A short and convenient way to produce the Taxol A-ring utilizing the Shapiro reaction

Article abstract of DOI:10.1016/S0040-4020(02)00089-3

The Shapiro reaction was utilized in an efficient route to a Taxol A-ring building block. Commercially available 2-methyl-1,3-cyclohexanedione was converted in three simple steps to various arenesulfonylhydrazones and then to the target molecule with the Shapiro reaction. Remarkable differences were observed in the reactivity and stability of different hydrazones and their dianions in the Shapiro reaction. This pathway is the shortest one reported to give the target molecule in good overall yield. The use of different electrophiles in the final Shapiro reaction step allows alternative ways to prepare the target alcohol.

Full text of DOI:10.1016/S0040-4020(02)00089-3

TETRAHEDRON
Pergamon
Tetrahedron 58 22002) 2175±2181
A short and convenient way to produce the Taxole A-ring
utilizing the Shapiro reaction
a
Olli P. Tormakangas, Reijo J. Toivola, Esko K. Karvinen and Ari M. P. Koskinen
b
c
a,p
È
È
aLaboratory of Organic Chemistry, Helsinki University of Technology, P.O. Box 6100, FIN-02015 Espoo, HUT, Finland
bDepartment of Chemistry, University of Oulu, P.O. Box 3000, FIN-90014 Oulu, Finland
cNeste Chemicals Oy, Technology Center, P.O. Box 310, FIN-06101 Porvoo, Finland
Received 22 October 2001; accepted 24 January 2002
AbstractÐThe Shapiro reaction was utilized in an ef®cient route to a Taxole A-ring building block. Commercially available 2-methyl-1,3-
cyclohexanedione was converted in three simple steps to various arenesulfonylhydrazones and then to the target molecule with the Shapiro
reaction. Remarkable differences were observed in the reactivity and stability of different hydrazones and their dianions in the Shapiro
reaction. This pathway is the shortest one reported to give the target molecule in good overall yield. The use of different electrophiles in the
®nal Shapiro reaction step allows alternative ways to prepare the target alcohol. q 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
Our retrosynthetic strategy for Taxole is shown in
Scheme 1. We have earlier reported our entries to the side
chain 26 and the C-ring precursor 4.7 Compound 5 has been
utilized successfully as an A-ring precursor in the total
synthesis of Taxole by Nicolaou et al.8
Synthetic preparation of Taxol 1 is still under intensive
investigation due to its extremely low availability from
nature 2bark of Paci®c yew tree Taxus brevifolia) and
growing shortage in treatment against a number of mam-
malian cancers.1 Semisynthesis of Taxole from 10-
deacetylbaccatin III, readily available from the needles of
Taxus baccata, has provided some amelioration against lack
of Taxol.2 Several different strategies to Taxole A-ring
fragments have been developed utilizing Diels±Alder
reaction,3 modi®cation of cyclohexanones4 and ene-
reaction5 as the most common methods.
The Shapiro reaction is an ef®cient way to create a new C±C
bond to the carbonyl carbon of ketones simultaneously
introducing a vinylic moiety into the product. In the Shapiro
reaction the ketone derived hydrazone is converted to a
dianion using an alkyl lithium base and then decomposed
directly to the vinyl anion.9 The vinyl anion can also be
alkylated to introduce a substituent to the neighboring
X
AcO
O
OH
NHBz
CO2R
OHC
MeO2C
OHC
BzHN
Ph
O
O
Ph
OH
O
O
O
OAc
3
4
HO OBz
2
OH
1
O
RO
5
Scheme 1. Retrosynthetic analysis of Taxole and the role of A-ring building block.
Keywords: Shapiro reaction; Taxol; shortest synthetic pathway.
Corresponding author. Tel.: 1358-9-451-2526; fax: 1358-9-451-2538; e-mail: ari.koskinen@hut.®
p
0040±4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020202)00089-3

È È
O. P. Tormakangas et al. / Tetrahedron 58 42002) 2175±2181
2176
carbon atom. The lithiated vinyl anion reacts readily with
electrophiles allowing an easy entry to the ®nal product. The
use of intramolecular electrophiles allows preparation of
cyclic products with high stereoselectivity.10
distilled 2103±1048C; 13 mmHg) in order to obtain suf®-
cient purity. Dimethylation of cyclohexane-1,3-dione
directly to 7 was also attempted but the yield was rather
low 240%) and more side products were observed.
In this paper we report the shortest and simplest synthesis of
a Taxole A-ring fragment utilizing the Shapiro reaction as
the key step. The route involves only four steps and
proceeds with high yields.
Selective monoketalization of pure 2,2-dimethyl-1,3-cyclo-
hexadione 7 was achieved by treatment with 2,2-dimethyl-
1,3-propanediol and 1 mol% of p-TsOH in CH2Cl2 with
azeotropic removal of water.12 Traces of two side products
were observed but no diketalized dione.
In the beginning of this work, we wanted to study the
reactivity of different electrophiles in the Shapiro reaction.
The aim was to use tosyl 12a and trisylhydrazone 12b
2Scheme 3) as model compounds. Ketone 8 was ®rst
converted 2LDA/CH3I) to the methylated ketone 11 in
66% yield. However, preparation of the arylsulfonylhydra-
zones from ketone 11 proved to be impossible or extremely
slow. No product was observed in the case of 12b and only
traces of product was formed in the case of 12a even after
24 h. The steric hindrance caused by the methyl groups in
the a-position obviously retards the reaction. With HCl as
acid catalyst, only deketalization of 11 was observed.
