New insights into the mechanism and an expanded scope of the Fe(III)-mediated vinblastine coupling reaction

Article abstract of DOI:10.1021/ja306229x

A definition of the scope of aromatic substrates that participate with catharanthine in an Fe(III)-mediated coupling reaction, an examination of the key structural features of catharanthine required for participation in the reaction, and the development of a generalized indole functionalization reaction that bears little structural relationship to catharanthine itself are detailed. In addition to providing insights into the mechanism of the Fe(III)-mediated coupling reaction of catharanthine with vindoline suggesting the reaction conducted in acidic aqueous buffer may be radical mediated, the studies provide new opportunities for the preparation of previously inaccessible vinblastine analogs and define powerful new methodology for the synthesis of indole-containing natural and unnatural products.

Full text of DOI:10.1021/ja306229x

Communication
pubs.acs.org/JACS
New Insights into the Mechanism and an Expanded Scope of the
Fe(III)-Mediated Vinblastine Coupling Reaction
Hiroaki Gotoh,† Justin E. Sears,† Albert Eschenmoser,†,‡ 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, United States
‡Laboratory of Organic Chemistry, ETH Zurich, Honggerberg Wolfgang-Pauli-Strasse 10, CHI H309, CH-8093, Zurich, Switzerland
̈
̈
̈
S
* Supporting Information
ABSTRACT: A definition of the scope of aromatic
substrates that participate with catharanthine in an
Fe(III)-mediated coupling reaction, an examination of
the key structural features of catharanthine required for
participation in the reaction, and the development of a
generalized indole functionalization reaction that bears
little structural relationship to catharanthine itself are
detailed. In addition to providing insights into the
mechanism of the Fe(III)-mediated coupling reaction of
catharanthine with vindoline suggesting the reaction
conducted in acidic aqueous buffer may be radical
mediated, the studies provide new opportunities for the
preparation of previously inaccessible vinblastine analogs
and define powerful new methodology for the synthesis of
indole-containing natural and unnatural products.
inblastine (1) is the most widely recognized member of the
V
Vinca alkaloids as a result of its clinical use as an antitumor
drug (Figure 1).1,2 Originally isolated in trace amounts from
Cantharanthus roseus,1 its biological activity was among the first
to be shown to arise from inhibition of microtubule formation
and mitosis that is still regarded as one of the more successful
oncology drug targets.2−4 Complementary to earlier efforts,4 we
recently reported the total synthesis of vinblastine and its
unnatural enantiomer5 enlisting a biomimetic Fe(III)-promoted
coupling of vindoline (3) with catharanthine (4) and its
extension to the total synthesis of a series of related natural
products including vincristine (2) and key analogs,6,7 Figure 1.
Although insights into this coupling reaction have been
disclosed in earlier studies since its introduction by Kutney,8
there are features of its mechanism that are not well understood.
The Fe(III)-promoted coupling reaction displays a superb
stereochemical selectivity, producing exclusively the natural C16′
diastereomer at 25 °C (90%, Figure 1), compared to the
traditional Polonovski fragmentation9−11 (5:1, −78 °C and 1:1, 0
°C, eq 1) or chloroindolenine-based couplings.12,13 The lack of a
Figure 1. Natural products and Fe(III)-promoted coupling.
the recovery of catharanthine from the reaction mixture in the
absence of vindoline (2 h, NaBH4 workup), and the different
substrate scope of the reactions6,9 indicate that the azabenzful-
vene intermediate central to the Polonovski fragmentation and
related couplings may not be a participant in the Fe(III)-
promoted coupling reaction as previously assumed8 (eq 1).
Herein, we disclose studies conducted to probe the scope and
reaction mechanism of the Fe(III)-promoted coupling reaction.
