Total synthesis of complestatin: Development of a Pd(0)-mediated indole annulation for macrocyclization

Article abstract of DOI:10.1021/ja102304p

Full details of the initial development and continued examination of a powerful intramolecular palladium(0)-mediated indole annulation for macrocyclization closure of the strained 16-membered biaryl ring system found in complestatin (1, chloropeptin II) and the definition of factors impacting its intrinsic atropodiastereoselectivity are described. Its examination and use in an alternative, second-generation total synthesis of complestatin are detailed in which the order of the macrocyclization reactions was reversed from our first-generation total synthesis. In this approach and with the ABCD biaryl ether ring system in place, the key Larock cyclization was conducted with substrate 36 (containing four phenols, five secondary amides, one carbamate, and four labile aryl chlorides) and provided the product 37 (56%) exclusively as a single atropisomer (>20:1, detection limits) possessing the natural (R)-configuration. In this instance, the complexity of the substrate and the reverse macrocyclization order did not diminish the atropodiastereoselectivity; rather, it provided an improvement over the 4:1 selectivity that was observed with the analogous substrate used to provide the isolated DEF ring system in our first-generation approach. Just as significant, the atroposelectivity represents a complete reversal of the diasteroselectivity observed with analogous macrocyclizations conducted using a Suzuki biaryl coupling.

Full text of DOI:10.1021/ja102304p

Published on Web 05/17/2010
Total Synthesis of Complestatin: Development of a
Pd(0)-Mediated Indole Annulation for Macrocyclization
Hiroyuki Shimamura, Steven P. Breazzano, Joie Garfunkle, F. Scott Kimball,
John D. Trzupek, and Dale L. Boger*
Department of Chemistry and the Skaggs Institute for Chemical Biology, The Scripps Research
Institute, 10550 North Torrey Pines Road, La Jolla, California 92037
Received March 18, 2010; E-mail: boger@scripps.edu
Abstract: Full details of the initial development and continued examination of a powerful intramolecular
palladium(0)-mediated indole annulation for macrocyclization closure of the strained 16-membered biaryl
ring system found in complestatin (1, chloropeptin II) and the definition of factors impacting its intrinsic
atropodiastereoselectivity are described. Its examination and use in an alternative, second-generation total
synthesis of complestatin are detailed in which the order of the macrocyclization reactions was reversed
from our first-generation total synthesis. In this approach and with the ABCD biaryl ether ring system in
place, the key Larock cyclization was conducted with substrate 36 (containing four phenols, five secondary
amides, one carbamate, and four labile aryl chlorides) and provided the product 37 (56%) exclusively as
a single atropisomer (>20:1, detection limits) possessing the natural (R)-configuration. In this instance,
the complexity of the substrate and the reverse macrocyclization order did not diminish the atropodiaste-
reoselectivity; rather, it provided an improvement over the 4:1 selectivity that was observed with the
analogous substrate used to provide the isolated DEF ring system in our first-generation approach. Just
as significant, the atroposelectivity represents a complete reversal of the diasteroselectivity observed with
analogous macrocyclizations conducted using a Suzuki biaryl coupling.
Introduction
Complestatin (1, chloropeptin II) was first disclosed as an
inhibitor of the alternate pathway of human complement in
1980,1 although it was 9 years later that its original isolation2
from Streptomyces laVendulae was reported along with its partial
structure elucidation (connectivity and partial stereochemistry)
by Seto (Figure 1).3 Shortly thereafter, the first report of its
activity against HIV infectivity and its cytopathic effects was
disclosed.4 Later in 1994, Omura reported the isolation of
chloropeptin I (2) and chloropeptin II (1) from Streptomyces
sp. WK-3419 as inhibitors of HIV gp120-CD4 binding, estab-
lished that chloropeptin II and complestatin are identical, and
determined their partial stereochemistry.5 A subsequent detailed
NMR analysis provided the full structural and stereochemical
assignment for chloropeptin I (2) including that of the axial
Figure 1. Natural products.
atropisomer chirality.6 Additional isolations of the natural
products have been reported at Merck8 and Schering-Plough9
that have not only described related members of the family but
have also defined further insights into their biological properties
including their additional ability to inhibit HIV-1 integrase. In
(1) (a) Kaneko, I.; Fearon, D. T.; Austen, K. F. J. Immunol. 1980, 124,
1194–1198. (b) Tachikawa, K.; Hasumi, K.; Endo, A. Thromb.
Haemost. 1997, 77, 137–142. (c) Tachikawa, K.; Hasumi, K.; Endo,
A. Thrombosis Res. 1997, 87, 571–576.
(2) Kaneko, I.; Kamoshida, K.; Takahashi, S. J. Antibiot. 1989, 42, 236–
241.
(3) Seto, H.; Fujioka, T.; Furihata, K.; Kaneko, I.; Takahashi, S.
Tetrahedron Lett. 1989, 30, 4987–4990.
