 region of the polyketide.61 This suggests that
configurational or structural changes in this region of the
molecule do not dramatically interfere with F-actin binding. On
the other hand, ꢀ-alanine analogues E-31e and E-31f suffered
a dramatic drop in activity (Table 5, entries 5 and 6), possibly
due to the loss of stabilizing hydrophobic and/or hydrogen-
bonding interactions at the putative actin-binding site.19a
Alternatively, the loss of one stereocenter leading to greater
macrocycle flexibility and the breakdown of a potential type-II
ꢀ-turn may be the determining factor.32
To correlate cytotoxicity with the F-actin polymerizing
properties of these compounds, immunofluorescence cell imag-
ing was then performed by treating BSC-1 cells with 2, 3c, and
synthetic analogues E-31c and E-31e (Figure 5). In DMSO
(Figure 5A), a well-differentiated actin cytoskeleton including
stress fibers can be clearly observed that becomes disorganized
in the presence of the compounds (Figure 5B-E). At the half
maximal toxic concentration (IC50), a similar phenotype effect
is observedsthe stress fibers completely disappear, and actin
lump formation occurs, preferentially in the perinuclear region.
In summary, these results appear to indicate that, while the
phenolic group is indispenable for activity, 2 can be structurally
simplified in the case of E-31b, E-31c, and E-31d without
significant loss in cytotoxicity and actin specificity.
Figure 3. Previously reported cytotoxicity data for synthetic and natural
analogues of 260 and chondramide C (3c).19b
potent than 2.60 Finally, the relative inactivity of OTBS- and
OTIPS-protected analogues Z-28a and E-28d (Table 4, entries
3 and 4) strongly suggest that substituting the phenolic OH with
sterically bulky groups is not well tolerated. In summary, these
results indicate the extent by which the polyketide configuration
dictates the biological activity of 3c. The stereochemical
arrangement at C2 appears to be particularly sensitive to
variation, while the configuration at C6 and C7, N-methylation,
and even the olefin geometry adopt a more fine-tuning role,
leading in some cases to equally potent conformational
manifolds.19b
Molecular Modeling Studies. From the biological data it was
clear that all analogues show the phalloidin phenotype, which
is caused by stabilizing the F-actin fibers through binding at
the contact point of three independent monomers (Figure 6).
To better explain some of the observed SAR trends, we
(60) (a) Gala, F.; D’Auria, M. V.; De Marino, S.; Zollo, F.; Smith, C. D.;
Copper, J. E.; Zampella, A. Tetrahedron 2007, 63, 5212. (b) Gala,
F.; D’Auria, M. V.; De Marino, S.; Sepe, V.; Zollo, F.; Smith, C. D.;
Copper, J. E.; Zampella, A. Tetrahedron 2008, 64, 7127. (c) Gala, F.;
D’Auria, M. V.; De Marino, S.; Sepe, V.; Zollo, F.; Smith, C. D.;
Keller, S. N.; Zampella, A. Tetrahedron 2009, 65, 51.
(61) Inman, W.; Crews, P. J. Am. Chem. Soc. 1989, 111, 2822.
(62) Ebada, S. S.; Wray, V.; de Voogd, N. J.; Deng, Z.; Lin, W.; Proksch,
P. Mar. Drugs 2009, 7, 435.
Jasplakinolide Analogues. Jasplakinolide (2) and five simpli-
fied analogues were screened for cytotoxicity toward MCF-7
and HT-29 cancer cell lines as well (Table 5). The desbromo
(59) (a) Schwab, P.; France, M. B.; Ziller, J. W.; Grubbs, R. H. Angew.
Chem. 1995, 107, 2179. Angew. Chem., Int. Ed. 1995, 34, 2039. (b)
Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc. 1996,
118, 100. (c) Belderrain, T. R.; Grubbs, R. H. Organometallics 1997,
16, 4001.
(63) Tannert, R.; Hu, T.-S.; Arndt, H.-D.; Waldmann, H. Chem. Commun.
2009, 1493.
9
3072 J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010

Structure-Function Studies of Cyclodepsipeptides
A R T I C L E S
Table 5. Cytotoxicity Data for Jasplakinolide/Jaspamide 2 and Analogues 31
performed an overlay of a higher-resolution structure of F-actin
in the ‘unbound’ state reported recently27 with our previously
published binding model of 3c (the ‘bound’ state) derived from
phalloidin-bound F-actin (Figure 6).19a The overall structures
of the actin polymers appear quite similar (Figure 6A), but local
differences exist that could possibly result from ligand binding.
In particular, the Tyr-198 and Phe-200 residues present on
G-actin unit X (Figure 6) would appear to ‘move’ to enable
favorable interactions with the hydrophobic indole moiety of
the macrocycle in this flexible loop region.64 However, more
detailed conclusions must await higher-resolution structural data.
To our delight, our previously proposed binding mode for
3c is supported by the bioactivity trends observed (Figure 6, B
and C). The R-substituted carboxyl region (C1-C3) of the
macrocycle polyketide is located close to the protein surface
(Figure 6, B and C, red), whereas the C4-C5 olefin is apparently
afforded more space. It could be argued, therefore, that the
polyketide is of equal importance to the peptide for molecular
recognition at the binding cleft and not just for tethering the
peptide in the biologically active conformation. For example,
the inactivity (>15 µM) of chondramide C analogues E-29a and
E-29c (Table 4, entries 5 and 8, respectively), both bearing the
unnatural (2R)-configuration in combination with the natural
E-olefin geometry, might be better explained by unfavorable
steric interactions rather than by the population of undesirable
peptide conformations that stem from this single modification.
