A rapid stereocontrolled entry to the ABCD tetracyclic core of neotuberostemonine

Article abstract of DOI:10.1002/anie.200250507

Four out of the five rings of the alkaloid neotuberostemonine (1) are present in the advanced precursor 2. The tetracyclic compound 2 has now been prepared in a short, linear route via lactone acid 3. A key step in the synthesis was the cuprate-mediated S_N2′ ring-opening desymmetrization of the C₂-symmetric bislactone 4.

Full text of DOI:10.1002/anie.200250507

Communications
Natural-Product Synthesis
A Rapid Stereocontrolled Entry to the ABCD
Tetracyclic Core of Neotuberostemonine**
Kevin I. Booker-Milburn,* Paul Hirst,
Jonathan P. H. Charmant, and Luke H. J. Taylor
Neotuberostemonine (1), stenine (2), and tuberostemonine
(3) are three of a number of complex polycyclic alkaloids
isolated from the roots of stemonaceous plants. Extracts of
these plants have long been used in China and Japan as
O
O
O
H
H
H
D
H
H
H
O
H
O
H
O
H
H
H
H
H
C
A
E
N
N
N
O
O
O
O
H
H
B
H
H
H
H
3
2
1
human cough remedies and antihelminthics in domestic
animals.[1] X-ray crystallography and 2D NMR spectroscopic
studies have confirmed that neotuberostemonine differs from
2 and 3 at three of the contiguous stereocentres on the ACD
ring junctures.[2] A number of syntheses of 2 have now been
reported reported since the first successful total synthesis by
Hart and Chen,[3] and include contributions from the groups
of Wipf,[4] Morimoto,[5] and Padwa.[6] Morimoto et al.[7] have
described a full account of their total synthesis of 2, from
which they aim to develop a strategy towards 3. Very recently,
Wipf et al.[8] described the first total synthesis of 3. To date no
synthesis of 1 has been reported.
Previously we demonstrated that N-alkenyl maleimide
derivatives undergo a very powerful [5þ2] photocycloaddi-
tion reaction to yield the hexahydroazaazulene ring system
common to all the Stemona alkaloids. For example, irradi-
ation of the maleimide 4 gave the tricyclic azepine 6 in
excellent yield. The reaction is thought to proceed through
the fragmentation of the zwitterionic [2þ2] cycloadduct 5
formed initially (Scheme 1a).[9] Recently, we showed that this
reaction can be used to construct the tricyclic core of the
[*] Dr. K. I. Booker-Milburn, Dr. P. Hirst
School of Chemistry, University of Bristol
Cantock's Close, Bristol, BS8 1TS (UK)
Fax: (+44)117-929-8611
E-mail: k.booker-milburn@bristol.ac.uk
Dr. J. P. H. Charmant, L. H. J. Taylor
Structural Chemistry Laboratory
School of Chemistry, University of Bristol
Cantock's Close, Bristol, BS8 1TS (UK)
[**] This work was supported by a grant (GR/R02382/01) from the
Engineering and Physical Sciences Research Council (EPSRC).
Supporting information for this article (NMR data and experimental
details) is available on the WWW under http://www.angewandte.org
or from the author.
1642
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.50507
Angew. Chem. Int. Ed. 2003, 42, 1642 – 1644

