Enantioselective synthesis of (R)-1-azaspiro-[4.4]nonane-2,6-dione ethylene ketal, key chiral intermediate in the elaboration of (-)-cephalotaxine

Article abstract of DOI:10.3987/COM-05-S(K)3

An enantioselective synthesis of (R)-1-azaspiro[4.4]nonane-2,6-dione ethylene ketal (4), chiral relay in the elaboration of (-)-cephalotaxine (1b), starting from (R)-1-(2-methoxycarbonylethyl)-2-oxo-cyclopentanecarboxylic acid methyl ester ((R)-10) (8 chemical operations, 26% overall yield, ee ≥ 95%) is described. The Curtius rearrangement of acyl azide (12) was used as the key strategic element to install with a perfect stereochemical fidelity and a high efficiency an α-nitrogen substituent at the quaternary carbon center in 10.

Full text of DOI:10.3987/COM-05-S(K)3

HETEROCYCLES, Vol. 66, 2005
31
HETEROCYCLES, Vol. 66, 2005, pp. 31 – 37. © The Japan Institute of Heterocyclic Chemistry
Received, 13th April, 2005, Accepted, 9th June, 2005, Published online, 10th June, 2005. COM-05-S(K)3
ENANTIOSELECTIVE SYNTHESIS OF (R)-1-AZASPIRO-
[4.4]NONANE-2,6-DIONE ETHYLENE KETAL, KEY CHIRAL
INTERMEDIATE IN THE ELABORATION OF (-)-CEPHALOTAXINE
Mathieu Pizzonero, Françoise Dumas,* and Jean d’Angelo*
Department of Organic Chemistry, BioCIS, UMR 8076 CNRS, School of
Pharmacy, Paris XI University, 5, rue J.-B. Clément, F-92296 Châtenay-Malabry,
France. francoise.dumas@cep.u-psud, jean.dangelo@cep.u-psud.fr
Abstract – An enantioselective synthesis of (R)-1-azaspiro[4.4]nonane-2,6-dione
ethylene ketal (4), chiral relay in the elaboration of (-)-cephalotaxine (1b), starting
from (R)-1-(2-methoxycarbonylethyl)-2-oxo-cyclopentanecarboxylic acid methyl
ester ((R)-10) (8 chemical operations, 26% overall yield, ee ≥ 95%) is described.
The Curtius rearrangement of acyl azide (12) was used as the key strategic
element to install with a perfect stereochemical fidelity and a high efficiency an
α-nitrogen substituent at the quaternary carbon center in 10.
INTRODUCTION
Besides highly promising antileukemic properties exhibited by several naturally occurring esters
exemplified by homoharringtonine, HHT, 1a,1 (-)-cephalotaxine (1b) has become an attractive target for
the development of new synthetic methodology because of its unique backbone architecture,consisting of a
1-azaspiro[4.4]nonane unit fused to a benzazepine nucleus.2 As a result of more than 30 years of intensive
research, numerous approaches to (±)-1b have been developed.3 However, structure-activity relationships
of esters of cephalotaxine have demonstrated the stringent stereospecificity requirements for their
antitumor potency. Regarding the cephalotaxine moiety, it has been established that esters of “unnatural”
(+)-cephalotaxine (ent-1b) are devoid of antileukemic activity,4 a factor that has stimulated interest in the
enantioselective synthesis of 1b.5 One of the most efficient strategies in the area has featured the use of
derivatives of (R)-1-azaspiro[4.4]nonane-2,6-dione (2) harboring the future stereogenic center C-5 of 1b.
Since the correct stereochemistry at the two further stereogenic centers C-3 and C-4 in 1b has been readily
secured during the completion of the synthesis,6 the elaboration of CD subunit (R)-2 constitutes an enantio-
This paper is dedicated to the memory of the Emeritus Professor Kenji Koga of Tokyo University

