En route to the total synthesis of tashironin: On the exercise of stereochemical control by a methyl group in mediating remote cyclization reactions

Article abstract of DOI:10.1016/j.tetlet.2004.11.148

The synthesis of the [2.2.2]-bicyclic core (23) of the neurotrophic factor 11-O-debenzoyltashironin (1) has been achieved by an oxidative dearomatization-transannular Diels-Alder cascade. We have shown that the reaction sequence is also valuable for the efficient construction of related, complex [2.2.2]-bicyclic compounds (vide infra).

Full text of DOI:10.1016/j.tetlet.2004.11.148

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 843–847
En route to the total synthesis of tashironin: on the exercise
of stereochemical control by a methyl group in mediating
remote cyclization reactions
Silas P. Cook,a Christoph Gaul and Samuel J. Danishefskya,b,
b
*
aDepartment of Chemistry, Columbia University, Havemeyer Hall, New York, NY 10027, USA
bLaboratory for Bioorganic Chemistry, The Sloan-Kettering Institute for Cancer Research, 1275 York Avenue,
New York, NY 10021, USA
Received 2 November 2004; revised 29 November 2004; accepted 29 November 2004
Available online 18 December 2004
Abstract—The synthesis of the [2.2.2]-bicyclic core (23) of the neurotrophic factor 11-O-debenzoyltashironin (1) has been achieved
by an oxidative dearomatization-transannular Diels–Alder cascade. We have shown that the reaction sequence is also valuable for
the efficient construction of related, complex [2.2.2]-bicyclic compounds (vide infra).
Ó 2004 Elsevier Ltd. All rights reserved.
Neurotrophic factors, or neurotrophins, are agents that
can prevent neuronal death (neurotrophism) or promote
axonal growth (neurotropism). To date, all widely stud-
ied neurotrophins (cf. NGF, BDNF, GDNF, NT4/5,
NT6) are naturally occurring polypeptides or pro-
teins.1–3 Given their potential to treat neurodegenerative
disorders, it is not surprising that neurotrophic factors
have been the focus of considerable interdisciplinary re-
search since the discovery of the first neurotrophin,
NGF, by Montalcini and Hamburger.2 Indeed, peptidyl
neurotrophic factors have been extensively evaluated in
animal models for their ability to treat neurodegenera-
tive disease. However, due to unfavorable drug delivery
and pharmacokinetic characteristics, in vivo evaluation
of these neurotrophins requires direct microinjection
into the brain.3,4 Clearly, drug availability problems
associated with these polypeptidic structures are a seri-
ous impediment to their development in prospective hu-
man settings.
do we yet insist on a validated molecular target. In keep-
ing with earlier patterns, we prefer to focus on agents de-
rived from natural sources. Among naturally derived
prospects, we tend to favor compounds of novel archi-
tecture that invite provocative proposals for their syn-
thesis. Indeed, through the medium of total synthesis,
we have begun to develop a collection of small-molecule
neurotrophically-active agents, including tricycloillici-
none, merrilactone A, and jiadifenin.5 The repertoire
of evaluatable samples is further augmented by molecu-
lar modifications arising from the total synthesis efforts.
In 1995, Fukayama et al. reported the isolation and
structural characterization of the neurotrophically inac-
tive sesquiterpenoid tashironin (2) from the wood of
illicium tashiroi.6 More recently, Fukuyama and co-
workers reported on the isolation and structure elucida-
tion of 11-O-debenzoyltashironin (1), which promotes
neurite growth at concentrations as low as 0.1 lmol.7
In debenzoyltashironin, we recognized a molecular tar-
get with both a promising neurotrophic profile and an
interesting structural framework that would challenge
our laboratory from a synthetic standpoint. We set out
to accomplish a free-standing synthetic route that could
provide an alternative to the complexities of obtaining
either 1 or 2 from natural sources. Equally important,
diverted total synthesis could provide the means to ex-
plore broader questions of chemical space in the tashiro-
nin file (Fig. 1).
It is from this vantage point that our laboratory has
been seeking to explore the potential of small-molecule
agents that might, themselves, exhibit neurotrophic
activity. At this very early stage of the inquiry, we do
not yet discriminate as to the mechanism of action nor
*
Corresponding author. Tel.: +1 212 6395502; fax: +1 212 7728691;
e-mail: s-danishefsky@mskcc.org
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.11.148

844
S. P. Cook et al. / Tetrahedron Letters 46 (2005) 843–847
HO
Diels–Alder reaction to afford 5. It seemed that progres-
15
O
OR
sion from 5 to 1 could be accomplished. We were well
aware that the success of this strategy would be predi-
cated on the ability of the methyl stereocenter in the tan-
dem precursor 4 to exert diastereofacial control in the
oxidative dearomatization step.
