C-H amination in synthesis: An approach to the assembly of the B/C/D ring system of aconitine

Article abstract of DOI:10.1021/ol702375r

A strategy for the preparation of aconitine is described that attempts to exploit chemoselective C-H amination and the electrophilic reactivity of oxathiazinane N,O-acetals for assembling the complex, polycyclic carbon framework of the natural product.

Full text of DOI:10.1021/ol702375r

ORGANIC
LETTERS
2007
Vol. 9, No. 26
5465-5468
C−H Amination in Synthesis: An
Approach to the Assembly of the B/C/D
Ring System of Aconitine
Rosemary M. Conrad and J. Du Bois*
Department of Chemistry, Stanford UniVersity, Stanford, California 94305-5080
jdubois@stanford.edu
Received September 29, 2007
ABSTRACT
A strategy for the preparation of aconitine is described that attempts to exploit chemoselective C−H amination and the electrophilic reactivity
of oxathiazinane N,O-acetals for assembling the complex, polycyclic carbon framework of the natural product.
Extracts from the Aconitum genus of plants have been
recognized since antiquity for their potent analgesic and
antipyretic properties. The efficacy of these herbal medicines
can be traced, in part, to the complex alkaloid aconitine 1
(Figure 1).1 This topologically intricate natural product is
talatisamine, chasmanine, and 13-desoxydelphonine are the
only reports detailing the assembly of associated aconane
products.4 From our perspective, aconitine and related
alkaloids present an ideal forum to examine methods and to
devise strategies for crafting nitrogen-based heterocycles
through selective C-H amination.5 Additionally, an aconitine
synthesis presents opportunities for reaction development to
access highly oxygenated carbocycles, as exemplified by the
D ring of the target. Guided by these two principal aims,
we have attempted to exploit the versatile chemistry of
oxathiazinane N,O-acetals in a synthesis of the natural
product. This unique family of heterocycles can be generated
through the use of selective C-H amination reactions
developed in our laboratory. Under the action of a Lewis
acid, intramolecular arene addition to an oxathiazinane N,O-
acetal would enable simultaneous construction of the A and
Figure 1. A poisonous constituent of the Aconitum genus.
(3) (a) Wiesner, K.; Go¨tz, M.; Simmons, D. L.; Fowler, L. R.; Bachelor,
F. W.; Brown, R. F. C.; Bu¨chi, G. Tetrahedron Lett. 1959, 1 (2), 15. (b)
Codding, P. W. Acta Crystallogr. 1982, B38, 2519.
(4) (a) Wiesner, K.; Tsai, T. Y. R.; Huber, K.; Bolton, S. E.; Vlahov, R.
J. Am. Chem. Soc. 1974, 96, 4990. (b) Wiesner, K. Pure Appl. Chem. 1975,
41, 93. (c) Wiesner, K.; Tsai, T. Y. R.; Nambiar, K. P. Can. J. Chem. 1978,
56, 1451. (d) Wiesner, K. Pure Appl. Chem. 1979, 51, 689. (e) Wiesner,
K. Tetrahedron 1985, 41, 485.
one member of a large family of related structures that share
a common C19 diterpenoid frame.2 Although the molecular
form of aconitine has been known for close to 50 years, no
chemical preparation of the alkaloid has been described.3
Elegant works by Wiesner and co-workers in syntheses of
(5) (a) Wehn, P. M.; Du Bois, J. J. Am. Chem. Soc. 2002, 124, 12950.
(b) Hinman, A.; Du Bois, J. J. Am. Chem. Soc. 2003, 125, 11510. (c)
Fleming, J. J.; Du Bois, J. J. Am. Chem. Soc. 2006, 128, 3926. (d) Fleming,
J. J.; McReynolds, M. D.; Du Bois, J. J. Am. Chem. Soc. 2007, 129, 9964.
(1) Reviewed in: Ameri, A. Prog. Neurobiol. 1998, 56, 211.
(2) Amiya, T.; Bando, H. The Alkaloids; Brossi, A., Ed.; Academic Press,
Inc.: San Diego, CA, 1988; Vol. 34, Chapter 3, p 95.
10.1021/ol702375r CCC: $37.00
© 2007 American Chemical Society
Published on Web 12/01/2007

