Stereo- and regioselective synthesis of squalene tetraepoxide

Article abstract of DOI:10.1021/jo901920d

(Chemical Equation Presented) Squalene tetraepoxide, a putative biosynthetic precursor to a variety of oxacyclic triterpenoid natural products, has been efficiently synthesized by anionic coupling of two farnesol-derived diepoxides, which have arisen from

Full text of DOI:10.1021/jo901920d

pubs.acs.org/joc
recent development of enantio- and regioselective epoxida-
tion methods3 indicated that the efficient and stereoselective
synthesis of squalene tetraepoxide 1 could now be possible.
Stereo- and Regioselective Synthesis of Squalene
Tetraepoxide
Rongbiao Tong, Matthew A. Boone, and
Frank E. McDonald*
Department of Chemistry, Emory University, Atlanta,
Georgia 30322
fmcdona@emory.edu
Received September 4, 2009
Squalene tetraepoxide, a putative biosynthetic precursor
to a variety of oxacyclic triterpenoid natural products,
has been efficiently synthesized by anionic coupling of
two farnesol-derived diepoxides, which have arisen from
electronic control of the regioselectivity in organocataly-
tic enantioselective epoxidations.
FIGURE 1. Squalene tetraepoxide and representative polycyclic
ether natural products.
Our strategy was based on the regio- and enantioselective
epoxidation of farnesol derivatives. We anticipated that the
Shi enantioselective epoxidation of the alkene nearest to the
heteroatom substituent would be disfavored by the presence
of an electron-withdrawing allylic substituent.4,5 Farnesyl
acetate 6a gave reasonable but variable yields of the
6,7:10,11-diepoxide 8a (Table 1),6 and this product was
always accompanied by substantial quantities of the triep-
oxide. Farnesyl benzoate 6b was less reactive (probably due
to poor solubility in the polar solvent mixture), but the
electron-withdrawing and polar nitrobenzoate ester in 6c
gave superior results (entry 3),7 providing primarily the
diepoxide 8c along with mixtures of the monoepoxides,
which could be converted into additional quantities of 8c
by another cycle of the Shi epoxidation. The toluenesulfonyl
substituent8 in substrate 7 (entry 4) completely suppressed
epoxidation at the 2,3-alkene, to afford diepoxide 9 as the
major product.9,10
An isomer of squalene tetraepoxide (1, Figure 1) has been
proposed to be the biogenetic precursor to several oxacyclic
triterpenoid natural products, some of which have interest-
ing biological activities.1 Representative members of this
family of natural products, such as abudinols A and B
(2 and 3), are isomeric to 1 with the formula C30H50O4.
Other closely related natural products including raspacionin
(4) and sodwanone M (5) all display highly compact and
complex molecular architectures, and may also arise from
squalene tetraepoxide (1). We have been intrigued by the
structural diversity of these compounds, and thus became
interested in developing a synthetic route to squalene tetra-
epoxide as the all-R-enantiomer (ent-1), to test the biomi-
metic cyclization behavior and potentially gain access to
these complex natural products.
(4) (a) Regioselectivity of the Shi epoxidation of geraniol: Shi, Z.-X.; Shi,
Y. J. Org. Chem. 1998, 63, 3099. (b) Shi epoxidation of farnesol: McDonald,
F. E.; Bravo, F.; Wang, X.; Wei, X.; Toganoh, M.; Rodrıguez, J. R.; Do, B.;
Neiwert, W. A.; Hardcastle, K. I. J. Org. Chem. 2002, 67, 2515.
(5) The Shi epoxidation of geranyl acetate is apparently unknown. For
regioselective (racemic) epoxidations of geranyl acetate, see: (a) Dodd, D. S.;
Oehlschlager, A. C.; Georgopapadakou, N. H.; Polak, A.-M.; Hartman,
P. G. J. Org. Chem. 1992, 57, 7226. (b) Grocock, E. L.; Marples, B. A.; Toon,
R. C. Tetrahedron 2000, 56, 989.
(6) Tong, R.; McDonald, F. E. Angew. Chem., Int. Ed. 2008, 47, 4377.
