Total synthesis of the proposed structure of aldingenin B

Article abstract of DOI:10.1021/ol3007259

The first enantioselective total synthesis of the proposed structure of aldingenin B is reported in 16 steps from known compounds. The stereochemistry at C5 and C6 were established by an asymmetric acetal aldol. Following a ring-closing metathesis, a selective, substrate-controlled hydrogen bond-mediated dihydroxylation provided control of the C2 and C3 stereocenters. Discrepancies in the spectroscopic data of the synthetic and natural material led to the conclusion that the structure of the natural sample was misassigned.

Full text of DOI:10.1021/ol3007259

ORGANIC
LETTERS
2012
Vol. 14, No. 8
2168–2171
Total Synthesis of the Proposed Structure
of Aldingenin B
Michael T. Crimmins* and Colin O. Hughes
Kenan, Caudill, Venable, and Murray Laboratories of Chemistry, University of North
Carolina at Chapel Hill, Chapel Hill, North Carolina 27599, United States
crimmins@email.unc.edu
Received March 21, 2012
ABSTRACT
The first enantioselective total synthesis of the proposed structure of aldingenin B is reported in 16 steps from known compounds. The
stereochemistry at C5 and C6 were established by an asymmetric acetal aldol. Following a ring-closing metathesis, a selective, substrate-
controlled hydrogen bond-mediated dihydroxylation provided control of the C2 and C3 stereocenters. Discrepancies in the spectroscopic data of
the synthetic and natural material led to the conclusion that the structure of the natural sample was misassigned.
The Laurencia red algae are a prolific source of
halogenated secondary metabolites.1 These marine natural
products often serve as taxonomic markers for the deter-
mination of species within the Laurencia genus. It was thus
Scheme 1. Retrosynthetic Analysis of the Proposed Structure of
Aldingenin B (1)
of great interest when secondary metabolites of a Brazilian
species identified as Laurencia aldingensis were character-
ized, as this species was previously unknown outside its
native habitat off the coast of South Australia.2
Of the four novel sesquiterpene-derived natural pro-
ducts isolated, aldingenin B2b (1) was targeted for synthesis
due to its compact and complex molecular architecture.3
Its proposed structure, which was postulated based on
extensive NMR studies, possesses a stereochemically dense
tetracyclic framework, containing 7 stereocenters, 4 of
which are contiguous, a brominated tetrahydropyran, and
a cyclic ketal.
Our retrosynthesis is outlined in Scheme 1. It was
envisioned that the proposed structure of 1 would arise
(1) (a) Marine Natural Products; Scheuer, P. J., Ed.; Academic:
New York, 1978; Vol. 1. (b) Marine Natural Products; Scheuer, P. J.;
Ed.; Academic: New York, 1983; Vol. 5. (c) Fenical, W.; Norris, J. N.
J. Phycol. 1975, 11, 104. (d) Fenical, W. J. Phycol. 1975, 11, 245. (e)
Masuda, M.; Abe, T.; Suzuki, T.; Suzuki, M. Phycologia 1996, 35, 550.
(2) (a) de Carvalho, L. R.; Fujii, M. T.; Roque, N. F.; Kato, M. J.;
Lago, J. H. G. Tetrahedron Lett. 2003, 44, 2637. (b) de Carvalho, L. R.;
Fujii, M. T.; Roque, N. F.; Lago, J. H. G. Phytochemistry 2006, 67, 1331.
(3) Previous synthetic approach to the bicyclic ketal: Yang, J.;
Tummatorn, J.; Slegeris, R.; Tlais, S. F.; Dudley, G. B. Org. Lett.
2011, 13, 2065.
from a late stage bromoetherification of tricyclic ketal 2.