(4) Momota, K.; Kaneko, I.; Kimura, S.; Mitamura, K.; Shimada, K.
Biochem. Biophys. Res. Commun. 1991, 179, 243–250.
(5) (a) Matsuzaki, K.; Ikeda, H.; Ogino, T.; Matsumoto, A.; Woodruff,
H. B.; Tanaka, H.; Omura, S. J. Antibiot. 1994, 47, 1173–1174. (b)
Tanaka, H.; Matsuzaki, K.; Nakashima, H.; Ogino, T.; Matsumoto,
A.; Ikeda, H.; Woodruff, H. B.; Omura, S. J. Antibiot. 1997, 50, 58–
65. (c) Matsuzaki, K.; Ogino, T.; Sunazuka, T.; Tanaka, H.; Omura,
S. J. Antibiot. 1997, 50, 66–69.
(6) Gouda, H.; Matsuzaki, K.; Tanaka, H.; Hirono, S.; Omura, S.;
McCauley, J. A.; Sprengeler, P. A.; Furst, G. T.; Smith, A. B. J. Am.
Chem. Soc. 1996, 118, 13087–13088.
(7) (a) Jayasuriya, H.; Salituro, G. M.; Smith, S. K.; Heck, J. V.; Gould,
S. J.; Singh, S. B.; Homnick, C. F.; Holloway, M. K.; Pitzenberger,
S. M.; Patane, M. A. Tetrahedron Lett. 1998, 39, 2247–2248. (b)
Hegde, V. R.; Dai, P.; Patel, M.; Gullo, V. P. Tetrahedron Lett. 1998,
39, 5683–5684.
9
7776 J. AM. CHEM. SOC. 2010, 132, 7776–7783
10.1021/ja102304p  2010 American Chemical Society

Pd(0)-Mediated Indole Annulation for Macrocyclization
A R T I C L E S
these studies, an acid-catalyzed rearrangement (TFA, 50 °C,
>90%) of chloropeptin II (1, complestatin) to the less strained
chloropeptin I (2) was discovered1c and shown to proceed with
retention of the atropisomer stereochemistry, establishing the
full stereochemical assignments for 1.7
The chloropeptins have attracted considerable interest as a
result of their structural complexity and their equally important
HIV activity mediated through two complementary and unique
sites of action. In addition to the importance of their effects at
the two individual biological targets themselves, the fact that
they act at two sites suggests viral resistance to the agents would
be more difficult to emerge. Although structurally similar to
the glycopeptide antibiotics, one of the characteristic biaryl ether
linkages is replaced with a biaryl linkage to C6 or C7 of the
indole of a (R)-tryptophan embedded in the macrocyclic core
adopting a single atropisomer stereochemistry (R) that is not
capable of thermal interconversion. Snapper and Hoveyda
reported the first total synthesis of a member of this class of
natural products, chloropeptin I (2), confirming the structural
and stereochemical assignments.10 Their approach enlisted a late
stage intramolecular biaryl Stille coupling for closure of the
right-hand macrocyclic ring system (38-42%, three steps) and
provided exclusively the natural (R)-atropisomer of 2 when
conducted on substrate containing the intact left-hand macro-
cycle.10 Remarkably, it was later found that extending this
approach to chloropeptin II (1), using an analogous late-stage
intramolecular biaryl Suzuki coupling, provided exclusively
(63%) the unnatural (S)-atropisomer (isocomplestatin),11 whereas
closure of the isolated ring system proceeded in a non-
atroposelective manner (52%, 1:1 R:S). Repeating this work,
Zhu confirmed these observations again 3 years later and
suggested that the atropodiastereoselectivity of the biaryl ring
closure was dependent on the stereochemistry found in the
preformed left-hand ring system.12a Complementary to these and
related ongoing efforts,12,13 we recently reported the first total
synthesis of chloropeptin II (1, complestatin), the more strained
and challenging of the natural products.14 Intrinsic in the design
(8) (a) Singh, S. B.; Jayasuriya, H.; Salituro, G. M.; Zink, D. L.; Shafiee,
A.; Heimbuch, B.; Silverman, K. C.; Lingham, R. B.; Genilloud, O.;
Teran, A.; Vilella, D.; Felock, P.; Hazuda, D. J. Nat. Prod. 2001, 64,
874–882. (b) Singh, S. B.; Jayasuriya, H.; Hazuda, D. L.; Felock, P.;
Homnick, C. F.; Sardana, M.; Patane, M. A. Tetrahedron Lett. 1998,
39, 8769–8770.
Figure 2. First-generation approach and key macrocyclization reactions.14
(9) (a) Hegde, V. R.; Puar, M. S.; Dai, P. D.; Patel, M.; Gullo, V. P.;
Chan, T.-M.; Silver, J.; Pramanik, B. N.; Jenh, C.-H. Bioorg. Med.