32
HETEROCYCLES, Vol. 66, 2005
and diastereoselective approach to (-)-cephalotaxine in a formal sense. However, as the implementation of
the synthesis requires the coupling of a N-metallated derivative of 2 with an arylethyl tosylate (installation
of ring Atogether with two carbon atoms of future ring B), the very reactive cyclopentanone carbonyl of 2
needs to be protected. For that reason, the two strategies already reported for the synthesis of (R)-2 have
both focused on its ketal derivatives: ketal (3) was elaborated via HPLC separation of diastereomeric ketals
of racemic ketone (2),5b and ketal ((R)-4) was synthesized according to a semipinacolic
rearrangement-based enantioselective approach, employing (S)-1-(1-naphthyl)ethylamine as the chiral
inducer5e (Chart 1).
Chart 1. Structures of homoharringtonine (1a), cephalotaxine (1b) and key chiral spiro subunits 2-4.
O
O
HO
HO
CO
CO Me
B
A
N
O
H
N
H
N
H
N
O
O
O
O
5
1a: R =
C
4
O
O
O
H
D
2
3
RO
1b: R = H
2
3
4
H
OMe
RESULTS AND DISCUSSION
In this paper, we report an efficient enantioselective synthesis of cornerstone ketal ((R)-4), founded on the
asymmetric Michael reaction involving chiral imines/secondary enamines derived from enantiopure
1-phenylethylamine. This methodology, outlined in Scheme 1, has played a paramount role in asymmetric
synthesis, particularly for the enantioselective elaboration of quaternary carbon centers.7
Scheme 1. Asymmetric Michael reaction involving chiral imines/ secondary enamines.
Ph
H
Me
N
O
R
EWG
R = alkyl, OR, SPh, CO2R, CN
R
EWG = COR, CO2R, CN, NO2, SO2R
EWG
ee = 90-98%
then H3O+
5
6
However, while this reaction tolerates the presence of α-alkoxy8 or α-phenylthio9 substituents in starting
imine (5), all efforts devoted to the synthesis of adducts of type (6) bearing an α-nitrogen linkage at the
quaternary carbon center, utilizing either a variety of α-nitrogen substituents in imine (5) or DEAD as the
Michael acceptor, were fruitless.7a,b In light of this vexing outcome, we turned to an alternative “indirect”
route based on the Curtius rearrangement of an acyl azide.10
The starting material in the present approach, keto diester ((R)-10), was elaborated in 85% yield and an ee
≥ 95% through Michael addition of enamino ester ((R)-9) to methyl acrylate.11 Our initial synthetic efforts

HETEROCYCLES, Vol. 66, 2005
33
were focused on the key Curtius rearrangement. However, since the planned rearrangement sequence
necessarily proceeds via the intermediacy of a carboxylic acid linked to the quaternary carbon center, two
critical problems should first be resolved: the discrimination between the two carbomethoxy groups of 10
and the masking of its carbonyl group to circumvent the generation of a β-keto acid species, of notorious
thermal instability.
These two prerequisites were simultaneously satisfied through subjection of 10 to sequential reduction
(NaBH4, CeCl3.7H2O in MeOH, -78 °C)12 and saponification (LiOH in MeOH/H2O, followed by acidic
treatment) to provide an intermediary hydroxy diacid, the spontaneous lactonization of which furnished in
75% yield diastereomerically pure lactone acid (11), a stereochemical homogeneity revealed by the
presence of a single set of nine resonance peaks in its proton-decoupled 13C NMR spectrum. It should be
pointed out that the above lactonization process achieved the concomitant “internal protection” of the
secondary hydroxyl function and propanoic carboxyl group of the transient hydroxy diacid, hence
implicitly ensuring a clear-cut chemoselective discrimination between the ester groups of adduct (10).
Although the configuration of the newly created stereogenic center in bicyclic lactone (11) could not be
ascertained spectroscopically, the depicted cis ring junction may be reasonably assigned, assuming that
the reduction step involves a transition state structure based on a ceriumIII-coordinated β-keto ester
complex, in which the incoming hydride ion is delivered on the less hindered π-face of the carbonyl group,
anti to the vicinal propanoate appendage.13 The unusually facile cyclization of the hydroxy diacid
precursor of lactone (11), which occurred at 20 °C, furnishes an additional argument in favor of the cis
ring junction assignment. Several compelling reports in the field have indeed established that the ease of
formation of fused bicyclic lactones of type (11) from parent hydroxy acids is strongly dependent on the
stereochemistry of their ring junction. By way of illustration, the hydroxy acid progenitor of cis-fused
dihydronepetalactone cyclized at 25 °C, while elaboration of the related trans-isomer required prolonged
heating in refluxing xylene.14
At this juncture, we stood ready to perform the critical Curtius rearrangement sequence. In the event, acid
(11) was first converted into acyl azide (12) [(PhO)2PON3, Na2CO3, cat. DMAP]15 which was directly
rearranged into benzyl carbamate (13) (BnOH in refluxing toluene, 75% yield from 11). Conversion of
fused bicyclic lactone (13) into a derivative possessing the requisite 1-azaspiro[4.4]nonane core was
examined next.
It was our original hope that amine (14), issuing from the hydrogenolysis of 13, might undergo an
intramolecular nucleophilic rearrangement to provide spiro lactam (15). However, hydrogenolysis of 13
(5 bar of H2, 10% Pd/C, MeOH) produced 15 in only meager yield (<10%), along with high molecular
weight material. Although there was no firm evidence that self-condensation of transient amine (14),
leading to polyamides, competes with the formation of spiro lactam (15), the low efficiency of conversion