This ketal would be obtained from cycloketalization of
dithiane 3 and further manipulation to install the C14
methyl. Dithiane 3 would result from a nucleophilic
r
10.1021/ol3007259
Published on Web 04/03/2012
2012 American Chemical Society

Scheme 2. Synthesis of the Bicyclic Ketal Precursor
addition to aldehyde 4, the C2ÀC3 syn-diol installed via a
syn-dihydroxylation of appended cyclohexene 5. Cyclo-
hexene 5 would be formed through a ring-closing metath-
esis of aldol adduct 6, derived from an acetal aldol between
thiazolidinethione 7 and dibenzyl acetal 8.
of harnessing the architectural features of the molecule,
inspired by the work of Donohue,10 an unprecedented and
completely diastereoselective dihydroxylation was accom-
plished utilizing the H-bonding between primary alcohol 5
and the Lewis basic OsO4/TMEDA complex to direct the
oxidant to the proper face of the olefin afforded triol 10.
It should be noted this fortuitous result was possible only
after dihydroxylations utilizing the corresponding aldehyde
led to decomposition. Since the selective oxidation of triol 10
to the corresponding aldehyde proved problematic, the syn-
diol was initially protected as an acetonide, but later-stage
removal of the protecting group proved difficult. Thus, the
more acid-labile cyclopentylidene 11 was prepared. The
structure of acetal 11 was confirmed by the lack of NOE
signal between Me15ÀH5, and the coupling constants of
The synthesis began by treating 3-methyl-3-butenal4 to
nonisomerizing acetalizing conditions5 to provide dibenzyl
acetal 8 (Scheme 2). The requisite C6ÀC5 stereochemistry
of aldol adduct 6was set with a chiral auxiliary mediated acetal
aldol6 addition of thiazolidiethione 7 and dibenzyl acetal 8.
Reductive removal of the auxiliary upon exposure to
lithium borohydride provided alcohol 9. A ring-closing
metathesis reaction7 formed the cyclohexene 5. Initial
dihydroxylation of the alkene utilizing standard Upjohn
conditions8 led to a 3:1 inseparable mixture of isomers
favoring the desired diol. Modest improvement was seen
with the use of AD-mix β9 (6:1); however the isomers could
not be separated after subsequent reactions. Withthe focus
axial H5 (J5ax.-6ax. = 9.0 Hz, J5ax.-4eq. = 3.0 Hz, J5ax.-4ax
=
10.8 Hz). The alcohol was then converted to aldehyde 4
utilizing mild TPAP/NMO oxidation conditions11 to avoid
decomposition experienced during Swern oxidation.
All initial efforts toward the nucleophilic C8ÀC7 bond
forming event between prenyl-dithiane 12 and aldehyde
4 resulted in decomposition of the starting material, pre-
sumably via an undesired enolization.12 Cerium trichloride
was explored as an additive to decrease the effective basicity
while maintaining the nucleophilicity of the metalated
dithiane; however, results proved inconsistent. Numerous
known methods of preparing cerium trichloride13 for nu-
cleophilic additions were attempted; however all were found
to be inadequate for the substrate. Following extensive
(4) Crimmins, M. T.; Kirincich, S. J.; Wells, A. J.; Choy, A. L. Synth.
Commun. 1998, 28, 3675. Crimmins, M. T.; Choy, A. L. J. Am. Chem.
Soc. 1999, 121, 5663.
(5) Tsunoda, T.; Suzuki, M.; Noyori, R. Tetrahedron Lett. 1980, 21,
1357.
(6) Cosp, A.; Romea, P.; Talavera, P.; Urpı, F.; Vilarrasa, J.; Font-
Bardia, M.; Solans, X. Org. Lett. 2001, 3, 615. (b) Cosp, A.; Larrosa, I.;
Vilasis, I.; Romea, P.; Urpı, F.; Vilarrasa, J. Synlett 2003, 1109. (c)
ꢀ
ꢀ
Baiget, J.; Cosp, A.; Galvez, E.; Gomez-Pinal, L.; Romea, P.; Urpı, F.
Tetrahedron 2008, 64, 5637.
(7) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1,
953.
