Asymmetric total synthesis of solandelactone E: Stereocontrolled synthesis of the 2-ene-1,4-diol core through a lithiation-borylation-allylation sequence

Article abstract of DOI:10.1002/anie.201003236

A highly stereoselective, 13-step synthesis of solandelactone E is reported which employs the lithiation-borylation-allylation sequence as the key step (see scheme). This synthetic method solves the problem of poor stereocontrol at C11 that had dogged many previous syntheses of this class of molecules.

Full text of DOI:10.1002/anie.201003236

Angewandte
Chemie
DOI: 10.1002/anie.201003236
Natural Product Synthesis
Asymmetric Total Synthesis of Solandelactone E: Stereocontrolled
Synthesis of the 2-ene-1,4-diol Core through a Lithiation–Borylation–
Allylation Sequence**
Anna Robinson and Varinder K. Aggarwal*
The solandelactones comprise a novel class of oxygenated
fatty acids that contain an eight-membered lactone linked to a
trans cyclopropane, and were first isolated by Shin and co-
workers in 1996[1] from the hydroid Solanderia secunda. These
compounds belong to a wider group of marine oxylipins that
contain the linked lactone and cyclopropane unit, including
halicholactone[2] and the constanolactones.[3] Initial assign-
ment of structure and absolute stereochemistry was based
primarily on NMR studies, which led to two sets of
compounds depending on their configuration at C11: R con-
figuration for A, C, E, and G; S configuration for B, D, F, and
H. However, the experimental data for synthetic solandelac-
tone “F”, which had been synthesized by Martin and co-
workers,[4] matched the data for solandelactone E, thus
indicating that the two sets of compounds had been incor-
rectly assigned. Further syntheses by White et al.[5] as well as
Pietruszka and Rieche[6] confirmed that the configuration at
C11 in all of the solandelactones A–H should be depicted as
in Scheme 1.
predominantly by the configuration at C14 and not by the
configuration of the cyclopropane ring.[6,9] Martinꢀs synthesis
provides a unique strategy to access this class of compounds,
which features high levels of stereocontrol.[4] His strategy
avoids the C11–C12 disconnection but this inevitably makes
the synthesis more linear.
Our research group has recently described a new method
for the stereocontrolled synthesis of anti-1,4-diols[12] which
seemed to offer an ideal solution to achieving high levels of
stereocontrol at C11 while maintaining a convergent syn-
thesis. This method comprised of the coupling of a lithiated
carbamate, b-silyl vinyl borane 1, and an aldehyde, which gave
an allyl silane in high yield with high enantio- and diastereo-
selectivity. Subsequent epoxidation and acid-catalyzed elim-
ination gave 2-ene-1,4-diols[13] with high enantio- and diaste-
reoselectivity (Scheme 2). It was essential to use an unhin-
Scheme 1. Revised structure of solandelactones A–H.
Scheme 2. New method for synthesis of allyl silanes and 2-ene-1,4-
diols. mCPBA=meta-chloroperoxybenzoic acid.
Almost all syntheses of the solandelactones, constanolac-
tones,[7,8] and halicholactone[9–11] introduce the aliphatic side
chain (e.g. C12–C22 in solandelactone) through a Nozaki–
Hiyama–Kishi reaction between a vinyl organometallic
reagent and the cyclopropyl aldehyde. Unfortunately, this
reaction occurs with very poor stereocontrol. In fact, it has
been shown that the configuration at C11 ( ꢀ 2:1) is controlled
dered diamine or diamine-free conditions in the reaction of
the lithiated carbamate with the borane to achieve high
enantioselectivity.[12a] Herein we describe the total synthesis
of solandelactone E in a concise manner with high levels of
control over the configuration of all stereogenic centers, thus
establishing our methodology as a robust reaction for the
synthesis of complex molecules.
Our aim was to utilize our three-component coupling
reaction to install the C12–C22 aliphatic chain. Thus, dis-
connection across the C11–12 and C13–14 bonds would lead
to aldehyde 2, b-silyl vinyl borane 1, and carbamate 3
(Scheme 3). It was envisaged that aldehyde 2 could be
obtained through 4, which itself could be accessed from 5
through epoxidation and Taber cyclopropanation.[14]
Synthesis of aldehyde 2 was initiated with indium-
mediated coupling of allyl bromide and propargylic alcohol
6 (Scheme 4).[15] Subsequent Sharpless epoxidation[16] gave
[*] A. Robinson, Prof. Dr. V. K. Aggarwal
School of Chemistry, University of Bristol
Cantock’s Close, Bristol, BS81TS (UK)
Fax: (+44)117 925 1295
E-mail: v.aggarwal@bristol.ac.uk
[**] We thank the EPSRC for support of this work. V.K.A. thanks the Royal
Society for a Wolfson Research Merit Award and the EPSRC for a
Senior Research Fellowship.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201003236.