2. Results and discussion
The synthesis plan for the A-ring precursor 10 is shown in
Scheme 2. We have earlier reported the ®rst two steps in a
complementary and longer route to the Taxole A-ring
block.4d Herein we optimized those steps and the product,
monoketal 8, was used as the starting material in the
preparation of the hydrazones 9a±c. The Shapiro reaction
of 9a±c and related hydrazones 12a±b was investigated
carefully.
Methylation of commercially available 2-methyl-1,3-cyclo-
hexanedione 6 was carried out with K2CO3/CH3I in
acetone.11 The product mixture contained 85% of the
desired 7 and 15% of 3-methoxy-2-methyl-cyclohex-2-
enone as the side product 2ratios based on GC analysis).
The crystalline side product was ®ltered off and recycled
to the enol form of the starting material 6 by treatment with
2 M HCl in CH2Cl2. Diketone 7 was isolated in 99% purity
when CH2Cl2 was used as the solvent in the extraction. If
toluene was used instead of CH2Cl2 the product had to be
We decided to use sterically less hindered hydrazones 9a±c
2of ketone 8) as model compounds 2Scheme 2) in order to
uncover the limitations in the Shapiro reaction. Addition-
ally, the use of hydrazones 9a±c instead of hydrazones
12a±b provides one step shorter reaction pathway. All
stable hydrazones 9a±c were prepared in excellent yields.
Typically hydrazones are prepared under concentrated
conditions so that all starting material and reagent hardly
O
O
O
O
O
O
a
b
O
6
7 (69%)
8 (94%)
c
R'
R'
O
S
d
H
O
O
R
N
N
HO
O
O
O
(92%)
(79%)
(94%)
9a R=CH3, R'=H
b R=R'=CH3
c R=R'=CH(CH3)
10 (62% from 9a)
Scheme 2. Synthesis of Taxole A-ring block via tosylhydrazone 9a. Reagents and conditions: 2a) K2CO3, CH3I, acetone, rfx, 8 h; 2b) CH2Cl2,
Me2C2CH2OH)2, p-TsOH£H2O, CH2Cl2, rfx, 7 h; 2c) EtOH 2THF in 9b), hydrazide, 1408C 16 min, then rt 6 h; 2d) THF, 2508C, n-BuLi, 30 min, CH3I,
30 min, n-BuLi, from 2508C to rt, 25 min, paraformaldehyde, 30 min.
R'
O
H
O
O
b
a
R
S
N
N
O
8
O
O
O
R'
12 a R=CH3, R'=H
b R=R'=CH(CH3)2
11
Scheme 3. Reagents and conditions: 2a) LDA, THF, 2788C, CH3I, 08C then rt; 2b) hydrazide, EtOH, 2HCl).

È È
O. P. Tormakangas et al. / Tetrahedron 58 42002) 2175±2181
2177
dissolve in the solvent.13 Also a small excess of hydrazine
2105±110 mol%) is usually required to run the reaction to
completion. If the ketone used is sterically hindered,
addition of a catalytic amount of HCl, re¯uxing the reaction
mixture and longer reaction time are sometimes needed.14
Here, under optimized reaction conditions to prepare 9a±c,
the reaction mixture is ®rst heated up to 130 to 1408C until
complete dissolution is achieved and then allowed to react at
room temperature 2too high reaction temperatures caused
partial decomposition of the product even at 1508C). The
use of acid catalyst was not necessary. Absolute EtOH was
found to be the best solvent in the preparation of tosyl-
2508C gives dianion 15 which can be observed as a slightly
orange color. When this dianion is heated to room
temperature it slowly decomposes to vinyl anion 16. The
decomposition can be easily observed as the formation of
small bubbles when N2 is liberated. An electrophile must be
added immediately when the gas formation has ceased
especially if THF is used as the solvent in order to avoid
possible protonation of vinyl anion 16 by THF.
The strength of the used alkyl lithium base in the
Shapiro reaction is also crucial.15 Stronger bases like
t-BuLi are sometimes required in the deprotonation. In
our case the formation of dianion 15 from 14 can be carried
out easily with n-BuLi. However, t-BuLi gave similar
results.
hydrazone 9a and trisylhydrazone 9c allowing
a
spontaneous crystallization of the hydrazones from the
reaction mixture. The preparation of 9a was carried out
also in THF and MeOH successfully but in lower yields.
Additionally, the reaction was slightly slower and the
hydrazone did not crystallize out from the reaction mixture.
Mesitylhydrazone 9b was prepared only in THF.
Differences in the stabilities of hydrazones 9a±c during
storage were also observed. Tosylhydrazone 9a seems to
be very stable and can be stored at 148C for a few years
under argon. Trisylhydrazone 9c decomposed partly under
similar conditions giving yellowish color in a few months.
The stability of mesitylhydrazone 9b during the storage was
close to that of tosylhydrazone 9a.
