Formal total synthesis of (±)-vindoline by tandem radical cyclization

Article abstract of DOI:10.1021/ol0171618

(Equation presented) A formal total synthesis of (±)-vindoline 1 has been achieved featuring the tandem cyclization of radicals produced from the iodoaryl azide 19a.

Full text of DOI:10.1021/ol0171618

ORGANIC
LETTERS
2002
Vol. 4, No. 3
443-445
Formal Total Synthesis of (±)-Vindoline
by Tandem Radical Cyclization
Sheng-ze Zhou, Sacha Bommezijn, and John A. Murphy*
Department of Pure and Applied Chemistry, UniVersity of Strathclyde,
295 Cathedral Street, G1 IXL Glasgow, Scotland
john.murphy@strath.ac.uk.
Received December 2, 2001
ABSTRACT
A formal total synthesis of (±)-vindoline 1 has been achieved featuring the tandem cyclization of radicals produced from the iodoaryl azide
19a.
Vindoline 1 attracts wide attention among synthetic chemists
because of its presence in the clinically important anticancer
“Vinca dimers” vinblastine 2 and vincristine 3, which are
produced in extremely small quantities in the plant Vinca
rosea.1 Our plans to synthesize modified analogues of
vinblastine or vincristine required that we undertake the
synthesis of vindoline. To date, total syntheses have been
achieved by a number of research teams.2 Although the
presence of the condensed ring system makes vindoline an
attractive target for a route based on radical cyclization, none
of the early synthetic approaches to vindoline was based on
the chemistry of free radicals. The recent approach of
Fukuyama et al.2j uses a radical approach to synthesize an
indole AB ring precursor, but we now present a tandem
radical cyclization employing an iodoaryl azide3,4 as the key
step which assembles the ABCE tetracycle in a formal total
synthesis of (()-vindoline (Scheme 1).
Bu¨chi’s seminal synthesis2a proceeds through a crucial
cyclization step, affording tetracycle 4. Although the aromatic
A ring in vindoline carries a methoxy group (R′ ) Me), the
reactivity of the A ring meant that Bu¨chi was required to
install a group which was substantially less electron-releasing
in this position (R′ ) Ts) for the cyclization reaction; the
tosyloxy group was later converted to the desired methoxy
group in 5. Since then, tetracycle 5 has become a key
intermediate in vindoline syntheses as shown, for example,
by the approach of Ban et al.2c
Bu¨chi’s problems with substitution of ring A by an
electron-donating methoxy group arose from generation of
highly electrophilic intermediates during the synthesis. The
generality of this problem is seen in the fact that other
syntheses have also employed sulfonate esters for the A
ring.2j In our case, the use of a methoxy-substituted arene
from the outset should not pose problems since we intended
to proceed via radical intermediates. Moreover, our strategy
for linking the A and C rings would involve Mitsunobu
(1) Gorman, M.; Neuss, N.; Biemann, K. J. Am. Chem. Soc. 1962, 84,
1058. Moncrief, J. W.; Lipscomb, W. N. Acta Crystallogr. 1966, 21, 322.
Neuss, N.; Gorman, M.; Hargrove, W.; Cone, N. J.; Biemann, K.; Bu¨chi,
G.; Manning, R. E. J. Am. Chem. Soc. 1964, 86, 1440.
(2) (a) Ando, M.; Bu¨chi, G.; Ohnuma, T. J. Am. Chem. Soc. 1975, 97,
6880. (b) Takano, S.; Shishido, K.; Matsuzaka, J.; Sato, M.; Ogasawara,
K. Heterocycles 1979, 13, 307. (c) Ban, Y.; Sekine, Y.; Oishi, T.
Tetrahedron Lett. 1978, 151. (d) Kutney, J. P.; Bunzli-Trepp, U.; Chan, K.
K.; de Souza, J. P.; Fujise, Y.; Honda, T.; Katsube, J.; Klein, F. K.;
Leutwiller, A.; Morehead, S.; Rohr, M.; Worth, B. R. J. Am. Chem. Soc.
1978, 100, 4220. (e) Danieli, B.; Lesma, G.; Palmisano, G.; Riva, R. J.
Chem. Soc., Chem. Commun. 1984, 909. (f) Andriamialisoa, R. Z.; Langlois,
N.; Langlois, Y.; J. Org. Chem. 1985, 50, 961. (g) Danieli, B.; Lesma, G.;
Palmisano, G.; Riva, R. J. Chem. Soc., Perkin Trans. 1 1987, 155. (h)
Feldman, P. L.; Rapoport, H.; J. Am. Chem. Soc. 1987, 109, 1603. (i)
Kuehne, M. E.; Podhorez, D. E.; Mulamba, T.; Bornmann, W. G. J. Org.
Chem. 1987, 52, 347. (j) Kobayashi, S.; Ueda, T.; Fukuyama, T. Synlett
2000, 883.
(3) (a) Kizil, M.; Patro, B.; Callaghan, O.; Murphy, J. A.; Hursthouse,
M. B.; Hibbs, D. J. Org. Chem. 1999, 64, 7856. (b) Kizil, M.; Murphy, J.
A. J. Chem. Soc., Chem. Commun. 1995, 1409 (c) Patro, B.; Murphy, J. A.
Org. Lett. 2000, 2, 3599.
(4) For key work on aliphatic iodoazides, see: Kim, S.; Joe, G. H.; Do,
J. J. Am. Chem. Soc. 1994, 116, 5521.
10.1021/ol0171618 CCC: $22.00 © 2002 American Chemical Society
Published on Web 01/09/2002