(8) VanReenen, V.; Kelly, R. C.; Cha, D. Y. Tetrahedron Lett. 1976,
17, 1973.
(9) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. Rev.
1994, 94, 2483.
(10) (a) Donohoe, T. J. Synlett. 2002, 8, 1223. (b) Donohoe, T. J.;
Blades, K.; Moore, P. R.; Waring, M. J.; Winter, J. J. G.; Helliwell, M.;
Newcombe, N. J.; Stemp, G. J. Org. Chem. 2002, 67, 7946.
(11) Ley, S. V.; Norman, J.; Griffith, W. P.; Marsden, S. P. Synthesis
1994, 639.
(12) Smith, A. B., III; Zhu, W.; Shirakami, S.; Sfouggatakis, C.;
Doughty, V. A.; Bennett, C. S.; Sakamoto, Y. Org. Lett. 2003, 5, 761.
Org. Lett., Vol. 14, No. 8, 2012
2169

The analogue 21 was synthesized by the nucleophilic
addition of dithiane 2016 to aldehyde 4 (Scheme 4). Dithiane
21 was converted to ketone 22, which when exposed to
perchloric acid, serendipitously cleaved the silyl ether in
addition to the cyclopentylidene resulting in cyclization
to ketal 23. The structure of the major diol isomer was
Scheme 3. Control for the Formation of the Tricyclic Ketal
1
confirmed by H, 13C NMR, a NOE crosspeak between
H7ÀH5, and a COSY signal between H6ÀH7. Unfortu-
nately, on larger scale reactions (>20 mg) these conditions
proved very inconsistent, a result attributed to the immis-
cibility of the dichloromethane with aqueous perchloric
acid. It was found that ultrasonic irradiation was necessary
to ensure proper mixing of the biphase on larger scales.17
Allaying concerns of lactonization, bis-oxidation under
Swern conditions18 provided keto-aldehyde 24. Exploiting
the hindered steric environment of the ketone, exposure of
aldehyde 24 to isopropylidene triphenylphosphorane formed
the desired trisubstituted olefin 25. Axial methyl lithium
addition provided the necessary stereochemistry at C7 (as
confirmed by NOESY correlation between Me14ÀH1β),
forming tertiary alcohol 2 in low yield (17%) as well as C9
elimination products. Again, the use of cerium trichloride
proved important in suppressing elimination, as its inclusion
in the axial methyl lithium addition formed only tertiary
alcohol 2 without any side products.
Formation of prenyl 25 with isopropylidene triphenyl-
phosphorane proved very inconsistent, necessitating the
development of an alternative strategy to forge trisubsti-
tuted olefin 25. Both “salt-free” methylene Wittig condi-
tions19 and Tebbe olefination20 provided terminal olefin
26, however in low yields (<22%). It was discovered that
addition of keto-aldehyde 24 to a solution of Nysted
reagent21 reliably produced terminal olefin 26, which when
exposed to cross-metathesis conditions in neat 2-methyl-2-
butene22 provided trisubstituted olefin 25. Bromoetherifi-
cation of alcohol 2 using 2,4,4,6-tetrabromocyclohexa-2,5-
dienone (TBCO)23,24 afforded bromo-tetrahydropyran 27
along with a mixture of bromo-tetrahydrofurans. The
stereochemistry of 27 was identified by the characteristic
axial H10 J-values of 13.2 and 4.2 Hz and NOE signals
between H9βÀMe13, H9βÀMe14, and Me13ÀMe14.
Removal of the benzyl group via standard palladium
hydroxide promoted hydrogenation afforded the proposed
structure of aldingenin B (1).
investigation, a new procedure was developed for purifying
and solubilizing cerium trichloride,14 allowing for the con-
sistent formation of an inconsequential diastereomeric mix-
ture of alcohol 3 in good yields.
