Total synthesis of panaginsene with structural revision

Article abstract of DOI:10.1021/ol500849m

A facile total synthesis of the reported structure for panaginsene through a trimethylenemethane (TMM) diyl mediated tandem cycloaddition reaction revealed that the spectroscopic data of the synthesized structure did not match with the data of the natural

Full text of DOI:10.1021/ol500849m

Letter
pubs.acs.org/OrgLett
Total Synthesis of Panaginsene with Structural Revision
Sujeong Geum and Hee-Yoon Lee*
Department of Chemistry, Korea Advanced Institute of Science and Technology (KAIST), Daejeon 305-701, Korea
S
* Supporting Information
ABSTRACT: A facile total synthesis of the reported structure
for panaginsene through a trimethylenemethane (TMM) diyl
mediated tandem cycloaddition reaction revealed that the
spectroscopic data of the synthesized structure did not match
with the data of the natural product. The total synthesis of the
stereoisomer of the reported structure confirmed that the correct
structure of panaginsene was the 11-epi stereoisomer of the
originally proposed structure of panaginsene.
mong the hydrocarbons isolated from the roots of Panax
ginseng C.A. Meyer,1 a sesquiterpene hydrocarbon with an
Scheme 1. Plausible Biosynthetic Route to 1
A
angularly fused tricyclic moiety, panaginsene 1 caught our
attention, as it belongs to the senoxydane family of angularly
fused triquinanes (Figure 1). The only other reported natural
Figure 1. Proposed structures of panaginsene and senoxydene.
product with a senoxydane framework was senoxydene 2,2a and
its structure was proved to be incorrect through the total
syntheses.2 Although the actual structure of senoxydene is still
unknown, it is noteworthy that the plausible biosynthetic route
to the reported structure of senoxydene was introduced to
construct its relative stereostructure. The structure of
panaginsene 1 was assigned based on the NMR data, and
when we adopted this biosynthetic analysis to panaginsene 1,
we could postulate that the C-11 methyl epimer of panaginsene
1′ might also be the natural product. The stereochemistry of
the methyl group at the C-11 of panaginsene might be
generated through a stereospecific 1,2-hydride shift as shown in
the plausible biosynthesis (Scheme 1). To confirm the structure
of a senoxydane natural product, panaginsene 1, the total
synthesis of 1 was accomplished through the tandem
cycloaddition reaction via trimethylenemethane (TMM) diyl.3
The retrosynthetic analysis of 1 (Scheme 2) suggested that
angularly fused triquianane 1 could be obtained in one step
stereoselectively by TMM diyl mediated intramolecular tandem
[2 + 3] cycloaddition reaction4 of the diazotized linear
precursor 8. The linear precursor 8 could readily be prepared
from the propargylic bromide 9, which would be derived from
(+)-citronellal. This tandem cycloaddition reaction strategy was
just right for the total synthesis of 1 since the reaction has been
applied successfully to the total synthesis of various
polyquinane natural products.5 The tandem cycloaddition
reaction would produce the angularly fused triquinane
stereoselectively, as the C-11 carbon center should control
the relative stereochemistry of the triquinane framework4 along
with the preset tetrasubstituted olefin.
The total synthesis of 1 was accomplished in 7 steps from 10
that was readily prepared from (+)-citronellal through
conversion of the aldehyde of citronellal into the corresponding
gem-dibromide6 (Scheme 3). The gem-dibromide of 10 was
transformed into the propargyl alcohol 11 through the two-step
protocol of base treatment in refluxing petroleum ether
followed by hydroxymethylation with paraformaldehyde at
room temperature of the acetylide anion generated at 0 °C by
the treatment of n-BuLi.6a The propargyl alcohol of 11 was
converted into the allenyl precursor 8 by Cu(I)-catalyzed SN2′
reaction of the corresponding bromide 9 using Grignard
reagent 127 in THF at 0 °C followed by deprotection of the
acetal group to the ketone using p-toluenesulfonic acid and
aqueous formaldehyde in THF at room temperature. Finally,
formation of the hydrazone 13 from the ketone of 8 using N-
Received: March 21, 2014
Published: April 17, 2014
© 2014 American Chemical Society
2466
dx.doi.org/10.1021/ol500849m | Org. Lett. 2014, 16, 2466−2469