B rings of aconitine. Our efforts to implement this strategy
on an advanced model system of the natural product are
highlighted herein.
table product mixture. By contrast, silyl ether 8 furnished
the desired spirocycle 11 in 68% yield and as a single
diastereomer, as determined by proton NMR of the un-
purified reaction mixture (Figure 3). Other Lewis acids tested,
Prior work from our lab has demonstrated that chemo-
selective C-H insertion of a Rh-stabilized nitrene into an
R-ethereal center can be utilized to generate oxathiazinane
N,O-acetal heterocycles such as 3 (Figure 2). In the presence
Figure 3. A spirodienone is formed from an oxathiazinane N,O-
acetal.
Figure 2. A C-H amination route to the aconitine skeleton.
including Sc(OTf)3 and AlCl3, gave similar results, albeit at
a cost to the isolated yield of 11. While we did not explore
alternative phenolic protecting groups at this stage of
planning, these two data points were quite informative and
suggest that deblocking of the phenol must prestage the
addition event.
of a Lewis acid and a nucleophilic agent, these novel N,O-
acetal derivatives function as electrophiles to make available
substitutionally complex oxathiazinane products.6 To capital-
ize on these findings for an aconitine synthesis, a late-stage
C-H amination reaction was envisioned that would afford
3. Treatment of 3 with a Lewis acid would result in the
subsequent addition of the pendant aromatic ring to generate
spirodienone 2. The successful implementation of this
strategy would make possible installation of both the C17
nitrogen and C11 tetrasubstituted stereocenters (aconitine
numbering) as well as the B ring of the carbon skeleton in
a single C-C bond-forming event. Importantly, the oxa-
thiazinane moiety serves as a controlling element to ensure
the proper stereochemical outcome in the iminium ion
addition (vide infra). While prior studies from our lab have
shown that allyl silanes, silyl enol ethers, and alkynyl zinc
species are effective nucleophiles when employed in coupling
reactions with oxathiazinane N,O-acetals, at the onset of this
study, we were uncertain if electron-rich arene groups could
function analogously.7 To test the viability of this plan, two
model substrates have been examined.
The isolation of 11 as a single diastereomer out of a
possible mixture of four is rather striking and deserves
comment.8 We have found that heating 11 in warm MeOH
results in C-C bond scission to return the N,O-acetal and
re-aromatized side chain (i.e., 11f9). On the basis of this
observation, we speculate that BF3‚OEt2/CH2Cl2 provides a
medium for reversible arene-iminium ion addition and that
thermodynamic biases are responsible for controlling the
product stereochemistry in 11.9 For the purpose of utilizing
this reaction sequence in the aconitine synthesis, the relative
stereochemistry between C7 and C17 in 11 is, however,
incorrect. Accordingly, we wondered whether specific geo-
metrical constraints could be imposed on the linker between
the N,O-acetal heterocycle and the arene ring to force the
requisite cis-stereochemistry at positions C7 and C17. The
challenge of assembling the rigid C/D ring skeleton of
aconitine was therefore assumed. To facilitate the testing of
these ideas, a partially deoxygenated form of the C/D bicycle
was targeted.
The bicyclooctane frame in 12 restricts the various
trajectories by which the arene ring may attack the reactive
iminium ion. Based on an analysis of computational models,
we reasoned that only one face of the iminium π-bond would
be properly aligned with the aryl moiety to enable successful
C-C bond formation. Reaction through this conformer, 12,
affords the requisite cis-oxathiazinane 13. Rotation around
the C7-C8 bond (i.e., 15) exposes the alternative dia-
stereoface of the iminium ion to the nucleophilic arene. In
The first oxathiazinane N,O-acetal having a tethered
aromatic group was formed in straightforward fashion from
an available pentanoic acid derivative. Although such a
substrate lacks both the C and D ring elements, ease of
synthesis rendered it an optimal starting point for our
investigations. Oxidative intramolecular C-H insertion of
either 5 or 6 under the action of 2 mol % Rh2(OAc)4 and
1.1 equiv of PhI(OAc)2 afforded the corresponding N,O-
acetals in quantitative conversion and with outstanding
chemoselectivity. Subsequent exposure of 7 and 8 to
BF3‚OEt2 (CH2Cl2, -78 to 23 °C), however, resulted in
considerably disparate outcomes. N,O-Acetal 7, having both
phenolic oxygens blocked as methyl ethers, gave an intrac-
(8) The structure and stereochemistry of 11 have been assigned through
1H and 13C HMBC, as well as 1H NOE studies, see the Supporting
Information for details.
(6) Fleming, J. J.; Fiori, K. W.; Du Bois, J. J. Am. Chem. Soc. 2003,
125, 2028.
(7) Fiori, K. W.; Fleming, J. J.; Du Bois, J. Angew. Chem., Int. Ed. 2004,
43, 4349.
(9) If the addition reaction were under kinetic control, it is likely that a
cis-configuration between C7 and C17 would be favored in accord with
predictive models for nucleophilic additions to cyclic iminium ions: Stevens,
R. V. Acc. Chem. Res. 1984, 17, 289.
5466
Org. Lett., Vol. 9, No. 26, 2007