We even found moderately active examples (Table 4, entries 6
and 7) where the unnatural Z-olefin geometry apparently permits
a ‘better fit’ of the (2R)-configured polyketide. Furthermore,
the activity for analogues carrying the natural (2S)-configuration
and the unnatural Z-olefin geometry (Table 4, entries 9 and 10)
was only attenuated by one-order-of-magnitude, hinting at a
greater tolerance of structural change at this position (Figure 6,
B and C, orange). Crucially, the dimethylhydroxyl portion of
the polyketide (C6-C7) can be seen extending out into the cleft
and is the region where structural change is tolerated most
(Figure 6, B and C, green). For analogues of 3c, changes in
stereochemistry in this region will force the macrocycle into
different conformational manifolds that lead to unfavorable
binding. While this may be the case, we (Table 4, entries 9-11)
and others (Figure 3)19b have found such analogues to be still
quite activesin general, only 10-fold less potent than 3c, and
in one case, resulting in an equipotent compound.19b For
analogues of 2, the potency of E-31c and E-31d (Table 4, entries
3 and 4) is supported well by this binding model. The slight
loss in activity observed in the case of E-31d might be explained
by an increase in macrocycle flexibility. This model also offers
an explanation for the fact that differences in ring size effective
in this region of the polyketide [for example between jasplaki-
nolide (2, 19-membered ring) and chondramide C (3c, 18-
membered ring)] do not strongly impact on activity.
Activity trends can also be rationalized in the peptide region
of the macrocycle. For example, the phenolic group of the
ꢀ-tyrosine residue of 3c is found protruding into a spacious
cavity, though not fully exposed, with the phenolic hydroxyl
group in close proximity (1.9-2.8 Å) to Asp-286 on G-actin
monomer Y, suggesting a favorable stabilizing hydrogen-
bonding interaction (see Supporting Information). In view of
this, it is of no surprise that substituting the phenolic OH of 3c
with a bulky silyl-protecting group significantly impairs activity
(Table 4, compare entries 1 and 2 with 4). The weak activity
(IC50 > 15 µM) of jasplakinolide analogues E-31e and E-31f
(64) All pictures were rendered and analyzed using PyMol, V0.99; DeLano
Scientific Ltd.: Palo Alto, CA.
9
J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010 3073

A R T I C L E S
Tannert et al.
Figure 5. Representative immunofluorescence microscopy images of
BSC-1 cells after 24 h of co-culture with different cyclodepsipeptides close
to their respective IC50 concentrations (magnification 20×, insert 40×). Actin
is stained red (FITC-phalloidin, Invitrogen) and DNA blue (DAPI, Sigma-
Aldrich). (A) DMSO; (B) 2 (25 nM); (C) 3c (100 nM); (D) E-31c (50 nM);
(E) E-31e (15 µM).
bromination or N-methylation was shown to correlate well with
relevant data for natural analogues. We have identified the key
role played by the phenolic ring of the ꢀ-tyrosine moiety for
compound activity. By comparing analogues of chondramide
C (3c), we indicate how changes in polyketide stereochemistry
determine cytotoxicity and rationalize structure-function trends
in terms of a unified binding mode. In this way, we can explain
the promiscuity of the C6-C7 region with respect to stereo-
chemical variation. In support of this argument, we identified
analogues of 2 lacking methyl substituents in this region of the
polyketide that retained potent cytotoxicity and the characteristic
actin-interfering phenotype profile. On the synthetic plane, the
reduced complexity of potent analogues E-31c and E-31d has
greatly facilitated the preparation by improving the E/Z-
selectivity, yield and reaction conditions (40 °C, 5 mol %
catalyst) of the pivotal RCM-mediated macrocycle-forming step.
Furthermore, the total number of discrete steps from commercial
starting materials has been reduced from 32 to 22 (comparing
our synthesis of 2 with that of E-31c and E-31d). The longest
linear sequence has been reduced to 15 steps when counting
all the trivial operations on solid phase. Overall, the synthetic
status reached is of considerable effectiveness and is expected
to facilitate future syntheses of more extensive macrocycle
libraries and of structurally simplified actin-modulator tool
compounds.
Figure 4. Literature data for chondramides A-D (3a-d)18,20 and jaspa-
mides F and H (32c,d).60
(Table 5, entries 5 and 6) might then be rationalized by a loss
of the putative hydrogen bond to Asp-286, fewer hydrophobic
interactions and an increase in macrocycle flexibility that all
systematically weaken binding (see Supporting Information).
Conclusion
Using a novel approach, a focused collection of structurally
and stereochemically diverse cyclodepsipeptides has been
prepared including the total synthesis of jasplakinolide/jaspamide
(2) and chondramide C (3c). The synthesis is highlighted by an
efficient solid phase-based preparation of the linear peptide-
based chain followed by a rewarding ruthenium-catalyzed ring-
closing metathesis of the macrocycle. In this case, metathesis
reactivity and E/Z-selectivity was shown to strongly depend on
the stereochemical configuration of the polyketide region of the
molecule, especially the influence of an olefinic R-methyl
substitutent. The activity of synthetic analogues lacking indole
Experimental Procedures
Example Steglich Esterification Conditions: L-Pea-Ala-
D-MeTrp-ꢀTyr(OTIPS)-O-(2S)-Hex (23n). The peptide acid 11d
(100 mg, 140 µmol - see Supporting Information for details of
synthesis) was dissolved under an argon atmosphere in CH2Cl2 (2.5
mL) and DMF (0.25 mL). At 23 °C, (S)-5-Hexen-2-ol 12k (42 mg,
0.42 mmol, 4.0 equiv), DMAP (34 mg, 0.28 mmol, 2.0 equiv), DIPEA
9
3074 J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010

Structure-Function Studies of Cyclodepsipeptides
A R T I C L E S
(49 µL, 0.28 mmol, 2.0 equiv), and finally EDC ·HCl (53 mg, 0.28
mmol, 2.0 equiv) were added sequentially to the colorless solution.
The mixture was stirred for 19 h, diluted with AcOEt (20 mL), and
washed with a saturated aqueous solution of NH4Cl (2 × 10 mL).
The aqueous layer was extracted with AcOEt (20 mL), and the
combined organic extracts were washed with brine (20 mL), dried
(Na2SO4), filtered, and concentrated. Column chromatography (AcOEt/
cyclohexane) of the crude material yielded the desired peptide diene
23n (95 mg, 120 µmol, 85%) as a clear colorless wax.