Organic Letters
Letter
spectroscopic data of the synthesized 1 did not match with
those of the natural product. Though the NMR spectroscopic
data of 1 did not match with the reported data, those data from
the natural product and the synthetic one were close enough to
each other to convince us that the actual structure of
panaginsene might be the C-11 methyl epimer, 1′. Therefore,
we decided to synthesize 1′ to confirm the actual structure of
the natural panaginsene.
Scheme 2. Retrosynthetic Analysis of 1
Retrosynthetic analysis of 1′ indicated that an extra
functionality was necessary that can epimerize the C-11 methyl
stereochemistry of 17 (Scheme 4). The precursor for the
Scheme 4. Retrosynthetic Analysis of 1′
Scheme 3. Total Synthesis of 1
tandem cycloaddition, 18, can be prepared from 19 in the same
manner as before, and 19 could be synthesized from dithiane
20 prepared from 2-methylpropane-1,3-diol, 21.
The total synthesis of the epimeric compound 1′ of the
reported structure of panaginsene started from 2-methyl-
propane-1,3-diol 21. Though enantiomerically pure protected
3-hydroxy-2-methylpropanal was available, it turned out that
the absolute stereochemistry of the methyl group was lost
during the dithiane formation of the aldehyde. That observation
led us to use 21 as the starting material for the synthesis
(Scheme 5). Monoprotection of the diol of 21 using 1 equiv of
the reagents in THF followed by Swern’s oxidation produced
the aldehyde that was converted into dithiane 20 using
propane-1,3-dithiol with magnesium bromide as a Lewis acid
at 0 °C. After isoprenylation at −40 °C of the anion of 20
generated by treatment with n-BuLi at 0 °C, deprotection of
the silyl group with tetrabutylammonium fluoride afforded the
alcohol 22. Oxidation of the alcohol of 22 that contained the
dithiane functionality was most effective with Doering’s
reagent,10 and subsequent treatment of the produced aldehyde
with Corey−Fuchs reagent11 produced 23. Since direct
conversion of the dibromoolefin of 23 into the propargylic
alcohol 19 using the same approach as in the synthesis of 11
from 10 was not successful, 19 was synthesized in a two-step
procedure of carboxylation of the acetylenic anion generated
from 23 by treatment of n-BuLi with ethyl chloroformate at
−40 °C followed by DIBAL-H reduction. The precursor for the
tandem cycloaddition reaction 18 was prepared in the same
manner as in the total synthesis of 1. Tandem cycloaddition
reaction of the allenyl diazo intermediate, generated from the
allenyl ketone 18, produced tricyclic compound 17 in 80% yield
over two steps. To our surprise, the tandem cycloaddition
proceeded much more efficiently from 18 than 13 in a selective
manner. The Thorpe−Ingold effect exerted by the dithiane
group must have facilitated the cycloaddition reaction toward
tosylhydrazide in methanol under 10 °C set the stage for the
tandem cycloaddition reaction. Generation of the diazo
compound 14 in refluxing toluene from the anion of 13
generated by treatment with NaH at 0 °C in toluene triggered
the tandem cycloaddition reaction through TMM 16 to
produce 1.
Due to the volatile and nonpolar nature of 1 and other
isomeric products, purification of the products was challenging.
While there were five isomeric products other than 1 observed
by the GC-MS, none of them could be purified and thus their
structures could not be identified.8 Only 1 was separated
successfully in pure form using silver nitrate impregnated silica
9
1
gel chromatography. As we suspected, the H and 13C NMR
2467
dx.doi.org/10.1021/ol500849m | Org. Lett. 2014, 16, 2466−2469