34
HETEROCYCLES, Vol. 66, 2005
[13 → 15] may be reasonably interpreted on this basis, invoking that the expected transannular
lactamization process [14 → 15] suffers from unfavorable steric factor, on account of a severely
constrained transition state structure, in which the angle of attack of the amino group with respect to the
plane of the lactone carbonyl (ca. 90 °)16 markedly differs from “ideal” Bürgi-Dunitz trajectory (109 °).17
In view of the above result, an alternative plan was devised, in which the order of reactions was inverted,
so that the lactone ring of 13 would be opened prior to the hydrogenolysis of the benzyl carbamate moiety.
To achieve this aim, 13 was subjected to sequential oxidation (1 equiv. of aqueous NaOH/cat.
RuO2/NaIO4, followed by acidic work-up)18 and esterification (CH2N2) to afford keto ester ((R)-16) (70%
yield from 13). This was then converted into ketal ((R)-17) upon reaction with (TMSOCH2)2 in the
presence of a catalytic amount of TMSOTf (THF, -78 °C → 20 °C).19
Finally, hydrogenolysis of crude 17 (5 bar of H2, Pd(OH)2, MeOH, 50 °C) delivered our goal (R)-4 (65%
yield from 16): mp 79-80 °C, [α]D -48.6 (c 1.4, CHCl3); lit.,5e: mp 75-76 °C, [α]D -49.4 (c 1.0, CHCl3)
(Scheme 2).
Scheme 2. Synthesis of spiro lactam (4).
O
Ph
H
Me
N
H
O
O
O
H
CO2Me
CO2Me
CO2Me
Ph
H
Me
CO2Me
c
b
a
+
NH2
CO2H
9
10
11
7
8
8
O
O
O
O
OH
O
O
H
H
N
H
H
e
d
f
O
11
CON3
NHCbz
NH2
13
12
14
15
g
O
H
O
O
O
O
NHCbz
CO2Me
h
NHCbz
CO2Me
i
N
O
17
16
4
Reagents and conditions: (a) cyclohexane, 80 °C, 12 h, 90%. (b) (i) 1.05 equiv. MgBr2 (0.5 M in Et2O),
1.5 equiv. methyl acrylate, THF, 20 °C, 24 h; (ii) 20% aq. AcOH, THF, 40 °C, 24 h (85%, ee >95%). (c)
(i) CeCl3.7H2O, MeOH 20 °C, then NaBH4 -78 °C; (ii) 6 M HCl, 20 °C; (iii) LiOH, H2O, 50 °C; (iv) 6 M
HCl (75 % for four steps). (d) Na2CO3, cat. DMAP, (PhO)2PON3, CH2Cl2, 20 °C (crude 12 was used in
the next step). (e) BnOH, refluxing toluene (75% for steps d and e). (f) 5 bar of H2, 10% Pd/C, MeOH,
20 °C (<10%). (g) (i) 1 equiv. 0.5 M NaOH, 50 °C; (ii) cat. RuO2, NaIO4, 20 °C; (iii) 6 M HCl; (iv)
CH2N2 (70% for four steps). (h) (TMSOCH2)2, cat. TMSOTf, CH2Cl2, -78 °C → 20 °C (crude 17 was
used in the next step). (i) 5 bar of H2, Pd(OH)2, MeOH, 50 °C (65% for steps h and i).