Angew. Chem. Int. Ed. 2010, 49, 6673 –6675
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6673

Communications
products but Raney–nickel with sodium hypophosphite[18]
gave 2 as a single product in high yield. Although aldehyde
2 had previously been synthesized, our route represents a
significant decrease in the number of steps (10 steps vs.
13 steps[5,19] or 14 steps[6]), with high stereocontrol and only a
single protecting-group manipulation.
In accordance with our synthetic method, our initial plan
was to treat carbamate 3 (synthesized by carbamoylation of
the commercially available alcohol) directly with sBuLi in the
presence of OꢀBrienꢀs (+)-sparteine surrogate[20] to give the
desired enantioenriched lithiated carbamate. However, treat-
ment of 3 with sBuLi and TMEDA resulted predominantly in
elimination of the carbamate (allylic deprotonation) rather
than deprotonation adjacent to the carbamate moiety.
Although carbamates bearing unsaturation at more remote
positions have been successfully employed, there is clearly a
limitation in the use of homoallylic carbamates. This discov-
ery could have represented a significant stumbling block for
the scope of our method. However, we were able to eliminate
the problem by making use of a stannane as a latent lithiated
species.[21] Thus, tin–lithium exchange of 14 was expected to
furnish the same lithiated carbamate required for the key
three-component coupling reaction.
Scheme 3. Retrosynthetic route to solandelactone E.
Stannane 14 was prepared as shown in Scheme 5. After
the conversion of the commercially available alcohol 11 into
carbamate 12, substrate-directed lithiation and trapping with
Bu3SnCl, as described by Hoppe and co-workers,[22] furnished
13 as an inseparable mixture with d.r. 10:1. Installation of the
required aliphatic chain was achieved by a sequence of
removal of the acetal protecting group, oxidative cleavage,
and Wittig olefination to give stannane 14.
With the three components for our coupling reaction in
hand, we subjected them to our standard lithiation–boryla-
tion–allylboration procedure.[12] However, 15 was only
obtained in 15% yield. The major impurity was formed by
attack of the lithiated carbamate on aldehyde 2, thus
indicating that the reaction with the borane was incomplete.
Visual examination of the reaction mixture following lithia-
tion showed it to be highly viscous at low temperature and so
we suspected that the problem lay in incomplete mixing with
the borane, thus resulting in competing addition to the
aldehyde. We were able to solve this problem by adding
TMEDA after nBuLi. This led to a much more mobile
reaction mixture at low temperature and after the addition of
borane 1 and aldehyde 2, gave 15 as a separable mixture with
Scheme 4. Synthesis of aldehyde 2. Reagents and conditions: a) allyl
bromide, In, THF, RT, 77%; b) (À)-DET, Ti(OiPr)4, tBuOOH, CH2Cl2,
À208C, 75%, e.r. 97:3; c) PhSO2Cl, Et3N, DMAP, CH2Cl2, RT, 90%;
d) NaHMDS, CH3CN, THF, À788C–08C, 91% d.r. 5.6:1 (separable);
e) DHP, PPTS, CH2Cl2, RT, 94%; f) K2OsO4·2H2O, NaIO4, 2,6-lutidine,
H2O/dioxane (1:3, RT; g) HO2C(CH2)3PPh3Br, NaHMDS, THF, À788C–
RT, Z/E >9:1; h) HCl (aq), RT, 59% over three steps; i) 1. 1,3,5-
trichlorobenzoyl chloride, Et3N, THF, RT; 2. DMAP, toluene, 908C–
608C, 79%; j) Raney–Ni, NaH2PO2·H2O, pyridine/H2O/AcOH (2:1:1),
RT, 87%. DET=diethyl tartrate, DHP=3,4-dihydro-2H-pyran,
DMAP=4-dimethylaminopyridine, HMDS=bis(trimethylsilyl)amide,
PPTS=pyridinium p-toluenesulfonate, THF=tetrahydrofuran,
THP=tetrahydropyranyl.
epoxide 6 in excellent enantiomeric ratio (e.r.). Notably, the
configuration of the epoxide is utilized to establish the three
stereogenic centers in aldehyde 2. Thus, after activation with a
benzenesulfonyl group, treatment of epoxide
7 with
NaHMDS and two equivalents of acetonitrile, through a
modification of Taberꢀs method,[14] gave cyclopropane 4 as a
separable mixture of diastereomers (5.6:1) in 91% yield.