Organic Letters
Letter
the elimination product. Since the unexpected reactivity of the
alcohol of 27 was thought to be due to the steric hindrance of
the alcohol with the β-configuration, inversion of the alcohol of
27 to the more exposed α-configuration was anticipated to
increase the reactivity of the alcohol and its derivatives. To
drive the epimerization reaction of 27, we adopted the reaction
catalyzed by the ruthenium complex 28 developed for the
dynamic kinetic resolution of racemic alcohols.16
Scheme 5. Synthesis of Angularly Fused Triquinane, 17
When a solution of 27 in the presence of ruthenium catalyst
28 was irradiated with a fluorescent lamp, an epimerization
reaction proceeded to produce a mixture of epimeric alcohols
27 and 29 with a 1:4 ratio. Fortunately, the desired alcohol 29
was thermodynamically more stable and was isolated in a 64%
yield. As anticipated, Mitsunobu type sulfide formation at 110
°C to 30 was successful, though the reaction was sluggish.
Finally a phenylsulfenyl group was removed by a reductive
desulfurization reaction17 in the presence of lithium in
ethylamine at low temperature to form 1′. To our delight,
the spectroscopic data of 1′ were in complete agreement with
the data reported for panaginsene.18
With both epimeric methyl isomers of panaginsene in hand,
the relative stereochemistry of the C-11 stereogenic center of 1
1
and 1′ was firmly confirmed by H−1H NOE experiments
(Figure 2).
the desired triquinane structure.12 Another surprise was that the
reaction produced 17 almost stereospecifically, as the C-11
methyl epimer of 17 was not detected.
For the completion of the total synthesis of 1′, the C-11
methyl group of 17 had to be inverted and the dithiane masked
carbonyl functionality had to be removed.
Deprotection of the dithiane of 17 with iodine and sodium
bicarbonate at 0 °C produced the corresponding ketone 25
(Scheme 6).13 Base mediated epimerization of the C-11 methyl
1
Figure 2. H−1H NOE experiments of 1 and 1′.
Scheme 6. Completion of the Synthesis of 1′
While the W. A. Konig group determined the relative
̈
1
stereochemistry of the C-11 methyl group by H−1H NOE
correlation of the protons of the C-15 with the proton H3, it
was observed that both protons of the C-15 of 1 and 1′ showed
a correlation with the proton H3. Furthermore, the protons of
C-15 of 1′ showed a correlation with the proton H6, though
weak. On the other hand, the protons of the C-15 of 1 showed
no correlation with the proton H6. It is quite probable that the
weak NOE between protons of the C-15 and the H6 mislead
the original stereochemical assignment of the natural product.
In conclusion, we have achieved the first total synthesis of the
proposed structure of panaginsene 1 and the actual structure of
panaginsene 1′ through the intramolecular TMM diyl-mediated
tandem cycloaddition reaction with high efficiency and
stereoselectivity. Through the total synthesis, we confirmed
that the originally assigned structure 1 must be revised to the
correct structure 1′.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedure and spectral data for all new
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
group produced an inseparable mixture of 25 and its C-11
methyl epimer in a 1:1 ratio. Reduction of the epimeric mixture
using NaBH4 at 0 °C proceeded stereoselectively to produce a
diastereomeric mixture of 26 and 27. After the desired isomer
27 was separated, various attempted deoxygenation reactions
failed to provide the reduction product.14,15 Mitsnobu type
functionalization reactions of the alcohol of 27 only produced
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: leehy@kaist.ac.kr.
2468
dx.doi.org/10.1021/ol500849m | Org. Lett. 2014, 16, 2466−2469