They were undertaken inspired by the hypothesis that the
Fe(III)-promoted coupling reaction conducted in aqueous buffer
may be radical mediated. In addition to new experimental
observations that may be interpreted to support such a view, the
studies demonstrate new opportunities for the preparation of
previously inaccessible vinblastine analogs, represent powerful
nucleophilic solvent (H2O) trap of reactive intermediates under
the aqueous buffer conditions of the Fe(III)-promoted reaction,
Received: June 26, 2012
Published: August 2, 2012
© 2012 American Chemical Society
13240
dx.doi.org/10.1021/ja306229x | J. Am. Chem. Soc. 2012, 134, 13240−13243

Journal of the American Chemical Society
Communication
methodology for the synthesis of indole-containing natural
products, and provide new insights into the biosynthesis of the
natural products.
examined to establish whether the tertiary amino group was in
fact required, as well as probe whether the reportedly important
indole free NH8c was critical for the coupling reaction (Figure 3).
To our knowledge, no systematic study of the Fe(III)-
promoted coupling of catharanthine with simpler substrates has
been reported.14 As a result, the Fe(III)-promoted reaction of
catharanthine (4) with candidate aromatic substrates was
examined (Figure 2). Whereas strongly electron-rich aromatic
Figure 3. Fe(III)-promoted coupling reactions of simpler indoles.
Notably, the indole NH would be required of a reaction
proceeding through a neutral azabenzfulvene and protonation of
the tertiary amino group under the reaction conditions may
preclude its otherwise competitive single electron oxidation.
Impressively and contrary to the expectations based on prior
mechanistic proposals, a range of simplified indoles were found
to participate in an Fe(III)-mediated coupling reaction with
vindoline (Figure 3). Moreover, each of the substrates 7a−7d
was also found to couple effectively with N,N-dimethyl-3-
methoxyaniline, establishing the generality of the observations.
Thus, the removal of the tertiary amino group as well as indole N-
methylation did not preclude coupling with vindoline. In
contrast, indole N-acylation (e.g., N-Boc and N-CO2Me
derivatives, not shown) did prevent participation in the coupling
reaction. These results indicate that an Fe(III)-mediated indole
oxidation15 is sufficient for the coupling reaction to occur and
reveal that the reaction does not require the presence of the
indole free NH8c or the participation of the catharanthine tertiary
amino group.8 Since the reaction no longer requires the
reduction of the catharanthine-derived iminium ion in the
reaction product, NaBH4 is no longer required for reductive
workup of the reactions. In these cases, the reaction typically
provided a mixture of diastereomers, confirming that C16′
stereochemistry set in the vinblastine Fe(III)-promoted coupling
reaction is principally controlled by catharanthine.
Figure 2. Catharanthine coupling with aromatic substrates.
substrates were found to participate in the coupling reaction
effectively, less electron-rich substrates (anisole, thioanisole, 3-
methoxythioanisole), neutral aromatic substrates (benzene and
thiophene), and a series of electron-deficient aromatic substrates
(e.g., methyl benzoate) failed to couple with catharanthine. In
each instance of a reaction, the coupling provided a single
diastereomer, possessing the natural C16′ stereochemistry found
in vinblastine. This observation with simple substrates indicates
that the stereochemistry of the Fe(III)-promoted coupling
reaction can be controlled by catharanthine alone and may be
independent of the structure of its reaction partner (e.g.,
vindoline). Moreover, each successful reaction produced a single
product regioisomer that corresponds to the site of aromatic
substitution by electrophiles or radicals. Although the scope of
the more traditional Potier/Kutney−Polonovski fragmentation
coupling has not been as extensively examined, it is reported not
only to display an analogous regioselectivity and a similar
requirement for electron-rich aromatic substrates but also to
exhibit a temperature-dependent stereochemical outcome,
typically producing predominately the natural C16′ stereo-
chemistry at low temperatures (−40 °C) and the unnatural C16′
stereochemistry at higher temperatures (0 °C).9−11
In order to probe whether the initial oxidation of catharanthine
might occur within the indole15 and not require oxidation of the
tertiary amino group as initially suggested by Kutney and
subsequently by others,8 we first found that tertiary amine
additives to the coupling reaction, including N-Boc-catharan-
thine (5 equiv) and quinuclidine (5 equiv), do not impact the
conversions. As a result, a series of simplified substrates were
The generality of the reaction was established further by
examining the coupling of 7a with the series of aromatic
substrates that react with catharanthine (Figure 4). The indole 7a
displayed the identical reactivity pattern observed with
catharanthine itself, coupling with the strongly electron-rich
aromatic substrates, but failing to react with anisole, 3-
methoxythioanisole, benzene, or methyl benzoate. Significantly,
13241
dx.doi.org/10.1021/ja306229x | J. Am. Chem. Soc. 2012, 134, 13240−13243

Journal of the American Chemical Society
Communication
Figure 4. Fe(III)-promoted couplings of 7a.