Following protection of the alcohol as its THP ether 8, the
alkene was subjected to oxidative cleavage, Wittig olefina-
tion, and removal of the THP group. Yamaguchi lactonization
of acid 9 then gave the eight-membered lactone 10.[6] This
sequence was considerably more rewarding than esterifica-
tion of 4 with 5-pentenoic acid and subsequent ring-closing
metathesis.[17] Finally, we needed to convert the nitrile into an
aldehyde in the presence of an olefin and a lactone—a
procedure that initially presented some challenges. Attempts
using diisobutylaluminum hydride led to a mixture of
Scheme 5. Route to chiral stannane carbamate 14. Reagents and
conditions: a) N,N-diisopropyl carbamoyl chloride, NaH, THF, reflux,
85%; b) sBuLi, Et2O; then Bu3SnCl, À788C, 42% (84% brsm),
d.r. 10:1 (inseparable); c) 1m HCl (aq), MeOH, RT; d) NaHCO3,
NaIO4, RT; e) C6H13PPh3Br, NaHMDS, THF, À788C–RT, 80% over
three steps. brsm=based on recovered starting material.
6674 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 6673 –6675

Angewandte
Chemie
[5] a) J. D. White, W. H. C. Martin, C. Lincoln, J. Yang, Org. Lett.
d.r. 10:1 in 73% yield. As the lithiated carbamate had an e.r.
of 10:1, the diastereoselectivity observed shows that the
sequence of lithiation–borylation–allylation sequence oc-
curred with complete retention of configuration in the
borylation step (as a consequence of using an unhindered
diamine complexed to the lithiated carbamate[12a]) and
complete control of the configuration in the allylboration
step.[23]
The penultimate step of the synthesis required selective
epoxidation of the allyl silane without oxidation of the two
other olefin groups present. This was achieved by utilizing the
directing effect of the homoallylic alcohol with Ti(OiPr)4/
tBuOOH[24] which, after acid-catalyzed rearrangement, gave
solandelactone E as a single diastereoisomer in 68% yield
over two steps (Scheme 6). The spectroscopic data matched
exactly with those of natural and synthetic solandelacto-
ne E.[1,4–6]
2007, 9, 3481 – 3483; b) J. D. White, C. Lincoln, J. Yang, W. H. C.
Martin, D. B. Chan, J. Org. Chem. 2008, 73, 4139 – 4150.
[6] J. Pietruszka, A. C. Rieche, Adv. Synth. Catal. 2008, 350, 1407 –
1412.
[7] J. White, M. S. Jensen, J. Am. Chem. Soc. 1995, 117, 6224 – 6233.
[8] G. Kumaraswamy, M. Padmaja, J. Org. Chem. 2008, 73, 5198 –
5201.
[9] D. J. Critcher, S. Connolly, M. Wills, J. Org. Chem. 1997, 62,
6638 – 6657.
[10] T. Takahashi, H. Watanabe, T. Kitahara, Heterocycles 2002, 58,
99 – 104.
[11] C.-Y. Zhu, X.-Y. Cao, B.-H. Zhu, C. Deng, X.-L. Sun, B.-Q.
Wang, Q. Shen, Y. Tang, Chem. Eur. J. 2009, 15, 11465 – 11468.
[12] a) M. Binanzer, G. Y. Fang, V. K. Aggarwal, Angew. Chem. 2010,
122, 5268 – 5271; Angew. Chem. Int. Ed. 2010, 49, 4264 – 4268, for
other related work on lithiated carbamates, see: b) J. L. Stym-
iest, G. Dutheuil, A. Mahmood, V. K. Aggarwal, Angew. Chem.
2007, 119, 7635 – 7638; Angew. Chem. Int. Ed. 2007, 46, 7491 –
7494; c) G. Dutheuil, M. P. Webster, P. A. Worthington, V. K.
Aggarwal, Angew. Chem. 2009, 121, 6435 – 6437; Angew. Chem.
Int. Ed. 2009, 48, 6317 – 6319; d) M. Althaus, A. Mahmood, J.
Ramꢁn Suꢂrez, S. P. Thomas, V. K. Aggarwal, J. Am. Chem. Soc.
2010, 132, 4025 – 4028; for a review on carbonyl allylborations,
see: e) H. Lachance, D. G. Hall in Organic Reactions, Vol. 73
(Ed.: S. E. Denmark), Wiley, New York, 2008.
[13] W. R. Roush, P. T. Grover, Tetrahedron Lett. 1990, 31, 7567 –
7570.
[14] D. F. Taber, G. Bui, B. Chen, J. Org. Chem. 2001, 66, 3423 – 3426.
Taber used nBuLi as the base but we found that NaHMDS gave
higher yields.
[15] B. C. Ranu, A. Majee, Chem. Commun. 1997, 1225 – 1226.
