Total Synthesis of (?)-Cephalotaxine and (?)-Homoharringtonine via Furan Oxidation–Transannular Mannich Cyclization

Article abstract of DOI:10.1002/anie.201902174

Homoharringtonine and its congener cephalotaxine were synthesized. Oxidative ring-opening of a furan unveils an amine-tethered dicarbonyl, which undergoes spontaneous transannular Mannich cyclization. The cascade builds the full cephalotaxine substructure in a single operation in 60 % yield. A Noyori reduction enabled synthesis of the title compounds with excellent enantioselectivity (k_rel=278).

Full text of DOI:10.1002/anie.201902174

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Total Synthesis of (–)-Cephalotaxine and (–)-Homoharringtonine
via Furan Oxidation-Transannular Mannich Cyclization
Authors: Xuan Ju and Christopher Beaudry
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201902174
Angew. Chem. 10.1002/ange.201902174
Link to VoR: http://dx.doi.org/10.1002/anie.201902174
http://dx.doi.org/10.1002/ange.201902174

10.1002/anie.201902174
Angewandte Chemie International Edition
COMMUNICATION
Total Synthesis of (–)-Cephalotaxine and (–)-Homoharringtonine
via Furan Oxidation-Transannular Mannich Cyclization
Xuan Ju and Christopher M. Beaudry*
Abstract: Homoharringtonine and its congener, cephalotaxine, were
synthesized. Oxidative ring-opening of a furan unveils an amine-
tethered dicarbonyl, which undergoes spontaneous transannular
Mannich cyclization. The cascade builds the full cephalotaxine
substructure in a single operation in 60% yield. A Noyori reduction
enabled synthesis of the title compounds with excellent
eantioselectivity (krel = 278).
cephalotaxine substructure.10 We decided to develop an efficient
synthesis of these natural cephalotaxus alkaloids and synthetic
congeners.
Cephalotaxine (1) has been prepared by borohydride
reduction of cephalotaxinone (2). The β-aminoketone functional
group in 2 suggested preparation through a Mannich cyclization
of intermediate 3 (Scheme 1). Retrosynthetic simplification of
the iminium ion leads to 1,4-diketone 4. Unsaturated diketone 4
could be made by oxidative opening of the corresponding
macrocyclic furan 5. 11 Strategically, use of the furan was
appealing because its aromaticity would mask the reactive
carbon atoms of the unsaturated 1,4-diketone prior to the key
Mannich cyclization event.
(–)-Homoharringtonine (HHT, Figure 1) is
a polycyclic
alkaloid isolated from the plum yew Cephalotaxus harringtonii.1
HHT binds to the peptidyl transferase center of the human
ribosome and inhibits protein translation. 2 (–)-HHT shows
nanomolar cytotoxcity against leukemia cells, and after a long
evaluation as a chemotherapeutic, it was approved for use in the
US in 2012 for the treatment of chronic myeloid leukemia.3
Total syntheses of HHT have been reported4; however, HHT
is prepared commercially by esterification of (–)-cephalotaxine
5
(1)
, a related alkaloid also found in C. harringtonii.
Cephalotaxine is more abundant in the plant (ca. 50% of the
alkaloid content), but it lacks the ester sidechain of HHT and is
biologically inactive. Cephalotaxine was first prepared by
Weinreb in 1972,6 and it has been prepared dozens of times
during the past decades.7,8 Despite such efforts, commercial
HHT is still prepared from plant-derived cephalotaxine.
Scheme 1. Synthetic Plan for Cephalotaxinone.
Intermediate 5 is a 12-membered macrocycle, and synthesis
of such medium-sized rings is not trivial. However, we
hypothesized that the same factors complicating medium ring
synthesis, namely ring strain and transannular interactions,
would increase the propensity of the substrate to undergo a
dearomatization and subsequent transannular Mannich reaction.
Figure 1. Cephalotaxus alkaloids.
We envisioned formation of the C8–N bond (cephalotaxine
numbering) arising from a phenethylamine and an oxygenated
carbon; either by a substitution reaction or by a reductive
amination. However, an optimal method to construct the
diarylmethylene group was less obvious. We considered
preparing 5 through construction of either the C3–C4 or C4–C13
bond. Initially, we decided on pursuing formation of the C4–C13
bond because structures 6 were well-known, and coupling
partner 7 resembled molecules Wasserman and coworkers
reported in a single-operation furan synthesis.12
Many side-chain analogs of HHT are known and have been
evaluated,9 a fact that is perhaps unsurprising considering the
semi-synthetic providence of commercial HHT, and the
dependence of biological activity on the ester side chain. Some
HHT ester congeners have exciting biological activities.