Rf ) 0.25 (AcOEt/cyclohexane 1:1); LC: tR ) 12.04 min (Method
A - see Supporting Information); [R]D ) +20.4 (c ) 1.0 in CHCl3);
1H NMR (400 MHz, CDCl3): mixture of rotamers (ratio > 7:1),
3
major rotamer: δ 8.09 (s, 1H; Trp2-NHH), 7.59 (d, J ) 7.8 Hz,
3
1H; Trp4-H), 7.32 (d, J ) 8.0 Hz, 1H; Trp7-H), 7.19-7.09 (m,
3H; Trp6-H, Trp5-H, ꢀTyrꢀ-NH), 7.09 (d, 3J ) 8.5 Hz, 2H; ꢀTyr2-
H, ꢀTyr6-H), 6.94 (s, 1H; Trp2-H), 6.76 (d, 3J ) 8.5 Hz, 2H; ꢀTyr3-
H, ꢀTyr5-H), 6.34 (d, 3J ) 6.1 Hz, 1H; AlaR-NH), 5.71 (ddt, 3J )
3
16.9/10.2/6.6 Hz, 1H; Hex5-H), 5.56 (dd, J ) 10.3/5.9 Hz, 1H;
3
3
TrpR-H), 5.36 (dd, J ) 14.8/7.1 Hz, 1H; ꢀTyrꢀ-H), 4.94 (dd, J
2
3
) 17.2 Hz, J ) 1.6 Hz, 1H; Hex6-Htrans), 4.92 (d, J ) 10.1 Hz,
1H; Hex6-Hcis), 4.86-4.77 (m, 1H; Hex2-H), 4.77 (s, 1H; Pea5-
Ha), 4.70 (s, 1H; Pea5-Hb), 4.61 (dq, 3J ) 6.8/6.8 Hz, 1H; AlaR-H),
3
2
3
3.46 (dd, J ) 5.9 Hz, J ) 15.7 Hz, 1H; Trpꢀ-Ha), 3.21 (dd, J )
10.5 Hz, 2J ) 15.7 Hz, 1H; Trpꢀ-Hb), 2.94 (s, 3H; Trp-NCH3), 2.90
3
2
3
(dd, J ) 7.4 Hz, J ) 15.2 Hz, 1H; ꢀTyrR-Ha), 2.75 (dd, J ) 6.2
Hz, 2J ) 15.2 Hz, 1H; ꢀTyrR-Hb), 2.46-2.36 (m, 1H; Pea2-H), 2.36
(dd, 3J ) 6.3 Hz, 2J ) 13.9 Hz, 1H; Pea3-Ha,), 2.03 (dd, 3J ) 7.6 Hz,
2J ) 13.7 Hz, 1H; Pea3-Hb), 1.99-1.86 (m, 1H; Hex4-H2), 1.68 (s,
3H; Pea7-H3), 1.52-1.40 (m, 1H; Hex3-Ha), 1.29-1.16 (m, 4H; Hex3-
Hb, TIPS), 1.11 (d, 3J ) 6.1 Hz, 3H; Hex1-H3), 1.08 (d, 3J ) 7.2 Hz,
18H; TIPS), 1.10-1.06 (m, 3H; Pea6-H3), 0.95 (d,3J ) 6.8 Hz, 3H;
Alaꢀ-H3) ppm; 13C NMR (100 MHz, CDCl3): δ 176.0 (Pea-1), 173.9
(Ala-CdO), 170.4 (Trp-CdO), 168.9 (ꢀTyr-CdO), 155.4 (ꢀTyr-4),
142.8 (Pea-4), 137.6 (Hex-5), 136.0 (Trp-7a), 132.9 (ꢀTyr-1), 127.5
(2 ×, ꢀTyr-2, ꢀTyr-6), 127.3 (Trp-3a), 122.0 (2 ×, Trp-2, Trp-6), 119.7
(2 ×, ꢀTyr-3, ꢀTyr-5), 119.4 (Trp-5), 118.5 (Trp-4), 114.9 (Hex-6),
112.4 (Pea-5), 111.1 (Trp-ꢀ), 111.0 (Trp-7), 70.7 (Hex-2), 56.8 (Trp-
R), 49.4 (ꢀTyr-ꢀ), 45.7 (Ala-R), 41.7 (Pea-3), 40.8 (ꢀTyr-R), 38.7 (Pea-
2), 34.8 (Hex-3), 31.0 (Trp-NCH3), 29.4 (Hex-4), 23.3 (Trp-ꢀ), 22.2
(Pea-7), 19.8 (Hex-1), 17.9 (6 ×, TIPS), 17.2 (Ala-ꢀ), 16.9 (Pea-6),
12.6 (3 ×, TIPS) ppm; IR (neat): ν˜ ) 3306, 3062, 2942, 2867, 1731,
1639, 1510 cm-1; HR-MS (ESI): calcd for C46H69N4O6Si [M + H]+:
801.4981, found 801.4989.
Synthesis of Chondramide C Analogues E-28a, Z-28a,
E-29a, and Z-29a. Peptide diene 23b (26 mg, 34 µmol - see
Supporting Information for details of synthesis) was placed in an
oven-dried 100 mL two-neck flask equipped with a stirring bar and
dissolved in PhMe (30 mL, 1.0 mM). Argon was bubbled through the
colorless solution Via canula at 23 °C for 10 min. A condenser was
then placed on one neck, the flask was placed in an oil bath at 110
°C, and argon bubbling was continued for a further 10 min. Grubbs’
second-generation catalyst (25) (7 mg, 8.5 µmol, 25 mol %) was
dissolved in PhMe (1 mL), and the dark-red solution was added rapidly
via syringe to the diene solution. The resulting dark solution was stirred
for 2 h at 110 °C oil bath temperature with continuous argon purging,
after which the solution was cooled down to 23 °C and concentrated
to yield the crude product. Two successive column chromatographies
(i) 50% AcOEt in cyclohexane, (ii) 33% AcOEt in cyclohexane)
of the crude material yielded cyclo-[Ala-D-MeTrp-ꢀTyr(OTBS)-
(2R,4E,6S,7R)-Hto] E-28a (4.4 mg, 6.0 µmol, 18%) and cyclo-[Ala-
D-MeTrp-ꢀTyr(OTBS)-(2R,4Z,6S,7R)-Hto] Z-28a (6.4 mg, 8.8 µmol,
26%) as colorless waxes.