Chem. Lett. 2003, 13, 573–575. (b) Hegde, V. R.; Puar, M. S.; Dai,
P.; Patel, M.; Gullo, V. P.; Pramanik, B. N.; Jenh, C.-H. Tetrahedron
Lett. 2002, 43, 5339–5341.
of the approach and by virtue of the single-step acid-catalyzed
conversion of 1 to chloropeptin I (2), the route also provided a
total synthesis of 2. Key to the approach was the development
of an intramolecular Larock indole synthesis15 for the initial
macrocyclization, adopting conditions that permit the use of a
2-bromoaniline16 and incorporating a large removable terminal
alkyne substituent (-SiEt3) to sterically dictate the indole
cyclization regioselectivity,15-17 Figure 2. Not only did this
cyclization provide the fully functionalized right-hand ring
system of 1 in excellent conversion (89%) and good atropodi-
astereoselectivity (4:1), but it also represented the first reported
(10) Deng, H.; Jung, J.-K.; Liu, T.; Kuntz, K. W.; Snapper, M. L.; Hoveyda,
A. H. J. Am. Chem. Soc. 2003, 125, 9032–9034.
(11) Shinohara, T.; Deng, H.; Snapper, M. L.; Hoveyda, A. H. J. Am. Chem.
Soc. 2005, 127, 7334–7336.
(12) (a) Jia, Y.; Bois-Choussy, M.; Zhu, J. Angew. Chem., Int. Ed. 2008,
47, 4167–4172. (b) Jia, Y.; Bois-Choussy, M.; Zhu, J. Org. Lett. 2007,
9, 2401–2404. (c) Yamada, Y.; Akiba, A.; Arima, S.; Okada, C.;
Yoshida, K.; Itou, F.; Kai, T.; Satou, T.; Takeda, K.; Harigaya, Y.
Chem. Pharm. Bull. Jpn. 2005, 53, 1277–1290. (d) Smith, A. B., III;
Chruma, J. J.; Han, Q.; Barbosa, J. Bioorg. Med. Chem. Lett. 2004,
14, 1697–1702. (e) Elder, A. M.; Rich, D. H. Org. Lett. 1999, 1, 1443–
1446. (f) Roussi, G.; Zamora, E. G.; Carbonnelle, A.-C.; Beugelmans,
R. Heterocycles 1999, 51, 2041–2063. (g) Kai, T.; Kajimoto, N.;
Konda, Y.; Harigaya, Y.; Takayanagi, H. Tetrahedron Lett. 1999, 40,
6289–6292. (h) Gurjar, M. K.; Tripathy, N. K. Tetrahedron Lett. 1997,
38, 2163–2166.
(15) (a) Larock, R. C.; Yum, E. K. J. Am. Chem. Soc. 1991, 113, 6689–
6690. (b) Larock, R. C.; Yum, E. K.; Refvik, M. D. J. Org. Chem.
1998, 63, 7652–7662.
(16) Shen, M.; Li, G.; Lu, B. Z.; Hossain, A.; Roschangar, F.; Farina, V.;
Senanayaka, C. H. Org. Lett. 2004, 6, 4129–4132.
(13) Wang, Z.; Bois-Choussy, M.; Jia, Y.; Zhu, J. Angew. Chem., Int. Ed.
2010, 49, 2018–2022.
(17) (a) Ma, C.; Liu, X.; Li, X.; Flippen-Anderson, J.; Yu, S.; Cook, J. M.
J. Org. Chem. 2001, 66, 4525–4542. (b) Jeschke, T.; Wensbo, D.;
Annby, U.; Gronowitz, S.; Cohen, L. A. Tetrahedron Lett. 1993, 34,
6471–6474.
(14) Garfunkle, J.; Kimball, F. S.; Trzupek, J. D.; Takazawa, S.; Shimamura,
H.; Tomishima, M.; Boger, D. L. J. Am. Chem. Soc. 2009, 131, 16036–
16038.
9
J. AM. CHEM. SOC. VOL. 132, NO. 22, 2010 7777

A R T I C L E S
Shimamura et al.
