Synthesis of the 10-azatricyclo[3.3.2.04,8]decane core of C 20-diterpenoid alkaloid racemulsonine via iodine(III) promoted transannular aziridination reaction

Article abstract of DOI:10.1021/ol400755x

The functionalized A/E/F ring system of C₂₀-diterpenoid alkaloid racemulsonine has been efficiently synthesized. The Key steps involved a diastereoselective Au(I)-catalyzed annulation to form cis-fused cyclopentene and a PIDA promoted transannular aziridination of primary amine followed by regio- and stereoselective ring cleavage of bridged aziridine.

Full text of DOI:10.1021/ol400755x

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Synthesis of the
10-Azatricyclo[3.3.2.04,8]decane Core of
C20-Diterpenoid Alkaloid Racemulsonine
via Iodine(III) Promoted Transannular
Aziridination Reaction
Ru-Huai Mei,† Zhi-Gang Liu,† Hang Cheng, Liang Xu,* and Feng-Peng Wang
Key Laboratory of Drug Targeting, Ministry of Education, and Department of
Chemistry of Medicinal Natural Products, West China College of Pharmacy,
Sichuan University, No. 17, Duan 3, Renmin Nan Road, Chengdu 610041, P. R. China
liangxu@scu.edu.cn
Received March 21, 2013
ABSTRACT
The functionalized A/E/F ring system of C20-diterpenoid alkaloid racemulsonine has been efficiently synthesized. The Key steps involved a
diastereoselective Au(I)-catalyzed annulation to form cis-fused cyclopentene and a PIDA promoted transannular aziridination of primary amine
followed by regio- and stereoselective ring cleavage of bridged aziridine.
Aconitium species, commonly known as monkshood or
wolf’s bane, are the very famous plants that have been used
extensively in traditional medicine throughout Asia and
Europe as painkillers and produce various poisons and
compounds of medicinal importance.1 The fascinating
bioactivities of these compounds are contributed by C18-,
C19-, and C20-diterpenoid alkaloids, which have long
attracted scientists’ strong interest in their phytochemistry,
synthesis, and medicinal chemistry.2 In 2000, our
group reported the isolation of a structurally unique C20-
diterpenenoid alkaloid racemulsonine from Aconitium
racemulosum var. penzhounense,3 which represents the
latest example of skeletal novelty of diterpenoid alkaloids
generated by nature (Figure 1).1d This alkaloid shows an
unprecedented hexacyclic ring framework containing
a bridged A/E/F tricyclic subunit of 10-azatricyclo-
[3.3.2.04,8]decane that features an unusual five-membered
ring A, while all other naturally occurring diterpenoid
alkaloids uniformly possess a six-membered ring A. Bio-
genetically, racemulsonine may be consideredas the results
of the sophisticated A-nor and B-homo-C-nor rearrange-
ments of the known denudatine-type diterpenoids, one of
the major groups of C20-diterpenoids.3
Despite the intriguing biosynthetic and structural prop-
erties of the unique azahexacyclic skeleton, there was
only a single report that described efforts to synthesize
this compound, resulting in a useful methodology for the
construction of an azatricyclic A/B/E-ring system bearing
a bridged azabicyclo[3.2.1]octane subunit.4 In combina-
tion with our extensive and long-standing interest in the
chemistry of diterpenoid alkaloids,5 and intrigued by the
fascinating novel structural features and potential biologi-
cal activities of this compound, we attempted to establish
a methodology for the construction of a highly bridged
† These authors contributed equally.
(1) Pelletier, S. W.; Page, S. W. Nat. Prod. Rep. 1986, 3, 451. (b)
Yunusov, M. S. Nat. Prod. Rep. 1993, 10, 471. (c) Atta-ur-Rahman;
Choudhary, M. I. Nat. Prod. Rep. 1999, 16, 619. (d) Wang, F. P.; Chen,
Q. H.; Liu, X. Y. Nat. Prod. Rep. 2010, 27, 529.
(2) (a) Wang, F. P.; Chen, Q. H. The Alkaloids 2010, 69, 1–577. (b)
Wang, F. P.; Liang, X. T. The Alkaloids 2002, 59, 2–280. (c) Wang, F. P.;
Chen, Q. H.; Liang, X. T. The Alkaloids 2009, 67, 1–78.
(3) Wang, F.-P.; Peng, C.-S.; Yu, K.-B. Tetrahedron 2000, 56, 7443.
(4) Kraus, G. A.; Kesavan, S. Tetrahedron Lett. 2005, 46, 1111.
r
10.1021/ol400755x
XXXX American Chemical Society

