A novel and efficient total synthesis of cephalotaxine

Article abstract of DOI:10.1021/ol035098b

(Matrix presented) Total synthesis of cephalotaxine (CET), the parent member of a class of structurally unique antileukemia Cephalotaxus alkaloids, was accomplished on the basis of a conceptually novel strategy featuring transannular reductive skeletal rearrangements as the key transformations for the construction of the pentacyclic ring skeleton of CET. The synthetic potential of the designated Clemmensen-Clemo-Prelog-Leonard reductive rearrangement was demonstrated for the first time in a facile synthesis of the benzazepine subunit of CET. A novel endocyclic enamine (cyclopentenone) annulation was discovered and rationalized as an unusual azo-Nazarov-type cyclization.

Full text of DOI:10.1021/ol035098b

ORGANIC
LETTERS
2003
Vol. 5, No. 16
2931-2934
A Novel and Efficient Total Synthesis of
Cephalotaxine
Wei-Dong Z. Li* and Yong-Qiang Wang
National Laboratory of Applied Organic Chemistry, Lanzhou UniVersity,
Lanzhou 730000, China
wdli@fas.harVard.edu; liwd@lzu.edu.cn
Received June 16, 2003
ABSTRACT
Total synthesis of cephalotaxine (CET), the parent member of a class of structurally unique antileukemia Cephalotaxus alkaloids, was
accomplished on the basis of a conceptually novel strategy featuring transannular reductive skeletal rearrangements as the key transformations
for the construction of the pentacyclic ring skeleton of CET. The synthetic potential of the designated Clemmensen−Clemo−Prelog−Leonard
reductive rearrangement was demonstrated for the first time in a facile synthesis of the benzazepine subunit of CET. A novel endocyclic
enamine (cyclopentenone) annulation was discovered and rationalized as an unusual azo-Nazarov-type cyclization.
Cephalotaxine (1, CET), the parent member of the Cepha-
lotaxus alkaloids,1 processes a unique benzazepine-bearing
pentacyclic ABCDE-ring skeleton. Naturally occurring esters
of CET (harringtonine and homoharringtonine) have been
found to be highly effective for the treatment of acute human
leukemia and are currently undergoing advanced clinical
trials.2 Homoharringtonine is also a potent agent against
strains of chloroquinine-resistant Plasmodium f. malaria
parasite in vitro.3 The unique structure and the therapeutic
potential of this group of alkaloids have stimulated much
synthetic research that has produced several elegant total
syntheses of CET1b,4 and numerous studies on the construc-
tion of the pentacyclic ring system.1b,5 Nonetheless, an even
more efficient and practical synthesis of CET is desirable in
light of the potential pharmaceutical needs. We describe
herein a novel and highly efficient synthesis of CET.
The unique heterocyclic ABCDE-ring system of CET has
continued to be a proving ground for new strategy and
synthetic methods.1b Our strategic thinking for elaborating
the key pentacyclic precursor, ketone 26 (Figure 1), was based
(1) For reviews, see: (a) Huang, L.; Xue, Z. In The Alkaloids; Brossi,
A., Ed.; Academic Press: New York, 1984; Vol. 23, pp 157-226. (b) Jalil
Miah, M. A.; Hudlicky, T.; Reed, J. W. In The Alkaloids; Brossi, A., Ed.;
Academic Press: New York, 1998; Vol. 51, pp 199-269.
Figure 1. Strategic plan.
(2) For a review, see: Kantarjian, H. M.; Talpaz, M.; Santini, V.; Murgo,
A.; Cheson, B.; O’Brien, S. M. Cancer 2001, 92, 1591.
on a transannular skeletal rearrangement (path a) of cyclic
enone 3 that parallels the biogenesis of CET postulated by
Parry et al.7 Enone 3 could in turn be derived from cyclic
ketone 5. A more conventional strategy is outlined as path
b, which involves a transannular rearrangement of ketone 5
to benzazepine precursor 4.
(3) Whaun, J. M.; Brown, N. D. Ann. Trop. Med. Parasitol. 1990, 84,
229. See also: Chem. Abstr. 1990, 113, 165042.
(4) For recent examples, see: (a) Suga, S.; Watanabe, M.; Yoshida, J.
J. Am. Chem. Soc. 2002, 124, 14824. (b) Koseki, Y.; Sato, H.; Watanabe,
Y.; Nagasaka, T. Org. Lett. 2002, 4, 885. (c) Tietze, L. F.; Schirok, H. J.