1
4
9
6
HO
5
O
R = H: Debenzoyltashironin (1)
R = Bz: Tashironin (2)
Our synthesis commenced with commercially available
vanillyl alcohol (3a). Compound 3a was treated with
catalytic amounts of tosyl acid in methanol to afford
3b, presumably, via a p-quinone methide intermediate
(Scheme 2). The aromatic methyl group was then
installed regiospecifically at the more hindered site
through directed lithiation followed by methylation to
afford 6. A subsequent DDQ-mediated oxidation gave
rise to 2-methylvanillin (7) in 44% overall yield from
3a. Crotylation of the phenolic hydroxyl group, fol-
lowed by a Claisen rearrangement produced phenol 8.
For greater flexibility in later transformations, the phe-
nol was protected either as a benzyl ether (9a) or as a
MOM ether (9b). Baeyer–Villiger oxidation of these
compounds followed by hydrolysis afforded phenols
10a and 10b in 60% and 89% overall yields, respectively,
Figure 1. Structure of debenzoyltashironin 1 and tashironin 2.
In our initial approach to constructing the structural
backbone of debenzoyltashironin, we hoped to employ
a
tandem oxidative dearomatization-transannular
Diels–Alder reaction to rapidly generate the highly
substituted tetracyclic core (5) from a relatively simple
precursor (4) (Scheme 1). According to our design, the
key substrate (4) might be prepared in relatively short
order from a suitably protected aromatic precursor (3).
We anticipated that, under suitable reaction conditions,
4 might undergo intramolecular oxidative dearomatiza-
tion, following the protocol of Pelter et al. and Tamura
et al.,8 to give rise to an intermediate cyclized species
that would be configured to undergo a transannular
OH
OH
OMe
RO
OMe
oxidative
OH
dearomatization
OTs
O
OH
3
4
O
HO
OMe
O
O
OMe
OH
O
Transannular
Diels-Alder
RO
RO
HO
OTs
O
OTs
1
5
Scheme 1. Synthetic strategy.
OH
OH
OH
OH
OMe
OMe
OMe
OMe
f
c
d, e
b
O
O
OR
3a R = H
OMe
6
7
8
a
3b R = Me
OR
OR
OR
OR
OMe
OMe
OMe
OMe
j
i
g, h
O
OTs
OTs
OH
O
10a R = Bn
10b R = MOM
11a R = Bn
11b R = MOM
9a R = Bn
9b R = MOM
12a R = Bn
12b R = MOM
Scheme 2. Synthesis of aldehydes 12a and 12b. Reagents and conditions: (a) MeOH, p-TSA, rt, 100%; (b) BuLi, THF, À15 to 0 °C, then À10 °C,
MeI; (c) DDQ, CH2Cl2/H2O 19:1, rt, 44% for two steps; (d) crotyl bromide, K2CO3, acetone, reflux; (e) neat, 185 °C, 81% for two steps; (f) MOMCl,
DIEA, CH2Cl2, rt, 98% or BnBr, K2CO3, acetone, reflux, 93%; (g) m-CPBA, CH2Cl2, 0 °C; (h) Et3N , CHCl2/MeOH 1:1, rt, 61% (10a) and 96% (10b)
2
over two steps; (i) TsCl, Et3N , CHCl2, rt, 88% (11a) and 91% (11b); (j) Rh(CO)2(acac), Billig ligand, CO/H2 1:1, toluene, 60 °C, 90% (12a) and 82%
2
(12b).