Angewandte
Chemie
knowledge this acid-catalyzed bislactonization procedure
is novel and may well prove to have application in the
synthesis of other C2-symmetric bislactones. Helmchen and
Bergner described the successful SN2’ ring-opening of
cyclohexene-fused lactones using an organocopper species
derived from Grignard reagents and CuBr·Me2S.[14]
Unfortunately, the bislactone 11 proved to be remarkably
insoluble in most organic solvents and as a result the initial
cuprate-mediated SN2’ ring-opening experiments based on
the literature conditions met with failure. Eventually it was
found that the use of 10 equivalents of the organocopper
reagent derived from EtMgBr–CuBr·Me2S in a THF/Me2S
solvent mixture (2:1) led to the ring-opened lactone acid 10
in 89% yield as a separable mixture of isomers (anti/syn =
8.5:1). Use of less organocopper reagent led to very slow
and incomplete conversion into product. Interestingly, ring
opening of 11 with the less stereochemically demanding
MeMgBr led to better anti/syn ratios (11:1). Selective
reduction of the acid with the NaBH4/mixed anhydride
technique described by Zoutani et al.[16] gave the alcohol
15 in good yield. Mitsunobu coupling with maleimide,
dimethylmaleimide, and dichloromaleimide then gave
three photocyclization precursors 16 in low to moderate
yields. Very pleasingly, irradiation of the dimethyl deriv-
ative 16b was found to proceed without event and gave the
key tetracycle 17b[17] in 65% yield as a single diastereoomer
(Scheme 2). The correct relative stereochemistry of 17b was
confirmed by X-ray crystallographic analysis (Figure 1). In
Scheme 1. a) The formation of the tricyclic azepine 6 in excellent yield by
irradiation of maleimide 4 is thought to proceed via zwitterionic cycload-
duct 5. b) Retrosynthetic analysis of neotuberostemonine (1).
antileukemia alkaloid cephalotaxine.[10] As the azepine 6
contains three of the five rings of neotuberostemonine (1) we
undertook a total synthesis program with the aim of
synthesizing 1, using our [5þ2] photocycloaddition as the
key step. The retrosynthetic analysis outlined in Scheme 1b
illustrates our overall strategy. It was envisaged that the
E ring could be installed in 7 by using elements of the elegant
iminium ion chemistry developed both by Martin et al.[11] and
by Morimoto et al.[12] in their independent studies on the
synthesis of Stemona alkaloids. It is anticipated that the
pyrrolidine ring of reduced derivatives of 8 will undergo
selective alkoxylation by use of the MnIII salen chemistry
developed by Katsuki and Punniyamurthy.[13] The tetracycle 8
could, in turn, be assembled in a key step from the maleimide
lactone 9 by using our [5þ2] photocycloaddition. It was
envisaged that the key cycloaddition precursor could be
derived from the lactone acid 10 after chemoselective
reduction and Mitsunobu coupling with maleimide. In one
of a number of strategies it was conceived that 10 could
derived from the C2-symmetric bislactone 11 by way of an
appealing cuprate-mediated SN2’ ring-opening desymmetriza-
tion sequence. Anti-selective ring-opening reactions of a
monolactone related to 11 were recently reported by Helm-
chen and Bergner.[14]
Multigram quantities of the racemic diol 12 were easily
prepared from furan and fumaryl chloride according to the
Diels–Alder/reduction sequence described by Paquette
et al.[15] Mesylation of 12 followed by treatment of the labile
bismesylate with KCN gave the dinitrile 13 in 75% overall
yield (Scheme 2). Compound 13 was easily converted into the
diacid 14 by hydrolysis under basic conditions. Diacid 14 was
then cyclized to the C2-symmetric bislactone 11 by an acid-
catalyzed tandem furan-ring-opening–lactonization sequence.
This same sequence was also possible in a single hydrolytic
step from 13, albeit at the expense of overall yield. To our
O
CN
OH
OH
CO2H
c
H
H
O
H
a
b
O
O
O
O
O
75%
85%
93%
H
11
14
CN
O
CO2H
13
12
d, 45%
e
89%
O
O
H
H
H
O
H
O
H
O
H
CO2H
g
f
O
N
OH
O
84%
10
15
R
16
8.5:1 anti/syn
R
a: R = H, 35%
b: R = Me, 31%
c: R = Cl, 63%
h
O
O
H
H
H
H
O
O
H
H
i
N
N
O
17
H
O
H
H
O
H
86%
O
a: R = H, 11%
b: R = Me, 65%
c: R = Cl, 60%
R
18
R
Scheme 2. Synthesis of the ABCD core of 1: a) MsCl, Et3N, Et2O, 2 h,
then KCN, DMSO, 1008C, 5 h; b) KOH, EtOH, H2O; c) pTSA, toluene,
reflux; d) H2SO4 (6m), heat, 2 h; e) EtMgBr (10 equiv), CuBr·Me2S
(10 equiv), THF/Me2S (2:1), À208C; f) EtOCOCl, Et3N, then NaBH4;
g) maleimide/dimethylmaleimide/dichloromaleimide, DIAD, PPh3,
THF, À788C!RT, 24 h; h) hn, pyrex, MeCN, 30–120 min; i) Zn, AcOH,
room temperature, 1.5 h. Ms=methanesulfonyl, DMSO=dimethyl
sulfoxide, pTSA=para-toluenesulfonic acid, DIAD=diisopropylazodi-
carboxylate.
Angew. Chem. Int. Ed. 2003, 42, 1642 – 1644
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1643