Am. Chem. Soc. 1999, 121, 10264.
10.1021/ol035098b CCC: $25.00 © 2003 American Chemical Society
Published on Web 07/04/2003

Scheme 1. Total Synthesis of Cephalotaxine
As shown in Scheme 1, the synthesis commenced from
tBuOH, 40 °C) to form the pentacyclic enone 3 (76%). Upon
short exposure to zinc dust in hot (100 °C) glacial acetic
acid, enone 3 was transformed into the desired pentacyclic
ketone 2 (mp 129-131 °C) in 65% isolated yield11 along
with ca. 5% of enone conjugate reduction product (mp 173-
174 °C).6a This remarkably facile transformation can be
rationalized as a transannular reductive rearrangement via a
transient bridged aziridinium12 intermediate as shown in
Scheme 2. This type of skeletal rearrangement can be found
in Bu¨chi’s synthesis of Iboga alkaloids.13
â-(3, 4-methylenedioxy)phenethylamine (6), which was
converted to 7 in quantity in 62% overall yield by the
following sequence: (1) Bischler-Napieralski cyclization
of corresponding oxalamide, (2) catalytic hydrogenation, and
(3) N-alkylation with ethyl 4-bromobutyrate.8 Reaction of
diester 7 with allyl bromide in DMSO at 35 °C for 48 h
followed by treatment with KOtBu in DMSO-THF (1:1) at
-15 °C f rt gave the allylation product 9 in 78% yield,9
presumably via a facile 2,3-sigmatropic rearrangement10 of
the corresponding N-ylide of ammonium salt 8. Dieckmann
cyclization of diester 9 (KOtBu, toluene, reflux) and subse-
quent decarboxylation (CaCl2, DMSO, 150 °C) produced the
cyclic ketone 5 in 59% yield. Wacker oxidation of 5 (10
mol % PdCl2, CuCl2, O2, 0.2 N HCl-DMF, 65 °C) furnished
the diketone 10 in 82% yield, which was cyclized (KOtBu,
Scheme 2. Transannular Rearrangement Pathway for 3 f 2
(5) For recent examples, see: (a) Worden, S. M.; Mapitse, R.; Hayes,
C. J. Tetrahedron Lett. 2002, 43, 6011. (b) Booker-Milburn, K. I.; Dudin,
L. F.; Anson, C. E.; Guile, S. D. Org. Lett. 2001, 3, 3005. (c) Kim, S.-H.;
Cha, J. K. Synthesis 2000, 2113. (d) Beall, L. S.; Padwa, A. Tetrahedron
Lett. 1998, 39, 4159. (e) Molander, G. A.; Hiersemann, M. Tetrahedron
Lett. 1997, 38, 4347. (f) De Oliveira, E. R.; Dumas, F.; D’Angelo, J.
Tetrahedron Lett. 1997, 38, 3723.
(6) (a) Dolby, L. J.; Nelson, S. J.; Senkovich, D. J. Org. Chem. 1972,
37, 3691. (b) Weinreb, S. M.; Auerbach, J. J. Am. Chem. Soc. 1975, 97,
2503. (c) Weinreb, S. M.; Semmelhack, M. F. Acc. Chem. Res. 1975, 8,
158. (d) Weinstein, B.; Craig, A. R. J. Org. Chem. 1976, 41, 875.
(7) Parry, R. J.; Chang, M. N. T.; Schwab, J. M.; Foxman, B. M. J. Am.
Chem. Soc. 1980, 102, 1099 (cf. path a below).
Attempted direct hydroxylation of the corresponding
enolate of cyclic ketone 2 using Bu¨chi’s procedure (KOtBu,
tBuOH, O2, P(OEt)3)14 resulted in a facile dehydrogenation
of 2 to an enone 17 (vide infra). Subjecting 2 to Moriarty
oxidation (PhI(OAc)2, KOH, MeOH, 0 °C)15 gave a hydroxy
dimethylketal as the sole product (mp 74-75 °C) in 76%
(11) The putative pentacyclic ketone 2 had not been actually synthesized
prior to this work; cf. refs 1b and 6.
(12) For a review, see: Nagata, W. In Lectures in Heterocyclic Chemistry;
Castle, R. N., Elslager, E. F., Eds.; J. Heterocyclic Chemistry, Inc.: Orem,
1972; Vol. 1, pp S29-S37.
(8) See the Supporting Information for details.
