Synthesis of the reported structure of pogostol and a total synthesis of (±)-kessane without the use of protecting groups

Article abstract of DOI:10.1021/ol035373u

(Matrix presented) A short racemic synthesis of kessane from 4-hydroxy-4-methyl-2-cyclohexenone is described using a route that also resulted in the synthesis of the reported structure of pogostol. The key step involves an Fe(III)-mediated tandem radical

Full text of DOI:10.1021/ol035373u

ORGANIC
LETTERS
2003
Vol. 5, No. 18
3309-3312
Synthesis of the Reported Structure of
Pogostol and a Total Synthesis of
(±)-Kessane without the Use of
Protecting Groups
,†
Kevin I. Booker-Milburn,* Helen Jenkins,† Jonathan P. H. Charmant,‡ and
Peter Mohr§
School of Chemistry, UniVersity of Bristol, Cantock’s Close, Bristol, BS8 1TS, UK,
Structural Chemistry Laboratory, School of Chemistry, UniVersity of Bristol,
Cantock’s Close, Bristol, BS8 1TS, UK, and F. Hoffmann-La Roche AG,
Pharmaceuticals DiVision, DiscoVery Chemistry, CH-4070 Basel, Switzerland
k.booker-milburn@bristol.ac.uk
Received July 23, 2003
ABSTRACT
A short racemic synthesis of kessane from 4-hydroxy-4-methyl-2-cyclohexenone is described using a route that also resulted in the synthesis
of the reported structure of pogostol. The key step involves an Fe(III)-mediated tandem radical ring-expansion/cyclization of the cyclopropylsilyl
ether 9. No protecting groups are used in the entire sequence. Comparison of the NMR data of synthetic pogostol to that in the literature
indicates that the structure originally proposed is incorrect.
Previously, we reported that cyclopropyl silyl ethers undergo
tandem radical ring-expansion/cyclization on treatment with
Fe(III) salts.1 For example, treatment of the cyclopropylsilyl
ether 1 with ferric chloride in DMF leads to the [5.3.0]
bicyclic chloro-ketone 3 by way of the ring-expanded â-keto
radical 2. At the same time, independent studies by Narasaka2
and co-workers showed that treatment of cyclopropanols with
Mn(pic)3, in conjunction with a variety of trapping agents,
could be used to effect similar transformations and add
functionality to the resulting bicyclic products.3 Since then,
Simpkins4 and Cha5 have reported applications of this
Fe(III)/Mn(III) chemistry in synthesis as well as development
of related methodology. As this strategy allows rapid
diastereoselective access to bicyclic ketones from simple
enones, it can be envisioned that this would be particularly
useful as a key step in the synthesis of [5.3.0] terpene-based
natural products. To that end, Narasaka et al. were able to
apply their Mn(III) methodolgy to a total synthesis of 10-
isothiocyanatoguaia-6-ene.3 Recently, we described the use
of our Fe(III) methodology in an approach to the terpenoids
dictyol-C and R-eudesmol.6 Herein, we describe the applica-
tion of the Fe(III)-mediated ring expansion-cyclization
† School of Chemistry, University of Bristol.
‡ Structural Chemistry Laboratory, School of Chemistry, University of
Bristol.
(3) Iwasawa, N.; Funahashi, M.; Hayakawa, S.; Ikeno, T.; Narasaka, K.
Bull. Chem. Soc. Jpn. 1999, 72, 85.
§ F. Hoffmann-La Roche AG.
(1) (a) Booker-Milburn, K. I. Synlett 1992, 809. (b) Booker-Milburn,
K. I.; Thompson, D. F. J. Chem. Soc., Perkin Trans. 1 1995, 2315. (c)
Booker-Milburn, K. I.; Barker, A.; Brailsford, W.; Cox, B.; Mansley, T. E.
Tetrahedron 1998, 54, 15321.
(2) Iwasawa, N.; Funahashi, M.; Hayakawa, S.; Narasaka, K. Chem. Lett.
1993, 545.
(4) Blake, A. J.; Highton, A. J.; Majid, T. N.; Simpkins, N. S. Org. Lett.
1999, 1787.
(5) Lee, H. B.; Sung, M. J.; Blackstock, S. C.; Cha, J. K. J. Am. Chem.
Soc. 2001, 123, 11322.
(6) Booker-Milburn, K. I.; Cox, B.; Grady, M.; Halley, F.; Marrison,
S. Tetrahedron Lett. 2000, 39, 4651.
10.1021/ol035373u CCC: $25.00 © 2003 American Chemical Society
Published on Web 08/08/2003