Figure 6. (A) Graphical overlay of the ‘bound’ (blue, lower resolution) and
‘unbound’ (green, higher resolution) structures of F-actin and the phalloidin
binding cleft. In the unbound state, the three actin subunits are assigned three
different shades of green. Amino acid residues Tyr-198 and Phe-200 are colored
differently in the bound (yellow and pale yellow) and unbound (red and dark
red) states. (B) Expanded space-fill model of the bound F-actin structure with
bound 3c.19a The three independent actin subunits (X, Y, and Z) are labeled in
three different shades of blue. (C) Close-up of the binding cleft: Regions where
structural changes lead to a complete loss of activity are labeled red; regions
that weaken but do not extinguish bioactivity are colored orange; changes that
are tolerated are shaded green. The Tyr-198 and Phe-200 residues of the actin
monomer unit X are highlighted yellow and pale yellow.64
cyclo-[Ala-D-MeTrp-ꢀTyr(OTBS)-(2R,4E,6S,7R)-Hto] E-28a. Rf
) 0.16 (cyclohexane/AcOEt 1:2); LC: tR ) 11.76 min (Method C -
see Supporting Information); [R]D ) +38.2 (c ) 0.90 in CHCl3); 1H
NMR (400 MHz, CDCl3): δ 8.00 (s, 1H; Trp2-NH), 7.58 (d, 3J )
7.9 Hz, 1H; Trp4-H), 7.31 (d, 3J ) 8.0 Hz, 1H; Trp7-H), 7.18 (ddd,
3
4
4J ) 1.2 Hz, J ) 8.1/7.1 Hz, 1H; Trp6-H), 7.12 (ddd, J ) 1.1
Hz, 3J ) 7.9/7.1 Hz, 1H; Trp5-H), 6.99 (d, 3J ) 8.5 Hz, 2H; ꢀTyr2-
9
J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010 3075

A R T I C L E S
Tannert et al.
H, ꢀTyr6-H), 6.91 (d, 3J ) 2.3 Hz, 1H; Trp2-H), 6.72 (d, 3J ) 8.6
Hz, 2H; ꢀTyr3-H, ꢀTyr5-H), 6.54 (d, 3J ) 6.6 Hz, 1H; AlaR-NH),
Rf ) 0.10 (CH2Cl2/MeOH 25:1); LC: tR ) 8.51 min (Method C -
see Supporting Information); [R]D ) +26.5 (c ) 0.60 in MeOH);
1H NMR (500 MHz, [d6]DMSO): δ 10.76 (s, 1H; Trp2-NH), 9.27
(br s, 1H; ꢀTyr4-OH), 8.38 (d, 3J ) 9.1 Hz, 1H; ꢀTyrꢀ-NH), 7.54
(d, 3J ) 7.9 Hz, 1H; Trp4-H), 7.29 (d, 3J ) 8.1 Hz, 1H; Trp7-H),
7.10-7.00 (m, 4H; Trp6-H, ꢀTyr2-H, ꢀTyr6-H, Trp2-H), 6.95 (t,
3
3
6.36 (d, J ) 9.2 Hz, 1H; ꢀTyrꢀ-NH), 5.49 (dd, J ) 9.5/7.0 Hz,
3
1H; TrpR-H), 5.48-5.41 (m, 1H; ꢀTyrꢀ-H), 5.08 (d, J ) 10.1
3
Hz, 1H; Hto5-H), 4.79 (dq, J ) 6.1/2.9 Hz, 1H; Hto7-H), 4.64
3
3
2
(dq, J ) 6.7/6.7 Hz, 1H; AlaR-H), 3.29 (dd, J ) 7.1 Hz, J )
3
3J ) 7.3 Hz, 1H; Trp5-H), 6.85 (d, J ) 6.7 Hz, 1H; AlaR-NH),
3
2
15.3 Hz, 1H; Trpꢀ-Ha), 3.21 (dd, J ) 9.6 Hz, J ) 15.4 Hz, 1H;
Trpꢀ-Hb), 2.91 (s, 3H; Trp-NCH3), 2.72-2.64 (m, 3H; ꢀTyrR-H2,
6.66 (d, 3J ) 8.5 Hz, 2H; ꢀTyr3-H, ꢀTyr5-H), 5.52 (dd, 3J ) 9.6/
6.4 Hz, 1H; TrpR-H), 5.19 (t, 3J ) 10.4 Hz, ꢀTyrꢀ-H), 5.00 (d, 3J
3
Hto6-H), 2.38 (dqd, J ) 9.8/7.0/2.8 Hz, 1H; Hto2-H), 2.28 (dd,
3
3J ) 4.5 Hz, 2J ) 13.4 Hz, 1H; Hto3-Ha), 2.08 (dd, 3J ) 13.1 Hz,
) 9.7 Hz, 1H; Hto5-H), 4.67 (dd, J ) 6.2/3.0 Hz, 1H; Hto7-H),
4.47 (dq, 3J ) 6.7/6.7 Hz, 1H; AlaR-H), 3.05 (dd, 3J ) 9.9 Hz, 2J
2J ) 13.1 Hz, 1H; Hto3-Hb), 1.60 (s, 3H; Hto10-H3), 1.20 (d, 3J )
3
2
3
) 15.0 Hz, 1H; Trpꢀ-Ha), 2.97 (dd, J ) 6.0 Hz, J ) 15.0 Hz,
7.1 Hz, 3H; Alaꢀ-H3), 0.98 (s, 9H; TBS), 0.94 (d, J ) 6.2 Hz,