the reported procedure for the methyl 4,4-bis(tert-
butoxymethyl)-2-oxocyclopentane carboxylate,8 dibenzyl
oxide 3 was converted to β-ketoester 4 in three steps.
Oxidation of 3 in aqueous KMnO4 gave the corresponding
diacid. After conversion of the diacid to the corresponding
diester, Dieckmann cyclization of diester with potassium
tert-butoxide in THF afforded the β-ketoester 4 in 52%
overall yield. Oxidation of 4 with DDQ in THF gave the
conjugated enone 5smoothly.9 Then the incorporation ofa
cis-fused cyclopentene onto the R,β-unsaturated β-ketoester
5 was next addressed via a two-step sequence. First, a
conjugate propargylation of 5 with allenyltriphenylstan-
nane in presence of TiCl4 afforded the alkyne 6 in 80%
yield;10 subsequently, the Au(I) catalyzed annulation of
6 under the modified Toste conditions,7 using 1,2-dichlor-
oethane as the solvent at an elevated temperature (50 °C),
proceeded efficiently to deliver the cis-fused cyclopentene 7
as a single diastereomer in 76% yield.
Figure 1. Structures of the hexacyclic diterpenoid alkaloids.
azatricyclic A/E/F-ring system I, which might serve as a
crucial point for the total synthesis of racemulsonine.
Herein, we report our endeavor to access the functiona-
lized 10-azatricyclo[3.3.2.04,8]decane skeleton as one of
the core bridged ring systems of racemulsonine through
an intramolecular transanular aziridination followed by
regio- and stereoselective aziridine ring cleavage.
Our strategy toward the synthesis of the cage-like aza-
tricyclic skeleton was attempted to utilize Nagata’s intra-
molecular aziridination reaction6 of unsaturated primary
amine III to deliver a bridged aziridine II (Scheme 1),
which could be expected to undergo a regio- and stereo-
selective aziridine ring cleavage to give I. The crucial
unsaturated primary amine III could be prepared by facile
founctional group interconversions from IV, which could
arise from the R, β-unsaturatedβ-ketoesterV followedbya
known gold-catalyzed cyclopentene annulation procedure.7
Scheme 2. Synthesis of Bicyclic A/F-Ring Precursor 7
Scheme 1. Retrosynthetic Analysis of Tricyclic Amine I
With the fused bicyclic A/F-ring precursor 7 in place, we
then turned our attention to the installation of a methoxy
group at C-3 and realization of the C-1 quaternary carbon
center with the right configuration by desymmetrization of
the prochiral diol (Scheme 3). Thus, exclusive reduction
of the carbonyl group of 7 with NaBH4 in methanol at
À40 °C, followed by methylation with MeI/Ag2O,11 gave
the C-1R methoxy product 8 as a single stereomer in 80%
yield over two steps. The stereochemistry of 8 was assigned
on the basis of NOE difference experiments which show
through-space interactions between H-3 and H-8 (4.26 and
2.16ppm). After selectiveremoval ofthe benzyl groups of8
by treatment with BBr3 in dichloromethane, desymmetri-
zation of the resulting 1,3-diol 9 was effected by treatment
with TBSCl and imidazole in dicholomethane,12 delivering
Our synthesis began with the protection of the known
diol 28 with a benzyl group (Scheme 2). According to
(5) (a) Wang, F. P.; Chen, Q. H.; Li, B. G. Tetrahedron 2001, 57, 4705.
(b) Xu, L.; Chen, Q. H.; Wang, F. P. Tetrahedron 2002, 58, 4267. (c)
Chen, Q. H.; Xu, L.; Wang, F. P. Tetrahedron 2002, 58, 9431. (d) Zou,
C. L.; Cai, L.; Ji, H.; Xie, G. B.; Wang, F. P.; Jian, X. X.; Song, L.; Liu,
X. Y.; Chen, D. L.; Chen, Q. H. Tetrahedron 2008, 64, 7594. (e) Tang, P.;
Wang, L.; Chen, Q. H.; Wang, F. P. Tetrahedron 2011, 67, 1076. (f) Liu,
Z. G.; Xu, L.; Chen, Q. H.; Wang, F. P. Tetrahedron 2012, 68, 159. (g)
Cheng, H.; Xu, L.; Chen, D. L.; Chen, Q. H.; Wang, F. P. Tetrahedron
2012, 68, 1171.
(9) Chou, H.-H.; Wu, H.-M.; Wu, J.-D.; Ly, T. W.; Jan, N.-W.; Shia,
K.-S.; Liu, H.-J. Org. Lett. 2007, 10, 121.
(10) Haruta, J. I.; Nishi, K.; Matsuda, S.; Akai, S.; Tamura, Y.; Kita,
(6) Nagata, W.; Hirai, S.; Kawata, K.; Aoki, T. J. Am. Chem. Soc.
1967, 89, 5045.
Y. J. Org. Chem. 1990, 55, 4853.
(7) Staben, S. T.; Kennedy-Smith, J. J.; Toste, F. D. Angew. Chem.,
Int. Ed. 2004, 43, 5350.
(8) Peelen, T. J.; Chi, Y.; English, E. P.; Gellman, S. H. Org. Lett.
(11) Werz, D. B.; Seeberger, P. H. Angew. Chem., Int. Ed. 2005, 44,
6315.
(12) Mahapatra, T.; Das, T.; Nanda, S. Tetrahedron: Asymmetry
2008, 19, 2497.
2004, 6, 4411.
B
Org. Lett., Vol. XX, No. XX, XXXX