sequence to a short total synthesis of the reported structures
of pogostol 4, pogostol O-methyl ether 5, and the tricyclic
sesquiterpene kessane 6. Pogostol is one of the constituents
of the musky oil isolated from the patchouli plant, which
has been used in traditional Chinese medicine and has
antiemetic properties.7 A full structural assignment of pog-
ostol by NMR was described by Weyerstahl in 2000.8 In
1998, Waterman described the isolation and NMR structural
assignment of pogostol O-methyl ether 5 from the stem bark
of Artabotrys stenopetalus (Annonaceae).9 Kessane, conceiv-
ably related to 4 by cyclization, was isolated10 from Valeriana
officinalis in 1967, and a detailed structural assignment was
made by van Beek et al.11 in 1993.
of epimers.14 Tebbe methylenation15 of these yielded a
partially separable 2.5:1 mixture of the reported structure
for (()-pogostol 4 and epi-pogostol 13 in quantitative overall
yield.
However, at this point we were dismayed to find that
1
detailed comparison of the literature H and 13C NMR data
of pogostol with both synthetic (()-4 and the epimer (()-
13 unequivocally demonstrated that neither of them were
consistent with the data reported for the natural product.
Fearing that an earlier step in our synthesis had proceeded
to give an incorrect stereocenter, we sought to resolve the
issue by X-ray crystallography. Reduction of 10 with NaBH4
produced a 1:1 mixture of the corresponding epimeric
alcohols. Both these alcohols were then converted to
separable epimeric p-nitrobenzoates. Careful crystallization
of one of the esters 14 gave suitable crystals for X-ray
crystallography, which proved that both the conjugate
addition and the ring expansion-cyclization sequence had
proceeded as anticipated to give the same relative stereo-
chemistry reported for pogostol (Figure 1).16 We can only
Initially we set out to synthesize pogostol O-methyl ether
5 from 4-methoxy-4-methyl-2-cyclohexenone 7 (R ) Me).
Conjugate addition of 3-butenylmagnesium bromide to 7 (R
) Me) using our previously developed conditions1b produced
the corresponding TMS enol-ether as a single diastereoiso-
mer, but unfortunately the butenyl group was delivered to
the enone from the opposite side of the methoxy group.
About this time, Csaky et al.12 reported that Grignard
reagents could be added in a syn manner to 4-hydroxy-2-
cyclopentenones without the need for copper(I) catalysis.
With this in mind, 4-hydroxy-4-methyl-2-cyclohexenone 7
(R ) H) was treated with 3 equiv of 3-butenylmagnesium
bromide and an excess of TMSCl followed by the addition
of Et3N to afford the enol-ether 8 as a single diastereoisomer.
It is interesting to note that if Et3N was omitted, then the
intermediate magnesium enolate was inert to reaction with
TMSCl and only the corresponding ketone could be isolated
on workup. As soon as Et3N was added, an instantaneous
reaction was triggered and TLC indicated the formation of
the enol-ether. Furthermore, the tertiary hydroxyl group in
8 proved to be so inert to reaction with TMSCl under these
conditions that it was carried on through subsequent steps
without protection. Cyclopropanation with Et2Zn/CH2I2 gave
an excellent yield of the hydroxy-cyclopropane 9. It should
be noted that the use of PhMe as a solvent, instead of the
usual Et2O, resulted in greatly enhanced rates and yields of
this and related cyclopropanations. With quantities of 9 in
hand, we then investigated the key Fe(III)-mediated ring
expansion-cyclization reaction. With Fe(NO3)3 and 1,4-
cyclohexadiene as a hydrogen atom donor, it was found that
the bicyclic ketone 10 could be formed in 57% yield as a
single diastereoisomer and without elimination of the tertiary
hydroxyl group. Introduction of the 2-propenyl moiety was
achieved by first treating 10 with ethyl triphenyl phosphorane
to generate the corresponding alkene 11 as a 1:1 E/Z mixture.
Hydroboration of this followed by in situ oxidation13 with
PCC generated the ketone 12 in good yield as a 2:1 mixture
Figure 1. X-ray of p-nitrobenzoate 14.
conclude, therefore, that the structure originally proposed for
the natural product is now untenable and that structural
revision will be necessary.
Initial attempts to form (()-pogostol O-methyl ether by
methylation of both (()-4 and (()-13 under standard
conditions with MeI and base were unsuccessful presumably
due to the extremely hindered nature of the tertiary hydroxyl
group. However, methylation under more forcing conditions
yielded the reported structure for pogostol O-methyl ether
(13) Rao, C. G.; Kulkarni, S. U.; Brown, H. C. J. Organomet. Chem.
1979, 172, C20-C22.
(7) Yang, Y.; Kinoshita, K.; Koyama, K.; Takahashi, K.; Tai, T.;
Nunoura, Y.; Watanabe, K. Phytomedicine 1999, 6, 89.
(8) Weyerstahl, P.; Marschall, H.; Splittgerber, U.; Wolf, D. FlaVour
Fragr. J. 2000, 15, 153.
(9) Fleischer, T. C.; Waigh R. D.; Waterman, P. G. J. Nat. Prod. 1997,
60, 1054.
(10) Hikino, H.; Hikino, Y.; Takeshita, Y.; Shirata, K.; Takemoto, T.
Chem. Pharm. Bull. 1967, 15, 321.
(11) van Beek, T. A.; Lelyveld, G. P.; Gijsen, H. J. M. J. Essent. Oil.
Res. 1993, 5, 169.
(14) Stepwise hydroboration and oxidation of the purified mixture of
alcohols with DMP yielded the ketones 12 as a 1:1 mixture.
(15) (a) Tebbe, F. N.; Parshall, G. W.; Reddy, G. S. J. Am. Chem. Soc.
1978, 3611. (b) Clawson, L.; Buchwald, S. L.; Grubbs, R. H. Tetrahedron
Lett. 1984, 25, 5733-5736.
(16) X-ray data were collected on a Bruker APEX single-crystal
diffractometer. Crystal data for 14: C19H25NO5, M ) 347.40, orthorhombic,
a ) 12.774(7) Å, b ) 10.955(4) Å, c ) 25.097(9) Å, U ) 3512(3) Å3, T
) 100 K, space group Pbca (no. 61), Z ) 8, µ(Mo KR) ) 0.095 mm-1
,
Rint ) 8.0% (for 18382 data), wR2 ) 10.1% (for all 2922 unique data), R1
) 4.7% [for 2161 data with I > 2s(I)].
(12) Csaky, A. G.; Mba, M.; Plumet, J. J. Org. Chem. 2001, 66, 9026.
3310
Org. Lett., Vol. 5, No. 18, 2003