1H; Trpꢀ-Hb), 2.91 (s, 3H; Trp-NCH3), 2.76 (dd, 3J ) 11.5 Hz, 2J
3H; Hto9-H3), 0.82 (d, 3J ) 5.4 Hz, 3H; Hto11-H3), 0.81 (d, 3J )
3
2
5.3 Hz, 3H; Hto8-H3), 0.19 (d, J ) 0.7 Hz, 6H; TBS) ppm; 13C
3
) 13.4 Hz, 1H; ꢀTyrR-Ha), 2.56 (dd, J ) 2.3 Hz, J ) 13.4 Hz,
1H; ꢀTyrR-Hb), 2.52-2.46 (m, 1H; Hto2-H), 2.44-2.34 (m, 1H;
Hto6-H), 2.15 (app d, 2J ) 13.4 Hz, 1H; Hto3-Ha), 2.00 (dd, 3J )
12.8 Hz, 2J ) 12.8 Hz, 1H; Hto3-Hb), 1.53 (s, 3H; Hto10-H3), 1.10
NMR (100 MHz, CDCl3): δ 174.1 (Hto-1), 174.0 (Ala-CdO), 170.9
(ꢀTyr-CdO), 168.8 (Trp-CdO), 154.9 (ꢀTyr-4), 136.1 (Trp-7a),
133.9 (Hto-4/ꢀTyr-1), 133.8 (ꢀTyr-1/Hto-4), 127.1 (Trp-3a), 126.8
(2 ×, ꢀTyr-2, ꢀTyr-6), 126.5 (Hto-5), 122.2 (Trp-2/Trp-6), 122.2
(Trp-6/Trp-2), 120.1 (2 ×, ꢀTyr-3, ꢀTyr-5), 119.6 (Trp-5), 118.4
(Trp-4), 111.1 (Trp-7), 110.5 (Trp-3), 75.0 (Hto-7), 56.4 (Trp-R),
50.0 (ꢀTyr-ꢀ), 46.7 (Ala-R), 45.7 (Hto-3), 41.7 (ꢀTyr-R), 41.4 (Hto-
2), 37.0 (Hto-6), 30.1 (Trp-NCH3), 25.6 (3 ×, TBS), 23.7 (Trp-ꢀ),
19.2 (Ala-ꢀ), 18.4 (Hto-9), 18.2 (Hto-11), 18.2 (TBS), 17.7 (Hto-
8), 15.0 (Hto-10), -4.4 (2 ×, TBS) ppm; IR (neat): ν˜ ) 3320,
2958, 2930, 1722, 1671, 1635, 1459, 1263, 1207, 914, 840, 804,
782, 741 cm-1; HR-MS (ESI): calcd for C41H59N4O6Si [M + H]+:
731.4198, found 731.4202.
cyclo-[Ala-D-MeTrp-ꢀTyr(OTBS)-(2R,4Z,6S,7R)-Hto] Z-28a. Rf
) 0.39 (cyclohexane/AcOEt 1:2); LC: tR ) 11.99 min (Method C -
see Supporting Information); [R]D ) +9.1 (c ) 0.90 in CHCl3);
1H NMR (500 MHz, CDCl3): δ 7.97 (s, 1H; Trp2-NH), 7.58 (d, 3J
) 7.9 Hz, 1H; Trp4-H), 7.33 (d, 3J ) 8.1 Hz, 1H; Trp7-H),
7.19-7.13 (m, 2H; ꢀTyrꢀ-NH, Trp6-H), 7.12 (t, 3J ) 7.5 Hz, 1H;
Trp5-H), 6.94 (d, 3J ) 8.5 Hz, 2H; ꢀTyr2-H, ꢀTyr6-H), 6.83 (d, 3J
) 6.6 Hz, 1H; AlaR-NH), 6.78 (d, 3J ) 1.9 Hz, 1H; Trp2-H), 6.70
3
3
(d, J ) 7.1 Hz, 3H; Hto9-H3), 0.83 (d, J ) 6.2 Hz, 3H; Hto11-
H3), 0.80 (d, 3J ) 6.9 Hz, 3H; Hto8-H3), 0.71 (d, 3J ) 6.7 Hz, 3H;
Alaꢀ-H3) ppm; 13C NMR (125 MHz, CDCl3): δ 172.7 (Hto-1),
171.9 (Ala-CdO), 171.2 (ꢀTyr-CdO), 168.4 (Trp-CdO), 156.0
(ꢀTyr-4), 135.9 (Trp-7a), 133.2 (Hto-4), 132.1 (ꢀTyr-1), 127.1 (2
×, ꢀTyr-2, ꢀTyr-6), 126.9 (Trp-3a), 125.9 (Hto-5), 123.1 (Trp-2),
120.7 (Trp-6), 118.2 (Trp-5), 118.0 (Trp-4), 114.8 (2 ×, ꢀTyr-3,
ꢀTyr-5), 111.0 (Trp-7), 109.0 (Trp-3), 73.8 (Hto-7), 54.8 (Trp-R),
49.5 (ꢀTyr-ꢀ), 45.4 (Ala-R), 44.4 (Hto-3), 41.3 (ꢀTyr-R), 40.6 (Hto-
2), 36.0 (Hto-6), 29.5 (Trp-NCH3), 24.1 (Trp-ꢀ), 18.4 (Ala-ꢀ), 17.5
(Hto-9), 17.5 (Hto-11), 17.3 (Hto-10), 14.8 (Hto-8) ppm; HR-MS
(ESI): calcd for C35H45N4O6 [M + H]+: 617.3334, found 617.3332.
cyclo-[Ala-D-MeTrp-ꢀTyr-(2R,4Z,6S,7R)-Hto] Z-29a. The des-
ilylation of Z-28a (11 mg, 15 µmol) was performed according to
the procedure described for the desilylation of E-28a. Purification
of the crude material by column chromatography (67% cyclohexane
in AcOEt, then 3% MeOH in CH2Cl2) yielded Z-29a as a colorless
wax (7.8 mg, 13 µmol, 88%, 23% over two steps), which was
purified further by preparative HPLC (Method E - see Supporting
Information) prior to cytotoxocity studies to yield a colorless powder
after lyophilization.