Communications
[1] For a comprehensive review on the Stemona alkaloids, see: R. A.
Pilli, M. C. Ferreira de Oliveira, Nat. Prod. Rep. 2000, 17, 1 1 7 –
127.
[2] C. N. Dao, P. Luger, P. T. Ky, V. N. Kim, N. X. Dung, Acta.
Crystallogr. Sect. C 1994, 50, 1612 – 1615.
[3] C. Y. Chen, D. J. Hart, J. Org. Chem. 1990, 55, 6236; C. Y. Chen,
D. J. Hart, J. Org. Chem. 1993, 58, 3840 – 3849.
[4] P. Wipf, Y. Kim, D. M. Goldstein, J. Am. Chem. Soc. 1995, 117,
11106 – 11112.
[5] Y. Morimoto, M. Iwahashi, K. Nishida, Y. Hayashi, H. Shir-
ahama, Angew. Chem. 1996, 108, 968 – 970; Angew. Chem. Int.
Ed. Engl. 1996, 35, 904 – 906.
[6] J. D. Ginn, A. Padwa, Org. Lett. 2002, 4, 1515 – 1517.
[7] Y. Morimoto, M. Iwahashi, T. Kinoshita, K. Nishida, Chem. Eur.
J. 2001, 7, 4107 – 4116.
[8] P. Wipf, S. R. Rector, H. Takahashi, J. Am. Chem. Soc. 2002, 124,
14848 – 14849.
[9] K. I. Booker-Milburn, C. E. Anson, C. Clissold, N. J. Costin,
R. F. Dainty, M. Murray, D. Patel, A. Sharp, Eur. J. Org. Chem.
2001, 1473 – 1482.
[10] K. I. Booker-Milburn, L. F. Dudin, C. E. Anson, S. D. Guile,
Org. Lett. 2001, 3, 3005 – 3008.
Figure 1. Molecular structure of the tetracyclic [5þ2] photocycloadduct
17b (R=Me). Ellipsoids set at the 50% probability level.
[11] S. F. Martin, K.J. Barr, J. Am. Chem. Soc. 1996, 118, 3299 – 3300.
[12] Y. Morimoto, K. Nishida, Y. Hayashi, Tetrahedron Lett. 1993, 34,
5773 – 5776.
[13] T. Punniyamurthy, T. Katsuki, Tetrahedron 1999, 55, 9439 – 9454.
[14] E. J. Bergner, G. Helmchen, Eur. J. Org. Chem. 2000, 419 – 423.
[15] L. A. Paquette, T. M. Kravetz, P. Charumilind, Tetrahedron 1986,
42, 1789 – 1795.
light of our previous experience with substrates derived from
the parent maleimide, it was perhaps not too surprising to
[9]
find that 16a (R = H) gave only a low yield (11%) of the
corresponding photoadduct 17a; the majority of the product
underwent further photoreactions as it was formed. Fortu-
nately the dichloro derivative 16c underwent photolysis to
give the [5þ2] cycloadduct 17c in 60% yield. Finally,
reduction of the alkene, with concomitant dehalogenation,
yielded the ketoamide 18 in excellent yield upon treatment of
17c with zinc in acetic acid under our previously developed
conditions.[9] This ketoamide has all the functionality neces-
sary for completion of the synthesis of 1. After reduction of
the ketone and deoxygenation, we will investigate the
installation of the methyl group of the lactone either by
direct methylation or by a methylenation–hydrogenation
sequence. Alkoxylation and lactone formation will be
attempted as described in Scheme 1, after which thioamide
formation and reduction with Raney nickel[3] should complete
the synthesis of 1.
In summary, an original route to the tetracyclic core of
neotuberostemonine (1) has been described which highlights
the application of the maleimide [5þ2] photocycloaddition in
natural-product synthesis. This linear route is notable for its
brevity, partly a result of the fact that no protecting groups are
used anywhere in the sequence—six steps to 17 from 12 by the
shortest sequence. Also of note is a novel cuprate-mediated
ring-opening desymmetrization of a C2-symmetric bislactone,
which may prove useful for the synthesis of other highly
functionalized cyclohexene-fused butyrolactones. Present
work is concerned with exploring the reactivity of 18 and
ultimate implementation of this as an advanced intermediate
for the total synthesis of 1.
[16] M. A. N. Zoutani, A. Pancrazi, J. Ardisson, Synlett 2001, 769 –
772.
[17] Crystal dimensions: 0.75 ” 0.60 ” 0.10 mm3, orthorhombic, space
group Pbca, a = 11.7444(16), b = 7.6959(10), c = 34.758(6) Š,
V= 3141.5(8) Š3, 1calcd = 1.342, 2qmax = 55.08, MoKa radiation
(0.71073 Š). Data were collected at 100 K as a series of
frames, each covering 0.38 in w, and integrated software
(SAINT, Bruker 2001) to give a Lorentz polarization corrected
data set of 18721 measured and 3579 independent reflections.
An absorption correction (SADABS, Sheldrick 2001) was
applied (m = 0.094,
Tmin = 0.772, Tmax = 1.000). Solution and
refinement (SHELXTL, Bruker-AXS 2001) against j F2 j used
211 parameters with all hydrogen atoms constrained to ideal
geometries and refined with isotropic displacement parameters
equal to 1.5 ” (methyl) or 1.2 ” (all other hydrogen atoms) the
equivalent isotropic displacement parameter of their parent
atom. Final residuals: R1 [2715 data with I > 2s(I)] = 4.4%, wR2
(all data) = 12.5%. Max/min residual electron density: + 0.310,
À0.265 eŠÀ3. CCDC 196931 contains the supplementary crys-
tallographic data for this paper. These data can be obtained free
of charge via www.ccdc.cam.ac.uk/conts/retrieving.html (or from
the Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB21EZ, UK; fax: (+ 44)1223-336-033; or deposit@
ccdc.cam.ac.uk).
Received: November 8, 2002 [Z50507]
Keywords: asymmetric synthesis · cuprates · lactones ·
.
natural products · photocyclization · total synthesis
1644
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2003, 42, 1642 – 1644