S. P. Cook et al. / Tetrahedron Letters 46 (2005) 843–847
845
OBn
OTs
OBn
OBn
OMe
O
OMe
c
OMe
a, b
O
OEt
O
O
OTs
OTs
b
a
12a
13
12a
OBn
OBn
OBn
OTs
OH
OMe
OMe
MsO
OMe
MsO
OMe
d, e
f
OEt
OEt
OEt
OH
O
OTs
O
OTs
O
OTs
18
19
14
15
c, d
c, d
O
O
OMe
O
OH
OH
Transannular
Diels-Alder
O
OMe
OMe
OMe
MsO
MsO
OH
OH
OTs
OTs
16
10-membered acetal
OTs
20
OTs
17
21
5:1 mixture of diastereomers
e
e
Scheme 3. Probing the key step. Reagents and conditions: (a) ethyl
diazoacetate, SnCl2, CH2Cl2, rt, 87%; (b) MeI, K2CO3, acetone, reflux,
O
O
OMe
OMe
1
1
93%; (c) MsCl, Et3N , CHCl2, rt, 48%; (d) DIBAL-H, CH2Cl2,
2
15
15
O
O
11
11
À78 °C; (e) 5% Pd/C, H2EtOAc, rt, 66% for two steps; (f) PIDA,
toluene/MeCN, rt, 60%.
OTs
OTs
21a
20a
from 8. The resultant phenolic hydroxyl group in each
compound was then tosylated to produce 11a and 11b
in the yields shown. These compounds were hydro-
formylated with Rh(CO)2(acac) and the Billig bis-organo-
phosphite ligand according to BuchwaldÕs protocol, to
give exclusively the desired linear aldehydes 12a and
12b in 90% and 82% yields, respectively.9
O
O
15
15
1
O
11 OMe
O
11 OMe
1
OTs
OTs
23
22
Scheme 4. Successful examples of the oxidative dearomatization/
TADA sequence. Reagents and conditions: (a) phosphonate,
KHMDS, 18-crown-6, THF, À78 °C, 52%; (b) methyl phosphonate,
KHMDS, 18-crown-6, THF, À78 °C, 68%; (c) BBr3, CH2Cl2, À78 °C;
(d) DIBAL-H, CH2Cl2, À78 °C, 87% (20) and 69% (21) over two steps;
(e) PIDA, toluene, rt, 66% (22) and 60% (23).
At this juncture, we were prepared to test the feasibility
of our key transformation. Toward this end, subjection
of aldehyde 12a to Roskamp conditions followed by
methylation of the resulting b-ketoester gave rise to 13
(Scheme 3).10 Treatment of 13 with mesyl chloride led
to the formation of the desired (E)-tetrasubstituted ole-
fin11 14 in modest yield. Finally, reduction of the ethyl
ester with DIBAL-H, followed by benzyl deprotection
gave the key substrate, 15.
ination conditions with two different phosphonates
(Scheme 4).14 Disubstituted olefin 18 and trisubstituted
olefin 19 were obtained in 52% yield and 68% yield, both
as 10:1-mixtures of (Z)/(E)-geometric isomers. Cleavage
of the corresponding benzyl protecting group followed
by DIBAL-H reduction of the ester functionalities pro-
ceeded smoothly to provide the key substrates 20 and 21
in 87% and 69% yield, respectively.
In the event, treatment of 15 with phenyliodine(III)
diacetate (PIDA, PhI(OAc)2) in a toluene/acetonitrile
solvent system gave rise cleanly12 to a new product,
16,13 which was quite unstable. As shown in Scheme 3,
compound 16 failed to undergo the expected transannu-
lar Diels–Alder reaction (see target structure 17) under
several sets of conditions. The surprising lability of 16
limited our options for permuting reaction conditions
to accomplish transannular Diels–Alder reaction.
In the event, treatment of phenol 20 with PIDA in tolu-
ene over several hours at room temperature resulted in
the formation of the desired transannular Diels–Alder
adduct 22 in 66% yield as a single diastereomer, presum-
ably via the corresponding 10-membered acetal. Simi-
larly, when phenol 21, containing a trisubstituted
double bond, was subjected to identical conditions, the
transannular Diels–Alder adduct 23 was formed in a
stereoselective fashion in 60% yield. NMR analysis
(NOE) of compound 23 revealed the stereochemistry
at C-1, with the C-15 methyl group residing in the unde-
sired a position. In other words, this C-15 methyl had
orchestrated the sense of the oxidative cyclization
We postulated that the stereoelectronic complexity cou-
pled with the high steric demand required of the tetra-
substituted diene in reacting with a tetrasubstituted
dienophile might be responsible for the resistance of
16 to undergo transannular Diels–Alder reaction. To
probe this hypothesis, we sought to examine the behav-
ior of less complicated putative Diels–Alder substrates.