Following numerous unsuccessful attempts at convert-
ing dithiane 3 into the tricyclic ketal core of 1 (e.g., through
formation of the diol and then treatment of the dithiane to
various oxidants), conditions were developed to convert
the dithiane into carbonyl 13 without conjugation of the
olefin. Unfortunately, upon exposure to perchloric acid,
the desired tricyclic intermediate 14 was not formed, but
rather tetrahydropyran 15 was formed, presumably result-
ing from acid-promoted tertiary carbocation formation
at C11, followed by intramolecular trapping with the C7
alcohol.
Cyclization attempts utilizing milder acids led to no
reaction, and efforts to avoid alcohol-carbocation trap-
ping by mitigating the nucleophilicity of the C7 oxygen
through protecting groups or oxidation to the ketone were
met with no reaction and decomposition, respectively.
As a control experiment, saturated dithiane 16 was
coupled with aldehyde 4, followed by dithiane removal,
whereupon exposure to perchloric acid cleanly afforded
tricycle 19 after oxidation of the secondary alcohol
1
(Scheme 3). It is notable that the H NMR spectrum of
tricycle 19 is analogous to that of similar compounds
prepared by Dudley during studies toward the synthesis
of aldingenin B.15 The knowledge gained from this experi-
ment indicated that the C10À11 olefin in 13 was respon-
sible for failure to elaborate dithiane 3 into alcohol 14.
Therefore, efforts were focused toward the synthesis of an
analogue of prenyl-dithiane 3 containing a masked olefin
as a way to circumvent these problems.
(16) Gaunt, M. J.; Hook, D. F.; Tanner, H. R.; Ley, S. V. Org. Lett.
2003, 5, 4815.
(17) Crimmins, M. T.; Bankaitis-Davis, D. M.; Hollis, W. G. J. Org.
Chem. 1988, 53, 652. Crimmins, M. T.; O’Mahony, R. J. Org. Chem.
1990, 55, 5894.
(18) Mancuso, A. J.; Huang, S.-L.; Swern, D. J. Org. Chem. 1978, 43,
2480.
(19) Schmidba, H.; Vornberg, W.; Stuhler, H. Chem. Ber. 1972, 105,
1084.
(20) Tebbe, F. N.; Parshall, G. W.; Reddy, G. S. J. Am. Chem. Soc.
1978, 100, 3611.
(21) (a) Nysted, L. N. U.S. Patent 3,865,848, 1975; Chem. Abstr. 1975,
83, 10406q. (b) Matsubara, S.; Sugihara, M.; Utimoto, K. Synlett 1998,
313.
(13) (a) Imamoto, T.; Kusumoto, T.; Tawarayama, Y.; Sugiura, Y.;
Mita, T.; Hatanaka, Y.; Yokoyama, M. J. Org. Chem. 1984, 49, 3904. (b)
Dimitrov, V.; Kostova, K.; Genov, M. Tetrahedron Lett. 1996, 37, 6787.
(c) Krasovskiy, A.; Kopp, F.; Knochel, P. Angew. Chem., Int. Ed. 2006,
45, 497. (d) Metzger, A.; Gavryushin, A.; Knochel, P. Synlett 2009, 9,
1433.
(22) Chatterjee, A. K.; Sanders, D. P.; Grubbs, R. H. Org. Lett. 2002,
4, 1939.
(14) See Supporting Information for details.
(15) Compound 3 in ref 3 above as well as two additional similar
compounds provided in personal communications.
(23) Hashimoto, M.; Kan, T.; Nozaki, K.; Yanagiya, M.; Shirahama,
H.; Matsumoto, T. J. Org. Chem. 1990, 55, 5088.
(24) Kato, T.; Ichinose, I.; Hosogai, T. Chem. Lett. 1976, 1187.