Crystallographic studies suggest that the amine-substituted
benzodioxole portion of HHT (i.e. the “cephalotaxine portion”) is
implicated in binding to the ribosome.2a However, relatively few
congeners of HHT are known that vary the polycyclic
Wasserman’s strategy was extended to compound 7, which
contains a primary alcohol (Scheme 2). Alcohol 7 could be
coupled with known amine 8 through reductive amination,
forming 9. However, conditions to give macrocycle 5 could not
be identified. Interestingly, the ester functionality in 9 was
surprisingly resistant to any types of hydride reduction or
nucleophilic additions. Similarly, the ester in 7 (or alcohol
protected congeners) was resistant to reduction or
intermolecular addition of nucleophilic reagents derived from 8.
Potentially, the lack of electrophilicity of 7 and 9 arises from the
[a]
X. Ju and Prof. C. M. Beaudry
Department of Chemistry
Oregon State University
153 Gilbert Hall
Corvallis, OR 97333, USA
E-mail: christopher.beaudry@oregonstate.edu
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/anie.201902174
Angewandte Chemie International Edition
COMMUNICATION
lone pairs on the methoxy group and the furanyl oxygen, which
can donate to the ester carbonyl, decreasing reactivity.
suspected that polar non-nucleophilic solvents would stabilize
the charged reactive intermediates without solvolytic trapping,
and could increase the reaction yield. In the event, we
discovered that trifluoroethanol markedly increased the reaction
yield to 60%.
Scheme 2. Wasserman-based Furan Synthesis.
Scheme 4. Racemic Synthesis of Cephalotaxine.
We next decided to take an alternative approach involving
formation of the C4–C3 bond by way of a Friedel–Crafts
Mechanistically, single electron oxidation of 5 gives a radical
cation 16. Transannular trapping of the radical oxocarbenium ion
by the tethered amine results in 17. Oxidation of 17 gives
oxocarbenium ion 18. Fragmentation of 18 releases iminium ion
3, which can proceed via Mannich cyclization to give (±)-
cephalotaxinone (2) as a single diasteromer.
13
alkylation
hydroxypentanone 10,
of
a
furan. Beginning with TBS-protected
14
Claisen-type condensation and
cyclization gave furanone 11.15 Methylation gave methoxyfuran
12. Subjection of furan 12 to known benzyl chloride 1316
(prepared from homopiperonylamine) gave smooth bond
formation yielding 14. The protected alcohol was activated as
the corresponding mesylate (15). Treatment of 15 with aqueous
base led to cyclization with concomitant removal of the TFA
group to give macrocycle 5.17,18 This final cyclization was quite
robust; no dimer or oligomers were detected during the
formation of the 12-membered ring. Moreover, the reaction was
conducted on scales as large as 650 mg with good isolated
yields of 5.
Synthesis of non-racemic cephalotaxinone could, at least in
principle, be accomplished by rendering the key transformation
(5 à 2) enantioselective; however, it was unclear which step of
the mechanism would need to be controlled. Considering the
plausible mechanism shown, the enantiodetermining step may
be the final Mannich event. Alternatively, if hypothetical
intermediate 3 has a conformation that is stable on the timescale
of the Mannich step, then there would be a possibility of
conformational chirality of the undecatrienone ring. In such a
scenario, the initial transannular addition to the furanyl radical
cation 16 could be enantiodetermining. Finally, such a scenario
would require good transfer of point-to-conformational chirality,
and non-reversibility of the mechanistic steps.
In these scenarios, the enantiodeterming step is an addition to
a
cationic
species.
The
counter
anion
is
dichlorodicyanohydroquinone anion. Efforts to control this
process by introducing a chiral counter anion were unsuccessful,
despite adding several chiral protic acids in multiple solvents. No
measureable enantioenrichment of cephalotaxinone was ever
observed.