We initially studied the Shapiro reaction with trisylhydra-
zone 9c which was treated with t-BuLi at 2788C in THF and
formation of the ®rst dianion 13c was observed as a red
color 2Scheme 4). The dianion of trisylhydrazone was
found to be too labile even at 2788C and decomposition
was observed. Methylation of the dianion was carried out at
2788C as well as the preparation of the second dianion 15c,
followed by heating to 08C and quenching by D2O. In the
product mixture there was only 7.5% of compound 17 where
methylation proceeded successfully at the a-carbon and
then decomposed to vinyl anion and captured with
deuterium. The main product was 19 257% of product
mixture) which indicates that the dianion had not reacted
with CH3I at all but decomposed to vinyl anion and was
trapped later with deuterium. The presence of product 20
was a clear evidence of premature decomposition of dianion
13c to its vinyl anion. N2 evolution was also observed
already at 2458C which indicates the decomposition of
the dianion to the vinyl anion.
Detailed description of the Shapiro reaction is described in
Scheme 4. Hydrazone 9 was ®rst treated with 220 mol% of
n-BuLi in order to prepare dianion 13. The ®rst hydrogen
removed is the one located on nitrogen and this monoanion
is usually colorless. Addition of another equivalent of butyl
lithium gives a beautiful deep red color. The dianion was
methylated quantitatively to 14 with CH3I at 2508C
2internal temperature). The second addition of n-BuLi at
R'
O
R'
O
H
O
a
O
R
S
N
N
R
S
N N
O
O
O
O
R'
R'
13a-c
9a-c
b
R'
R'
R'
O
O
S
c
O
O
R
S
N
N
R
N N
O
O
O
O
R'
14a-c
15a-c
d
O
Li
O
16
Scheme 4. Detailed steps of Shapiro reaction from hydrazones to vinyl anion 16. Reagents and reaction conditions: 2a) n-BuLi 2220 mol%), THF, 2558C,
30 min; 2b) CH3I 2250 mol%), 2508C, 30 min; 2c) n-BuLi 2400 mol%), THF, 2508C, 30 min; 2d) rt, 25 min.

È È
O. P. Tormakangas et al. / Tetrahedron 58 42002) 2175±2181
2178
16 can occur either by reaction with the solvent16 or due to
O
O
H
D
D
the MeOH liberated during monomerization of para-
formaldehyde. Thus, other forms of formaldehyde were
also investigated. Gaseous formaldehyde was generated
from paraformaldehyde by heating at 1308C and then led
into the reaction with an argon ¯ow.17 This method gave
28% isolated yield of 10 at best. The use of excess 1,3,5-
trioxane in THF gave 22% isolated yield.
O
O
18
17
19
O
O
H3C
O
O
These results with Shapiro reaction of 9a±c are in
accordance with the results reported by Shapiro et al.
originally. They observed that the ortho-position of the
aromatic ring of tosylhydrazones can be deprotonated with
strong alkyllithium bases to give a trianion. Vinyl anions
related to 16 have also been reported to be basic enough to
deprotonate the aromatic ortho-position.18 The problem of
vagrant deprotonation could be overcome by using
300±400 mol% of BuLi in the preparation of dianions.
However, we observed no difference in the product distribu-
tion when the amount of used BuLi was varied from 220 to
400 mol%. Furthermore, dianions of tosylhydrazones
decompose extremely slowly to the vinyl anion as compared
to the corresponding trisylhydrazones.19 Therefore decom-
position of the dianion should be fast and the electrophile
should be added immediately after complete decomposition
of dianion.
20
The ®nal proof of the premature decomposition of the
dianion of 9c was obtained when dianion 13c was prepared
at 2788C and quenched after 45 min with D2O. Both
deuterated trisylhydrazone 21 285%) and deuterated vinyl
anion 19 215%) were observed in the product mixture
2Scheme 5). At higher temperature 22488C) and shorter
stirring before quench 215 min) the product mixture
contained 96% of 19 with H/D ratio 22:78.
Mesitylhydrazone 9b was assumed to be a better choice
because a possible deprotonation of aromatic ortho-protons
is avoided and the vinyl anion could be stable enough for
methylation. Also the decomposition of dianion 15b could
be fast enough in order to avoid protonation of 16 by the
solvent. The reaction proceeded as described in Scheme 4.
Methylation was complete but the step 15b!16 was slow
and the vinyl anion was protonated by THF.
A few additional experiments with carefully dried electro-
philes were conducted to ®nd out whether the protonation
was caused by the solvent 2THF) or by moisture. The
reaction of benzylchloromethylether with 16 to the BOM
protected alcohol 10 was unsuccessful. Methylchloro-
formate and dimethylcarbonate were also examined to
create an ester functionality which could be reduced to the
target compound, alcohol 10. However, the yield of 22 was
very low with both electrophiles and again, protonated vinyl
anion was obtained as the main product.
Tosylhydrazone 9a proved to be the best choice. Steps
9a!14a 2Scheme 4) were carried out uneventfully. After
formation of the second dianion 15a the solution was
warmed to room temperature. The decomposition 2evolu-
tion of N2) was complete in 25 min. A shorter reaction
time or lower decomposition temperature gives a signi®cant
amount of hydrazone as a side product. Paraformaldehyde
210 mol equiv.) was added to give allylic alcohol 10 in 62%
isolated yield. Protonated vinyl anion was still obtained as a
side product 2,10%). Immediate protonation of vinyl anion
When DMF was employed as the electrophile 2Scheme 6)
1) t-BuLi
-78˚C
2) D2O
O
D
O
S
N
N
D
9c
19
+
O
O
21
Scheme 5.
O
O
CH3OCOCl
or
LiAlH4
LiAlH4
H3CO
O
(CH3O)2CO
O
HO
22
O
16
DMF
r.t.