(()-5 and the epimer, (()-15. These same two compounds
were also synthesized via an alternative strategy involving
methylation of 11 to give the ether 16. This was followed
by hydroboration to the alcohols 17 and Dess-Martin
oxidation to the ketones 18. Finally, Tebbe methylenation
afforded a mixture of (()-5 and (()-15, which were identical
to those prepared above (Scheme 3). Unfortunately, NMR
Scheme 3. Total Synthesis of (()-Pogostol O-Methyl Ether
and Epimera
Scheme 1. Fe(III)-Mediated Ring Expansion-Cyclization
a Reagents: (a) Me3OBF4, proton sponge, CH2Cl2; (b) MeOTf,
2,6-di-tert-butylpyridine, CH2Cl2; (c) MeOTf, 2,6-di-tert-butylpy-
ridine, CH2Cl2, heat; (d) BH3‚THF, THF then NaOH, H2O2; (e)
Dess-Martin periodinane, CH2Cl2; (f) Tebbe reagent (Cp2TiCl2,
AlMe3), CH2Cl2.
analysis showed that neither (()-5 or (()-15 possessed
spectral data consistent with that reported by Waterman et
via two different synthetic routes, from a ketone 10 whose
structure is unambiguous, we can only conclude that the re-
ported structure for pogostol O-methyl ether is also incorrect.
Although this was disappointing, it occurred to us that
“epi-pogostol” (()-13 may prove to be useful as an inter-
mediate in a racemic synthesis of the sesquiterpene kessane
6, which was previously synthesized by Gijsen et al.17 It was
envisaged that the pyran ring in 6 could be constructed by
an electrophilic cyclofunctionalization reaction of epi-pog-
ostol (()-13. Initial attempts using mercuric acetate gave
only starting material. Treatment of racemic (()-13 (drawn
in Scheme 4 with the same relative stereochemistry as 6)
Scheme 2. Total Synthesis of (()-Pogostol and epi-Pogostola
Scheme 4. Total Synthesis of (()-Kessanea
a Reagents: (a) 3-butenylmagnesium bromide, TMSCl, DMPU,
THF -78 °C, 1 h then Et3N; (b) Et2Zn, CH2I2, PhMe, 0 °C; (c)
Fe(NO3)3, 1,4-cyclohexadiene, DMF, 10 min; (d) 3 equiv Ph3Pd
CHCH3, THF, rt to 50 °C; (e) BH3‚THF, THF then PCC, THF; (f)
Tebbe reagent (Cp2TiCl2, AlMe3, PhMe), THF.
a Reagents: (a) N-iodosuccinimide, CH2Cl2, rt, 4 h; (b) Pd/C/
H2, EtOAc, 6 h.
al. for pogostol O-methyl ether. As with pogostol, there were
simply too many differences in key 13C signals for a satis-
factory match. As both (()-5 and (()-15 have been derived,
with N-iodosuccinimide18 generated the labile diastereomeric
iodides 19 and 20 in essentially quantitative yield. Attempts
Org. Lett., Vol. 5, No. 18, 2003
3311