We next pursued a resolution-based strategy for the synthesis
of enantioenriched cephalotaxine. No enantioselective reduction
of 2 has been reported. Gratifyingly, Noyori’s conditions19 gave
excellent results, producing (–)-cephalotaxine (1) with very high
Scheme 3. Synthesis of the Macrocycle.
enantioselectivity (97% ee) at 50% conversion (Scheme 5, krel
=
278). Spectroscopic data and optical rotation for (–)-
cephalotaxine matched those previously reported.4–7 The major
enantiomer of 1 is predicted by Noyori’s model and was verified
by comparison to previously reported optical rotation data. (–)-
Cephalotaxine was then advanced to (–)-HHT following the
procedure of Gin.4 Spectral data and optical rotation for (–)-HHT
also matched those from previous reports.4,5
Key intermediate 5 was first subjected to DDQ in tert-butanol
as reported by Sayama (Scheme 4).11 Gratifyingly, the reaction
gave cephalotaxinone, albeit in low yield (~10%). Oxidative furan
openings are responsive to the reaction solvent, and the
oxidations often result in incorporation of the alcoholic solvent to
the furan. Sayama found that sterically hindered (i.e. non-
nucleophilic) solvents tend to give less solvolysis products. We
This article is protected by copyright. All rights reserved.

10.1002/anie.201902174
Angewandte Chemie International Edition
COMMUNICATION
[1] R. G. Powell, D. Weisleder, C. R. Smith, W. K. Rohwedder Tetrahedron
Lett. 1970, 11, 815–818.
[2] (a) N. G. de Loubresse, I. Prokhorova, W. Holtkamp, M. V. Rodnina, G.
Yusupova, M. Yusupov Nature 2014, 513, 517–523, (b) M. Fresno, A.
Jiménez, D. Vázquez Eur. J. Biochem. 1977, 72, 323–330; (c) R. M.
Tujebajeva, D. M. Graifer, G. G. Karpova, N. A. Ajtkhozhina FEBS Lett.
1989, 257, 254–256.
[3] F. Alvandi, V. E. Kwitkowski, C.-W. Ko, M. D. Rothmann, S. Ricci, H. Saber,
D. Ghosh, J. Brown, E. Pfeiler, E. Chikhale, J. Grillo, J. Bullock, R.
Kane, E. Kaminskas, A. T. Farrell, R. Pazdur Oncologist 2014, 19, 94–
99.
[4] (a) J. D. Eckelbarger, J. T. Wilmot, M. T. Epperson, C. S. Thakur, D. Shum,
C. Antczak, L. Tarassishin, H. Djaballah, D. Y. Gin Chem. Eur. J. 2008,
14, 4293–4306; (b) M. Marguerit, G. Little, Y. Wang, L. He, S. Allwein, J.
Reif, J. Rossi, R. Roemmele, R. Bakale Eur. J. Org. Chem. 2015, 36,
8003–8010.
Scheme 5. Kinetic Resolution of Cephalotaxine.
With the undesired (+)-2 in hand, attempts were made to
racemize this material. Previously, Mori
[5] (a) Z. Zhao, Y. Xi, H. Zhao, J. Hou, J. Zhang, Z. Wang Acta Pharmaceutica
Sinica 1980, 15, 46–49; (b) Y.-K. Wang, Y.-L. Li, X.-F. Pan, S.-B. Li,
W.-K. Huang Acta Chimica Sinica 1985, 43, 161–167; (c) S. Hiranuma,
T. Hudlicky Tetrahedron Lett. 1982, 23, 3431–3434; (d) S. Hiranuma, M.
Shibata, T. Hudlicky J. Org. Chem. 1983, 48, 5321–5326; (e) J.-C.
Cheng, J.-H. Zhang, Q.-B. Zhang, J. Yang, L. Huang Yaoxue Xuebao,
1984, 19, 178–183; (f) J.-P. Robin, R. Dhal, G. Dujardin, L. Girodier, L.
Mevellec, S. Poutot Tetrahedron Lett. 1999, 40, 2931–2934; (g) G.-Z.
Ma, P.-F. Li, L. Liu, W.-D. Z. Li, L. Chen Org. Lett. 2017, 19, 2250−2253.
[6] (a) J. Auerbach, S. M. Weinreb J. Am. Chem. Soc. 1972, 94, 7172–7173;
(b) S. M. Weinreb, J. Auerbach J. Am. Chem. Soc. 1975, 97, 2503–
2506.