Thus, aldehyde 12a was subjected to Still–Gennari olef-

846
S. P. Cook et al. / Tetrahedron Letters 46 (2005) 843–847
Scheme 5. Synthesis of acetal 26. Reagents and conditions: (a) Bestmann reagent, K2CO3, MeOH, rt, 65%; (b) HCl, MeOH, rt, 90%; (c) PIDA,
MeOH, rt, then toluene, reflux, 85%.
reactions of 20 and 21 so as to produce 20a and 21a.
These intermediates must, per force, produce the ob-
served transannular Diels–Alder products.
ring system through an intermolecular trapping of an
allenyl alcohol by oxidative dearomatization and a sub-
sequent Diels–Alder reaction with the internal olefin of
the allenol. This proposed IMDA cycloaddition raised
obvious issues of regioselectivity. Although we were
confident that the internal olefin of the allene would re-
act preferentially due to the geometry of the transition
state, predictions as to whether the reaction would give
the desired five-membered acetal Diels–Alder adduct
(30) or the six-membered acetal (ÔtwistedÕ) product (29)
were less obvious (Scheme 6). In practice, phenol 27
was constructed in nine steps from the commercially
available vanillyl alcohol 3a in 20% overall yield. Fol-
lowing treatment of 27 with PIDA in the presence of
5 equiv of allenyl alcohol 28,17 the undesired ÔtwistedÕ
Diels–Alder adduct 29 was obtained (as shown by
NOE studies) (Scheme 6). Although this interesting out-
come disqualified this route as an approach toward tash-
ironin, it has led to studies in our laboratory that are still
in progress to determine the source of the preference for
the ÔtwistedÕ cycloaddition.
Having shown that both the di- and trisubstituted ole-
fins do indeed undergo stereospecific oxidative dearoma-
tization followed by transannular Diels–Alder, in
contrast to the tetrasubstituted system in 16, it seemed
prudent to reduce the complexity of the key sequence:
instead of constructing all four rings of tashironin in a
single transformation, we imagined building three of
the four rings by the oxidative dearomatization-trans-
annular Diels–Alder cascade and then closing the fourth
ring in subsequent transformations. Following this
logic, the [2.2.2] bicyclic ring system could be formed
concomitantly either with the fused cyclopentane ring
or with the five-membered acetal. We have explored
each approach in turn.
In the first approach, aldehyde 12b was converted to ter-
minal alkyne 24 in 65% yield using the Bestmann
reagent (Scheme 5).15 Phenol 25 was revealed in 90%
yield after HCl-mediated removal of the MOM group.
Exposure of phenol 25 to PIDA in MeOH at room tem-
perature presumably led to the formation of the corre-
sponding masked o-quinone, which underwent the
desired transannular Diels–Alder reaction upon heating
to afford 26 diastereoselectively in 85% yield (ds >95%).
O
O
Br
O
OMe
O
OH
Br
OMe
28
•
nOe
OTs
Happily, we were able to obtain single crystals of 26. X-
ray analysis led to a decisive structural verification,
revealing the relative configuration of the two quarter-
nary stereocenters C-6/C-9 and the tertiary stereocenter
C-1 as shown in Scheme 5.16 The completion of the syn-
thesis of debenzoyltashironin through this route would
require the installation of functional handles in the
Diels–Alder precursor to allow for the inversion of
the stereochemistry at C-1 as well as for the closing
of the five-membered acetal.
H
OTs
a
a
nOe
H
OH
29
Br
OMe
OTs
O
O
OMe
27
O
Br
OMe
Br
O
OTs
OTs
30
In an alternate approach, we sought to install the five-
membered acetal in conjunction with the [2.2.2] bicyclic
Scheme 6. ÔTwistedÕ Diels–Alder product 29. Reagents and conditions:
(a) allenyl alcohol 28, PIDA, toluene, rt, 69%.

S. P. Cook et al. / Tetrahedron Letters 46 (2005) 843–847
847
Carcache, D. A.; Tian, Y.; Li, Y. M.; Danishefsky, S. J. J.
Am. Chem. Soc. 2004, 126, 14358.
6. Fukuyama, Y.; Shida, N.; Kodama, M. Tetrahedron Lett.
1995, 36, 583.
7. Huang, J. M.; Yokoyama, R.; Yang, C. S.; Fukuyama, Y.
J. Nat. Prod. 2001, 64, 428.
8. (a) Pelter, A.; Elgendy, S. Tetrahedron Lett. 1988, 29, 677;
(b) Tamura, Y.; Yakura, T.; Haruta, J.; Kita, Y. J. Org.