3
3
(d, J ) 8.6 Hz, 2H; ꢀTyr3-H, ꢀTyr5-H), 5.44 (dd, J ) 8.2/8.2
3
Hz, 1H; TrpR-H), 5.23 (dd, J ) 8.9/3.4 Hz, 1H; ꢀTyrꢀ-H), 4.87
3
(d, J ) 8.9 Hz, 1H; Hto5-H), 4.75-4.65 (m, 2H; Hto7-H, AlaR-
Rf ) 0.17 (CH2Cl2/MeOH 25:1); LC: tR ) 8.80 min (Method C -
see Supporting Information); [R]D ) +3.2 (c ) 0.60 in MeOH);
1H NMR (500 MHz, [d6]DMSO): δ 10.81 (s, 1H; Trp2-NH), 9.25
(s, 1H; ꢀTyr4-OH), 8.60 (d, 3J ) 8.3 Hz, 1H; ꢀTyrꢀ-NH), 7.56 (d,
3
2
H), 3.33 (dd, J ) 7.9 Hz, J ) 15.3 Hz, 1H; Trpꢀ-Ha), 3.17 (dd,
3J ) 8.5 Hz, 2J ) 15.1 Hz, 1H; Trpꢀ-Hb), 2.98 (s, 3H; Trp-NCH3),
2.77-2.71 (m, 1H; Hto6-H), 2.70 (dd, 3J ) 3.4 Hz, 2J ) 15.1 Hz,
1H; ꢀTyrR-Ha), 2.67-2.53 (m, 3H; ꢀTyrR-Hb, Hto2-H, Hto3-Ha),
3
3J ) 7.9 Hz, 1H; Trp4-H), 7.31 (d, J ) 8.1 Hz, 1H; Trp7-H),
3
2.11-2.03 (m, 1H; Hto3-Hb), 1.71 (s, 3H; Hto10-H3), 1.21 (d, J
7.11-6.99 (m, 4H; Trp6-H, ꢀTyr2-H, ꢀTyr6-H, Trp2-H), 6.97 (t,
) 6.6 Hz, 3H; Alaꢀ-H3), 1.15 (d, 3J ) 6.6 Hz, 3H; Hto9-H3), 1.02
3
3J ) 7.4 Hz, 1H; Trp5-H), 6.85 (d, J ) 6.7 Hz, 1H; AlaR-NH),
3
3
6.66 (d, 3J ) 8.5 Hz, 2H; ꢀTyr3-H, ꢀTyr5-H), 5.44 (dd, 3J ) 9.2/
6.8 Hz, 1H; TrpR-H), 5.15-5.06 (m, 1H; ꢀTyrꢀ-H), 4.83 (d, 3J )
9.7 Hz, 1H; Hto5-H), 4.64 (dq, 3J ) 6.5/4.3 Hz, 1H; Hto7-H), 4.51
(dq, 3J ) 6.6/6.6 Hz, 1H; AlaR-H), 3.07 (dd, 3J ) 9.4 Hz, 2J ) 15.3
Hz, 1H; Trpꢀ-Ha), 2.98 (s, 3H; Trp-NCH3), 2.96 (dd, 3J ) 6.3 Hz,
2J ) 15.1 Hz, 1H; Trpꢀ-Hb), 2.69-2.62 (m, 2H; ꢀTyrR-H2),
2.59-2.52 (m, 1H; Hto2-H), 2.48-2.35 (m, 2H; Hto6-H, Hto3-
(d, J ) 6.6 Hz, 3H; Hto11-H3), 0.98 (s, 9H; TBS), 0.79 (d, J )
6.8 Hz, 3H; Hto8-H3), 0.19 (s, 6H; TBS) ppm; 13C NMR (100 MHz,
CDCl3): δ 174.8 (Hto-1), 173.6 (Ala-CdO), 170.3 (ꢀTyr-CdO),
168.8 (Trp-CdO), 154.8 (ꢀTyr-4), 136.0 (Trp-7a), 135.1 (Hto-4),
133.6 (ꢀTyr-1), 128.6 (Hto-5), 127.1 (Trp-3a), 127.0 (2 ×, ꢀTyr-
2, ꢀTyr-6), 122.2 (Trp-2/Trp-6), 122.1 (Trp-6/Trp-2), 119.9 (2 ×,
ꢀTyr-3, ꢀTyr-5), 119.5 (Trp-5), 118.4 (Trp-4), 111.0 (Trp-7), 110.4
(Trp-3), 74.1 (Hto-7), 56.1 (Trp-R), 49.6 (ꢀTyr-ꢀ), 45.9 (Ala-R), 41.5
(ꢀTyr-R), 40.2 (Hto-2), 37.8 (Hto-3), 34.0 (Hto-6), 30.1 (Trp-NCH3),
25.6 (3 ×, TBS), 23.4 (Trp-ꢀ), 22.8 (Ala-ꢀ), 18.4 (Hto-9), 18.2 (TBS),
16.9 (Hto-11), 12.7 (Hto-10), 12.5 (Hto-8), -4.3 (2 ×, TBS) ppm; IR
(neat): ν˜ ) 3325, 2932, 1678, 1632, 1510, 1460, 1261, 1205, 1173,
1066, 913, 840, 804, 782, 740 cm-1; HR-MS (ESI): calcd for
C41H59N4O6Si [M + H]+: 731.4198, found 731.4201.
cyclo-[Ala-D-MeTrp-ꢀTyr-(2R,4E,6S,7R)-Hto] E-29a. To a
solution of E-28a (11 mg, 15 µmol) in THF (0,75 mL, 20 mM) in
a 5 mL round-bottomed flask was added at 23 °C by micropipet
TBAF (1.0 equiv, 15 µmol) as a solution in THF (1.0 M, 15 µL).
The reaction was stirred for 30 min, and then all solvents were
removed by reduced pressure. Purification of the crude material
by column chromatography (67% cyclohexane in AcOEt, then 3%
MeOH in CH2Cl2) yielded E-29a as a colorless wax (7.1 mg, 12
µmol, 77%, 14% over two steps) that was further purified by
preparative HPLC (Method E - see Supporting Information) prior
to cytotoxocity studies to yield a colorless powder after lyophilization.