20
, 22
and Stoltz 21
observed racemization during the methylation of (–)-19 to give
(±)-2 (Scheme 6). They attributed the racemization to formation
of intermediate achiral cyclopentadienone 20. This suggested
racemization of (+)-2 should be possible, and it would allow
recycling of the unwanted cephalotaxinone enantiomer.
Gratifyingly, subjection of (+)-2 to Mori/Stoltz conditions resulted
in racemization, giving (±)-2. This allows simple recycling of the
unwanted enantiomer of cephalotaxinone.23
[7] For review of cephalotaxine syntheses from 1975–2011, see: H. Abdelkafi,
B. Nay Nat. Prod. Rep. 2012, 29, 845–869.
[8] (a) P. Xing, Z.-G. Huang, Y. Jin, B. Jiang A Synthesis 2013, 45, 596–600;
(b) M. G. Gonçalves-Martin, S. Zigmantas, P. Renaud Helv. Chim. Acta
2012, 95, 2502–2514; (c) Q.-W. Zhang, K. Xiang, Y.-Q. Tu, S.-Y. Zhang,
X.-M. Zhang, Y.-M. Zhao, T.-C. Zhang Chem. Asian J. 2012, 7, 894–
898; (d) Z.-W. Zhang, X.-F. Zhang, J. Feng, Y.-H. Yang, C.-C. Wang,
J.-C. Feng, S. Liu J. Org. Chem. 2013, 78, 786−790; (e) K.-J. Xiao, J.-
M. Luo, X.-E. Xia, Y. Wang, P.-Q. Huang Chem. Eur. J. 2013, 19,
13075–13086; (f) S.-H. Huang, X. Tian, X. Mi, Y. Wang, R. Hong
Tetrahedron Lett. 2015, 56, 6656–6658; (g) H. Liu, J. Yu, X. Li, R. Yan,
J.-C. Xiao, R. Hong Org. Lett. 2015, 17, 4444−4447; (h) P. Gouthami, R.
Chegondi, S. Chandrasekhar Org. Lett. 2016, 18, 2044−2046; (i) X.-Y.
Ma, X.-T. An, X.-H. Zhao, J.-Y. Du, Y.-H. Deng, X.-Z. Zhang, C.-A. Fan
Org. Lett. 2017, 19, 2965−2968; (j) Z.-W. Zhang, C.-C. Wang, H. Xue,
Y. Dong, J.-H. Yang, S. Liu, W.-Q. Liu, W.-D. Z. Li Org. Lett. 2018, 20,
1050−1053
Scheme 6. Racemization of Cephalotaxinone.
In summary, we have completed a total synthesis of (–)-
cephalotaxine
transformation involved an oxidative furan opening with
spontaneous transannular Mannnich reaction. Noyori
and
(–)-homoharringtonine.
Our
key
A
reduction converts (±)-cephalotaxinone to (–)-cephalotaxine with
excellent enantioselectivity. The unwanted enantiomer of
cephalotaxinone can be recycled by racemization in high yield.
Overall, our synthesis of (–)-cephalotaxine is 9 steps and occurs
in >5% chemical yield, which compares well with previous
syntheses.
[9] (a) D.-Z. Wang, G.-E. Ma, R.-S. Xu Acta Pharmaceutica Sinica 1992, 27,
178−184. (b) D. Gin, J. Wilmot, H. Djaballah US20110071097A1, Mar
24, 2011; (c) L. Chen, W.-D. Li CN102675327A, Sep 19, 2012; (d) F.
Rong, R. Xu, F. Xie, H. Lai US20140206669A1, Jul 24, 2014; (e) R. Xu,
F. Rong, F. Xie, H. Lai US20140303147A1, Oct 9, 2014.
[10] (a) J. M. Cassady, P.-Z. Cong, R. G. Cooks, R. Roush, C. J. Chang, R. G.
Powell Acta Pharmaceutica Sinica 1988, 23, 351–355; (b) I. Takano, I.
Yasuda, M. Nishijima, Y. Hitotsuyanagi, K. Takeya, H. Itokawa
Tetrahedron Lett. 1996, 37, 7053–7054; (c) H. Morita, M. Arisaka, N.
Yoshida, J. Kobayashi Tetrahedron 2000, 56, 2929–2934; (d) M.