O
10 (73%)
O
H
O
23 (61%)
Scheme 6.

È È
O. P. Tormakangas et al. / Tetrahedron 58 42002) 2175±2181
2179
aldehyde 23 was obtained in 61% isolated yield. Aldehyde
23 was easily reduced to alcohol 10 with LiAlH4 in THF in
73% isolated yield. The reactivity of N,N-disubstituted
amides with alkyl lithium compounds is known to be very
high giving fast reaction and good yields.20 Thus, due to the
high reactivity of the amide, protonation of 16 by THF does
not occur in signi®cant amounts.
starting material was observed as a white precipitation. The
reaction mixture was ®ltered, CH2Cl2 was evaporated, the
residue taken up in toluene 2300 ml) and stirred with 20 ml
of 2 M HCl for 2 h. The toluene phase was separated and the
water phase was washed with toluene. The toluene phases
were combined, dried over Na2SO4, ®ltered and evaporated
to dryness. The oily product was cooled and it crystallized
out as a white powder. Yield 38.54 g 269%), purity.98%
2GC). TLC 2MTBE/Hex, 80:20) Rfˆ0.25. Mp 39±408C. IR
2in KBr): 2977, 2940, 2876, 1732, 1702. 1H NMR
To avoid protonation of 16 by THF, DME and Et2O were
investigated. However, methylation 213a!14a) was incom-
plete in these experiments. Also, TMEDA/hexane was
impracticable here because of strong salt formation between
CH3I and TMEDA.21
3
2200 MHz) d 2.70 2t, 4H, Jˆ6.8 Hz), 1.85±2.03 2m, 2H),
1.31 2s, 6H, 2£CH3). 13C NMR 260 MHz) d 210.4, 61.6,
37.3, 22.1, 17.9. Calcd for C8H12O2: C 68.55, H 8.64; Found
C 68.34, H 8.46.
4.1.2. 3,3,7,7-Tetramethyl-1,5-dioxaspiro[5,5]undecan-
8-one -8).12 Dimethyldiketone 7 215.42 g; 110 mmol), 2,2-
dimethyl-1,3-propanediol 234.37 g; 330 mmol), p-TsOH
monohydrate 20.154 g; 1.5 mol%) and CH2Cl2 2220 ml)
were placed into the reaction ¯ask. This yellowish solution
was re¯uxed for 7 h with azeotropic water removal. CH2Cl2
was evaporated and the product precipitated out as white
crystals. The precipitate was dissolved in 250 ml of hexanes
and washed ®rst with 125 ml of 1 M NaHCO3 and then
water 22£125 ml). The organic phase was dried over
Na2SO4. After ®ltration the solvent was evaporated and
the product crystallized out as a white powder. The product
was dried under high vacuum 20.15 mmHg) for 4 h. Yield
23.4 g 294%), mp 62.5±66.08C. TLC 2Hex/MTBE, 80:20)
3. Conclusions
Taxole A-ring building block 10 was synthesized with a
novel and short method consisting of only four steps with
high 38% overall yield. Tosylhydrazone 9a was found to be
the best of the studied arylhydrazones in Shapiro reaction
allowing the formation of stable dianions and complete
methylation. Evidence of the possible effect of steric
hindrance was observed in the preparation and reactivity
of hydrazones. Protonation of the vinyl anion by THF
remained problematic to some extent giving always some
protonated vinyl anion. This can be minimized with rapid
decomposition of dianion 15a at room temperature followed
by immediate addition of the electrophile. The Shapiro reac-
tion as the ®nal step allows the use of different electrophiles
and thus the formation of various different functionalities to
the ®nal compound.
1
Rfˆ0.17. IR 2KBr disk) 3385, 2983, 2959, 2872, 1708. H
NMR 2400 MHz) d 3.63 2d, 2H, 2Jˆ10.5 Hz), 3.34 2dd, 2H,
4
3
2Jˆ10.5 Hz, Jˆ1.5 Hz), 2.42 2t, 2H, Jˆ7 Hz), 2.24±2.19
2m, 2H), 1.72±1.63 2m, 2H), 1.21 2s, 6H), 1.16 2s, 3H), 0.72
2s, 3H). 13C NMR 260 MHz) d 213.3, 101.9, 70.2, 55.4,
36.4, 29.8, 23.3, 22.3, 20.7, 19.4, 18.7.
4. Experimental
4.1. General
4.1.3.
3,3,7,7-Tetramethyl-8--tosylhydrazono)-1,5-di-
oxaspiro[5,5]undecane -9a). Ketone 8 26.79 g; 30 mmol),
tosylhydrazide 26.16 g; 33 mmol) and 18 ml of abs. ethanol
were placed into the reaction ¯ask. The mixture was heated
for 16 min at 1408C until all solid had dissolved. The solu-
tion was stirred for 6 h at room temperature and the product
crystallized out as a white precipitate. The solvent was
evaporated and the solid residue was puri®ed by means of
grinding with 24 ml of cold methanol/water 280:20) solu-
tion. The precipitate was ®ltered, washed with 15 ml of cold
methanol/water 275:25) and dried carefully to give 9a as a
white solid 210.86 g, 92%). Mp 140±144.58C. TLC 2Hex/
MTBE, 80:20) Rfˆ0.02. IR 2KBr disk): 3205, 2990, 2975,
2947, 2921, 2866, 1630. 1H NMR 2400 MHz) d 7.84 2d, 2H,
3Jˆ8.8 Hz), 7.28 2d, 2H, 3Jˆ8.8 Hz), 3.55 2d, 2H,
All solvents used were dry and distilled immediately before
use. Merck silica gel 60F 2230±400 mesh) plates were used
in TLC analyses. The TLC plates were stained with 1%
phosphomolybdic acid in ethanol. NMR spectra were
measured on Bruker AM 200 and Bruker DPX-400
instruments in CDCl3 with TMS as internal reference. Gas
chromatography was preformed on Perkin±Elmer Model
with OV-1701 column. Mass spectra were recorded on
Kratos MS80 RF Autoconsole. IR spectra were measured
with Perkin±Elmer Spectrum One instrument. All melting
points were measured with a digital Gallenkamp GMB
2capillary) apparatus.