these studies illustrate the unexpected generality of the reaction
to substrates that bear no close structural resemblance to
catharanthine, providing powerful methodology for the synthesis
of indole-containing products structurally distinct from vinblas-
tine. The identical substrate scope for the reactions of 7a and
catharanthine suggest that their couplings may proceed by
analogous reaction mechanisms. Like those of catharanthine, the
reactions of 7 are conducted at rt in acidic aqueous solution (0.05
N aq HCl/CF3CH2OH) without competitive nucleophilic
solvent (H2O) or counteranion (Cl−) participation, indicating
that the reactions are unlikely to proceed through an indole-
derived, neutral, or cationic azabenzfulvene.
We suggest this entails coupling of the electrophilic radical
generated by a single electron oxidation of the indole (eqs 2 and
3). Consistent with this formulation, we disclosed studies that
established the required electron-withdrawing properties of the
catharanthine C16′ substituent as well as the surprising impact of
C10′ indole substituents on its participation in the Fe(III)-
mediated coupling reaction.16 Consistent with an initiating single
electron indole oxidation, the latter studies revealed that
catharanthine C10′ electron-donating and neutral substituents
support the reaction, whereas electron-withdrawing substituents
progressively diminish both the rate and yield of the coupling
with vindoline. In further support of the direct coupling of the
initial single electron indole oxidation intermediate, we have now
confirmed that vindoline is susceptible to aromatic substitution
by electrophilic carbon radicals. Treatment of vindoline with the
radical derived from either ICH2CO2Et (Et3B/air, DMSO or
CF3CH2OH, 25 °C, 41%)17 or BrCH(CO2Et)2 (Na2S2O4, 1:1
CH3CN/H2O, 25 °C, 56%)18 provided the aromatic substitution
products arising from addition of the electrophilic radicals to the
vindoline aromatic ring followed by air oxidation for rear-
omatization (eq 4). This electrophilic radical substitution
reaction of vindoline in protic solvents suggests that the
Fe(III)-promoted reactions of catharanthine conducted in
water may be radical mediated, rather than involve an
intermediate electrophilic azabenzfulvene as previously sug-
gested for the Fe(III)-promoted reaction8 in analogy to other
coupling methods conducted in aprotic solvents.9−13 The
postulated coupling agent, the fragmented cation radical B,
illustrated in eq 3 is both captodatively stabilized19 and proposed
to be reversibly generated to account for the recovery of
catharanthine in the absence of vindoline (2 h) following a
reductive workup. In contrast, a committed azabenzfulvene
intermediate (see eq 1) is not expected to revert back to
catharanthine upon reductive workup. In the fragmented radical
B, an intramolecular one-electron-two-center bonding inter-
action between the radical site α to the carbomethoxy group and
the iminium carbon may stabilize a conformation in which the
upper face of the radical is sterically blocked, thus accounting for
the remarkably high stereoselectivity of the coupling reaction
even when conducted at rt. Combined, this proposal suggests
that the initial indole radical cation A and its fragmented cation
radical B may possess a unique blend of stability, persistence,
electrophilic character, and conformational properties to
effectively react selectively with vindoline.
In addition to the mechanistic implications on the Fe(III)-
mediated vinblastine coupling reaction, the synthetic possibilities
presented by the reactions of 7 are substantial, representing
methodology for quaternary center generation adjacent to indole
by a reaction mediated by Fe(III) and conducted in aqueous
buffer at rt. The scope of the reaction is such that vinblastine
analogs containing deep-seated changes to the upper velban-
amine subunit not previously envisioned may now be accessed.