[16] For Sharpless epoxidation of this substrate, see: a) Y. E. Raifeld,
A. A. Nikitenko, B. M. Arshava, I. E. Mikerin, L. L. Zilberg,
G. Y. Vid, S. A. Lang, Jr., V. J. Lee, Tetrahedron 1994, 50, 8613 –
8616; b) X. Ginesta, M. A. Pericas, A. Riera, Tetrahedron Lett.
2002, 43, 779 – 782; for seminal work by Sharpless on catalytic
asymmetric epoxidation, see: c) Y. Gao, J. M. Klunder, R. M.
Hanson, H. Masamune, S. Y. Ko, K. B. Sharpless, J. Am. Chem.
Soc. 1987, 109, 5765 – 5780; d) H. C. Kolb, M. S. Van Nieu-
wenhze, K. B. Sharpless, Chem. Rev. 1994, 94, 2483 – 2547.
[17] Problems with this strategy in the construction of eight-
membered lactones have been encountered before, see refer-
ences [5] and [6].
Scheme 6. Completion of solandelactone E. Reagents and conditions:
a) nBuLi, TMEDA, Et2O; then borane 1; then aldehyde 2, 73%,
d.r. 10:1 (separable); b) tBuOOH, Ti(OiPr)4, M.S. (4 ꢀ), CH2Cl2;
AcOH, MeOH, 68% over two steps. M.S. =molecular sieves, TME-
DA=N,N,N’,N’-tetramethylethylenediamine.
In conclusion, we have completed a short (13 steps,
longest linear sequence) and highly selective synthesis of
solandelactone E. The key steps included Sharpless epoxida-
tion, Taber cyclopropanation, chemoselective reduction of a
nitrile to an aldehyde, and our own lithiation–borylation–
allylation sequence on a highly functionalized substrate. Not
only does the completed synthesis establish the method for
application to complex targets, it also solves the problem of
poor stereocontrol at C11 that had dogged many previous
syntheses of this class of molecules.
[18] O. G. Backeberg, B. Staskun, J. Chem. Soc. 1962, 3961 – 3963.
[19] D. K. Mohapatra, G. S. Yellol, ARKIVOC 2003, 9, 21 – 33.
[20] M. J. Dearden, C. R. Firkin, J.-P. R. Hermet, P. OꢀBrien, J. Am.
Chem. Soc. 2002, 124, 11870 – 11871.
[21] a) W. C. Still, C. Sreekumar, J. Am. Chem. Soc. 1980, 102, 1201 –
1202; b) D. Hoppe, F. Marr, M. Brꢃggemann in Organolithiums
in Enantioselective Synthesis Topics Organomet. Chem., Vol. 5
(Ed.: D. M. Hodgson), Springer, Heidelberg, 2003, pp. 61 – 138.
[22] a) H. Helmke, D. Hoppe, Synlett 1995, 978 – 980; for a review of
Hoppeꢀs seminal contributions, see: b) D. Hoppe, T. Hense,
Angew. Chem. 1997, 109, 2473; Angew. Chem. Int. Ed. Engl.
1997, 36, 2376 – 2410; D. Hoppe, T. Hense, Angew. Chem. 1997,
109, 2376 – 2410; Angew. Chem. Int. Ed. Engl. 1997, 36, 2282 –
2316; c) E. Beckmann, V. Desai, D. Hoppe, Synlett 2004, 2275 –
2280.
[23] Nucleophilic additions to related trans cyclopropanes usually
occur without stereoinduction, see reference [9]. In our case, the
very high stereochemical control observed is dominated by the
stereochemistry of the allylborating agent and by the influence
of the 9-BBN moiety. 9-BBN = 9-borabicyclo[3.3.1]nonane.
[24] Y. Landais, L. Parra-Rapado, Tetrahedron Lett. 1996, 37, 1205 –
1208.
Received: May 28, 2010
Published online: August 3, 2010
Keywords: allyl silanes · 1,4-ene-diols · lithiation–borylation ·
.
natural products · total synthesis
[1] Y. Seo, K. W. Cho, J-R. Rho, J. Shin, Tetrahedron 1996, 52,
10583 – 10596.
[2] H. Niwa, K. Wakamatsu, K. Yamada, Tetrahedron Lett. 1989, 30,
4543.
[3] D. G. Nagle, W. H. Gerwick, Tetrahedron Lett. 1990, 31, 2995.
[4] a) J. E. Davoren, S. F. Martin, J. Am. Chem. Soc. 2007, 129, 510 –
511; b) J. E. Davoren, C. Harcken, S. F. Martin, J. Org. Chem.
2008, 73, 391 – 402.
Angew. Chem. Int. Ed. 2010, 49, 6673 –6675
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6675