2
4
2Jˆ10.5 Hz), 3.25 2dd, 2H, Jˆ10.5 Hz, Jˆ1.4 Hz), 2.42
2s, 3H), 2.26 2t, 2H, 3Jˆ6.8 Hz), 2.06±1.99 2m, 2H), 1.56±
1.47 2m, 2H), 1.13 2s, 3H), 1.12 2s, 6H), 0.69 2s, 3H). 13C
NMR 260 MHz) d 166.2, 143.6, 135.5, 129.2, 128.3, 100.7,
69.9, 49.1, 29.7, 23.3, 22.3, 22.0, 21.6, 20.9, 20.6, 19.0.
Calcd for C20H30O4N2S: C 60.89; H 7.66; N 7.10; Found
C 61.17; H 8.01; N 7.06.
4.1.1. 2,2-Dimethyl-1,3-cyclohexanedione -7). Acetone
2300 ml), 2-methyl-1,3-cyclohexanedione 250.5 g; 400
mmol) 6, K2CO3 2110.57 g; 800 mmol) and CH3I
262.3 ml; 1000 mmol) were placed into the reaction vessel
and re¯uxed for 8 h with vigorous stirring to avoid K2CO3
hardening on the walls. Acetone was evaporated, 200 ml of
CH2Cl2 was added and evaporated to dryness to remove
acetone traces. The product mixture was dissolved in
CH2Cl2 2400 ml) and extracted with 600 ml of water. The
water phase was washed with 100 ml of CH2Cl2. The
combined organic phase was stirred with 200 ml of 2 M
HCl for 4 h at room temperature and the enol form of the
4.1.4.
3,3,7,7-Tetramethyl-8--2,4,6-trimethylbenzene-
hydrazono)-1,5-dioxaspiro-[5,5]undecane -9b). Ketone 8
22.263 g; 10 mmol), 2,4,6-trimethylbenzene-sulfonyl-
hydrazide 22.362 g; 11 mmol) and 6 ml of abs. ethanol

È È
O. P. Tormakangas et al. / Tetrahedron 58 42002) 2175±2181
2180
were placed into the reaction ¯ask. The mixture was stirred
for 40 min at 1308C and then for 3.5 h at room temperature.
Ethanol was evaporated and the white precipitate was puri-
®ed by means of grinding with 8 ml of methanol/water
280:20) solution. The mixture was ®ltered and the precipi-
tate was washed with 2£25 ml of cold methanol/water
275:25) solution and dried carefully to give 9b as a white
solid 23.35 g, 79%). Mp 159.5±163.58C. TLC 2CH2Cl2/
MeOH, 90:10) Rfˆ0.75. IR 2KBr disk): 3252, 2963, 2871,
2735, 1651, 1603. 1H NMR 2200 MHz) d 6.93 2s, 2H), 3.54
2d, 2H, 2Jˆ11.3 Hz), 3.22 2d, 2H, 2Jˆ11.3 Hz), 2.65 2s, 6H),
2.29 2s, 3H), 2.25±2.14 2m, 2H), 2.08±2.02 2m, 2H), 1.60±
1.50 2m, 2H), 1.09 2s, 3H), 1.02 2s, 6H), 0.68 2s, 3H). 13C
NMR 260 MHz) d 163.3, 142.3, 140.2, 131.5, 100.6, 69.8,
49.0, 29.7, 23.2, 22.2, 21.3, 21.0, 20.8, 20.6, 19.1. Calcd for
C22H34O4N2S: C 62.53; H 8.11; N 6.63; Found C 62.38; H
8.10; N 6.46.
puri®cation 2Hex/MTBE, 45:55) gave 0.158 g 262%) of 10
in .98% purity 2NMR). The product can be further puri®ed
by recrystallization from methanol/water 270:30). Mp
111.0±113.18C. TLC 2Hex/MTBE, 45:55) Rfˆ0.25. IR
2KBr disk): 3532, 2975, 2958, 2930, 2901, 2870, 2729,
2700, 1660. 1H NMR 2400 MHz) d 4.15 2d, 2H,
Jˆ4.1Hz), 3.69 2d, 2H, 2Jˆ11.1 Hz), 3.37 2dd, 2H,
1.19 2s, 3H), 1.16 2s, 6H), 0.73 2s, 3H). 13C NMR 2100 MHz)
d 136.4, 131.7, 100.0, 70.3, 59.3, 43.7, 29.95, 29.87, 23.3,
22.5, 22.3, 19.2, 18.4. Calcd for C15H26O3: C 70.83, H
10.30, Found C 70.72, H 10.68.