(9) Direct allylation of diester 7 with allyl bromide by the action of
KOtBu in THF or DMF resulted in a low and irreproducible yield of 9.
(10) This was evidenced by the slowness and incompleteness of direct
methylation (MeI, KOtBu) as compared to allylation or crotylation of diester
7 or corresponding 3, 4-dimethoxy analogue. For a review, see: Bru¨ckner,
R. In ComprehensiVe Organic Synthesis; Trost, B. M., Fleming, I., Eds.;
Pergamon Press: Oxford, 1991; Vol. 6, pp 873-908.
(13) Bu¨chi, G.; Coffen, D. L.; Kocsis, K.; Sonnet, P. E.; Ziegler, F. E.
J. Am. Chem. Soc. 1966, 88, 3099.
(14) Bu¨chi, G.; Kulsa, P.; Ogasawara, K.; Rosati, R. L. J. Am. Chem.
Soc. 1970, 92, 999.
(15) Moriarty, R. M.; Hu, H.; Gupta, S. C. Tetrahedron Lett. 1981, 22,
1283. For a review, see: Moriarty, R. M.; Prakash, O. Org. React. 1999,
54, 273.
2932
Org. Lett., Vol. 5, No. 16, 2003

yield, which was deduced to be 11 by spectroscopic
analysis.16 Deketalization of 11 in THF followed by autoxi-
Scheme 4. Reductive Rearrangement Pathway for 5 f 14
t
dative dehydrogenation17 (KOtBu, BuOH, O2, P(OEt)3,
40 °C) of the resulting crude hydroxy ketone 12 (epi-
desmethylcephalotaxine) afforded desmethylcephalotaxinone
(13) (mp 105-107 °C) in 86% yield,6b identical in every
respect with a sample prepared from natural (-)-CET18 by
hydrolysis7 and autoxidative dehydrogenation as for 11 f
13. Since the conversion of desmethylcephalotaxinone (13)
to natural (-)-CET via methylation, optical resolution with
L-tartaric acid,19 and borohydride reduction has been
demonstrated,1b the sequence outlined in Scheme 1 consti-
tutes a total synthesis of CET with an overall yield of ca.
12% to desmethylcephalotaxinone (3) from diester 7 through
an eight-stage sequence.
The synthesis described above evolved from an initial
strategic plan (Figure 1, path b) that was based on an
intramolecular Mannich cyclization20 of the iminium inter-
mediate 4 for the E-ring formation. We envisioned that the
corresponding precursor to 4 would be generated from cyclic
ketone 5 by a Clemmensen reductive rearrangement. The
anomalous Clemmensen reduction of cyclic R-amino ketone
was first noticed by Clemo in 1931,21 clarified by Prelog,22
and later studied systematically by Leonard and co-workers.23
Surprisingly, the potential of this interesting reductive
rearrangement in organic synthesis has not been recognized.
We took advantage of this unique Clemmensen-Clemo-
Prelog-Leonard reductive rearrangement for constructing the
benzazepine-bearing ABCD-ring system of CET. To our
delight, standard Clemmensen reduction (Zn-Hg, concen-
trated HCl, reflux) of R-amino ketone 5 led to allyl
benzazepine derivative 14 in 76% isolated yield (Scheme
3). Furthermore, we found that anhydrous Clemmensen
This facile rearrangement can be generalized (Scheme 4)
as an acid-catalyzed transannular interaction26 of N:fCdOH+
(leading to a transient aziridinium intermediate), which would
facilitate the benzylic C-N bond reductive cleavage.27 The
electron-rich aromatic system certainly contributed to this
facile process.
Acidic Wacker oxidation (vide supra) of 14 gave methyl
ketone 15 in 74% yield. However, the attempted generation
of iminium intermediate 4 from 15 by the oxidative action28
of Hg(OAc)2 resulted in a slow decomposition of 15 under
a variety of reaction conditions. In case an alternative
Polonovski-Potier protocol29 might be more selective, the
corresponding amine oxide of 15 prepared by m-CPBA
oxidation was treated with excess trifluoroacetic anhydride
(TFAA) in CH2Cl2, and enamine ketone 16 was obtained as
the sole product in 70% yield, which was apparently formed
by isomerization of the initially generated iminium salt 4.