E/Z isomers.11 Selective ozonolysis of this intermediate
smoothly provided ketone 18 in good overall yield (75%,
two steps). Hydrogenation of this product mixture followed
by intramolecular epoxide ring-opening with catalytic Sc(OTf)3
afforded lactone 19 as a single diastereomer (70%, two steps).
The use of Rh/C in benzene for the reduction of alkene 18
was found to mitigate competing hydrogenolysis of the
epoxide.12 Subsequent alcohol methylation and enol triflate
formation proceeded without event.13
To further advance lactone 20, cross-coupling of the aryl
group was accomplished with 5 mol % Pd(PPh3)4 and CuI.14
Attempts to reduce the aryl olefin by using H2 and supported
metal catalysts were, however, unsuccessful and generally
returned starting material. Conversely, diimide, formed in
situ from the bis-carboxylate salt, reacted smoothly and with
perfect diastereocontrol to give the saturated product 22. At
this point, the remainder of the carbon framework could be
introduced through lactone R-alkylation with MeOCH2Cl.
Analogous to the diimide reduction, enolate alkylation was
highly selective, occurring exclusively from the less-hindered,
convex face of the tricyclic skeleton. Finally, LiAlH4
reduction of lactone 23 and phenol deprotection gave alcohol
24.
The flexible nature of intermediate 24 was exploited to
generate a small subset of sulfamate structures 25 having
different combinations of phenol and C15-OH protecting
groups (i.e., silylated phenol groups, esterification of C15-
OH). The ability to prepare such compounds was important
given our earlier experiences with the cyclization of 7 and 8
(see Figure 3). As we would subsequently determine, though
each of the sulfamate derivatives could be made to undergo
successful R-ethereal C-H insertion, only one such deriva-
tive, 26, afforded an isolable product following the attempted
iminium ion cyclization.
Figure 4. Stereochemical analysis of arene addition.
this orientation, the π-orbital of the iminium ion is situated
almost perpendicular to the π-face of the phenol (Figure 4).
Second, the approach of the aryl nucleophile appears to be
obstructed by C7-H irrespective the conformation of the
iminium ion. If such conjectures prove valid, generation of
the undesired C7/C17 trans-isomer 16 is precluded.
A second stereochemical concern that troubles the arene-
iminium ion addition involves selectivity at C11. Either face
of the prochiral aryl group can potentially attack the iminium
ion to furnish isomeric structures, 13 or 14 (or a mixture
thereof). Buoyed by the successful coupling of 8 to give a
single product, 11, we were hopeful for a similar outcome
to favor 13. Steric repulsion between the phenolic C1-OH
and C6 might serve as a controlling element in this case. As
a contigency, however, a symmetrically substituted arene
could be used without having to modify significantly the
overall synthetic plan.
Preparation of a sulfamate substrate containing the C/D
rings of the aconitine skeleton was accomplished starting
from ketone 17, a material that is readily accessed from
commercial 3-ethoxy-2-cyclohexen-1-one.10 As shown in
Scheme 1, ketone olefination on 17 was performed through
a modified Horner-Wadsworth-Emmons reaction to furnish
the unsaturated ester as an inconsequential 2:1 mixture of
Despite the complexity and degree of attendant functional
groups present in sulfamate 26, treatment with Rh2(esp)2,
PhI(OAc)2, and MgO furnished the N,O-acetal product 27
in high yield and with exquisite chemoselectivity (Figure
5). Other dinuclear Rh catalysts also promote this transfor-
mation, but none were as effective as Rh2(esp)2.15 The
Scheme 1
Org. Lett., Vol. 9, No. 26, 2007
5467