4
2Jˆ10.2 Hz, Jˆ1.4 Hz), 2.10±1.98 2m, 4H), 1.77 2s, 3H),
4.1.7. 3,3,7,7,9-Pentamethyl-1,5-dioxaspiro[5,5]undecan-
8-one -11). Diisopropylamine 20.79 ml; 5.4 mmol) and THF
210 ml) were placed into the reaction ¯ask under argon and
the solution was cooled down to 2788C. n-BuLi 25.0 mmol)
was added and the solution was stirred for 30 min at 2788C
and then 45 min at 08C. Ketone 8 21.132 g; 5.0 mmol) in
THF 27 ml) was added dropwise. The solution was stirred
for 30 min at 08C. CH3I 20.375 ml; 6.0 mmol) was added
and the solution was stirred at room temperature for 30 min
and quenched with 1ml of water. THF was evaporated and
the residue was taken up in 25 ml of hexanes. The organic
phase was washed with 9 ml of water, the layers were sepa-
rated and the organic phase was dried over Na2SO4, ®ltered
and evaporated. The white precipitate was puri®ed by
recrystallization 25 ml MeOH/0.9 ml H2O). The product
was washed with 4 ml of cold MeOH/H2O 270:30). Yield
0.792 g 266%). Mp 79.5±81.58C. TLC 2Hex/MTBE, 80:20)
Rfˆ0.36. IR 2KBr disk): 3407, 2960, 2929, 2867, 2785,
4.1.5.
3,3,7,7-Tetramethyl-8--trisylhydrazono)-1,5-di-
24.53 g;
oxaspiro[5,5]undecane -9c).4g Ketone
8
20 mmol), trisylhydrazide 26.27 g; 21mmol) and THF
260 ml) were stirred for 4 h at room temperature. THF was
evaporated and the white precipitate was ground with 50 ml
of methanol/water 285:15) solution, ®ltered, washed with
50 ml of cold methanol/water 280:20), ®ltered again and
dried carefully to give 9c as a white solid 29.52 g, 94%).
Mp 153.5±1568C. TLC 2CH2Cl2/MeOH 90:10) Rfˆ0.78. IR
1
2KBr disk): 3262, 2959, 2867, 1714, 1624, 1599, 1561. H
NMR 2400 MHz) d 7.14 2s, 2H), 4.16 2sept., 2H,
3Jˆ6.7 Hz), 3.37 2d, 2H, 2Jˆ11.6 Hz), 3.22 2d, 2H,
2Jˆ11.6 Hz), 2.90 2sept., 1H, 3Jˆ6.9 Hz), 2.212t, 2H,
3Jˆ6.8 Hz), 2.04±1.98 2m, 2H), 1.58±1.48 2m, 2H), 1.25
2d, 12H, 3Jˆ6.7 Hz), 1.24 2d, 6H, 3Jˆ6.9 Hz), 1.08 2s,
3H), 1.01 2s, 6H), 0.67 2s, 3H). 13C NMR 260 MHz) d
163.0, 152.9, 151.1, 124.0, 123.4, 100.6, 69.8, 48.9, 34.2,
29.9, 24.8, 23.6, 23.5, 23.2, 22.3, 21.5, 20.8, 20.5, 19.4,
18.9. Calcd for C28H46O4N2S: C 66.37; H 9.15; N 5.53;
Found C 66.18; H 9.20; N 5.46.
1
2
1714. H NMR 2400 MHz) d 3.72 2d, 1H, Jˆ11.3 Hz),
3.52 2d, 1H, 2Jˆ11.3 Hz), 3.37 2dd, 1H, 2Jˆ11.3 Hz,
4Jˆ2.7 Hz), 3.28 2dd, 1H, Jˆ11.3 Hz, Jˆ2.7 Hz), 2.76±
2
4
2.68 2m, 1H), 2.64 2dt, 1H, 2Jˆ12.9 Hz, 3Jˆ6.4 Hz), 1.89±
2
3
1.73 2m, 2H), 1.78 2dd, 1H, Jˆ14.3 Hz, Jˆ4.3 Hz), 1,75
2dd, 1H, 2Jˆ14.3 Hz, 3Jˆ4.3 Hz), 1.21 2s, 3H), 1.19 2s, 3H),
1.15 2s, 3H), 1.02 2d, 3H, Jˆ6.40 Hz), 0.72 2s, 3H). 13C
3
NMR 2100 MHz) d 214.0, 102.2, 70.5, 69.7, 55.0, 39.2,
29.8, 27.8, 23.2, 22.3, 21.1, 16.3, 14.9.
4.1.6. 3,3,7,7,9-Pentamethyl-8--hydroxymethyl)-1,5-di-
oxaspiro-[5,5]undec-8-ene -10). Tosylhydrazone 9a
20.398 g; 1.0 mmol) was dissolved in THF 25 ml) under
argon and the solution was cooled down to 2558C.
n-BuLi 21.40 M in pentane, 2.2 mmol) was added dropwise
in order to keep temperature stable and a dark red color was
observed. The solution was stirred for 60 min at 2508C and
CH3I 21.8 mmol) was added dropwise giving light yellow
color and small amount of white 2LiI) precipitate. The solu-
tion was stirred for 30 min at 2508C and the solution turned
colorless. Another n-BuLi 24.0 mmol) addition gave orange
color. The solution was warmed up to room temperature and
the dianion decomposed to its vinyl anion which was
observed as a gas formation 2starts even at 228C). The
solution was stirred at room temperature for 25 min in
order to complete the vinyl anion formation. Paraformalde-
hyde 28.4 mmol) was added into the reaction in one portion
to give a slightly exothermic reaction. The reaction was
stirred for 60 min and the solvents were evaporated. The
residue was taken up with 8 ml of hexane and washed
with 8 ml of water. The organic layer was separated and
the water phase was washed again with 8 ml of hexane.