Compound 16 had been previously synthesized6 by an
alternative enamine alkylation of the so-called Dolby-
Weinreb enamine,1b and it had been claimed6 by Weinreb
and Weinstein that cyclization to 2 (formally an endocyclic
enamine annulation) could not be realized under various
conditions, although Dolby et al. reported that acid-catalyzed
Scheme 3. Enamine Cyclopentenone Annulation Pathway to 2
(18) We thank Professor Xiao-Tian Liang of Beijing Institute of Materia
Medica for a gift of natural cephalotaxine.
(19) Cf. ref 1b, pp 219-220. See also: Chem. Abstr. 1995, 122, 56276.
(20) For an excellent review, see: Overman, L. E.; Ricca, D. J. In
ComprehensiVe Organic Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon
Press: Oxford, 1991; Vol. 3, pp 1007-1046.
(21) (a) Clemo, G. R.; Ramage, G. R. J. Chem. Soc. 1931, 437. (b)
Clemo, G. R.; Raper, R.; Vipond, H. J. J. Chem. Soc. 1949, 2095.
(22) Prelog, V.; Seiwerth, R. Chem. Ber. 1939, 72, 1638.
(23) (a) Leonard, N. J.; Wildman, W. C. J. Am. Chem. Soc. 1949, 71,
3089. (b) Leonard, N. J.; Ruyle, W. V. J. Am. Chem. Soc. 1949, 71, 3094.
(c) Leonard, N. J.; Barthel, E.; Jr. J. Am. Chem. Soc. 1949, 71, 3098. (d)
Leonard, N. J.; Curry, J. W.; Sagura, J. J. J. Am. Chem. Soc. 1953, 75,
6249 and previous articles in this series. (e) Wilson, W. Chem. Ind. (London)
1955, 200. (f) Brewster, J. H. J. Am. Chem. Soc. 1954, 76, 6364.
(24) Vedejs, E. Org. React. 1975, 22, 401.
(25) Studies along this line toward natural product synthesis are in
progress and will be reported in due course.
reduction conditions,24 with zinc dust in hot glacial acetic
acid, were also equally effective for this reductive rearrange-
ment. This convenient and mild procedure is well suited to
multifunctional substrates.25
(26) For a review, see: Leonard, N. J. Rec. Chem. Prog. 1956, 17, 243.
(27) Cf.: (a) Gaskell, A. J.; Joule, J. A. Tetrahedron 1968, 24, 5115. In
contrast, see: (b) Noe´, E.; Se´raphin, D.; Zhang, Q.; Djate´, F.; He´nin, J.;
Laronze, J.-Y.; Le´vy, J. Tetrahedron Lett. 1996, 37, 5701. (c) A¨ıt-Mohand,
S.; Noe´, E.; He´nin, J.; Laronze, J.-Y. Eur. J. Org. Chem. 1999, 3429.
(28) For leading references, see: (a) Leonard, N. J.; Hay, A. S.; Fulmer,
R. W.; Gash, V. W. J. Am. Chem. Soc. 1955, 77, 439. (b) Leonard, N. J.;
Fulmer, R. W.; Hay, A. S. J. Am. Chem. Soc. 1956, 78, 3457.
(29) For a review, see: Grierson, D. S.; Husson, H.-P. In ComprehensiVe
Organic Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press: Oxford,
1991; Vol. 6, pp 910-947.
(16) Cf.: (a) Yasuda, S.; Yamada, T.; Hanaoka, M. Tetrahedron Lett.
1986, 27, 2023. (b) Yasuda, S.; Yamamoto, Y.; Yoshida, S.; Hanaoka, M.
Chem. Pharm. Bull. 1988, 36, 4229.
(17) For example, see: Woodward, R. B. (work with Volpp, G.;
Gougoutas, J. Z.) The HarVey Lectures 1963, 31.
Org. Lett., Vol. 5, No. 16, 2003
2933

cyclization of 16 led to a rearranged product (saturated enone
3) in low yield.6a
Catalytic hydrogenation (H2, 10% Pd-C, 10% HOAc-
EtOH) of pentacyclic enone 17 gave the same pentacyclic
ketone 2 described above in 76% yield. Interestingly, the
action of zinc dust in hot HOAc produced smoothly ketone
2 (exclusively) without causing skeletal rearrangement
(Scheme 6), which provides further support for our mecha-
Although keto enamine 16 appeared not to be a promising
intermediate,6 a series of experiments on acidic treatment
of 16 was examined. To our surprise, a hot mixture of 40%
acetic acid and 16 in an open flask (exposure to air) generated
gradually a sole isolable product in ca. 30% yield whose
structure was verified as the pentacyclic enone 17 (spectro-
scopically).30 Further experimentation improved the yield in
this cyclization to 57% (glacial acetic acid as solvent, air,
and a stoichiometric amount of FeSO4). It is evident31 that
this cyclization is an acid-catalyzed O2-dependent process,
which we reasoned to be an unusual azo-Nazarov-type
cyclization32 (Scheme 5) initiated by an acid-catalyzed
autoxidation.