(7) Regio- and enantioselective Shi monoepoxidation of geranyl p-nitro-
benzoate: Neighbors, J. D.; Mente, N. R.; Boss, K. D.; Zehnder, D. W.;
Wiemer, D. F. Tetrahedron Lett. 2008, 49, 516.
A tetraepoxide of squalene has been claimed from the
epoxidation of squalene with 4 equiv of perbenzoic acid.2
The level of characterization available in that era did not
conclusively permit positional assignment of the epoxides.
Moreover, perbenzoic acid epoxidation would undoubtably
have provided a mixture of stereoisomers. However, the
ꢀ
(1) (a) Fernandez, J.; Souto, M.; Norte, M. Nat. Prod. Rep. 2000, 17, 235.
(8) Tong, R.; Valentine, J. C.; McDonald, F. E.; Cao, R.; Fang, X.;
Hardcastle, K. I. J. Am. Chem. Soc. 2007, 129, 1050.
(9) Regioselective (racemic) monoepoxidation of geranyl p-tolylsulfone:
Eren, D.; Keinan, E. J. Am. Chem. Soc. 1988, 110, 4356.
(b) Kashman, Y.; Rudi, A. Phytochem. Rev. 2004, 3, 309. (c) Domingo, V.;
Arteaga, J. F.; Quılez del Moral, J. F.; Barrero, A. F. Nat. Prod. Rep. 2009,
26, 115.
(2) van Duuren, B. L.; Schmitt, F. L. J. Org. Chem. 1960, 25, 1761.
(3) (a) Wang, Z.-X.; Tu, Y.; Frohn, M.; Zhang, J.-R.; Shi, Y. J. Am.
Chem. Soc. 1997, 119, 11224. (b) Wong, O. A.; Shi, Y. Chem. Rev. 2008, 108,
3958.
(10) Although epoxidations catalyzed by D-epoxone have provided the
all-(R)-epoxides 8 and 9 resulting in all-(R)-squalene tetraepoxide ent-1, the
(S)-antipodes of 8 and/or 9 can be prepared by using L-epoxone arising from
L-fructose: Zhao, M.-X.; Shi, Y. J. Org. Chem. 2006, 71, 5377.
DOI: 10.1021/jo901920d
r
Published on Web 10/14/2009
J. Org. Chem. 2009, 74, 8407–8409 8407
2009 American Chemical Society

Tong et al.
JOCNote
TABLE 1. Regioselectivity of Shi Epoxidations of Farnesyl Esters
Experimental Section
Farnesyl-p-tolyl Sulfone 7.8 To a solution of trans,trans-
farnesol (10.0 g, 45 mmol) in dry THF (0.22 M, 200 mL) was
added triphenylphosphine (PPh3, 14.7 g, 56 mmol) at 0 °C.
N-Bromosuccinimide (NBS, 9.23 g, 51.6 mmol) was then slowly
added in ten batches over 20 min. The light yellow reaction
mixture was stirred for 1.5 h at 0 °C until complete conversion
was achieved as monitored by TLC. Then, tetrabutylammo-
nium iodide (Bu4NI, 1.70 g, 4.5 mmol) and p-toluenesulfinic
acid sodium salt (NaSO2Tol, 12 g, 68 mmol) were subsequently
added. The light yellow suspension was warmed to room
temperature and stirred for 16 h. During this time, the reaction
became light brown in color. The reaction was quenched with
saturated NaHSO3 (200 mL). The layers were separated and the
organic layer was collected. The aqueous layer was extracted
with Et2O (2 ꢀ 100 mL). The combined organic fractions were
washed with saturated NaHCO3 (100 mL) and brine (100 mL),
then dried with anhydrous MgSO4. After filtration, the com-
bined solvents were removed under reduced pressure. Chroma-
tography (9:1 hexanes:EtOAc) gave 1-farnesyl p-tolyl sulfone 7
(10.8 g, 67%).
monoepox-
entry substrate ide yield %b
diepoxide
yield (%)
triepoxide
yield (%)
1
2
3
4
6a
6b
6c
7
trace
61
28
35-55
26
40
not observed
trace
not observed
49-61c
71
trace
aReaction conditions: D-epoxone (0.5 equiv), Oxone (2.8 equiv),
K2CO3 (8.0 equiv), Bu4NHSO4 (0.1 equiv), Na2EDTA, Na2B4O7,
DMM:MeCN (2:1), H2O,
10,11-monoepoxides. Isolated yields of 8c from one cycle of epoxida-
tion of 6c.