Scheme 1a
a Reagents and conditions: (a) DIBAL-H, toluene, rt; (b) CBr4, PPh3, 2,6-lutidine, THF, rt, 90%; (c) Mg, THF, then 4-(4-
methoxybenzyloxy)butyraldehyde, rt, 85%; (d) oxalyl chloride, DMSO, Et3N, DCM, -78 °C, 90%; (e) 1. OsO4 (0.5%), NMO, acetone,
H2O, rt; 2. NaIO4, acetone, H2O, rt, 85%; (f) NaOH (1%), H2O, 2-propanol, reflux. 84%; (g) CeCl3, methanol, NaBH4, -15 °C, 95%; (h)
p-NO2C6H4CO2H, DEAD, PPh3 then NaOH, MeOH, 100%; (i) DEAD, PMe3, THF, rt, 92%, 90% for 16, 16a; (j) DDQ, DCM, H2O, rt,
85%, 87% for 17, 17a; (k) MsCl, Et3N, DCM, rt, 93%, 90% for 18, 18a; (l) NaN3, DMF, 60 °C, 90%, 94% for 19, 19a; (m) TTMSS,
AIBN, benzene, ∆, 35%, 65% for 20, 20a; (n) H2, Pd/C, EtOAc, rt, 99%; (o) 2-nitrophenylselenocyanate, Bu3P, THF, rt, 95%; (p) NaIO4,
NaHCO3, MeOH, 50 °C, 86%; (q) OsO4 (1%), NMO, acetone, H2O, rt, 91%; (r) NH3, Na, 2-propanol; (s) HCHO (36% in H2O), NaCNBH3,
pH ) 3 (with HCl), MeOH/DCM/H2O, 85% from 23; (t) NaIO4, acetone, H2O, 84%.
chemistry where the key nitrogen of ring B would be present
as a sulfonamide. [The sulfonyl functionality would be more
than a protecting group: the acidity it imparts to the nitrogen
would be key to the success of the Mitsunobu reaction.] The
sulfonamide should modulate the electron-donating powers
of this nitrogen and could be converted to an N-methyl group
at the appropriate time.
Our plan envisaged a synthesis of 6, bearing a Q-
substituent, which could be converted to the desired ketone
in 5. In principle, Q might most conveniently be a protected
alcohol or ketone, but intermediates bearing such structures
would be highly unstable toward elimination from ring C.
Instead, a benzyloxymethyl group was chosen, because of
its stability under a broad range of experimental conditions.
The present synthesis started with reduction of diethyl
allylmalonate and protection of the resulting diol, affording
benzylidene acetal 7 without purification. Treatment with
DIBAL-H in toluene afforded the monoprotected diol 8 in
excellent yield (75% from diethyl allylmalonate). The free
hydroxyl group in 8 was transformed to bromide 9 (90%),
which was used in a Grignard reaction to couple with 4-(4-
methoxybenzyloxy)butyraldehyde to give adduct 10 (85%).
Swern oxidation of the secondary alcohol afforded ketone
11 in 90% yield. Oxidative cleavage provided aldehyde 12
(85%), which, on treatment with base, underwent aldol
condensation smoothly (84%), to afford cyclohexenone 13.
This was followed by Luche reduction, yielding the alcohol
14 (95%). Gratifyingly, the stereoselectivity of the reduction
was excellent (95/5), although the relative stereochemistry
at the two carbon centers is not yet determined. Alcohol 14
has now been prepared on a 30 g scale by this practical route.
Conversion to the cyclization precursor was straight-
forward. Coupling of sulfonamide 15 and alcohol 14 by
Mitsunobu reaction gave 16 in excellent yield (92%).
Deprotection of the 4-methoxybenzyl group in 16 by DDQ
(85%), mesylation of the resulting hydroxyl group (93%),
and transformation of the mesylate to azide 19 (90%) all
went smoothly. However, tandem radical cyclization with
tristrimethylsilylsilane (TTMSS) followed by acetylation
gave 35% yield of the desired tetracycle 20. Judging that
the stereochemistry of the benzyloxymethyl group in 19
might be the cause of this low yield, it was decided to invert
444
Org. Lett., Vol. 4, No. 3, 2002