Scheme 6. Zinc-Acid Reduction Pathway for 17 f 2
nistic rational (Scheme 2) for the transannular reductive
rearrangement of enone 3 to pentacyclic ketone 2. It is worthy
to note that the isomeric resonance structures i vs iv and ii
vs iii led to different reduction products as shown in Schemes
2 and 6, respectively.
Scheme 5. Oxidative Cyclization Pathway for 16 f 17
In conclusion, a short, efficient, and practical total
synthesis of CET has been developed based a conceptually
novel strategy featuring transannular reductive rearrange-
ments as key transformations for establishing the pentacyclic
ring skeleton. The efficiency and practicality of the synthesis
follow from its brevity (9 stages from diester 7), the ready
availability of starting materials, the operationally simple
reactions with cheap reagent chemicals, and good overall
yield. Finally, this synthesis of CET may be regarded as a
biomimetic synthesis35 in terms of pentacyclic ring construc-
tion.7 Further work on the asymmetric synthesis of CET
based on this strategy is underway in our laboratory.
Related acid-catalyzed, oxidative cyclizations can be found
in Woodward’s synthesis of chlorophyll-a33 and in an oxy-
Nazarov cyclization5c in Weinreb’s pioneering CET synthe-
sis.6b Since endocyclic enamine 16 is a direct enamine
alkylation product, this unusual cyclization can be regarded
as a novel endocyclic enamine cyclopentenone annulation.34
Such processes could have broad applications to the synthesis
of complex polycyclic natural alkaloids and other hetero-
cycles.
Acknowledgment. This paper is dedicated to Professor
E. J. Corey on the occasion of his 75th birthday. We thank
the National Natural Science Foundation (Distinguished
Youth Fund 29925204 and Fund 20021001), the Fok Ying
Tung Education Foundation (Fund 71012), and the Ministry
of National Education (Funds 99114 and 2000-66) for
financial support. The Cheung Kong Scholars program is
gratefully acknowledged.
(30) Dolby et al. had observed6a that, on treatment of keto enamine 16
with a hot mixture of HOAc-NaOAc, the formation of a new product with
practically identical IR maxima absorptions with that of 17, but did not
further characterize its structure.
(31) Strict exclusion of O2 from the reaction mixture led to no apparent
reaction even after refluxing for 24 h. Oxidants other than O2 can also be
used, but they are much less efficient. Since the additive Fe2+ appears to
accelerate the reaction with clean conversion, Fe2+ may act as a peroxide
reductant. In contrast, for an alternative likely radical autoxidation of
enamine catalyzed by Fe3+, see: Malhotra, S. K.; Hostynek, J. J.; Lundin,
A. F. J. Am. Chem. Soc. 1968, 90, 6565.
(32) Habermas, K. L.; Denmark, S. E. Org. React. 1994, 45, 1.
(33) Woodward, R. B.; Ayer, W. A.; Beaton, J. M.; Bickelhaupt, F.;
Bonnett, R.; Buchschacher, P.; Closs, G. L.; Dutler, H.; Hannah, J.; Hauck,
F. P.; Ito, S.; Langemann, A.; Le Goff, E.; Leimgruber, W.; Lwowski, W.;
Sauer, J.; Valenta, Z.; Volz, H. Tetrahedron 1990, 46, 7599.
(34) For a review on enamine annulations, see: Stevens, R. V. Acc.
Chem. Res. 1977, 10, 193.
Supporting Information Available: Experimental pro-
cedures and spectral data of compounds 2, 3, 5, 7, 9-11,
and 13-17. This material is available free of charge via the
Internet at http://pubs.acs.org.
OL035098B
(35) For attempted biomimetic synthesis, see: (a) Marino, J. P.; Samanen,
J. M. J. Org. Chem. 1976, 41, 179. (b) Kupchan, S. M.; Dhingra, O. P.;
Kim, C.-K. J. Org. Chem. 1978, 43, 4464.
2934
Org. Lett., Vol. 5, No. 16, 2003