at reductive deiodination under a number of conditions re-
sulted in either ring opening back to epi-pogostol (Zn/AcOH)
or no reaction (LiAlH4). Hydrogenation over Pd/C, however,
eventually afforded synthetic kessane in 66.5% overall yield
crystallographic, and subsequent synthetic studies, however,
have clearly demonstrated that the previously reported
structures for both these natural products are incorrect and
will require revision. It was possible to convert epi-pogostol
(()-13 into (()-kessane 6 in an additional two steps, thus
completing an eight-step linear synthesis of this tricyclic
sesquiterpene in which no protecting groups were used.
1
from (()-13. Detailed analysis of the H and 13C spectral
data of synthetic (()-6 showed excellent agreement ((0.2
ppm for all signals in 13C spectra) with that reported for the
natural product by van Beek et al.11
Acknowledgment. We thank the EPSRC and Roche
Pharmaceuticals for an Industrial CASE award.
In summary, a short racemic synthesis of the reported
structures of both pogostol and pogostol O-methyl ether has
been described using our previously developed Fe(III)-
mediated tandem radical ring expansion-cyclization chem-
istry of cyclopropylsilyl ethers. Detailed NMR, X-ray
Supporting Information Available: Experimental pro-
cedures and characterization for all new compounds prepared.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(17) Gijsen, H. J. M.; Wijnberg, J. B. P. A.; Stork, G. A.; deGroot, A.
Tetrahedron 1991, 47, 4409.
(18) Reitz, A. B.; Nortey, S. O.; Maryanoff, B. E. J. Org. Chem. 1987,
52, 4191.
OL035373U
3312
Org. Lett., Vol. 5, No. 18, 2003