Organic Letters
Letter
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported by the Basic Science Research
Program through the National Research Foundation of Korea
(NRF) funded by the Ministry of Education, Science and
Technology (KRF-2008-314-C00198), and by a 2011-2012
basic research grant from the KAIST.
REFERENCES
■
(1) Konig, W. A.; Koch, A.; Basar, S.; Richter, R. Phytochemistry 2005,
66, 2708−2713.
(2) (a) Bohlmann, F.; Zdero, C. Phytochemistry 1979, 18, 1747−
1748. (b) Ito, S.; Tsunoda, T.; Kabasawa, Y. Tetrahedron Lett. 1984,
25, 773−776. (c) Paquette, L. A.; Galemmo, R. A., Jr.; Springer, J. P. J.
Am. Chem. Soc. 1983, 105, 6975−6976. (d) Paquette, L. A.; Galemmo,
R. A., Jr.; Caille, J. C.; Valpey, R. S. J. Org. Chem. 1986, 51, 686−695.
(3) Lee, H.-Y.; Kang, T.; Kim, W.-Y.; Kim, B. G. J. Am. Chem. Soc.
2011, 133, 18050−18053.
(4) Lee, H.-Y.; Yoon, Y.; Kang, T. Chem.Asian J. 2011, 6, 646−651.
(5) (a) Lee, H.-Y.; Kim, Y. J. Am. Chem. Soc. 2003, 125, 10156−
10157. (b) Lee, H.-Y.; Shin, S.-S.; Kim, W.-Y. Chem.Asian J. 2012, 7,
2450−2456. (c) Lee, H.-Y.; Song, S.-B.; Kang, T.; Kim, Y. J.; Geum, S.
J. Pure Appl. Chem. 2013, 85, 741−753.
(6) (a) Kim, Y.-J.; Yoon, Y.; Lee, H.-Y. Tetrahedron 2013, 69, 7810−
7816. (b) Spaggiari, A.; Vaccari, D.; Davoli, P.; Torre, G.; Prati, F. J.
Org. Chem. 2007, 72, 2216−2219.
(7) Kalesse, M.; Stellfeld, T.; Stelmakh, A. Org. Lett. 2006, 5, 3485−
3488.
(8) The ratio of 1 and isomers was determinded by GC-MS (1/
isomers = 1.27:1)
(9) (a) Williams, C. M.; Mander, L. N. Tetrahedron 2001, 57, 425−
447. (b) Gupta, A. S.; Dev, S. J. Chromatogr. 1963, 12, 189−195.
(10) (a) Mohan, P.; Fuertes, M. J. Tetrahedron Lett. 2013, 54, 3919−
3925. (b) Parikh, J. R.; Doering, W. v. E. J. Am. Chem. Soc. 1967, 89,
5505−5507.
(11) Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 36, 3769−
3772.
(12) Jung, M. E.; Pilizzi, G. Chem. Rev. 2005, 105, 1735−1766.
(13) Renaud, P.; Schenk, K.; Lapointe, G. Org. Lett. 2011, 13, 4774−
4777.
(14) (a) Barton, D. H.; McCombie, S. W. J. Chem. Soc., Perkin Trans.
1 1975, 16, 1574−1585. (b) Zard, S. Z.; Quiclet-Sire, B. Tetrahedron
Lett. 1998, 39, 9435−9438.
(15) Ireland, R. E.; Muchmore, D. C.; Hengartner, U. J. Am. Chem.
Soc. 1972, 94, 5098−5100.
(16) Do, Y.; Hwang, I.-C.; Kim, M.-J.; Park, J. J. Org. Chem. 2010, 75,
5740−5742.
(17) Paquette, L. A.; Galatsis, P. J. Org. Chem. 1989, 54, 5039−5043.
(18) See Supporting Information.
NOTE ADDED AFTER ASAP PUBLICATION
Scheme 6 contained an error in the version published ASAP
April 17, 2014; the correct version reposted April 21, 2014.
■
2469
dx.doi.org/10.1021/ol500849m | Org. Lett. 2014, 16, 2466−2469