To explore this possibility, the Fe(III)-mediated coupling of 7
was extended to the preparation of a series of increasingly more
complex vinblastine analogs that contain changes in the upper
velbanamine subunit (Figure 5). In each case, the coupling
precursors 11a−17a were prepared as racemic mixtures and
chromatographically resolved (ChiralCel OD or AD) prior to
coupling each individual enantiomer with vindoline (one
enantiomer shown). Without optimization, the reaction of
substrates that possess a substituent adjacent to the reacting
center (11a−15a) proceeded selectively, generating a single
diastereomer resulting from coupling on the sterically more
13242
dx.doi.org/10.1021/ja306229x | J. Am. Chem. Soc. 2012, 134, 13240−13243

Journal of the American Chemical Society
Communication
AUTHOR INFORMATION
Corresponding Author
boger@scripps.edu
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We gratefully acknowledge the support of the National Institute
of Health (CA042056, CA115526). J.E.S. is a Skaggs Fellow.
■
REFERENCES
■
(1) (a) Noble, R. L.; Beer, C. T.; Cutts, J. H. Ann. N.Y. Acad. Sci. 1958,
76, 882. (b) Svoboda, G. H.; Nuess, N.; Gorman, M. J. Am. Pharm. Assoc.
Sci. Ed. 1959, 48, 659.
(2) (a) Neuss, N.; Neuss, M. N. In The Alkaloids; Brossi, A., Suffness,
M., Eds.; Academic: San Diego, 1990; Vol. 37, p 229. (b) Jordan, M. A.;
Wilson, L. Nat. Rev. Cancer 2004, 4, 253.
(3) (a) Pearce, H. L. In The Alkaloids; Brossi, A., Suffness, M., Eds.;
Academic: San Diego, 1990; Vol. 37; p 145. (b) Borman, L. S.; Kuehne,
M. E. In The Alkaloids; Brossi, A., Suffness, M., Eds.; Academic: San
Diego, 1990; Vol. 37, p 144.
(4) (a) Kuehne, M. E.; Marko, I. In The Alkaloids; Brossi, A., Suffness,
M., Eds.; Academic: San Diego, 1990; Vol. 37, p 131. (b) Potier, P. J.
Nat. Prod. 1980, 43, 72. (c) Kutney, J. P. Acc. Chem. Res. 1993, 26, 559.
(d) Yokoshima, S.; Ueda, T.; Kobayashi, S.; Sato, A.; Kuboyama, T.;
Tokuyama, H.; Fukuyama, T. J. Am. Chem. Soc. 2002, 124, 2137.
(5) Ishikawa, H.; Colby, D. A.; Boger, D. L. J. Am. Chem. Soc. 2008, 130,
420.
Figure 5. Previously inaccessible vinblastine analogs.
(6) Ishikawa, H.; Colby, D. A.; Seto, S.; Va, P.; Tam, A.; Kakei, H.; Rayl,
T. J.; Hwang, I.; Boger, D. L. J. Am. Chem. Soc. 2009, 131, 4904.
(7) (a) Va, P.; Campbell, E. L.; Robertson, W. M.; Boger, D. L. J. Am.
Chem. Soc. 2010, 132, 8489. (b) Sasaki, Y.; Kato, D.; Boger, D. L. J. Am.
Chem. Soc. 2010, 132, 13533. (c) Leggans, E. K.; Barker, T. J.; Duncan,
K. K.; Boger, D. L. Org. Lett. 2012, 14, 1428.
(8) (a) Vukovic, J.; Goodbody, A. E.; Kutney, J. P.; Misawa, M.
Tetrahedron 1988, 44, 325. (b) Szantay, C., Jr.; Balazs, M.; Bolcskei, H.;
Szantay, C. Tetrahedron 1991, 47, 1265. (c) Sundberg, R. J.; Hong, J.;
Smith, S. Q.; Sabat, M.; Tabakovic, I. Tetrahedron 1998, 54, 6259.
(9) Langlois, N.; Gueritte, F.; Langlois, Y.; Potier, P. J. Am. Chem. Soc.