2170
Org. Lett., Vol. 14, No. 8, 2012

Scheme 4. Completion of the Synthesis of the Proposed Structure of Aldingenin B
Additionally, the HMBC spectrum of the natural sample
does not display a H2ÀC8 crosspeak, whereas this key
HMBC signal is observed in synthetic 1. Moreover, COSY,
NOESY, HMQC, HMBC, and DEPT-135 spectra indicate
the structure of synthetic 1 is as shown, and spectral informa-
tion of all intermediates are as expected for such structures.
In summary, the first total synthesis of the proposed
structure of aldingenin B has been accomplished in 16
steps (LLS) in 2.7% overall yield. Key steps include a
diastereoselective acetal aldol, a ring-closing metathesis, a
substrate-controlled, hydrogen bond-mediated dihydrox-
ylation, a cycloketalization to construct the compact
tricycle, and a bromoetherification to complete the carbo-
cycle. The development of a novel way of drying and
solubilizing cerium trichloride for use in nonbasic nucleo-
philic additions was also achieved.
Table 1. 1H NMR Assignments of Natural and Synthetic 1
H
natural sample
C6D6, 500 MHz
synthetic 1
no.
C6D6, 600 MHz
1R
1β
2
1.65 m
1.55 ddd (15.0, 3.6, 1.8)
1.65 ddd (15.0, 3.6, 1.8)
3.62 dd (3.6, 1.8)
1.10 dd (13.8, 7.8)
2.21 dd (13.8, 7.8)
4.65 app. t (7.8)
1.53 app. s
3.99 dd (9.6, 6.3)
1.92 dd (14.5, 9.6)
2.16 dd (14.5, 4.7)
3.86 ddd (9.6, 8.4, 4.7)
1.44 dd (9.0, 8.4)
1.72 dd (13.5, 3.5)
2.19 t (13.5)
4β
4R
5
6
9R
9β
10
12
13
14
15
2.25 dd (13.2, 4.2)
2.50 t (13.2)
4.17 dd (13.5, 3.5)
1.36 s
4.55 dd (13.2, 4.2)
1.40 s
1.49 s
1.39 s
1.19 s
1.24 s
1.57 s
1.00 s
Acknowledgment. This work was supported by a re-
search grant from The National Institutes of General
Medical Sciences (GM60567). We thank the Department
of Education for a GAANN fellowhsip (C.O.H.). Thanks
Unfortunately, the NMR spectra of synthetic 1 differ
significantly from the data for the natural sample. Inspec-
tion of the assigned spectra of the natural sample raises
doubts as to the validity of its structural assignment. In
particular, several discrepanciesexist between the observed
couplings of H5, H6, and H2 in the natural sample and the
expected couplings of the assigned structure.
~
are due to Dr. Joao H. G. Lago (Universidade Federal de
Sao Paulo, Diadema, Brazil) for spectral data of the natural
~
sample as well as to Professor Greg Dudley (Florida State
University) for compounds similar to 19, 25, and26. Thanks
to Drs. Sohrab Habibi and Marc ter Horst (UNCÀCH) for
help with acquiring spectral data. Insightful conversations
with Dr. Marc D. Walter, Dr. Mark Mans, and Dr. Simon
Meek (UNCÀCH) are gratefully acknowledged.
Inspection of models of 1 reveals the H6ÀH5 dihedral
angle to be 90° ((2°); the expected coupling of such vicinally
orthogonal protons is <2 Hz. The natural sample displays
an 8.4 Hz coupling between these protons, while there is no
detected coupling between H5ÀH6 in synthetic 1.
Note Added after ASAP Publication. An incomplete SI
file was replaced on April 6, 2012.
Furthermore, the reported coupling constants for the
“bridgehead” protons H6 and H2 in the natural sample are
9.0, 8.4 and 9.6, 6.3 Hz respectively (Table 1). The expected
value of coupling constants of such bridgehead protons is
<4 Hz, as observed in the couplings of H2 (J = 3.6, 1.8 Hz)
and H6 (app. s) in the synthetic sample and examples
reported by Dudley.3,15
Supporting Information Available. Experimental de-
tails and spectral data for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 8, 2012
2171