Yoshinaga, H. Morita, T. Dota, J. Kobayashi Tetrahedron 2004, 60,
7861–7868.
Acknowledgements
We acknowledge funding from NSF (CHE-1465287) and Oregon
State University. This paper is dedicated to Professor Larry E.
Overman.
[11] (a) G. Piancatelli, A. Scettri, M. D'Auria Tetrahedron 1980, 36, 661–663;
(b) Y. Kobayashi, H. Katsuno, F. Sato Chem. Lett. 1983, 1771–1774;
(c) L. Lepage, Y. Lepage Synthesis 1983, 12, 1018–1019; (d) K.
Gollnick, A. Griesbeck Tetrahedron 1985, 41, 2057–2068; (e) C.
Dominguez, A. G. Csaky, J. Plumet Tetrahedron Lett. 1990, 31, 7669–
7670; (f) B. J. Adger, C. Barrett, J. Brennan, P. McGuigan, M. A.
McKervey, B. Tarbit J. Chem. Soc., Chem. Commun. 1993, 15, 1220–
1222; (g) J. Wahlen, B. Moens, D. E. De Vos, P. L. Alsters, P. A.
Keywords: homoharringtonine • cephalotaxine • alkaloid •
Mannich • total synthesis
This article is protected by copyright. All rights reserved.

10.1002/anie.201902174
Angewandte Chemie International Edition
COMMUNICATION
Jacobs Adv. Synth. Catal. 2004, 346, 333–338; (h) S. Sayama
Heterocycles 2005, 65, 1347–1358; (i) S. Sayama Synth. Commun.
2007, 37, 3067–3075.
[ 18 ] Tietze also formed this C8–N bond of cephalotaxine using an
intermolecular alkylation of a primary amine and a tosylate. See: L. F.
Tietze, H. Schirok, M. Wöhrmann, K. Schrader Eur. J. Org. Chem. 2000,
2433–2444.
[12] H. H. Wasserman, G. M. Lee Tetrahedron Lett. 1994, 35, 9783–9786.
[13] A brief solvent screen was performed for this reaction. See supporting
information.
[19] K. J. Haack, S. Hashiguchi, A. Fujii, T. Ikariya, R. Noyori Angew. Chem.
Int. Ed. 1997, 36, 285−288.
[14] L. Liu, P. E. Floreancig Org. Lett. 2009, 11, 3152−3155.
[15] J. D. Winkler, K. Oh, S. M. Asselin Org. Lett. 2005, 7, 387−389.
[16] Y. Onozaki, N. Kurono, H. Senboku, M. Tokuda, K. Orito J. Org. Chem.
2009, 74, 5486–5495.
[20] N. Isono, M. Mori J. Org. Chem. 1995, 60, 115–119.
[21] Stoltz and coworkers observed similar reactivity in the drupacine series,
which contains a C11 acetoxy group. Thus, the reversible conjugate
addition equilibrated diastereomers not enantiomers. See Ref. 22.
[22] Q. Liu, E. M. Ferreira, B. M. Stoltz J. Org. Chem. 2007, 72, 7352–7358.
[23] Preliminary investigations into combining the racemization of (+)-2 and the
Noyori reduction in a dynamic kinetic resolution were unsuccessful;
however, this work is ongoing and will be reported in due course.
[17] Presumably, the trifluoroacetamide is first removed under the conditions,
and a subsequent intramolecular substitution reaction occurs to form
the 12-membered ring.
If the reaction is conducted at lower
temperatures, the corresponding primary amine can be isolated, and
resubjection of this material to the cyclization conditions results in
formation of macrocyclic intermediate 5.
This article is protected by copyright. All rights reserved.

10.1002/anie.201902174
Angewandte Chemie International Edition
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 2:
COMMUNICATION
Xuan Ju and Prof. Dr. Christopher M.
Beaudry*
Page No. – Page No.
Total Synthesis of (–)-Cephalotaxine
and (–)-Homoharringtonine via Furan
Oxidation-Transannular Mannich
Cyclization
Oxidative ring-opening of a furan unveils an amine-tethered dicarbonyl, which
undergoes spontaneous transannular Mannich cyclization. The cascade builds the
full cephalotaxine substructure in a single operation in 60% yield. A Noyori
reduction enabled synthesis of the title compounds with excellent eantioselectivity
(krel = 278).
This article is protected by copyright. All rights reserved.