In summary, we have reported herein the highly concise
synthesis of the tetracyclic ring system that forms the
core of 11-O-debenzoyltashironin. The synthesis was
achieved by a Pelter–Tamura oxidation resulting in the
trapping of a tethered allylic alcohol, followed by a
transannular Diels–Alder reaction. We have also shown
that this reaction sequence is viable for the efficient con-
struction of related, rather complex [2.2.2]-bicyclic com-
pounds. Studies as to the application of these findings to
enable the total synthesis of the tashironins are well in
progress.
Chem. 1987, 52, 3927; (c) For
a recent review on
cyclohexadienone ketals, see: Magdziak, D.; Meek, S. J.;
Pettus, T. R. R. Chem. Rev. 2004, 104, 1383.
9. Cuny, G. D.; Buchwald, S. L. J. Am. Chem. Soc. 1993, 115,
2066.
10. Holmquist, C. R.; Roskamp, E. J. J. Org. Chem. 1989, 54,
3258.
Acknowledgements
11. Olefin geometry determined by NOE studies of reduced
product.
Support for this research was provided by the National
Institutes of Health (Grant HL-25848). We would like
to thank Celera for the generous fellowship awarded
to S.P.C. Postdoctoral support is gratefully acknowl-
edged by C.G. (Deutscher Akademischer Austauschd-
ienst, DAAD).
12. We note that formation of spiro ether i by ipso spirocyc-
lization is apparently not at all competitive with the meta
pathway leading from 15 to 16. Such an ipso pathway is
inherent in the Becker–Adler reaction and analogs thereof.
Evidently, the proclivity for attack at the electron-rich
methoxy-bearing carbon is dominant in the PIDA-medi-
ated cyclization.
References and notes
O
OH
O
OMe
O
1. Bennett, M. R.; Gibson, W. G.; Lemon, G. Auton.
Neurosci. 2002, 95, 1.
MsO
OMe
OMe
MsO
PIDA
O
OH
MsO
2. (a) Levi-Montalcini, R.; Hamburger, V. J. Exp. Zool. 1951,
116, 321–362; (b) Lu, P.; Blesch, A.; Tuszynski, M. H. J.
Comp. Neurol. 2001, 436, 456.
OTs
OTs
16
OTs
i
15
3. Kaneko, M.; Saito, Y.; Saito, H.; Matsumoto, T.; Matsuda,
Y.; Vaught, J. L.; Dionne, C. A.; Angeles, T. S.; Glicksman,
M. A.; Neff, N. T.; Rotella, D. P.; Kauer, J. C.; Mallamo, J.
P.; Hudkins, R. L.; Muraka, C. J. Med. Chem. 1997, 40,
1863.
4. (a) Kirik, D.; Georgievska, B.; Bjorklund, A. Nat. Neurosci.
2004, 7, 105; (b) Dawbarn, D.; Allen, S. J. Neuropathol.
Appl. Neurobiol. 2003, 29, 211–230; (c) Pollack, S. J.;
Harper, S. J. Curr. Drug Target CNS Neurol. Disord. 2002,
1, 59; (d) Gonzalez de Aguilar, J.-L.; Girlanda-Junges, C.;
Coowar, D.; Duportail, G.; Loeffler, J.-P.; Luu, B. Brain
Res. 2001, 920, 65; (e) Fournier, J. J. Pharm. Pharmacol.
1998, 50, 323.
13. Of course, in reality, for a particular antipode, the methyl
group defines the absolute configuration of the resultant
cage-like latticework in structure 16. For convenience
sake, we portray the intermediate by permuting the
relative stereochemistry of the methyl bearing carbon.
14. Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4405.
15. (a) Muller, S.; Liepold, B.; Roth, G. J.; Bestman, H. J.
Synlett 1996, 521; (b) Ohira, S. Synth. Commun. 1989, 19,
561.
16. Crystallographic data (excluding structural data) for
compound 26 have been deposited with the Cambridge
Crystallographic Data Centre (CCDC) as Deposition No
CCDC 230898.
5. (a) Pettus, T. R. R.; Chen, X.-T.; Danishefsky, S. J. J. Am.
Chem. Soc. 1998, 120, 12684; (b) Birman, V.; Danishefsky,
S. J. J. Am. Chem. Soc. 2002, 124, 2080; (c) Cho, Y. S.;
17. Isaac, M. B.; Chan, T.-H. J. Chem. Soc., Chem. Commun.
1995, 1003.