The hexane phases were combined and dried over
Na2SO4, ®ltered and evaporated. Column chromatographic
4.1.8. 3,3,7,7,9-Pentamethyl-1,5-dioxaspiro-[5,5]undec-
8-ene-8-carboxaldehyde -23). Tosylhydrazone 9a
20.790 g; 2.0 mmol) was dissolved in THF 210 ml) under
argon and the solution was cooled to 2558C. n-BuLi
24.4 mmol) was added dropwise in order to keep tempera-
ture stable and a dark red color was observed. The solution
was stirred for 30 min at 2508C and CH3I 20.250 ml;
4.0 mmol) was added dropwise giving light yellow color
and small amount of white 2LiI) precipitate. The solution
was stirred for 30 min at 25 08C followed with addition of
n-BuLi 28.0 mmol) giving orange color. The solution was
heated to 158C, stirred for 5 min and heated to rt 21238C).
The solution was vigorously stirred for 25 min at rt and the
decomposition of dianion to vinyl anion was observed as a
gas formation. DMF 21.5 ml; 19 mmol) was added to give
immediately a yellowish solution. After 30 min stirring the
solvent was evaporated and the white solid residue was
taken up in 15 ml of H2O and 15 ml of hexane. The layers
were separated and the water phase was washed with addi-
tional 15 ml of hexane. The combined hexane phases were
dried over Na2SO4, with ®ltered and evaporated to give a
yellowish oil. Column chromatography 2MTBE/hexane,

È È
O. P. Tormakangas et al. / Tetrahedron 58 42002) 2175±2181
2181
Prunet, J. Synlett 2000, 323±326. 2b) Nivlet, A.; Dechoux, L.;
Martel, J.-P.; Proess, G.; Mannes, D.; Alcaraz, L.; Harnett,
J. J.; Le Gall, T.; Mioskowski, C. Eur. J. Org. Chem. 1999,
3241±4249. 2c) Bourgeois, D.; Lallemand, J.-Y.; Pancrazi, A.;
Prunet, J. Synlett 1999, 1555±1558. 2d) Toivola, R. J.;
Koskinen, A. M. P. Synlett 1996, 1159±1161. 2e) Karvinen,
E. K.; Koskinen, A. M. P. Tetrahedron 1995, 51, 7555±7560.
2f) Di Grandi, M. J.; Jung, D. K.; Krol, W. J.; Danishefsky, S. J.
J. Org. Chem. 1993, 58, 4989±4992. 2g) Kress, M. H.; Ruel,
R.; Miller, W. H.; Kishi, Y. Tetrahedron Lett. 1993, 34, 5999±
6002.
40:60; 50 g of silica gel in column of 3 cm diameter) gave
0.310 g 261%) of oily 23 which slowly crystallized over-
night. Mp 65±70.58C. TLC 2Hex/MTBE, 60:40) Rfˆ0.32.
1
IR 2KBr disk) 3401, 2959, 2868, 1670, 1613. H NMR
2400 MHz) d 10.07 2s, 1H), 3.68 2d, 2H, 2Jˆ11.0 Hz),
3.38 2dd, 2H, 2Jˆ10.5 Hz, 4Jˆ1.4 Hz), 2.23 2t, 2H,
3
3Jˆ6.6 Hz), 2.10 2t, 2H, Jˆ6.6 Hz), 2.09 2s, 3H), 1.32 2s,
6H), 1.21 2s, 3H), 0.74 2s, 3H). 13C NMR 2100 MHz) d
192.3, 153.0, 139.1, 99.9, 70.3, 42.5, 32.7, 29.9, 23.2,
22.3, 21.7, 19.0, 17.8, 14.2. HRMS 2EI) calcd for
M12C15H24O3): 252.1726, found 252.1733, Dˆ2.8 ppm.
5. 2a) Wennerberg, J.; Polla, M.; Frejd, T. J. Org. Chem. 1997,
62, 8735±8740. 2b) Polla, M.; Frejd, T. Tetrahedron 1993, 49,
2701±2710.
4.1.9. Reduction of aldehyde 23 to alcohol 10. Aldehyde
23 20.72 g; 2.85 mmol) was dissolved in THF 210 ml) under
argon. The solution was cooled to 08C and LiAlH4 20.262 g;
6.9 mmol) was added in few portions. The reaction was
stirred for 4 h at room temperature to reach the completion
2monitored with TLC; Hex/EtOAc, 75:25) and was
quenched at 08C by addition of 0.7 ml 239 mmol) of H2O
at 08C. The precipitate was ®ltered and washed with
5£10 ml of THF. The solvent was evaporated and the
product was recrystallized from 4 ml of hexanes to give
0.530 g of white crystals.
6. Koskinen, A. M. P.; Karvinen, E. K.; SiirilaÈ, J. P. J. Chem.
Soc., Chem. Commun. 1994, 21±22.
7. 2a) Lajunen, M.; Koskinen, A. M. P. J. Chem. Soc., Perkin
Trans. 1 2000, 1439±1443. 2b) Xia, Z. -Q.; Karvinen, E. K.;
Rissanen, K.; Koskinen, A. M. P. Unpublished.