BPF) and was conducted under conditions developed by Farina
and Senanayake that permit the use of 2-bromoanilines.16 In
these early studies, we found that the low mass recovery of
material, problematic substrate epimerization, and competitive
reduction of the aryl chlorides observed when using the
recommended inorganic base additives (NaHCO3 > KOAc >
K2CO3, Na2CO3)16 could be minimized or eliminated with the
use of organic base additives of which Et3N (or i-Pr2NEt) proved
satisfactory. Still, at this early stage, the conversions were low,
and there was a slight preference for formation of the unnatural
(S)-atropisomer. Substantial improvements in the conversions
were observed by enlisting microwave heating to achieve higher
reaction temperatures (130 vs 110 °C, NMP). The free aniline
5 was found to participate more effectively than the correspond-
ing trifluoroacetamide 3 that underwent hydrolysis under the
reaction conditions, and the use of a 2-bromoaniline proved
better than the corresponding 2-iodoaniline (e.g., 3 vs 4). This
latter observation was and remains especially interesting in light
of the original Larock cyclization disclosures where only the
latter, but not the former, substrates participate.15 As we
proceeded to advance the product 6 in the first-generation total
synthesis, we found that the unprotected and strained indole in
6 often displayed a competitive reactivity that could be masked
by acetylation. Consequently, we examined the use of the
acetamide 7 directly in the macrocyclization reaction and found
that it displayed an improved atropodiastereoselectivity, but
suffered from lower conversions than observed with the free
aniline 5. Progressive reoptimization of the reaction conditions
for the acetamide substrate 7 revealed that the use of increasingly
less polar solvents (toluene/MeCN (1:1) > MeCN > DMF >
NMP) incrementally improved the conversions, allowed the use
of lower reaction temperatures, enhanced the atropodiastereo-
selectivity, and permitted a switch back to conventional thermal
versus microwave reaction conditions. This latter development
substantially impacted the progress in the total synthesis of 1
that had become scale limited by the microwave vessel volumes
and the dilute reaction concentrations (1-4 mM) used to
promote macrocyclization. Representative major observations
chronicling some these improvements, which transpired over a
period of years, are summarized in eq 1.
Figure 3. Alternative approach.
example of what will be a useful Larock macrocyclization
strategy. Subsequent introduction of the left-hand ring system,
enlisting an aromatic nucleophilic substitution reaction for ring
closure with biaryl ether formation completed the assemblage
of the bicyclic structure of 1, and represented a macrocyclization
order complementary to that used by Snapper and Hoveyda.10,11
The success with the Larock macrocyclization reaction for
assemblage of the right-hand ring system of 1 and its tailored
atropodiastereoselectivity for generation of the natural atropi-
somer raised the question of whether its utilization might
overcome or even reverse the unnatural atropodiastereoselec-
tivity observed with a Suzuki macrocyclization when applied
to the reversed macrocyclization order (Figure 3). Herein, we
report the results of the implementation of this second, alterna-
tive total synthesis of complestatin (1, chloropeptin II), estab-
lishing the ramifications of reversing the order of the macro-
cyclizations, and key details of the development of the Larock
macrocyclization reaction.
Results and Discussion
Larock Macrocyclization. Early in our efforts and because
of the strain in the right-hand ring systems, we explored
approaches that would assemble the five-membered ring of the
indole during or following macrocyclization. Because com-
plestatin (1, chloropeptin II) can be directly converted to
chloropeptin I (2), we focused our efforts on the synthesis of 1
even though the strain in its right-hand ring system is greater.
A number of approaches were examined that included free
radical or Heck alkene addition reactions as well as lactamiza-
tions for macrocyclization with derigidified substrates, none of
which led to productive approaches.18 In the course of these
exploratory studies, we found that an intramolecular Larock
indole synthesis was capable of both providing the macrocy-
clization and forming the five-membered ring of the indole
simultaneously, albeit in initial modest conversions. Early studies
revealed that the aryl bromides 3 and 5, containing from the
onset a large triethylsilyl group at the alkyne terminus to
sterically dictate the cyclization regioselectivity,15 participated
in the intramolecular Larock cyclization promoted by Pd(OAc)2.
The reaction benefited from the use of the electron-rich bidentate
phosphine ligand 1,1′-bis(di-tert-butylphosphino)ferrocene (Dt-
In the interim since the completion of our first-generation
total synthesis,14 the impact of the aniline substituent on the
(18) Trzupek, J. D. Ph.D. Dissertation, 2006.
9
7778 J. AM. CHEM. SOC. VOL. 132, NO. 22, 2010

Pd(0)-Mediated Indole Annulation for Macrocyclization
A R T I C L E S
atropodiastereoselectivty of the macrocyclization reaction has
been further investigated by enlisting a series of substrates
prepared by single-step acylation of 5 (eq 2). Without optimiza-
tion of the individual reactions, a smooth trend of increasing
atropodiastereoselectivity is observed with both the nature of
the acyl group (COR > CO2R) and most notably as the size of
the aniline acyl group increases with the reactions proceeding
in conversions that approach those optimized for 7. Significantly,
the macrocyclization reactions of 11 (6.5:1 R:S) and especially
that of 13 (>20:1 R:S, detection limits) proceed with atropose-
lectivities that exceed what we reported with 7 (4:1 R:S)14
indicating that improvements in our first-generation total
synthesis of 1 are possible. Although speculative and based only
on the simple examination of molecular models, the aniline
amine and its attached acyl substituent appear to be directed
toward and lie over the peptide chain linking the reacting
partners in intermediates leading to the unnatural (S)-atropiso-
mer, whereas they are distal and directed away from the linking
peptide chain in intermediates leading to the natural (R)-
atropisomer. Thus, in retrospect it may not be surprising that
the atroposelectivity of the macrocyclization increases with the
steric size of the aniline substituent. The stereochemical
assignments were confirmed by extensive NMR analysis and,
as first defined by Hoveyda,10 the most recognizable and
diagnostic distinctions in the atropisomers that can be used
confidently to assign the steroechemistry are the NMR chemical
shifts, multiplicity, and NOE’s for Trp R-CH (CDCl3 or acetone-
d6: δ 3.54-3.71, t for R vs δ 5.01-5.07, m for S) and the
diastereotopic Trp ꢀ-CH2 (δ 2.96-3.14, d and 3.59-3.71, dd
for R vs δ 3.22-3.31, dd and 3.31-3.55, dd for S), both of
which are altered considerably relative to the starting precursors
(δ 4.19-4.31, m and 2.65-2.28, m, respectively). Similarly,
the E-ring R-CH (CDCl3: δ 5.25-5.29, d for R vs δ 4.57-4.80,
d for S) and the D-ring aryl CH that lies on the internal face of
the macrocycle (CDCl3: δ 5.43-5.57, s for R vs δ 5.13-5.20,
s for S) exhibit chemical shifts that are diagnostic of the
atropisomer stereochemistry.