1976, 98, 7017.
(10) Kutney, J. P.; Hibino, T.; Jahngen, E.; Okutani, T.; Ratcliffe, A. H.;
Treasurywala, A. M.; Wunderly, S. Helv. Chim. Acta 1976, 59, 2858.
(11) Sundberg, R. J.; Gadamasetti, K. G.; Hunt, P. J. Tetrahedron 1992,
48, 277.
(12) (a) Bornmann, W. G.; Kuehne, M. E. J. Org. Chem. 1992, 57, 1752.
(b) Schill, G.; Priester, C. U.; Windhovel, U. F.; Fritz, H. Tetrahedron
1987, 43, 3765. (c) Kuboyama, T.; Yokoshima, S.; Tokuyama, H.;
Fukuyama, T. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 11966.
(13) (a) Magnus, P.; Mendoza, J. S.; Stamford, A.; Ladlow, M.; Willis,
P. J. Am. Chem. Soc. 1992, 114, 10232. (b) Gunic, E.; Tabakovic, I.;
Gasic, M. J. J. Chem. Soc., Chem. Commun. 1993, 1496.
(14) Ngo, Q. A.; Roussi, F.; Thoret, S.; Gueritte, F. Chem. Biol. Drug
Des. 2010, 75, 284.
accessible face opposite the adjacent substituent. In contrast,
substrates whose added substituents are more removed (16a and
17a) typically provided two diastereomers in near equal
amounts. Like the reactions of 7 and in these unoptimized
reactions, recovered starting material (3) accounted for the
material balance in the absence of coupling (e.g., 12−15a).
These key studies further define easily understood fundamental
stereochemical features of the reaction, indicating that such non-
catharanthine-derived indoles, like catharanthine itself, can
participate in highly diastereoselective Fe(III)-promoted cou-
plings with vindoline. These may be used in conjunction with the
X-ray structure of vinblastine bound to tubulin to design
previously inaccessible vinblastine analogs containing effective
replacements for the upper velbanamine subunit.
Finally, in addition to a comparison of Fe(III) salts (FeCl3,
90% > Fe2(SO4)3, 71% > Fe2(ox)3, 0%), we examined additional
oxidants in our studies.8 Under near identical conditions (aq 0.05
N HCl/CF3CH2OH), Mn(OAc)3 (2.1 equiv, 2 h, 25 °C, 79%)
and Ce(NH4)2(NO3)6 (2 equiv, 16 h, 25 °C, 51%) were found to
support the coupling of catharanthine and vindoline to provide 5.
The extension of the studies to the preparation of such
vinblastine analogs, the use of the single electron oxidative
coupling in the synthesis of additional indole-containing natural
products, further studies on the mechanism of the Fe(III)-
promoted coupling of catharathine with vindoline and its
analogs, and their implications on the biosynthesis of the natural
products themselves are in progress and will be disclosed in due
course.
(15) Bergman, J.; Bergman, S.; Lindstrom, J.-O. Tetrahedron Lett.
1989, 30, 5337.
(16) (a) Tam, A.; Gotoh, H.; Robertson, W. M.; Boger, D. L. Bioorg.
Med. Chem. Lett. 2010, 20, 6408. (b) Gotoh, H.; Duncan, K. K.;
Robertson, W. M.; Boger, D. L. ACS Med. Chem. Lett. 2011, 2, 948.
(17) Baciocchi, E.; Muraglia, E. Tetrahedron Lett. 1993, 34, 5015.
(18) Dmowski, W.; Urbanczyk-Lipkowska, Z.; Wojcik, D. J. Fluorine
Chem. 2009, 130, 509.
(19) Viehe, H. G.; Janousek, Z.; Merenyi, R.; Stella, L. Acc. Chem. Res.
1985, 18, 148.
ASSOCIATED CONTENT
■
S
* Supporting Information
Full experimental details and spectra are provided. This material
is available free of charge via the Internet at http://pubs.acs.org.
13243
dx.doi.org/10.1021/ja306229x | J. Am. Chem. Soc. 2012, 134, 13240−13243