an easily separable mixture of TBS ether 10a and 10b in a
1.2:1 ratio and combined 90% yield. While the isomeric
ratio is fairly low, the undesired minor isomer 10b could be
cleanly recycled to diol 9 by treatment with TBAF in THF,
with minimal loss of material. The desired alcohol 10a was
then oxidized with PCC/Na2CO3 in dichloromethane to
furnish a labile aldehyde 11, which subsequently under-
went a reductive amination by treatment with ammonia
in ethanol and titanium(IV) isopropoxide, followed by
reduction in situ withNaBH4, affording the primaryamine
12 in 80% yield.13
iodine reagents could generate nitrenoids capable of inter-
molecular or intramlecular aziridination of olefins,17 we
attempted to modify Nagata’s method by employing
iodine(III) reagents as oxidants. To our delight, when 12
was treated with phenyliodine(III) diacetate (PIDA) and
K2CO3 in 1,2-dichloroethane at 55 °C for 1.5 h, 13 was
obtained in an encouraging 33% yield with no starting
material recovered (entry 5). After a series of experiments
(entries 6À9),18 we found that treatment of 12 with PIDA
and K2CO3 using silica gel as an additive gave the best
result, delivering 13 in up to 72% yield (entry 9). The use
of iodosobenzene (PhIO) or the more potent oxidant
phenyliodine(III) bistrifluoroacetate (PIFA) resulted in a
lower yield (entries 10 and 11). This efficient intramolecu-
lar aziridination reaction proceeding via an iminoiodinane
intermediate generated from a cycloalkenyl primary amine
without a metal catalyst perhaps represents a new method
to the bridged, polycyclic aziridine. It is noteworthy that
the bridged aziridine 13 is quite stable and could be stored
without decomposition in a refrigerator (0À4 °C) for more
than two weeks even at neat liquid state.6,14
Scheme 3. Synthesis of Bicyclic Primary Amine 12
Table 1. Optimization of Reaction Conditions to Transform 12
into 13
t
time yielda
entry
reagents
solvent
PhH
(°C)
(h)
(%)
With the key intermediate 12 in hand, our effort was
focused on the synthesis of the challenging bridged azir-
idine. As shown in Table 1, an initial attempt to apply
Nagata’s intramolecular aziridination method6,14 was car-
ried out by oxidation of the primary amine 12 with excess
Pb(OAc)4 in benzene buffered with powdered K2CO3,
providing the bridged aziridine 13 in a disappointing
15% isolated yield, along with 30% of recovered starting
material and some unidentified low polar byproducts
(entry 1). Further attempts using NCS6 or NBS15 as an
oxidant as well as the modified procedure employing NCS
and CuCl16 were uniformly unsuccessful (entries 2À4). In
light of some reported examples that oxidations of primary
carbamate and N-amino phthalimide with hypervalent
1
Pb(OAc)4, K2CO3
NCS
reflux
24
2
15b,d
0c,d
0c,d
0c,d
33
2
CH2Cl2 rt
CH2Cl2 rt
CH2Cl2 rt
3
NBS
2
4
NCS, CuCl
PIDA, K2CO3
PIDA, MgO
PIDA,Cu(OTf)2
PIDA, BF3 3 Et2O
2
5
DCE
DCE
DCE
DCE
55
55
55
55
55
55
55
1.5
1.5
1.5
2
6
24
7
30
8
5
9
PIDA, K2CO3, SiO2 DCE
1
72
10
11
PhIO, SiO2
DCE
DCE
1
51
PIFA, K2CO3, SiO2
1
48
a Isolated yields. b 30% of the amine 12 was recovered. c All of the
amine 12 was consumed. d Unidentified low polar byproducts were
observed.
Having achieved the synthesis of the bridged tetracyclic
aziridine, the selective aziridine ring cleavage to prepare
the desired A/E/F azatricyclic system was next investigated
(Scheme 4). Gratifyingly, simple treatment of 12 with acetic
anhydrate in CH2Cl2 at room temperature fortunately gave
(13) Miriyala, B.; Bhattacharyya, S.; Williamson, J. S. Tetrahedron
2004, 60, 1463.
(14) Portoghese, P. S.; Sepp, D. T. Tetrahedron 1973, 29, 2253. (b)
Roy, A.; Roberts, F. G.; Wilderman, P. R.; Zhou, K.; Peters, R. J.;
Coates, R. M. J. Am. Chem. Soc. 2007, 129, 12453. (c) Hu, H.; Faraldos,
J. A.; Coates, R. M. J. Am. Chem. Soc. 2009, 131, 11998.
(15) Carman, R. M.; Derbyshire, R. P. C. Aust. J. Chem. 2003, 56,
319.
(16) Harayama, T.; Ohtani, M.; Oki, M.; Inubushi, Y. Chem. Pharm.
Bull. 1975, 23, 1511.
(17) (a) Wirth, T. Angew. Chem., Int. Ed. 2005, 44, 3656. (b)Zhdankin,
V. V.; Stang, P. J. Chem. Rev. 2002, 102, 2523. (c) Dauban, P.; Dodd,
R. H. Synlett 2003, 1571. (d) Dauban, P.; Dodd, R. H. Org. Lett. 2000, 2,
2327. (e) Padwa, A.; Stengel, T. Org. Lett. 2002, 4, 2137. (f) Liang, J.-L.;
Yuan, S.-X.; Chan, P. W. H.; Che, C.-M. Org. Lett. 2002, 4, 4507.
(18) For selected examples of Lewis acids facilitating the conversion
of hypervalent iodine(III) reagents, see: (a) Kita, Y.; Tohma, H.;
Hatanaka, K.; Takada, T.; Fujita, S.; Mitoh, S.; Sakurai, H.; Oka,
S. J. Am. Chem. Soc. 1994, 116, 3684. (b) Correa, A.; Tellitu, I.;
Dominguez, E.; SanMartin, R. J. Org. Chem. 2006, 71, 8316. (c) Kita,
Y.; Morimoto, K.; Ito, M.; Ogawa, C.; Goto, A.; Dohi, T. J. Am. Chem.
Soc. 2009, 131, 1668.
Org. Lett., Vol. XX, No. XX, XXXX
C