of additional products. Clean protodesilylation of (R)-6 was
accomplished (eq 3), but required more controlled reaction
conditions (HF-pyr, THF-pyr, 25 °C, 2 h) for the clean
generation of (R)-15 (90%). As a consequence, work progressing
forward in our first-generation total synthesis, as well as herein,
was only conducted with intermediates bearing the natural (R)-
atropisomer stereochemistry. Additionally, and with the two
atropisomers (R)- and (S)-15 in hand, we were able to further
confirm the stereochemical assignments and establish for
ourselves that, as previously reported,11 they are not capable of
thermal interconversion.
Analogous to complestatin itself,7 the isolated DEFG ring
system of complestatin containing the free indole undergoes the
strain-releasing TFA-catalyzed rearrangement (50 °C, 15 min,
>98%) providing the C7 versus C6 indole biaryl link charac-
teristic of chloropeptin I, and this occurs with retention of the
atropisomer stereochemistry.11 Thus, it was of interest to
establish the behavior of the corresponding indole acetamide
derivatives 16a and16b toward this same rearrangement (Scheme
1). In the course of our work, we came to appreciate and
exploited the fact that acetylation of the indole tamed its
electrophilic reactivity providing intermediates that exhibited a
robust stability. Consistent with these observations but not
investigated in detail, we found that exposure of either 16a or16b
to TFA at room temperature had no effect on 16 (24 h), that no
rearrangement of either 16a or16b was observed at 50 °C over
2.5-5 h, and that longer exposures of 16a or16b to TFA at 50
°C (24 h) led to their slow consumption (decomposition) without
evidence of rearrangement (Scheme 1). By comparison, the
corresponding free indole 18, resulting from deacylation of 16a
with LiOH (96%), underwent rearrangement to 19 within 10
min at 50 °C in TFA (85%, Scheme 1). Thus, indole N-acylation
has a pronounced impact on the strained indole reactivity, and
its importance should not be underestimated.
Reactivity and Rearrangement of the DEF Ring System. In
initial studies, significant differences in the reactivity of the two
atropisomers of 6, bearing the free indole, were observed
indicating that work with the unnatural (S)-isomer might not
translate into analogous successful reactions with the natural
(R)-isomer. For example, acid-catalyzed removal of the TES
substituent with (S)-6 proved straightforward (5 equiv of TsOH,
DMF, 25 °C, 20 h) providing (S)-15 (96%) without event or
optimization (eq 3), whereas the analogous treatment of (R)-6
provided only minor amounts of (R)-15 along with a number
Second-Generation Total Synthesis of Complestatin. The
central D subunit was prepared following small modifications
in the approach previously detailed,14 allowing for its global
deprotection prior to incorporation into the route. Thus, asym-
metric aminohydroxylation [(DHQD)2PHAL]19 of the styrene
derived from 3-iodo-4,5-dimethoxybenzaldehyde (Ph3PdCH2,
99%) produced the aminoalcohol 20 in good yield (75%),
(19) Reddy, K. L.; Sharpless, K. B. J. Am. Chem. Soc. 1998, 120, 1207–
1217.
9
J. AM. CHEM. SOC. VOL. 132, NO. 22, 2010 7779

A R T I C L E S
Shimamura et al.
Scheme 1
Scheme 2
fluoro-3-nitrophenylalanine subunit central to the tripeptide 28.20
This set the stage for the first key macrocyclization reaction21
for formation of the 16-membered biaryl ether (Scheme 3).
Treatment of 29 with K2CO3 (10 equiv) in anhydrous THF (1
mM, 60 °C, 12 h) in the presence of 4 Å molecular sieves and
18-crown-6 proceeded exceptionally well, providing 30 in yields
as high as 95% as an inconsequential 5:1 mixture of atropiso-
mers. This represents an improvement over our first-generation
approach (81%, essentially a single atropisomer)14 where the
ABCD ring closure was conducted on a substrate containing
the intact DEF ring system. A trace of a separable product,
constituting the epimer at the terminal 4-hydroxyphenylglycine,
was detected, isolated (<5%), and characterized and represents
the only indication of competitive epimerization of the substrate
or product under these reaction conditions. Likely contributing
to this lack of epimerization at the C-terminus methyl ester is
the conduct of the cyclization with the A-ring free phenol that
is deprotonated under the reaction conditions. The use of DMSO
as solvent (1 mM, 25 °C, 36 h), while permitting the reaction
to proceed at room temperature (71% yield) and providing an
analogous 5:1 mixture of atropisomers, led to more significant
epimerization at this center (>10%). Moreover and like observa-
tions made in our first-generation approach, no products were
detected resulting from competitive participation of the ad-
ditional three unprotected phenols in nucleophilic substitution
reactions with the o-fluoronitroaromatic that would provide
either the more strained or more hindered 14-membered or 17-
membered biaryl ethers.
(20) Enzymatic hydrolysis (resolution) of readily available racemic N-acetyl-
4-fluoro-3-nitrophenylalanine using Aspergillus acylase I following a
procedure detailed for its isomer N-acetyl-3-fluoro-4-nitrophenylalanine
provided (S)-4-fluoro-3-nitrophenylalanine (>40%, >95-99% ee, >3
g scale). See: Groll, M.; Gotz, M.; Kaiser, M.; Weyher, E.; Moroder,
L. Chem. Biol 2006, 13, 607–614. Details are provided in the Sup-
porting Information.
(21) (a) Boger, D. L. Med. Res. ReV. 2001, 21, 356–381. (b) Crowley, B. M.;
Boger, D. L. J. Am. Chem. Soc. 2006, 128, 2885–2892. (c) Boger,
D. L.; Kim, S. H.; Mori, Y.; Weng, J.-H.; Rogel, O.; Castle, S. L.;
McAtee, J. J. J. Am. Chem. Soc. 2001, 123, 1862–1871. (d) Boger,
D. L.; Kim, S. H.; Miyazaki, S.; Strittmatter, H.; Weng, J.-H.; Mori,
Y.; Rogel, O.; Castle, S. L.; McAtee, J. J. J. Am. Chem. Soc. 2000,
122, 7416–7417. (e) Boger, D. L.; Miyazaki, S.; Kim, S. H.; Wu,
J. H.; Loiseleur, O.; Castle, S. L. J. Am. Chem. Soc. 1999, 121, 3226–
3227. (f) Boger, D. L.; Miyazaki, S.; Kim, S. H.; Wu, J. H.; Castle,
S. L.; Loiseleur, O.; Jin, Q. J. Am. Chem. Soc. 1999, 121, 10004–
10011. (g) Boger, D. L.; Castle, S. L.; Miyazaki, S.; Wu, J. H.; Beresis,
R. T.; Loiseleur, O. J. Org. Chem. 1999, 64, 70–80. (h) Boger, D. L.;
Borzilleri, R. M.; Nukui, S.; Beresis, R. T. J. Org. Chem. 1997, 62,
4721–4736. (i) Boger, D. L.; Borzilleri, R. M.; Nukui, S. Bioorg. Med.
Chem. Lett. 1995, 5, 3091–3096. (j) Boger, D. L.; Borzilleri, R. M.
Bioorg. Med. Chem. Lett. 1995, 5, 1187–1190. (k) Boger, D. L.;
Nomoto, Y.; Teegarden, B. R. J. Org. Chem. 1993, 58, 1425–1433.
regioselectivity (5:1), and enantioslectivity (>98% ee) (Scheme
2).14 Protection of the primary alcohol as its TBS ether 21 (94%)
followed by conversion of the aryl iodide to the corresponding
aryl boronic acid and its Suzuki coupling with 2-bromo-5-
iodoaniline14 (22, 99%, two steps) provided 23. Aniline acety-
lation (97%), deprotection of 24 to provide the primary alcohol
25 (Bu4NF, 97%), and its single-step oxidation (TEMPO/
NaOCl) to the carboxylic acid provided 26 (90%). Deprotection
of the aryl methyl ethers (BBr3) under conditions that also
remove the Boc group was followed by its reinstallation (Boc2O)
providing 27 (74%).
Coupling (EDCI/HOAt, THF, 0 °C) of 27 with the tripeptide
2814 provided 29 in good yield (73%). Although not essential
for the progression of the work, we also took this opportunity
to improve the synthesis (scalability) of the (S)-N-methyl-4-
9
7780 J. AM. CHEM. SOC. VOL. 132, NO. 22, 2010

Pd(0)-Mediated Indole Annulation for Macrocyclization
A R T I C L E S
Scheme 3
Figure 4. Diagnostic spectroscopic NOE’s.