3
2
Ha), 1.94 (dd, J ) 4.1 Hz, J ) 12.7 Hz, 1H; Hto3-Hb), 1.66 (s,
3H; Hto10-H3), 1.11 (d, 3J ) 6.7 Hz, 3H; Alaꢀ-H3), 1.02 (d, 3J )
6.6 Hz, 3H; Hto9-H3), 0.89 (d, J ) 6.6 Hz, 3H; Hto11-H3), 0.68
3
(d, 3J ) 6.8 Hz, 3H; Hto8-H3) ppm; 13C NMR (125 MHz,
[d6]DMSO): δ 172.6 (Hto-1), 172.0 (Ala-CdO), 168.9 (ꢀTyr-
CdO), 168.5 (Trp-CdO), 155.9 (ꢀTyr-4), 136.0 (Trp-7a), 133.7
(Hto-4), 132.5 (ꢀTyr-1), 128.0 (Hto-5), 126.9 (2 ×, ꢀTyr-2, ꢀTyr-
6), 126.8 (Trp-3a), 122.7 (Trp-2), 120.8 (Trp-6), 118.1 (Trp-5),
118.0 (Trp-4), 114.8 (2 ×, ꢀTyr-3, ꢀTyr-5), 111.0 (Trp-7), 109.1
(Trp-3), 72.2 (Hto-7), 54.8 (Trp-R), 48.9 (ꢀTyr-ꢀ), 44.8 (Ala-R),
44.4 (Hto-3), 41.8 (ꢀTyr-R), 40.0 (Hto-2), 37.7 (Hto-6), 33.8 (Trp-
NCH3), 24.0 (Trp-ꢀ), 22.7 (Ala-ꢀ), 17.8 (Hto-9), 17.6 (Hto-11),
12.8 (Hto-10), 12.5 (Hto-8) ppm; HR-MS (ESI): calcd for
C35H45N4O6 [M + H]+: 617.3334, found 617.3331.
Synthesis of Simplified Jasplakinolide/Jaspamide Analogues
E-30c and E-31c. cyclo-[Ala-D-MeTrp-ꢀTyr-(2S,4E,8S)-Htn]
(E-31c). Conditions A. Peptide diene 23n (29 mg, 36 µmol) was
subjected to RCM according to the conditions described for the
9
3076 J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010

Structure-Function Studies of Cyclodepsipeptides
A R T I C L E S
RRCM of peptide diene 23b. Column chromatography (70% AcOEt
in cyclohexane) of the crude metathesis product yielded E-30c (16.9
mg, 21.9 µmol, 61%) as a pale brown wax.
7.60 (d, 3J ) 7.9 Hz, 1H; Trp4-H), 7.32 (d, 3J ) 8.0 Hz, 1H; Trp7-
3
4
H), 7.22 (d, J ) 8.2 Hz, 1H; ꢀTyrꢀ-NH), 7.16 (td, J ) 1.0 Hz,
3
3J ) 7.2 Hz, 1H; Trp6-H), 7.10 (td, J ) 7.2 Hz, 1H; Trp5-H),
3
Conditions B. Peptide diene 23n (6.8 mg, 8.5 µmol) was placed
in an oven-dried microwave tube equipped with a stirring bar and
dissolved in anhydrous CH2Cl2 (6.0 mL). Catalyst 25 (0.36 mg,
0.043 µmol, 10 mol %) was dissolved separately in anhydrous
CH2Cl2 (2.5 mL) and the red solution introduced rapidly to the
reaction via plastic syringe (total volume ) 8.5 mL, 1 mM). The
vessel was then hermetically sealed and heated in an oil bath (40
°C) with stirring for 2 h. The reaction was then cooled and the
solvent removed under reduced pressure. Column chromatography
(70% AcOEt in cyclohexane) of the crude metathesis product
yielded cyclo-[Ala-D-MeTrp-ꢀTyr(OTIPS)-(2S,4E,8S)-Htn] E-30c
(4.3 mg, 5.5 µmol, 65%) as a pale brown wax.
6.92 (d, J ) 8.4 Hz, 2H; ꢀTyr2-H, ꢀTyr6-H), 6.83 (s, 1H; Trp2-
H), 6.71-6.67 (m, 1H; AlaR-NH), 6.68 (d, 3J ) 8.2 Hz, 2H; ꢀTyr3-
3
H, ꢀTyr5-H), 5.65 (dd, J ) 8.9/7.2 Hz, 1H; TrpR-H), 5.19 (ddd,
3J ) 8.3/8.3/3.9 Hz, 1H; ꢀTyrꢀ-H), 5.03 (t, 3J ) 6.7 Hz, 1H; Htn5-
H), 4.85-4.77 (m, 2H; Htn8-H, AlaR-H), 3.42 (dd, 3J ) 6.9 Hz, 2J )
3
2
15.5 Hz, 1H; Trpꢀ-Ha), 3.16 (dd, J ) 9.3 Hz, J ) 15.4 Hz, 1H;
Trpꢀ-Hb), 2.96 (s, 3H; Trp-NCH3), 2.73 (dd, 3J ) 3.8 Hz, 2J ) 15.5
Hz, 1H; ꢀTyrR-Ha), 2.54 (dd, 3J ) 8.6 Hz, 2J ) 15.5 Hz, 1H; ꢀTyrR-
Hb), 2.45-2.40 (m, 2H; Htn2-H, Htn3-Ha), 1.90-1.80 (m, 3H; Htn3-
Hb, Htn6-H2), 1.59-1.51 (m, 1H; Htn7-Ha), 1.48 (s, 3H; Htn11-H3),
3
1.39-1.30 (m, 1H; Htn7-Hb), 1.13 (d, J ) 5.7 Hz, 3H; Htn10-H3),
1.09 (d, 3J ) 6.3 Hz, 3H; Htn9-H3), 0.99 (d, 3J ) 6.7 Hz, 3H; Alaꢀ-
H3) ppm;13C NMR (150 MHz, CDCl3): δ)174.7 (Ala-CdO), 174.7
(Htn-1), 170.5 (Trp-CdO), 169.1 (ꢀTyr-CdO), 155.7 (ꢀTyr-4), 136.2
(Trp-7a), 133.9 (Htn-5), 131.9 (ꢀTyr-1), 127.1 (2 ×, ꢀTyr-2, ꢀTyr-
6), 127.1 (Trp-3a), 124.5 (Htn-4), 122.1 (Trp-6), 122.0 (Trp-2), 119.4
(Trp-5), 118.5 (Trp-4), 115.6 (2 ×, ꢀTyr-3, ꢀTyr-5), 111.2 (Trp-7),
110.4 (Trp-3), 70.0 (Htn-8), 56.2 (Trp-R), 49.2 (ꢀTyr-ꢀ), 46.0 (Ala-
R), 43.4 (Htn-7), 40.0 (Htn-2), 40.0 (ꢀTyr-R), 35.4 (Htn-3), 30.4 (Trp-
NCH3), 23.2 (Htn-6), 23.2 (Trp-ꢀ), 20.2 (Htn-9), 19.7 (Htn-10), 18.1
(Ala-ꢀ), 16.2 (Htn-11) ppm; IR (neat): ν˜ ) 3308, 3055, 2930, 2854,
1726, 1631, 1510 cm-1; HR-MS (ESI): calcd for C35H45N4O6 [M +
H]+: 617.3334, found 617.3331.