HETEROCYCLES, Vol. 66, 2005
35
To conclude, the synthesis of ketal ((R)-4), crucial relay in the elaboration of natural (-)-cephalotaxine
(1b), has been completed in 26% overall yield and with an ee ≥ 95% from Michael adduct ((R)-10) by a
linear sequence of eight chemical operations (mean yield per step: 85%). Critical to the success of this
endeavor was the demonstration of the viability of the Curtius rearrangement as the key tactical element
to install with a perfect stereochemical fidelity and a high efficiency an α-nitrogen substituent at the
quaternary carbon center in Michael adducts of type (6). Also of major significance was the use of readily
available and inexpensive (R)-1-phenylethylamine (8) as the chiral auxiliary,20 which was easily and
nearly quantitatively recovered without any alteration of its stereochemical integrity. Regarding the
economical viewpoint, it should be emphasized that the present synthesis takes precedence over the
previously published approaches to CD subunits (3) and (4),21 thereby offering a decisive advantage for a
potential development of therapeutically important esters of (-)-cephalotaxine and analogs.
ACKOWLEGMENTS
This work was supported by a fellowship (BDI) to M.P. from the Centre National de la Recherche
Scientifique. We thank Dr J.-P. Robin (Oncopharm SA) for fruitful discussions. We also thank Dr M.
Ourevitch for NMR assistance and Mrs S. Mairesse-Lebrun for performing elemental analyses.
REFERENCES and NOTES
1. (a) M. A. Miah Jal, T. Hudlicky, and J. W. Reed, In The Alkaloids, ed. by F. M. Manske, Academic
Press, New York, 1998, Vol. 51, pp. 199-269. For an hemisynthesis of enantiopure
homoharringtonine, see: (b) J.-P. Robin, R. Dhal, G. Dujardin, L. Girodier, L. Mevellec, and S.
Poutot, Tetrahedron Lett., 1999, 40, 2931. For an enantioselective synthesis of the ester side-chain of
homoharringtonine, see: (c) L. Keller, F. Dumas, and J. d'Angelo, Eur. J. Org. Chem., 2003, 2488.
2. For a recent contribution in this area, see: M. Pizzonero, L. Keller, F. Dumas, M. Ourevitch, G.
Morgant, A. Spasojević-de Biré, G. Bogdanović, N. E. Ghermani, and J. d’Angelo, J. Org. Chem.,
2004, 69, 4336.
3. Total syntheses of (±)-cephalotaxine: W.-D. Z. Li and Y.-Q. Wang, Org. Lett., 2003, 5, 2931 and
references cited. (b) Recent formal synthesis of (±)-cephalotaxine: W.-D. Z. Li and B.-C. Ma, J. Org.
Chem., 2005; 70, 3277.
4. (a) K. L. Mikolajczak, C. R. Smith Jr., and D. Weisleder, J. Med. Chem., 1977, 20, 328. (b) J.-P.
Robin, R. Dhal, F. Drouye, J.-P. Marie, N. Radosevic, J. Robin, K. Souchaud, and P. Bataille, U.S.
Patent 6 579 869, 2003 (Chem. Abstr., 2004, 140, 94179).
5. (a) N. Isono and M. Mori, J. Org. Chem., 1995, 60, 115. (b) T. Nagasaka, H. Sato, and S.-i. Saeki,
Tetrahedron : Asymmetry, 1997, 8, 191. (c) L. F. Tietze and H. Schirok, J. Am. Chem. Soc., 1999,