Extensive NMR analysis of each atropisomer of 30 indicate
that both adopt a conformation containing a cis N-methyl amide
at the site linking the B- and C-residues, and that the major
product constitutes the (M)-atropisomer, whereas the minor
isomer possesses the (P)-configuration (Figure 4). Given the
ease and conversion for the cyclization of 29 relative to related
isolated 16-membered biaryl ethers that we have examined,21
it appears that the N-methyl amide facilitates ring closure
perhaps by promoting the adoption of more compact versus
extended conformations conducive to cyclization. In this regard,
it is interesting to note that the ABCD ring system lacking the
terminal 4-hydroxyphenylglycine residue (A ring) does not adopt
this cis N-methyl amide conformation and the corresponding
acyclic precursor would not be expected to cyclize with such
ease.12a Individually or as the atropisomer mixture, 30 was
subjected to a two-step procedure22 for removal of the nitro
group (SnCl2; t-BuONO, H3PO2, THF), providing 32 in excellent
conversions (90%, two steps, Scheme 3). Like 30, the fully
functionalized ABCD ring system 32 common to both com-
plestatin I and II was found by NMR to adopt a conformation
bearing a cis N-methyl amide linking the B- and C-residues
(Figure 4).
With 32 in hand, this set the stage for examination of the
key Larock macrocyclization reaction. N-Boc deprotection (4
N HCl in dioxane) and coupling (HATU, 2,6-lutidine, CH2Cl2/
THF) of the crude amine with (R)-FmocHN-3,5-Cl2-Hpg14 (33)
provided 34 in good yield (96%) under conditions devised to
completely suppress epimerization (Scheme 3). Fmoc depro-
tection (Bu4NF)23 followed by coupling of the liberated amine
with the alkyne bearing amino acid 3514,17 provided the required
(22) Crowley, B. M.; Mori, Y.; McComas, C. C.; Tang, D.; Boger, D. L.
J. Am. Chem. Soc. 2004, 126, 4310–4317.
(23) (a) Jiang, W.; Wanner, J.; Lee, R. J.; Bounaud, P.-Y.; Boger, D. L.
J. Am. Chem. Soc. 2002, 124, 5288–5290. (b) Jiang, W.; Wanner, J.;
Lee, R. J.; Bounaud, P.-Y.; Boger, D. L. J. Am. Chem. Soc. 2003,
125, 1877–1887.
9
J. AM. CHEM. SOC. VOL. 132, NO. 22, 2010 7781

A R T I C L E S
Shimamura et al.
Figure 5. Key spectroscopic NOE’s (THF-d8).
substrate 36 in quantitative yield. Unlike our first-generation
approach or the examples reported in eq 2, the macrocyclization
of 36 is conducted with a substrate now containing four
unprotected phenols (two of which are especially acidic), five
secondary amides, one carbamate, and four labile aryl chlorides.
Without optimization, exposure of 36, bearing the aniline
acetamide, to the conditions developed for conducting the
intramolecular Larock cyclization (1.3 equiv of Pd(OAc)2, 1.3
equiv of DtBPF, 5.0 equiv of Et3N, 1 mM in toluene/MeCN
1:1, 110 °C, 2 h) provided 37 (56%) exclusively as a single
atropisomer (>20:1, detection limits) possessing the natural (R)-
configuration. Thus, the complexity of the substrate and the
reverse macrocyclization order did not diminish the atropodi-
astereoselectivity, rather it provided an improvement over the
4:1 selectivity that we observed with the analogous acetamide
substrate 7 providing the isolated DEF ring system.14 More
impressively, it represents a complete reversal of the atropodi-
asteroselectivity observed with the Suzuki coupling first de-
scribed by Snapper and Hoveyda11 and later confirmed by
Zhu.12a Although limited efforts have been made to reduce this
reaction to one that is catalytic in palladium, which have been
only modestly successful providing reactions that proceed in
less dependable conversions, we note that even the efforts
enlisting a Stille or Suzuki coupling for macrocyclization of
the DEF ring systems have also employed stoichiometric
palladium to avoid substrate or product (peptide) sequestration
of the reagent.10,11,13 Extensive NMR characterization of 37
confirmed the atropisomer configuration exhibiting the charac-
teristic chemical shifts, multiplicity, and NOE’s for Trp R-CH
(THF-d8: δ 3.62, t), the diastereotopic Trp ꢀ-CH2 (δ 3.02, d
and 3.67, dd), the E-ring R-CH (δ 5.39, d) and the D-ring aryl
CH that lies on the internal face of the macrocycle oriented
toward the indole (δ 5.40, s) that are diagnostic of the
atropisomer stereochemistry (Figure 5). Acid-catalyzed removal
of the indole C2-triethylsilyl substituent (2 N HCl-dioxane/THF,
0 °C, 5 min) under conditions that were found to also partially
deprotect the N-terminus Boc group (∼30%), followed by its
reinstallation (Boc2O) provided 38 (84% overall) identical in
all respects to the material prepared in our first-generation total
synthesis.14 Subsequent deprotection of 38 or more direct
deprotection of 37 (4 N HCl in dioxane, 25 °C, 1 h), followed
by acylation of the N-terminus free amine with 2-(3,5-dichloro-
4-hydroxyphenyl)-2-oxoacetic acid (39, EDCI/HOAt, 0 °C, 2 h)
provided our prior penultimate precursor 4014 (60%) to com-
plestatin. As detailed earlier,14 deprotection of 40 was ac-
complished with LiOH (THF/H2O, 0 °C, 3 h, 60%) in a reaction
where the indole N-acetyl group is removed faster (<30 min)
than the methyl ester and provided complestatin (1, chloropeptin
Figure 6. Key elements of the first- and second-generation total syntheses
of complestatin.