0
°C. bCombined yield of 6,7- and
c
The diepoxide 8c was converted into the known diepoxy-
bromide 10.6 The 30-carbon chain of squalene tetraepoxide
was then constructed by addition of KO-t-Bu at low tem-
perature to a THF solution of the two 15-carbon synthons
diepoxysulfone 9 and diepoxybromide 10 (1.5 equiv). Palla-
dium-catalyzed reductive desulfonylation11 of the resul-
ting intermediate 11 provided squalene tetraepoxide ent-1
(Scheme 1).
p-Nitrobenzoyl Diepoxide 8c. trans,trans-Farnesyl p-nitro-
benzoate 6c (20 g, 54 mmol)13 was transferred into a three-
necked 3.0 L flask. Then DMM:MeCN (2:1, 0.10 M, 500 mL)
and Na2B4O7 (0.05 M solution in 4ꢀ10-4 M Na2EDTA, 0.15 M,
350 mL) were added, followed by the addition of Bu4NHSO4
(1.8 g, 5.4 mmol). D-Epoxone (7.0 g, 27 mmol) was added. The
flask was equipped with a mechanical stirrer and two addition
funnels. To one addition funnel was added Oxone (140 g,
220 mmol) dissolved in 4 ꢀ 10-4 M Na2EDTA (400 mL). To
the other addition funnel was added K2CO3 (112 g, 810 mmol)
dissolved in distilled H2O (400 mL). The flask was cooled to 0 °C
and the Oxone and K2CO3 solutions were added simultaneously
dropwise over a 1.25 h period. After the additions were com-
plete, EtOAc (500 mL) was added to the reaction and trans-
ferred to a 3.0 L separatory funnel. After the organic layer was
collected, the aqueous was extracted with EtOAc (750 mL). The
combined organic fractions were dried with MgSO4. After
filtration, the volatiles were removed under reduced pressure.
Chromatography on silica gel (4:1 f 2:1 hexanes:EtOAc) pro-
vided diepoxide 8c (dr=4: 1) as a pale yellow oil (10.6 g, 49%),
along with the monoepoxide (mixture of the 6,7- and 10,11-
epoxides) (5.93 g, 28%). Additional amounts of diepoxide could
be obtained by subjecting the monoepoxide to the same reaction
conditions, using only 2.0 equiv of Oxone and 8.0 equiv of
SCHEME 1. Anionic Coupling of 9 and 10, and Conversion to
Squalene Tetraepoxide ent-1a
K2CO3. [R]23 þ8.8 (c 1.40, CHCl3); 1H NMR (400 MHz,
D
CDCl3) δ 8.29 (d, J = 8.8 Hz, 2H), 8.22 (d, J = 8.8 Hz, 2H),
5.52 (t, J = 7.2 Hz, 1H), 4.90 (d, J = 6.8 Hz, 2H), 2.75 (t, J =
6.0 Hz, 1H), 2.71 (m, 1H), 2.24 (m, 2H), 1.81 (s, 3H), 1.79-1.56
(m, 7H), 1.31 (s, 3H), 1.29 (s, 3H), 1.27 (s, 3H); 13C NMR
(100 MHz, CDCl3) δ 164.9, 150.7, 142.5, 135.9, 130.9 (2ꢀ), 123.7
(2ꢀ), 118.5, 63.9, 62.8 (2ꢀ), 60.5, 58.6, 36.4, 35.4, 27.1, 25.0,
24.7, 18.4, 16.9, 16.8; IR (KBr) 2962, 1724, 1606, 1529, 1456,
1381, 1348, 1271, 1101, 1014, 874, 721 cm-1; HRMS (ESI) [Mþ
Hþ] calcd for C22H29N1O6 404.20676, found 404.20717.
aReagents and conditions: (a) K2CO3, MeOH, 15 min (88% yield);
(b) MsCl, Et3N, THF, -40 °C, 30 min; then LiBr, THF, 0 °C, 15 min
(96% yield); (c) KO-t-Bu, THF, -78 °C, 2 h (77% yield); (d) PdCl2(dppp)
(20 mol %), LiBEt3H, THF, 0 °C, 40 min (64% yield).