36
HETEROCYCLES, Vol. 66, 2005
121, 10264. (d) M. Ikeda, S. A. A. El Bialy, K-i. Hirose, M. Kotake, T. Sato, S. M. M. Bayomi, I. A.
Shehata, A. M. Abdelal, L. M. Gad, and T. Yakura, Chem. Pharm. Bull., 1999, 47, 983. (e) L. Planas,
J. Pérard-Viret, and J. Royer, J. Org. Chem., 2004, 69, 3087.
6. M. E. Kuehne, W. G. Bornmann, W. H. Parsons, T. D. Spitzer, J. F. Blount, and J. Zubieta, J. Org.
Chem., 1988, 53, 3439.
7. Reviews: (a) J. d'Angelo, D. Desmaële, F. Dumas, and A. Guingant, Tetrahedron: Asymmetry, 1992,
3, 459. (b) J. d'Angelo, C. Cavé, D. Desmaële, and F. Dumas, In Trends in Organic Synthesis; ed. by
S. G. Pandalai, Trivandrum, India, 1993, Vol. 4, pp. 555-615. For recent developments, see: (c) I.
Jabin, G. Révial, M. Pfau, and P. Netchitaïlo, Tetrahedron: Asymmetry, 2002, 13, 563 (d) C. Camara,
D. Joseph, F. Dumas, J. d'Angelo, and A. Chiaroni, Tetrahedron Lett., 2002, 43, 1445. (e) C.
Thominiaux, A. Chiaroni, and D. Desmaële, Tetrahedron Lett., 2002, 43, 4107. (f) J. Christoffers
and H. Scharl, Eur. J. Org. Chem., 2002, 1505. (g) Ref. 1c of the present paper. (h) C. Camara, L.
Keller, and F. Dumas, Tetrahedron: Asymmetry, 2003, 14, 3263. (i) C. Camara, L. Keller, K.
Jean-Charles, D. Joseph, and F. Dumas, Int. J. High Press. Res., 2004, 24, 149. (j) M. Danet, M.
Normand-Bayle, J. Mahuteau, J. d'Angelo, G. Morgant, and D. Desmaële, Eur. J. Org. Chem., 2004,
1911. (k) F. Hendra, M. Nour, I. Baglin, G. Morgant, and C. Cavé, Tetrahedron: Asymmetry, 2004,
15, 1027. (l) D. Desmaële, S. Delarue-Cochin, C. Cavé, J. d’Angelo, and G. Morgant, Org. Lett.,
2004, 6, 2421.
8. D. Desmaële and J. d'Angelo, Tetrahedron Lett., 1989, 30, 345.
9. M. Nour, K. Tan, C. Cavé, D. Villeneuve, D. Desmaële, J. d’Angelo, and C. Riche, Tetrahedron :
Asymmetry, 2000, 11, 2015.
10. See in relation: (a) M. Martín-Vilà, E. Muray, G. P. Aguado, A. Alvarez-Larena, V. Branchadell, C.
Minguillón, E. Giralt, and R. M. Ortuño, Tetrahedron: Asymmetry, 2000, 11, 3569. (b) A. Salgado, T.
Huybrechts, A. Eeckhaut, J. Van der Eycken, Z. Szakonyi, F. Fülöp, A. Tkachev, and N. De Kimpe,
Tetrahedron, 2001, 57, 2781.
1
11. Ee of 10 was determined using H NMR in the presence of Eu(hfc)3. (a) A. Guingant and H.
Hammami, Tetrahedron: Asymmetry, 1991, 2, 411. For a further investigation of this reaction, see:
(b) K. Tan, R. Alvarez, M. Nour, C. Cavé, A. Chiaroni, C. Riche, and J. d'Angelo, Tetrahedron Lett.,
2001, 42, 5021.
12. A. L. Gemal and J. L. Luche, J. Am. Chem. Soc., 1981, 103, 5454. However, when the reduction of
10 was performed at 20 °C, a complex mixture of products was obtained.
13. Réduction par les alumino- et borohydrures en synthèse organique ; ed. by J. Seyden-Penne,
Technique et Documentation-Lavoisier, Paris, 1988, pp. 61-64.
14. J. Ficini and J. d’Angelo, Tetrahedron Lett., 1976, 9, 687.

HETEROCYCLES, Vol. 66, 2005
37
15. C. Zhang, S. A. Lomenzo, C. J. Ballay II, and M. L. Trudell, J. Org. Chem., 1997, 62, 7888.
16. Value estimated on molecular model.
17. Stereoselective Synthesis. In Stereochemistry of Organic Compounds; ed. by E. L. Eliel and S. H.
Wilen, Wiley, New York, 1994, pp. 835-890.
18. P. E. Eaton, Y. S. Or, and S. J. Branca, J. Am. Chem. Soc., 1981, 103, 2134.
19. T. Tsunoda, M. Suzuki, and R. Noyori, Tetrahedron Lett., 1980, 21, 1357.
20. (R)-8 (99% ee) is commercially available at low cost: 73.7 € per 100 g (0.825 mole) [Supplier: Acrós
Organics, 2005].
21. (a) Preparation of 3 according to ref. 5b required tedious separation of diastereomers by HPLC. (b)
(S)-1-(1-naphthyl)ethylamine, the chiral inducer required for the synthesis of (S)-4 according to ref.
5e, is a very expensive chemical: 43.9 € per 1 g (0.0058 mole) [Supplier: Acrós Organics, 2005].
Furthermore, this inducer cannot be recovered, since further immolated through hydrogenolysis.