II), and its subsequent acid-catalyzed (50% TFA/H2O, 50 °C,
5 h) rearrangement provided chloropeptin 1 (2).24
Conclusions
Herein, we disclose full details of the development of a
powerful intramolecular Larock indole annulation for macro-
cyclization closure of the strained 16-membered biaryl ring
system found in complestatin (1, chloropeptin II),25 the definition
of factors impacting its intrinsic atroposelectivity, and its
examination in a second-generation, alternative total synthesis
of complestatin. The examination of this alternative approach,
in which the order of the macrocyclization reactions is reversed
and the Larock cyclization is conducted with a substrate (36)
containing four unprotected phenols, five secondary amides, one
carbamate, and four labile aryl chlorides, provided the product
37 (56%) exclusively as a single atropisomer (>20:1, detection
limits) possessing the natural (R)-configuration. Thus, the
complexity of this substrate or the reverse macrocyclization
order not only did not diminish the atropodiastereoselectivity,
but it also provided an improvement over the 4:1 selectivity
that we observed with the acetamide substrate 7 used to provide
the DEF ring system in our first-generation approach (Figure
6).14 More impressively, it represents a complete reversal of
(24) Abbreviations: DMF, N,N-dimethylformamide; DMSO, dimethylsul-
foxide; DtBPF, 1,1′-bis(di-tert-butylphosphino)ferrocene; EDCI, 1-[3-
(dimethylamino)propyl]-3-ethylcarbodiimide hydrochloride; HATU,
O-(7-azabenzotriazol-1-yl)-N,N,N′,N′-tetramethyluronium hexafluo-
rophosphate; HOAt, 1-hydroxy-7-azabenzotriazole; NMP, N-meth-
ylpyrrolidinone; TBS, tert-butyldimethylsilyl; TES, triethylsilyl; TFA,
trifluoroacetic acid; THF, tetrahydrofuran.
(25) For complementary palladium-catalyzed reactions introduced in the
course of our natural products total synthesis efforts, see: (a) Boger,
D. L.; Panek, J. S. Tetrahedron Lett. 1984, 25, 3175–3178. (b) Boger,
D. L.; Panek, J. S.; Duff, S. R.; Yasuda, M. J. Org. Chem. 1985, 50,
5782–5789, and 5790-5795. (c) Boger, D. L.; Patel, M. Tetrahedron
Lett. 1987, 28, 2499–2502. (d) Boger, D. L.; Patel, M. J. Org. Chem.
1988, 53, 1405–1415.
9
7782 J. AM. CHEM. SOC. VOL. 132, NO. 22, 2010

Pd(0)-Mediated Indole Annulation for Macrocyclization
A R T I C L E S
tripeptide 28, M. Tichenor for introducing improvements in the
Larock macrocyclization and subsequent elaboration to the DEF
ring system, Dr. J. Cottell for initiating studies on the ABCD ring
system, and D. S. Lewy, Drs. A. Pichota, L. Resnick, and W. Han
for exploration of early-stage routes to the DEF ring system in
which the syntheses of the amino acid subunits were first developed.
J.G. and J.D.T. were Skaggs fellows.
the atropodiasteroselectivity observed with analogous macro-
cyclizations conducted using a Suzuki coupling.11,12a Extensions
of these studies to natural products related to the chloropeptins,
the preparation of key analogues, and further studies on the
Larock macrocyclization are in progress and will be disclosed
in due course.
Acknowledgment. In memory of David S. Lewy. We gratefully
acknowledge the financial support of the National Institutes of
Health (CA041101), the Skaggs Institute for Chemical Biology,
and fellowship support from the National Science Foundation
(S.P.B.). We wish to thank Dr. S. Takizawa for the many
innovations introduced en route to the first-generation total syn-
thesis, Dr. M. Tomishima for large-scale preparations of the
Supporting Information Available: Full experimental details.
This material is available free of charge via the Internet at http://
pubs.acs.org.
JA102304P
9
J. AM. CHEM. SOC. VOL. 132, NO. 22, 2010 7783