Figure 5. Chemo- and regioselective C-H insertion gives the
desired N,O-acetal.
selective generation of N,O-acetal 27 in this process is quite
striking considering the placement of a 3° C-H bond at C8.
Such a level of positional control is testament to the
sensitivity of the Rh-mediated process to both steric and
electronic effects.
While sulfamates such as 26 could be smoothly converted
to the corresponding N,O-acetal, unfortunately, subsequent
attempts to induce Lewis acid-mediated cyclization resulted
in intractable product mixtures. Different combinations of
Lewis acid (BF3‚OEt2, Sc(OTf)3) and solvent (CH2Cl2,
CH3CN, CH3OH) failed to improve matters, until it was
determined that removal of both the tBuMe2Si- and CF3CO2-
groups followed by BF3‚OEt2 treatment could furnish an
isolable amount of a new product 33 (∼10%). Extensive 1H
and 13C NMR, and mass spectral analysis of this compound
indicated that the spirodienone had indeed formed. The
proximity of the 2° alcohol at C15, however, appears to
complicate matters by displacing the incipient oxathiazinane
(path a). In addition, alkylation of the phenolic-OH was
noted. An alternative pathway to 33 could proceed through
N,O-acetal 31 (path b). Activation of this intermediate by
BF3‚OEt2 would generate an oxocarbenium ion 32, which
is then intercepted by the arene group. Phenolic alkylation
as in path a completes this reaction cascade (Figure 6). Both
proposals suggest that protection of the C15-OH could
positively alter the outcome of this process. All attempts,
however, to perform the iminium ion cyclization in the
presence of an appropriate C15-OH blocking group did not
prove fruitful.
Figure 6. An unexpected rearrangement.
our synthetic plan to the alkaloid. Nevertheless, the success
of the C-H amination reaction when applied to structures
such as 26 (arguably the most complex tested to date) is
quite rewarding and gives evidence of the utility of this
method for the construction of complex molecules.
Acknowledgment. We are grateful to the National
Institutes of Health, the Camille and Henry Dreyfus Founda-
tion, the A. P. Sloan Foundation, Amgen, Boehringer-
Ingelheim, GlaxoSmithKline, Merck, and Pfizer for generous
support of this work. R.M.C is the recipient of Eli Lilly and
Amgen graduate fellowships.
While the isolated pentacycle 33 is structurally intriguing,
it is clearly imperfect as an intermediate toward aconitine.
Owing to these collective results, we are forced to re-evaluate
Note Added after ASAP Publication. Scheme 1 and Figure
5 contained an error in the version published ASAP on
December 1, 2007; the corrected version published on
December 4, 2007.
(10) Preparation of 17 is accomplished in five transformations, the details
of which are described in the Supporting Information.
(11) Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfeld, A. P.;
Masamune, S.; Roush, W. R.; Sakai, T. Tetrahedron Lett. 1984, 25, 2183.
(12) Conditions that were hampered by or led to exclusive epoxide
opening included the following: H2 and Pd/C, Pt/C or PtO2 in EtOAc, THF,
Et2O or C6H6.
(13) Comins, D. L.; Dehghani, A. Tetrahedron Lett. 1992, 33, 6299.
(14) Mee, S. P. H.; Lee, V.; Baldwin, J. E. Angew. Chem., Int. Ed. 2004,
43, 1132.
Supporting Information Available: General experimen-
tal protocols and characterization data for all new com-
pounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
(15) (a) Rh2(esp)2
) Rh2(R,R,R′,R′-tetramethyl-1,3-benzenedipro-
pionate)2. (b) Espino, C. G.; Fiori, K. W.; Kim, M.; Du Bois, J. J. Am.
Chem. Soc. 2004, 126, 15378. (c) Other catalysts that were examined
included Rh2(OAc)4, Rh2(O2CCPh3)4, Rh2(O2CCF3)4, and Rh2(O2CCMe3)4.
OL702375R
5468
Org. Lett., Vol. 9, No. 26, 2007