cyclo-[Ala-D-MeTrp-ꢀTyr(OTIPS)-(2S,4E,8S)-Htn] E-30c. Rf
) 0.26 (AcOEt/cyclohexane 7:3); LC: tR ) 12.20 min (Method B -
see Supporting Information); [R]D20 ) +32.5 (c ) 0.08 in CHCl3);
1H NMR (400 MHz, CDCl3): δ 8.09 (s, 1H; Trp2-NH), 7.62 (d, 3J
3
) 7.8 Hz, 1H; Trp4-H), 7.33 (d, J ) 8.0 Hz, 1H; Trp7-H), 7.18
4
3
(td, J ) 1.0 Hz, J ) 7.1 Hz, 1H; Trp6-H), 7.16-7.10 (m, 1H;
ꢀTyrꢀ-NH), 7.12 (td, 3J ) 7.1 Hz, 1H; Trp5-H), 7.04 (d, 3J ) 8.4
3
Hz, 2H; ꢀTyr2-H, ꢀTyr6-H), 6.94 (s, 1H; Trp2-H), 6.80 (d, J )
8.2 Hz, 2H; ꢀTyr3-H, ꢀTyr5-H), 6.69-6.67 (m, 1H; AlaR-NH),
3
3
5.68 (dd, J ) 8.8/7.3 Hz, 1H; TrpR-H), 5.24 (ddd, J ) 8.3/8.3/
3.8 Hz, 1H; ꢀTyrꢀ-H), 5.03 (t, 3J ) 6.8 Hz, 1H; Htn5-H),
4.85-4.74 (m, 2H; Htn8-H, AlaR-H), 3.38 (dd, 3J ) 7.0 Hz, 2J )
3
2
15.6 Hz, 1H; Trpꢀ-Ha), 3.23 (dd, J ) 9.3 Hz, J ) 15.4 Hz, 1H;
Acknowledgment. This work was supported by the Max
Planck Gesellschaft (to H.W.), the Deutsche Forschungsgemein-
schaft (Emmy-Noether young investigator award to H.D.A.), the
state of North-Rhine-Westphalia (ZACG Dortmund) and the
European Union (RTN Endocyte). T.S.H. and L.G.M. are grateful
to the Alexander von Humboldt-Stiftung for research fellowships.
We thank Prof. Dr. M. Kalesse for discussions and supply of a
sample of natural chondramide C. We also thank Prof. Dr. S.
Blechert and Dipl. Chem. D. Rost (TU Berlin) for support and
catalyst samples during our metathesis studies, and Dipl. Biol.
S. Baumann is thanked for helpful assistance with preparing
figure graphics.
3
2
Trpꢀ-Hb), 2.92 (s, 3H; Trp-NCH3), 2.78 (dd, J ) 3.9 Hz, J )
15.1 Hz, 1H; ꢀTyrR-Ha), 2.58 (dd, 3J ) 8.6 Hz, 2J ) 15.1 Hz, 1H;
ꢀTyrR-Hb), 2.45-2.40 (m, 2H; Htn2-H, Htn3-Ha), 1.90-1.80 (m,
3H; Htn3-Hb, Htn6-H2), 1.59-1.51 (m, 1H; Htn7-Ha), 1.50 (s, 3H;
Htn11-H3), 1.39-1.30 (m, 1H; Htn7-Hb), 1.30-1.16 (m, 6H; Htn10-
3
3
H3 and TIPS), 1.10 (d, J ) 7.4 Hz, 18H; TIPS), 1.08 (d, J ) 6.2
Hz, 3H; Htn9-H3), 0.97 (d, 3J ) 6.8 Hz, 3H; Alaꢀ-H3) ppm;13C NMR
(100 MHz, CDCl3): δ 174.7, 174.3, 170.3, 169.0, 155.4, 136.2, 134.0,
132.8, 127.3, 127.0, 124.4, 122.1, 121.9, 119.9, 119.8, 119.5, 118.6,
111.1, 111.0, 69.8, 55.8, 48.9, 46.0, 43.4, 39.8, 39.7, 35.5, 30.2, 29.6,
23.2, 22.9, 20.2, 19.8, 18.0, 17.9, 16.3, 12.6 ppm; IR (thin film): ν˜ )
3310, 2927, 2867, 1730, 1660, 1636, 1509 cm-1; HR-MS (ESI): calcd
for C44H65N4O6Si [M + H]+: 773.4673, found 773.4673; calcd for
C44H64N4NaO6Si [M + Na]+: 795.4493, found 795.4484.
cyclo-[Ala-D-MeTrp-ꢀTyr-(2S,4E,8S)-Htn] (E-31c). The desi-
lylation of E-30c (13 mg, 17 µmol) was performed according to
the procedure described for the desilylation of E-28a. Column
chromatography (AcOEt/cyclohexane) of the crude material yielded
E-31c as a colorless wax (10 mg, 16 µmol, 97%, 58% over two
steps), which was purified again prior to cytotoxicity studies by
preparative HPLC (Method E - see Supporting Information) to
afford a colorless powder after lyophilization.
Supporting Information Available: Detailed experimental
(synthetic and cell screening) procedures, comprehensive com-
1
pound characterization, including copies of H NMR and 13C
NMR spectral data, molecular modeling data, additional graphi-
cal representation and numbering convention for peptide diene
23n and macrocyclic structures E-28a, Z-28a, E-29a, Z-29a,
E-30c, and E-31c. This material is available free of charge via
the Internet at http://pubs.acs.org.
Rf ) 0.31 (AcOEt/cyclohexane 4:1); LC: tR ) 7.13 min (Method
A - see Supporting Information); [R]D ) +36.9 (c ) 0.79 in
1
MeOH); H NMR (500 MHz, CDCl3): δ 8.29 (s, 1H; Trp2-NH),
JA9095126
9
J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010 3077