8. Nicolaou, K. C.; Yang, Z.; Liu, J.-J.; Nantermet, P. G.;
Claiborne, C. F.; Renaud, J.; Guy, R. K.; Shibayama, K.
J. Am. Chem. Soc. 1995, 117, 645±652.
9. 2a) Adlington, R. M.; Barrett, A. G. M. Acc. Chem. Res. 1983,
16, 55±59. 2b) Shapiro, R. H. Org. React. 1976, 23, 405±506.
2c) Shapiro, R. H.; Heath, M. J. J. Am. Chem. Soc. 1967, 89,
5734±5735. 2d) March, J. Advanced Organic Chemistry:
Reactions, Mechanisms and Structures; 4th ed; Wiley: New
York, 1992; pp 1019±1021.
Acknowledgements
This work was ®nancially supported by the Finnish
È
Academy. We also wish to thank Paivi Joensuu at the
University of Oulu during the time we spent there for
excellent mass spectra.
10. Chamberlin, A. R.; Bloom, S. H.; Cervini, L. A.; Fotsch, C. H.
J. Am. Chem. Soc. 1988, 110, 4788±4796.
11. Spencer, T. A. J. Org. Chem. 1968, 33, 2269±3373.
12. Raju, N.; Rajagopalan, K.; Swaminathan, S. Tetrahedron Lett.
1980, 21, 1577±1580.
References
13. 2a) Lipton, M. F.; Shapiro, R. H. J. Org. Chem. 1978, 43,
1409±1413. 2b) Hutchins, R. O.; Kacher, M.; Rua, L.
J. Org. Chem. 1975, 40, 923±926.
1. Review on the chemistry and biology of Taxole:
2a) Nicolaou, K. C.; Dai, W.-M.; Guy, R. K. Angew. Chem.,
Int. Ed. Engl. 1994, 33, 15±44. Sources of paclitaxel and
related compounds: 2b) Rozendaal, E. L. M.; Lelyveld, G. P.;
van Beek, T. A. Phytochemistry 2000, 53, 383±389.
14. Hutchins, R. O.; Milewski, C. A.; Maryanoff, B. E. J. Am.
Chem. Soc. 1973, 95, 3662±3668.
15. Chamberlin, A. R.; Stemke, J. E.; Bond, F. T. J. Org. Chem.
1978, 43, 147±154.
2. 2a) Kingston, D. G. I. Pharmacol. Ther. 1991, 52, 1±34.
2b) Denis, J.-N.; Greene, A. E. J. Am. Chem. Soc. 1988, 110,
5917±5919.
16. 2a) Chamberlin, A. R.; Bloom, S. H. Org. React. 1990, 39, 1 ±
83. 2b) Stemke, J. E.; Bond, F. T. Tetrahedron Lett. 1975,
1815±1818.
3. 2a) Carballares, S.; Craig, D.; Lane, C. A. L.; MacKenzie,
A. R.; Mitchell, W. P.; Wood, A. Chem. Commun. 2000,
1767±1768. 2b) Funk, R. L.; Yost, III, K. J. J. Org. Chem.
1996, 61, 2598±2599. 2c) Frost, C.; Linnane, P.; Magnus, P.;
Spyvee, M. Tetrahedron Lett. 1996, 37, 9139±9142.
2d) Nicolaou, K. C.; Liu, J.-J.; Ueno, H.; Sorensen, E. J.;
Claiborne, C. F.; Guy, R. K.; Hwang, C.-K.; Nakada, M.;
Nantermet, P. G. J. Am. Chem. Soc. 1995, 117, 634±644.
2e) Queneau, Y.; Krol, W. J.; Bornmann, W. G.; Danishefsky,
S. J. J. Org. Chem. 1992, 57, 4043±4047. 2f) Lin, J.; Nikaido,
M. M.; Clark, G. J. J. Org. Chem. 1987, 52, 3745±3752.
2g) Bonnert, R. V.; Jenkins, P. R. J. Chem. Soc., Perkin
Trans 1 1989, 413±418. 2h) Shea, K. J.; Wise, S. J. Am.
Chem. Soc. 1978, 100, 6519±6521.
17. 2a) Ullmann's Encyclopedia of Industrial Chemistry, 5th
Completely Revised ed.; VCH: Weinheim, 1996, Vol. A11,
Formaldehyde, pp 619±651. 2b) Spence, R.; Wild, W.
J. Chem. Soc. 1935, 338±340.
18. 2a) Stemke, J. E.; Chamberlin, A. R.; Bond, F. T. Tetrahedron
Lett. 1976, 2947±2950. 2b) Shapiro, R. H.; Hornaman, E. C.
J. Org. Chem. 1974, 39, 2302±2303.
19. Bond, F. T.; Di Pietro, R. A. J. Org. Chem. 1981, 46, 1 31 5±
1318.
20. Evans, D. A. J. Chem. Soc. 1956, 4691. March, J. Advanced
Organic ChemistryÐReactions, Mechanisms and Structure;
4th ed; Wiley: New York, 1992; p 489.
21. Shapiro, R. H.; Lipton, M. F.; Kolonko, K. J.; Buswell, R. L.;
Capuano, L. A. Tetrahedron Lett. 1975, 1811±1814.
4. Latest examples: 2a) Bourgeois, D.; Maiti, G.; Pancrazi, A.;