Although preliminary studies on the Lewis acid-promoted
cyclizations of ent-1 have not yet proven productive, we
believe that our synthesis of tetraepoxide 1 will have value in
preparing substrates for biosynthetic feeding experiments.
Our synthetic route allows for incorporation of stable iso-
topes, for instance by double 13C-labeling in farnesol.12 In
addition, the unsymmetrical nature of our synthetic ap-
proach provides considerable flexibility in the nature of
polyepoxide synthons for preparation of epoxide diastereo-
mers of 1, and access to several other patterns of squalene
polyepoxides.
Sulfonyl Diepoxide 9. 1-Farnesyl-p-tolyl sulfone 7 (3.6 g, 10 mmol)
was transferred to a three-necked 1.0 L flask, to which was added
DMM:MeCN (2:1, 0.067 M, 150 mL), Na2B4O7 (0.05 M solution
in 4 ꢀ 10-4 M Na2EDTA, 0.091 M, 110 mL), Bu4NHSO4 (0.34 g,
1.0 mmol), and D-epoxone (1.3 g, 5.0 mmol) sequentially. The
solution was cooled to 0 °C and vigorously stirred. The flask was
equipped with two addition funnels. To one addition funnel was
added Oxone (17 g, 28 mmol) dissolved in 4 ꢀ 10-4 M Na2EDTA
(140 mL). To the other addition funnel was added K2CO3 (15 g, 110
mmol) dissolved in distilled H2O (140 mL). The Oxone and K2CO3
(11) Mohri, M.; Kinoshita, H.; Inomata, K.; Kotake, H. Chem. Lett.
1985, 451.
(12) Graziani, E. I.; Andersen, R. J. J. Am. Chem. Soc. 1996, 118, 4701.
€
(13) Gohler, S.; Stark, C. B. W. Org. Biomol. Chem. 2007, 5, 1605.
8408 J. Org. Chem. Vol. 74, No. 21, 2009

Tong et al.
JOCNote
solutions were added dropwise over a 2 h period. Upon completion
of the additions, the reaction was allowed to stir for an additional
20 min, at which time H2O(100mL) andEt2O (200 mL) were added.
The layers were separated. The aqueous layer was extracted with
Et2O (2 ꢀ 100 mL). The organic extracts were dried with MgSO4.
After filtration, the volatiles were removed under reduced pressure.
Chromatography on silica gel (2:1 f 1:1 hexanes:EtOAc) gave the
4.02 (d, J = 7.8 Hz, 2H), 2.73 (m, 2H), 2.24 (m, 1H), 2.18
(m, 1H), 1.76 (s, 3H), 1.68 (m, 3H), 1.61 (m, 3H), 1.32 (s, 3H),
1.29 (s, 3H), 1.28 (s, 3H); 13C NMR (150 MHz, CDCl3) δ 142.6,
121.3, 63.9, 62.7, 60.5, 58.7, 36.4, 35.3, 29.4, 26.9, 25.0, 24.7,
18.8, 16.9, 16.1; IR (KBr) 2962, 1655, 1454, 1381, 1203, 1122,
876 cm-1; HRMS (APCI) [M þ Hþ] calcd for C15H26O2Br1
317.11107, found 317.11115.
diepoxy allylic sulfone 9(dr=5: 1) as a yellow oil (2.8 g, 71%). [R]23
Tetraepoxy Allylic Sulfone 11. The diepoxy allylic bromide 10
(1.8 g, 5.7 mmol) and diepoxy allylic sulfone 9 (1.6 g, 4.0 mmol)
were dissolved in THF (0.05 M, 81 mL) and cooled to -78 °C.
Then KO-t-Bu (1.0 M solution in THF, 5.3 mL, 5.3 mmol) was
added to the solution via syringe pump over a 30 min period. The
reaction mixture was stirred for 2 h at -78 °C. Then saturated
NaHCO3 (200 mL) was added to quench the reaction. The
organic layer was collected and the aqueous layer was extracted
with Et2O (200 mL). The organic extracts were combined and
dried with MgSO4. After filtration, the volatiles were removed
under reduced pressure. Chromatography on silica gel (9:1 f
1.5:1 hexanes:EtOAc þ 0.5% Et3N) gave the tetraepoxy allylic
sulfone 11 as an oil (1.96 g, 77%). [R]23D þ13.8 (c 0.745, CHCl3);
1H NMR (400 MHz, CDCl3) δ 7.71 (d, J=7.6 Hz, 2H), 7.30 (d,
J=7.3 Hz, 2H), 5.01 (m, 2H), 3.73 (m, 1H), 2.69 (m, 4H), 2.44 (s,
3H), 2.40-2.24 (m, 2H), 2.20-2.00 (m, 4H), 1.80-1.50 (m,
12H), 1.62 (s, 6H), 1.31 (s, 3H), 1.30 (s, 3H), 1.28 (s, 3H), 1.27 (s,
3H), 1.26 (s, 3H), 1.35 (s, 3H); 13C NMR (100 MHz, CDCl3) δ
144.6 (2ꢀ), 137.8, 135.3, 129.6 (2ꢀ), 129.3 (2ꢀ), 119.5, 117.8,
64.9, 64.2, 64.0 (2ꢀ), 62.9, 62.7, 60.5, 58.6, 36.5, 35.8, 35.4, 35.3,
29.9, 27.5, 27.4, 27.0, 25.0 (3ꢀ), 24.8 (2ꢀ), 21.8, 18.9, 16.8 (2ꢀ),
16.6; IR (KBr) 2960, 2926, 2856, 1597, 1456, 1381, 1300, 1144,
1059, 1250, 874 cm-1; HRMS (APCI) [M þ Hþ] calcd for
C37H57O6S1 629.38704, found 629.38761.
(3R,6R,7R,18R,19R,22R)-Squalene Tetraepoxide (ent-1). To
a solution of tetraepoxy sulfone 11 (1.96 g, 3.1 mmol) in THF
(0.10 M, 31 mL) was added PdCl2(dppp) (370 mg, 0.62 mmol) at
0 °C. Lithium triethylborohydride (LiBEt3H, 1.0 M solution in
THF, 6.2 mL, 6.2 mmol) was then added dropwise to the
solution over a 15 min period. The reaction mixture was stirred
for an additional 40 min at 0 °C and then diluted with Et2O
(40 mL), followed by the addition of saturated NH4Cl (50 mL).
The organic layer was collected and the aqueous layer was
extracted with Et2O (50 mL). The combined organic fractions
were dried with MgSO4. After filtration, the volatiles were
removed under reduced pressure. Chromatography on silica
gel (9:1 f 1:1 hexanes:EtOAc þ 0.5% Et3N) gave squalene
tetraepoxide (ent-1) as a clear oil (944 mg, 64%) and recovered
11 (218 mg). [R]23D þ15.1 (c 0.81, CHCl3); 1H NMR (400 MHz,
CDCl3) δ 5.18 (br s, 2H), 2.72 (m, 4H), 2.16-2.08 (m, 4H), 2.02
(t, J = 2.8 Hz, 4H), 1.78 (m, 2H), 1.70-1.52 (m, 8H), 1.62 (s,
6H), 1.32 (s, 6H), 1.28 (s, 6H), 1.27 (s, 6H); 13C NMR (100 MHz,
CDCl3) δ 134.5, 125.0, 64.1, 63.2, 60.5, 58.6, 36.5, 35.5, 28.4,
27.5, 25.1, 24.9, 18.9, 16.9, 16.3; IR (KBr) 2960, 2926, 2858,
1452, 1379, 1323, 1250, 1120, 874 cm-1; HRMS (ESI) [MþHþ]
calcd for C30H51O4 475.37819, found 475.37829.
D
2.80 (c 1.01, CHCl3); 1H NMR (400 MHz, CDCl3) δ 7.72 (d, J =
8.0Hz,2H),7.31(d,J=8.4 Hz, 2H), 5.21 (t, J=8.0Hz,1H),3.78(d,
J=8.0 Hz, 2H), 2.69 (m, 2H), 2.43 (s, 3H), 2.14 (m, 2H), 1.78-1.50
(m, 6H), 1.38 (s, 3H), 1.30 (s, 3H), 1.27 (s, 3H), 1.26 (s, 3H); 13
C
NMR (100 MHz, CDCl3) δ 145.5, 144.7, 135.9, 129.8 (2ꢀ), 128.6
(2ꢀ), 111.2, 64.0, 62.7, 60.5, 58.6, 56.2, 36.5, 35.3, 27.1, 25.0, 24.7,
21.8, 18.8, 16.8, 16.4; IR (KBr) cm-1 2962, 2926, 1664, 1597, 1452,
1383, 1313, 1149, 1088, 744; HRMS (ESI) [M þ Hþ] calcd for
C22H33O4S1 393.20941, found 393.20941.
Diepoxy Allylic Bromide 10. To a solution of p-nitrobenzoate
diepoxide 8c (23 g, 57 mmol) dissolved in MeOH (0.50 M, 115 mL)
was added K2CO3 (3.9 g, 29 mmol) all at once. The reaction was
stirred for 15 min. After dilution with Et2O (100 mL), the reaction
was quenched by the addition of a saturated solution of NH4Cl
(250 mL). The organic layer was collected and the aqueous layer
was extracted with EtOAc (2ꢀ250 mL). The organic extracts were
combined and dried with MgSO4. After filtration, the volatiles
were removed under reduced pressure. Prior to chromatography
on silica gel, the mixture of diepoxy allylic alcohol product
containing the poorly soluble byproduct methyl p-nitrobenzoate
was dissolved in minimal EtOAc for loading onto the chromatog-
raphy column. Methyl p-nitrobenzoate eluted from the column
with 4:1 hexanes:EtOAc, and then flushing with 100% EtOAc
provided the polar diepoxy allylic alcohol as an oil (12.8 g, 88%).
[R]23D þ11.0 (c 0.965, CHCl3); 1H NMR (600 MHz, CDCl3) δ5.46
(m, 1H), 4.16 (d, J=6.6 Hz, 2H), 2.76-2.71 (m, 2H), 2.21 (m, 1H),
2.16 (m, 1H), 1.79 (m, 1H), 1.70 (s, 3H), 1.68 (m, 3H), 1.60 (m, 3H),
1.60 (m, 3H), 1.31 (s, 3H), 1.28 (s, 3H), 1.27 (s, 3H); 13C NMR
(150 MHz, CDCl3) δ 138.5, 124.3, 64.1, 62.9, 60.5, 59.4, 58.7, 36.4,
35.3, 27.0, 24.9, 24.7, 18.8, 16.9, 16.4; IR (KBr) 3437, 2924, 1666,
1454, 1385, 1250, 1119, 1011, 872 cm-1; HRMS (APCI) [MþHþ]
calcd for C15H27O3 255.19547, found 255.19552.
This diepoxy allylic alcohol intermediate (12.8 g, 50 mmol)
dissolved in THF (0.30 M, 170 mL) was cooled to -40 °C. Et3N
(10.5 mL, 76 mmol) was then added to the solution all at once,
followed by addition of MsCl (4.71 mL, 60 mmol) all at once.
The reaction was stirred for 30 min at -40 °C. After warming
to 0 °C, flame-dried LiBr (13.1 g, 150 mmol) dissolved in THF
(5.0 M, 30 mL) was added all at once. The reaction mixture was
stirred for an additional 15 min before being diluted with Et2O
(200 mL) and quenched with H2O (200 mL). The organic layer
was collected and the aqueous layer was extracted with Et2O
(100 mL). The organic extracts were combined and dried with
MgSO4. After filtration, the volatiles were removed under
reduced pressure. To the crude mixture was added hexanes
(100 mL), and the solids were filtered. After removal of the
volatiles under reduced pressure, the analytically pure allylic
Acknowledgment. This research was supported by the
National Science Foundation (CHE-0516793).
bromide 10 (15.3 g, 96%) was obtained.14 [R]23 þ4.9 (c 0.85,
D
CHCl3); 1H NMR (600 MHz, CDCl3) δ 5.59 (t, J=8.4 Hz, 1H),
Supporting Information Available: 1H and 13C NMR spec-
tra of new compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
(14) We elected not to subject this sensitive allylic bromide to chroma-
tography, as significant decomposition occurred (even with Et3N buffering).
Once prepared, the allylic bromide 10 was immediately used.
J. Org. Chem. Vol. 74, No. 21, 2009 8409