the stereochemistry of the hydroxy group in 14 by coupling
it with 4-nitrobenzoic acid under Mitsunobu reaction condi-
tions. Hydrolysis of the resulting benzoate gave 14a in
quantitative yield. This was efficiently converted to 19a as
above (through steps i-l) and subjected to the same
cyclization conditions. This time, the product 20a was
isolated in a more acceptable 65% yield.
The next tasks were to convert the N-Ms group to an N-Me
group and to introduce the ketone in ring C. In our initial
studies, the N-methanesulfonyl group in 20a was converted
to an N-methyl group, but the product proved to be highly
reactive. [For example, under Swern oxidation conditions,
the aromatic ring was observed to undergo electrophilic
substitution, recalling the problems of Bu¨chi et al.] Therefore
it was judged best to keep the methylsulfonamide protecting
group on the nitrogen to stabilize the aromatic ring, until
the final stages of the synthesis.
Accordingly our attention was directed to cleaving one
carbon from the hydroxymethyl group to introduce the ketone
group of the Bu¨chi intermediate. Hydrogenolysis at atmo-
spheric pressure removed the benzyl group in 20a (99%).
The resulting alcohol 21 reacted with 2-nitrophenylseleno-
cyanate and tributylphosphine in THF to give the phenyl-
seleno derivative in 95% yield. Oxidation with sodium
periodate in methanol gave the corresponding selenoxide,
which underwent facile elimination in situ to afford alkene
22 in excellent yield (86%). Dihydroxylation was performed
in acetone and water with a catalytic amount of osmium
tetroxide as oxidant and NMO as co-oxidant. The reaction
gave 23 (91%) as a mixture of diastereomers in excellent
yield. Deprotection of the methylsulfonyl group was then
performed with sodium in liquid ammonia with 2-propanol
(proton donor). The crude product was used directly without
purification in the next reaction, in which formaldehyde
solution in water and sodium cyanoborohydride were used
at pH < 3 to effect methylation on nitrogen in very good
combined yield (85% from 23). Finally sodium periodate
cleavage of the diol afforded Bu¨chi intermediate 5 in
excellent yield (84%). This approach therefore constitutes a
high-yielding synthesis of the crucial tetracyclic intermediate
5, and hence a formal total synthesis of (()-vindoline. This
synthesis promises to be adaptable to the synthesis of
enantiomerically pure (-)-vindoline by starting with a single
enantiomer5 of 8.
Acknowledgment. We thank EPSRC for funding and
EPSRC National Mass Spectrometry Service Centre, Swansea,
for mass spectrometric measurements. J.A.M. thanks the
Royal Society for a Leverhulme Senior Research Fellowship.
OL0171618
(5) Fukuyama, T.; Wang, C.-L.; Kishi, Y. J. Am. Chem. Soc. 1979, 101,
260. Wang, Y. F.; Sih, C. J. Tetrahedron Lett. 1984, 25, 4999.
Org. Lett., Vol. 4, No. 3, 2002
445