Scheme 4. Synthesis of Azatricyclic A/E/F-rings 14 and 15
Figure 2. Key NOESY relationship of 14a.
the same regio- and stereoslectivity. Direct introduction
of the N-ethyl group to form the tricyclic amine was also
investigated. Reaction of 13 with ethyl iodide in DMF at
room temperarure produced an ethyl quaternary amine
salt, which in turn was treated with sodium acetate in one
pot at 60 °C, delivering N-ethyl azatricyclic amine 15 in
75% yield.
In summary, we have achieved the construction of the
core bridged azatricyclic A/E/F-ring system of racemulso-
nine by using Au(I)-catalyzed annulation reaction and
PIDA-promoted transannularaziridination of theprimary
amine followed by regio- and stereoselective ring cleavage
as the key steps. The functionalized five-membered F-ring
with an OAc or Cl substituent at C-6 and the methoxy-
carbonyl group at C-4 could be a good relay for further
elaboration into the B-, C-, and D-rings of racemulsonine.
These investigations are underway in our laboratory and
will be published in due course.
the desired azatricyclo[3.3.2.0]decane compound 14a as a
singleregio- and stereoisomer in excellent yield, despite that
theoretically there might be another ring-opened regio-
isomer generated by nucleophilic attack at C-5 simulta-
neously . The high regioselectivity of this reaction might be
attributed to the relatively more steric congestion at C-5
caused by the adjacent C-4 quaternary center. The ring-
opened structure could be identified based on the 1H NMR
spectrum, which shows a double doublet (coupling with the
vicinal two protons at C-7) at 4.78 ppm that could only be
assigned to the proton at C-6 substituted with an OAc
group.19 The stereochemistry of the OAc group at C-6 was
established as an exo orientation on the basis of the
observation of the key NOESY relationship between H-6
and H-9 (Figure 2), which is compatible with an SN2
nucleophilic attack of aziridine by acetate. Finally the
whole stereochemical structure of 14a was unequivocally
confirmed based on the interpretation of its 2D NMR
spectra (see Supporting Information). Furthermore, acetyl
chloride was used to open the aziridine 13, yielding the
chlorine substituted tricyclic amide 14b in 88% yield with
Acknowledgment. Financial support for this work was
provided by the National Natural Science Foundation of
China (Nos. 30873147, 20902061, and 21272163).
Supporting Information Available. Experimentaldetails,
1H and 13C NMR spectra for all new compounds, the NOE
difference spectrum for compound 8, 2D NMR spectra and
their interpretations for compound 14a. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
(19) The double doublet displayed by the signal of the proton at C-6
of 14a at 4.78 ppm is due to the fact that the dihedral angel between H-6
and H-5 is ca. 90° based on observation of the molecular model of 14a.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX