Asymmetric synthesis of all eight seven-carbon dipropionate stereotetrads

Article abstract of DOI:10.1021/ja071217x

Enantiopure cycloheptadienyl sulfones 6 and 7 are diastereoselectively epoxidized to yield epoxyvinyl sulfones 8, 9, 14, and 16 in high yields and diastereomeric ratios. Syn and anti methylation of epoxides 8, 9, 14, and 16 enables access to all eight pos

Full text of DOI:10.1021/ja071217x

Published on Web 06/14/2007
Asymmetric Synthesis of All Eight Seven-Carbon
Dipropionate Stereotetrads
Ahmad El-Awa, Xavier Mollat du Jourdin, and Philip L. Fuchs*
Contribution from the Department of Chemistry, Purdue UniVersity,
West Lafayette, Indiana 47907
Received February 20, 2007; E-mail: pfuchs@purdue.edu
Abstract: Enantiopure cycloheptadienyl sulfones 6 and 7 are diastereoselectively epoxidized to yield
epoxyvinyl sulfones 8, 9, 14, and 16 in high yields and diastereomeric ratios. Syn and anti methylation of
epoxides 8, 9, 14, and 16 enables access to all eight possible diastereomeric stereotetrads, seven of which
are commonly found in polypropionate natural products. Anti methylations of the above epoxides are possible
by either the reaction of methyl organometallics promoted by copper(I), or via reaction with trimethylaluminum
to yield stereotetrads 11, 12, 22, and 24. Syn methylations are achieved via Lawton SN2′ reaction in the
case of stereotetrads 10, 15, and 38, while stereotetrad 13 is accessed by an oxidation/reduction alcohol
inversion sequence from stereotetrad 11. All stereotetrads were obtained in high diastereomeric ratios and
yields, and their relative stereochemistry was confirmed by X-ray crystallography. Oxidative cleavage of
the cyclic stereotetrads yields termini-differentiated acyclic heptanyl stereotetrads ready for use in building
larger fragments in the course of target syntheses.
Introduction
are ample examples of a family of natural products having
members that differ by alcohol or methyl stereochemistry, to
Polypropionates constitute a large family of natural products
that are biosynthesized by the condensation of two or more
propionic acid units.1
our knowledge, there are no examples where an entire ster-
eotetrad has been substituted. When one considers the steric
and electronic nature of the 1,3-dimethyl and 1,3-diol moieties
of polypropionates, it can be argued that these segments are
ideal conformational control elements, and their systematic
substitution might provide analogues that exhibit conformational
populations with improved binding to a target receptor. A
synthetic approach systematically exploring the structure-
activity relationships of a single stereotetrad would require 16
total syntheses if all enantiopure diastereomers were incor-
porated. Clearly, such an approach would rightly be disparaged
as a mindless ‘fishing expedition’. However, armed with
substrate-enzyme X-ray structural information combined with
in silico modeling, one should be empowered to select the most
optimal “unnatural” targets for synthesis and testing.
Many polypropionate natural products have been found to
possess medicinally relevant biological activities. Important
examples bearing the dipropionate stereotetrad include aplyro-
nine A (1),2 apoptolidine (2),3 discodermolide (3)4 (Figure 1),
erythronolide B,5 oleandolide,6 amphotericin B,7 and dictyosta-
tin.8 Of particular relevance to our research are the anti-actin
agent aplyronine A (1), the apoptosis inducer apoptolidine (2),
and the anti-tubulin compound discodermolide (3) (Figure 1).
All three compounds show high promise as anticancer agents,
with discodermolide (3) currently in clinical trials.9 While there
(1) (a) Davies-Coleman, M. T.; Garson, M. J. Nat. Prod. Rep. 1998, 15, 477.
(b) Koskinen, A. M. P.; Karisalmi, K. Chem. Soc. ReV. 2005, 34, 677.
(2) (a)Yamada, K.; Ojika, M.; Ishigaki, T.; Yoshida, Y.; Ekimoto, H.; Arakawa,
M. J. Am. Chem. Soc. 1993, 115, 11020. (b) Ojika, M.; Kigoshi, H.;
Ishigaki, T.; Tsukada, I.; Tsuboi, T.; Ogawa, T.; Yamada, K. J. Am. Chem.
Soc. 1994, 116, 7441.
(3) Daniel, P. T.; Koert, U.; Schuppan, J. Angew. Chem., Int. Ed. 2006, 45,
872 and references therein.
(4) (a)Kalesse, M. ChemBioChem 2000, 1, 171. (b) Paterson, I.; Florence, G.
J. Eur. J. Org. Chem. 2003, 2193 and references therein.
(5) (a) Corey, E. J.; Kim, S.; Yoo, S.; Nicolaou, K. C.; Melvin, L. S.; Brunelle,
D. J.; Falk, J. R.; Trybulski, E. J.; Lett, R.; Sheldrake, P. W. J. Am. Chem.
Soc. 1978, 100, 4620. (b) Oleinick, N. L. In Antibiotics; Corcoran, J. W.,
Hahn, F. E., Eds.; Springer-Verlag: New York, 1975; Vol III, pp 396-
419.
(6) (a) Sobin, B. A.; English, R. A.; Celmer, W. D Antibiot. Annu. 1955, 2,
827. (b) Hochstein, F. A.; Els, H.; Celmer, W. D.; Shapiro, B. L.;
Woodward, R. B. J. Am. Chem. Soc. 1960, 82, 3225. (c) Celmer, W. D. J.
Am. Chem. Soc. 1965, 87, 1797. (d) Ogura, H.; Furuhata, K.; Harada, Y.;
Iitaka, Y. J. Am. Chem. Soc. 1978, 100, 6733.
Before undertaking the (still considerable) effort of several
computationally inspired total syntheses, it was deemed prudent
to demonstrate our ability to deliver significant quantities of
all eight diastereomers potentially needed for the designer-SAR
study.
Only a few methods used for the construction of polypropi-
onate natural products are deemed of general utility, mostly
relying on asymmetric aldol10 and crotylation11 chemistry.
(10) For excellent reviews: (a) Mahrwald, R., Ed. Modern Aldol Reactions;
Wiley-VCH: New York, 2004. (b) Paterson, I.; Cowden, C. J.; Wallace,
D. J. In Modern Carbonyl Chemistry; Otera, J., Ed.; Wiley-VCH: New
York 2000; p 249.
(11) For key reviews: (a) Chemler, S. R.; Roush, W. R. In Modern Carbonyl
Chemistry; Otera, J., Ed.; Wiley-VCH: New York, 2000; p 403. (b)
Denmark, S. E.; Almstead, M. G. In Modern Carbonyl Chemistry; Otera,
J., Ed.; Wiley-VCH: New York, 2000; p 299.
(7) Vandeputte, J.; Wachtel, J. L.; Stiller, E. T. Antibiot. Annu. 1956, 587.
(8) (a) Pettit, G. R.; Cichacz, Z. A.; Gao, F.; Boyd, M. R.; Schmidt, J. M. J.
Chem. Soc., Chem. Commun. 1994, 9, 1111. (b) Paterson, I.; Britton, R.;
Delgado. O.; Wright, A. E. Chem. Commun. 2004, 6, 632.
(9) Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2004, 67, 1216.
9
9086
J. AM. CHEM. SOC. 2007, 129, 9086-9093
10.1021/ja071217x CCC: $37.00 © 2007 American Chemical Society

Asymmetric Synthesis of Dipropionate Stereotetrads
A R T I C L E S
Figure 1. Dipropionate unit in natural products.
Scheme 1
We envisioned a method for synthesis of the entire family
of diastereomeric stereotetrads based on our previous experience
with vinylsulfone chemistry.12 Pivotal to this strategy is the use
of environmentally friendly scalable synthetic methods. Hence,
we elected to rely on substrate control whenever possible, and
in cases where substrate control failed, asymmetric catalysts
rather than stoichiometric reagents and auxiliaries were adopted.13
Accordingly, the initial asymmetry was installed via catalytic
asymmetric Jacobsen epoxidation.14 All subsequent asymmetric
centers were established with achiral reagents relying on
substrate directed epoxidations, except epoxide 14 which was
accessible only by double stereoselective reagent control via a
second catalytic Jacobsen epoxidation. In seven of the eight
stereotetrads it was possible to develop direct methods for
stereoselectively installing the methyl group, while the eighth
target relied on epimerization. Details of construction of the
eight stereotetrads are discussed below.
Synthetic Plan
The synthetic plan begins with syn- and anti-dienyl sulfones
6 and 712c which are prepared from achiral dienyl sulfone 4,12a,15
available on the kilogram scale, wherein the first asymmetry is
introduced via Jacobsen catalytic asymmetric epoxidation
(Scheme 1).12a
Diastereoselective epoxidation of syn-dienyl sulfone 6 was
correctly anticipated to afford epoxide 8 or 9 as a function of
the reagent employed. Both 8 and 9 can be methylated from
the beta face to give stereotetrads 10 and 12 or from the alpha
face to give stereotetrads 11 and 13 respectively (Scheme 2).
Likewise, anti-dienyl sulfone 7 can be epoxidized from either
face followed by diastereoselective methylation of the resultant
epoxides to produce the remaining four stereotetrads.
Essential to this plan was the ability to control the diaster-
eoselectivity of each of the epoxidation and the methylation
events.
During a study targeting the total synthesis of aplyronine A
(1),16 we previously demonstrated the synthesis of diastereo-
meric stereotetrads 10 and 15 by 1,2-syn methylation of
epoxyvinyl sulfones 8 and 14 (Scheme 3). Furthermore, we have
generally shown12g that epoxyvinyl sulfones can be attacked
by methyl anion equivalents to give 1,2-anti addition products.
Hence, a campaign for synthesis of all eight stereoterads was
initiated. Moreover, since enantiopure epoxyvinyl sulfone 5 and
(12) (a) Park, T.; Torres, E.; Fuchs, P. L. Synthesis 2004, 11, 1895. (b) Tong,
Z.; Ma, S.; Fuchs, P. L. J. Sulfur Chem. 2004, 25, 1. (c) Torres, E.; Chen,
Y.; Kim, I.; Fuchs, P. L. Angew. Chem., Int. Ed. 2003, 42, 3124. (d) Chen,
Y.; Evarts, J.; Torres, E.; Fuchs, P. L. Org. Lett. 2002, 4, 3571. (e) Jiang,
W.; Lantrip, D. A.; Fuchs, P. L. Org. Lett. 2000, 2, 2181. (f) Evarts, J. B.;
Fuchs, P. L. Tetrahedron Lett. 1999, 40, 2703. (g) Hentemann, M.; Fuchs,
P. L. Tetrahedron Lett. 1999, 40, 2699. (h) Hentemann, M.; Fuchs, P. L.
Tetrahedron Lett. 1997, 38, 5615.
(13) Fuchs, P. L. Tetrahedron 2001, 57, 6855.
(14) Jacobsen, E. N.; Zhang, W.; Muci, A. R.; Ecker, J. R.; Deng, L. J. Am.
Chem. Soc. 1991, 113, 7063.
(15) Meyers, D. J.; Fuchs, P. L. J. Org. Chem. 2002, 67, 200.
(16) El-Awa, A.; Fuchs, P. L. Org. Lett. 2006, 8, 2905.
9
J. AM. CHEM. SOC. VOL. 129, NO. 29, 2007 9087

A R T I C L E S
Scheme 2
El-Awa et al.
Table 1. Epoxidation of Dienyl Sulfones 6 and 7
starting
entry
material
reagents
selectivity
yield (%)
1
7
Jacobsen’s
R,R-catalyst
Jacobsen’s
S,S-catalyst
CF3COOOH
14:16
68a
50:1a
2
3
4
5
6
7
7
7
6
6
6
6
14:16
1:10a
14:16
>1:50b
8:9
50:1a
70a
100b,c
72a
Jacobsen’s
R,R-catalyst
Jacobsen’s
S,S-catalyst
DMDO
no reaction
0
Scheme 3
8:9
100b,c
26:1b
8:9
CF3COOOH
100b
(71)d
1:10b
(1:30)d
a Yield and selectivity after flash column chromatography. b Crude yield
and selectivity. c The crude epoxide showed high purity and did not need
subsequent purification. d Yield and selectivity after crystallization.
epoxide 8 with a 26:1 selectivity (Table 1, entry 6).18 Con-
versely, trifluoroperacetic acid gave epoxide 9 in a 10:1
selectivity, which was easily improved to 30:1 after crystal-
lization from methanol (Table 1, entry 7).19 In the former case,
the absence of a hydrogen bond donor group in DMDO resulted
in dominance of steric control from the methyl moiety, while
the latter conditions featured hydrogen bonding of trifluoro-
peracetic acid to the silyl ether oxygen lone pair which
substantially exceeded the steric bias of the methyl group. When
using trifluoroperacetic acid with compound 7, both control
elements favored formation of epoxide 16, resulting in formation
of a single diastereomer in quantitative yield (Table 1, entry
3). Securing epoxide 14 via substrate-directed selectivity was
not possible, requiring double stereoselection via Jacobsen’s
catalyst (Table 1, entry 1).
its enantiomer are available on the mole scale12a using ∼1% of
commercially available Jacobsen salen-Mn catalysts, access to
all 16 possible stereotetrads is now possible.
Results and Discussion
Epoxidations. Previous investigations by our group have
indicated that anti-dienyl sulfone 7 is epoxidized with either
antipode of Jacobsen’s catalyst to give epoxides 14 and 16 in
high selectivity (Table 1, entries 1 and 2). On the other hand,
syn-dienyl sulfone 6 reacts with Jacobsen’s R,R-catalyst to
selectively give epoxide 8 (Table 1, entry 6). Perhaps not
surprisingly, 6 is completely inert toward the R-face specific,
Jacobsen S,S-catalyst (Table 1, entry 5).17
We envisioned that the existing asymmetric centers in 6 and
7 could be exploited in substrate-directed epoxidations with
achiral reagents since the allylic methyl groups in 6 and 7 exert
steric effects favoring anti epoxidations. Alternatively, the
homoallylic silyl ethers present in both substrates can direct
syn epoxidations by virtue of hydrogen bonding to the oxygen
lone pairs. Therefore, judicious selection of the epoxidizing
agent is critical in obtaining the desired epoxide in good
selectivity, especially in the case of 6, where both control
elements have opposed facial biases. In the event, epoxidation
of 6 with DMDO (dimethyldioxirane) quantitatively generated
Anti Methylations. With all four epoxides in hand, inves-
tigation of anti methylation chemistry was initiated. Several
literature reports have established that good 1,2-anti selectivity
can be achieved in reaction of vinyl epoxides with different
organometallic methylating agents.20 Among these reagents are
lithium tetramethylaluminate, lithium trimethylzincate,20a the
combination of methyllithium or methyl Grignard with boron
trifluoride,20b and a chiral copper(I) catalyst paired with
dimethylzinc.20c Those reports represented an appropriate point
(18) Unless otherwise indicated, all ratios reported in this publication are based
on 1H NMR integrations.
(19) The crude NMR showed a quantitative conversion and 10:1 epoxidation
selectivity.
(20) (a) Equey, O.; Vrancken, E.; Alexakis, A. Eur. J. Org. Chem. 2004, 2151.
(b) Alexakis, A.; Vrancken, E.; Mangeney, P.; Chemla, F. J. Chem. Soc.,
Perkin Trans. 1 2000, 3352. (c) Pineschi, M.; Moro, F. D.; Crotti, P.;
Bussolo, V. D.; Macchia, F. J. Org. Chem. 2004, 69, 2099.
(17) Torres, E. Ph. D. Thesis, Purdue University, May 2004. Also see ref 15.
9
9088 J. AM. CHEM. SOC. VOL. 129, NO. 29, 2007

Asymmetric Synthesis of Dipropionate Stereotetrads
A R T I C L E S
Table 2. 1,2-anti Methylation of Epoxide 8
Scheme 4
entry
reagent and conditions
products and selectivity
1
Me3Al/ rt/ DCM
11:10a
50:30
2
3
4
Li(Me4Al)/ rt/ THF
Li(Me3Zn)/ rt/ THF
MeMgBr/ rt/ THF
complex mixture
18
bromohydrin 11b
(10%)b,c
bromohydrin 11b
(60%)
of organocopper reagents and methyl metals. Before trying the
different combinations of organocopper reagents, it was deter-
mined that MeMgBr is very unreactive toward epoxide 8,
yielding 10% of the bromohydrin 11b22 along with recovery of
90% of the starting material after 2 days at room temperature
in tetrahydrofuran (Table 2, entry 4). Copper-catalyzed addition
of MeMgBr was very clean, but gave a 50:50 mixture of 1,2-
anti 11 combined with the unassigned 1,4 adduct 17 (Table 2,
entry 7). Varying the temperature between -60 °C to room
temperature, the copper(I) source or solvent had negligible effect
on the reaction. The 1,4-addition product 17 was formed as a
single diastereomer in some instances, and as a mixture of
epimers in others (Table 2).25 Switching to dimethylzinc gave
mostly 1,4-addition (Table 2, entry 8). Interestingly, the 1,4-
product from the dimethylzinc reaction was epimeric to that
from the Grignard reaction. Gilman’s cuprate26 gave the best
selectivity thus far: 8:1 in favor of 11 (Table 2, entry 9).
However, the reaction was accompanied with the formation of
dienyl sulfone 6, presumably by elimination of lithium alkoxide
from the electron rich π-allyl copper intermediate 19 (Scheme
4).
5
6
MeMgBr/ BF3‚OEt2
Et2O/ -78 °C
MeLi/ BF3‚OEt2
Et2O/ -78 °C
11:17d:18
63:25:12
11:17e
7
MeMgBr/ CuI (cat.)
Et2O/ -40 °C
1:1
8
Me2Zn/ Li2CuCl4 (cat.)
Tol/ rt
17f
9
MeLi/ CuI(2:1)
Et2O/ -20 °C
11:17d:6
77:10:13
11:17g
10
11
12
13
MeLi/CuCN (1:1)
THF/ -75 °C
11:89
MeLi/CuCN(2:1)
Et2O/ -78 °C
11:17d
63 :37
MeMgBr/CuI (2:1)
Et2O/ 0 °C
11:17d:6
46: 44: 10
11:17d:6
91:7:2
MeLi/MeMgBr (1:1)
CuI (cat)/ Et2O/ -40 °C
a The remaining 20% was an unidentified product. b 90% of the starting
material was recovered. c See ref 22. d 17 was formed as a mixture of
epimers. e 17 was formed as a single diastereomer. f 17 was formed as a
single diastereomer epimeric to that in entry 7. g 17 was formed as a single
epimer; the same as in entry 8.
of departure, although none of them had been applied to
epoxyvinyl sulfones.
Compound 20, bearing the allylic acetate, a much better
nucleofuge than lithium alkoxide, suffered â-elimination to
provide 21 with no methylation being observed.27
The results of screening different reagents for the 1,2-anti
methylation of epoxide 8 are summarized in Table 2. Reaction
of trimethylaluminum with epoxide 8 produced some 11, but
in low selectivity, along with syn methylation product 10 and
an unidentified side product (Table 2, entry 1), while tetra-
methylaluminate gave a complex mixture of products (Table 2,
entry 2). The zincate led mainly to base-promoted 1,4-
elimination,21 generating dienylic alcohol 18 (Table 2, entry 3).
The use of boron trifluoride in conjunction with methylmag-
nesium bromide or methyllithium delivered bromohydrin 11b22
in the former instance and low selectivity for 11 in the latter
(Table 2, entries 5 and 6). Although copper-promoted reactions
of organometallics are known to predominantly favor 1,4-
additions to vinyl epoxides,23 there were several reports in the
literature revealing that seemingly slight changes in reaction
conditions, or including additives, resulted in striking changes
in product distributions.24 Hence, we launched a screening study
Varying the reaction conditions (temperature, solvent, copper
source) gave variable product ratios, with the most selective
example shown in entry 9. Lower-order cyanocuprate gave good
selectivity toward 1,4-addition consistent with studies by Marino
et al.28 Both the Lipshutz cuprate29 and magnesiocuprate24a did
not offer any advantage over Gilman’s cuprate (Table 2, entries
11 and 12).
(24) (a) Lipshutz, B. In Organometallics in Synthesis: A Manual; Schlosser,
M., Ed.; Wiley: New York, 2002; p 665. (b) Taylor, R. J. K., Ed.
Organocopper Reagents a Practical Approach; Oxford University Press:
Oxford, 1994.
(25) The relative stereochemistries of the various 1,4-addition products were
not determined.
(26) (a) Gilman, H.; Jones, R. G.; Woods, L. A. J. Org. Chem. 1952, 17, 1630.
(b) Posner, G. H. Org. React. 1972, 19, 1.
(27) Similar eliminations are precedents in the literature, see: (a) Hegedus, L.
S.; Cross, J. J. Org Chem. 2004, 69, 8492. (b) Ibuka, T.; Chu, G.; Yoneda,
F. Tetrahedron Lett. 1984, 25, 3247. (c) Logusch, E. W. Tetrahedron Lett.
1979, 20, 3365. (d) Ruden, R. A.; Litterer, W. E. Tetrahedron Lett. 1975,
16, 2043. (e) Nilsson, A.; Ronlan, A. Tetrahedron Lett. 1975, 16, 1107.
(28) Marino, J. P.; Jaen, J. C. J. Am. Chem. Soc. 1982, 104, 3165.
(29) Lipshutz, B. H.; Sengupta, S. Org. React. 1992, 41, 135 and references
therein.
(21) (a) Thummel, R. P.; Rickborn, B. J. Am. Chem. Soc. 1970, 92, 2064. (b)
Thummel, R. P.; Rickborn, B. J. Org. Chem. 1971, 36, 1365.
(22) Bromohydrin structure:
(23) Marshall, J. A. Chem. ReV. 1989, 89, 1503.
9
J. AM. CHEM. SOC. VOL. 129, NO. 29, 2007 9089

A R T I C L E S
El-Awa et al.
Table 3. 1,2-anti Methylation of Epoxide 14
Table 4. 1,2-anti Methylation of Epoxide 16
entry
reagent and conditions
products and selectivity
1
Me3Al/ rt/ CH2Cl2
22:23
5:1
entry
reagent and conditions
products and selectivity
2
3
MeMgBr/ CuI (cat.)
Et2O/ -35 °C
MeLi/ CuI(2:1)
Et2O/ -70 °C
22:23
1:4
22:23
14:1
1
2
Me3Al/ rt/ CH2Cl2
MeMgBr/ Li2CuCl4 (cat.)
Et2O/ -35 °C
complex mixture
25
3
MeLi/ CuI(2:1)
24:25
Et2O/ -20 °C
10:1
Finally, when a 1:1 mixture of methyllithium and methyl-
magnesium bromide was cannulated to a solution of 8 in ether
at -40 °C with suspended copper(I) iodide, stereotetrad 11 was
obtained in 91:7:2 selectivity with respect to 17 and 6, with an
80% isolated yield of pure 11 after flash column chromatography
(Table 2, entry 13).30 The rationale for employing the MeLi-
MeMgBr combination was based upon the fact that the copper-
catalyzed Grignard reaction was clean and reproducible yet not
selective. However, Gilman’s cuprate showed good selectivity
for 11, but was neither clean nor uniformly reproducible.
Therefore, a reagent with hybrid properties was expected to
result in better selectivity and yield, which proved to be the
case. The actual species in the latter reaction is postulated to
be dimethylmagnesium, as it has been reported that dimethyl-
magnesium can be prepared by mixing MeLi and MeMgX.31
Scheme 5
one equivalent H2O, which gave 100:1 selectivity and 87%
isolated yield of 12 (eq 5).34
Having gained good experience with the anti methylation
reaction of 8, we next investigated the reaction of 14 starting
with those reagents that had performed well (Table 3). Treatment
of 14 with Me3Al resulted in a 5:1 mixture of 22 and 23
respectively.32 Cu(I)-catalyzed addition of MeMgBr furnished
a reversed selectivity pattern giving 22 and 23 in a 1:4 ratio.
Finally, Gilman’s reagent in ether at -70 °C gave 22/23 in
14:1 ratio, and a 62% isolated yield of pure 22 after flash column
chromatography (Table 3, eq 3).
Syn Methylations. The plan for achieving syn methylation
relied upon converting the epoxyvinyl sulfone to an allylic
alcohol with the double bond conjugated to the sulfone group.
This structural arrangement was expected to enable OH-directed
syn methylation to the vinyl sulfone. The desired result can be
achieved in two possible ways (A or B, Scheme 5). Treatment
of the epoxyvinyl sulfone 27 with a base should foster 1,4-
elimination of the epoxide moiety, leading to cross-conjugated
dienylic alcohol 28. Subsequent OH-directed methylation would
then give syn methylation product 30. Alternatively, 1,4-addition
to 27 with a hybrid species bearing nucleophilic/ nucleofugic
properties would give allylic alcohol 29. Subsequent reaction
of 29 with a methyl anion would yield 30 by a Lawton SN2′
reaction (Scheme 5).35
While Me3Al and Cu(I)-catalyzed MeMgBr reagent gave
unfavorable results, epoxide 16 reacted favorably with Gilman’s
reagent to give stereotetrad 24 in 10:1 selectivity with respect
to the 1,4-addition product 25 with no diene 7 observed, and in
71% yield of pure 24 after chromatographic purification (Table
4, eq 4).
Since copper(I)-promoted reactions of epoxides 8, 14, and
16 were satisfactory, it was surprising that epoxide 9 exclusively
gave 1,4-addition with both Gilman’s reagent and the Cu(I)-
catalyzed Grignard reagent. Me3Al reacted with 9 to give 10:1
selectivity in favor of the desired 1,2-anti addition product 12.
However, the ratio varied between 2:1 and 13:1 depending on
the number of equivalents of both Me3Al and H2O. After an
extended systematic study,33 it was determined that the optimum
protocol consists of four equivalents of Me3Al combined with
In practice, 28 proved to be unstable, giving silyl enolether
31 or ketone 32 under various conditions (eq 6). This isomer-
ization which involves a formal 1,5-hydrogen migration, was
quite fast, with a half-life of about 12 h under neutral conditions,
(30) Relative and absolute stereochemistries of all stereotetrads were determined
by X-ray crystallography of the stereotetrads themselves or derivatives.
See Supporting Information.
(34) For precedence to the necessity of H2O with Me3Al in the reaction with
epoxides see: Miyashita, M.; Hoshino, M.; Yoshikoshi, A. J. Org. Chem.
1991, 56, 6483. However, in that publication and the ensuing ones from
the same group 10 equivalents of Me3Al and 5 equivalents of H2O were
necessary.
(35) Brocchini, S. J.; Eberle, M.; Lawton, R. G. J. Am. Chem. Soc. 1988, 110,
5211 and references therein.
(31) Wakefield, B. J. Organomagnesium Methods in Organic Synthesis;
Academic Press: London, 1995; p 62.
(32) 23 was obtained as a single diastereomer, and its relative stereochemistry
was not determined.
(33) The results of that study will be published in due course.
9
9090 J. AM. CHEM. SOC. VOL. 129, NO. 29, 2007

Asymmetric Synthesis of Dipropionate Stereotetrads
A R T I C L E S
Scheme 6
Scheme 7
previously demonstrated with five- and six-membered epoxyvi-
nyl sulfones.40 Nevertheless, since dimethylamide anion is a poor
leaving group, it typically requires an additional nitrogen
alkylation or oxidation step to enable C-N bond scission.41
The search for a better nucleophile/nucleofuge was continued,
with the azole moiety being the optimal compromise. Although
pyrrole was totally unreactive, treatment of epoxide 8 with 3,5-
dimethylpyrazole resulted in a virtually quantitative yield of the
coveted Lawton SN2′ product 33 as a single diastereomer
(Scheme 6).16 Unfortunately, epoxide 9 reacted sluggishly with
3,5-dimethylpyrazole to give the desired adduct 34 in only 31%
yield after 15 h in toluene at 70 °C (Scheme 6). Poor selectivity
in the ensuing methylation step discouraged optimization of the
yield of 34.
and is much faster under basic conditions.36 Since generation
of 28 is performed under basic conditions, it was invariably
accompanied by 20-30% of 31 and/or 32.
Testing the syn methylation event was initiated with adduct
33. Methyllithium resulted in complex mixtures (Scheme 7).
While lithium tetramethylaluminate gave diastereomerically pure
product 10, the reaction failed to proceed beyond 50% conver-
sion, even employing a larger excess of reagent.42 Finally,
treatment of 33 with three equivalents of methylmagnesium
bromide in ether at room temperature gave 95% yield of
stereotetrad 10 with 20:1 syn selectivity (Scheme 7).16 The
selectivity of this reaction showed little solvent dependence,
allowing a single-pot operation as follows: Epoxide 8 is first
stirred with 3,5-dimethylpyrazole in toluene at 65 °C for 3 h,
and allowed to reach room temperature followed by slow
addition of MeMgBr and stirring at room temperature for an
additional hour to give 10 in 90% yield and 20:1 selectivity
(eq 7).
Moreover, addition of a methyl anion equivalent to 28 would
generate an allyl anion bearing a â-OTBS group. Therefore,
â-elimination may ensue, forming a conjugated diene with loss
of one of the asymmetric centers of the stereotetrad’s asymmetric
centers.37 Therefore, path A was abandoned (Scheme 5). Path
B requires the nucleophile undergoing addition to epoxyvinyl
sulfone 27 to possess several desirable properties: It should
add regioselectively in a 1,4-fashion, should be weakly basic
to avoid formation of 28, and should have good nucleofugacity,
so that it can be displaced simultaneously with the addition of
the methyl moiety to 29, i.e. the methyl group should perform
an SN2′ reaction to directly generate 30.
Screening to identify the optimal Lawton SN2′ nucleophile
was performed on epoxides 8, 9, and 14, with all three epoxides
showing similar trends. Ethylthiol, pyrrole, and acetone cyano-
hydrin all returned the starting material. Thiolates,38 imidazole,
and iodide exclusively gave 1,2-addition. Cyanide fostered base-
promoted elimination of the epoxide, yielding the dienylic
alcohol. Dimethylamine gave clean and quantitative 1,4-addition
to epoxide 9, affording 35 (Scheme 6),39 consistent with results
(36) While this reaction appears to involve an interesting competition between
two 1,5-H migrations, we have not investigated the specifics of this process.
We have previously observed a similar process in the 5-membered ring
series. See: Saddler, J. C.; Donaldson, R. E.; Fuchs, P. L. J. Am. Chem.
Soc. 1981, 103, 2110.
Epoxide 14 behaved similarly to 8 upon sequential treatment
of 3,5-dimethylpyrazole and MeMgBr.16 However, the MeMgBr
step showed an interesting temperature dependence. Conducting
the reaction at 0 °C afforded a 1:1 mixture of stereotetrads 15
and 22. Running the reaction at -45 °C favored 22 by a factor
of 4, while performing the reaction at 36 °C both reversed and
amplified the selectivity to 20:1 in favor of 15 (Scheme 8).43
(37) In fact, that pathway is used to install the first methyl group very early on
in the sequence used to make diad 13. See footnote 12c.
(38) Early experiments were performed either with benzyl or ethyl thiolates in
combination with Lewis acids and resulted in exclusive 1,2-addition.
o
However, later on we discovered that sodium thiolates in THF at -70 C
give good 1,4-selectivity.
(39) 33, 34, and 35 each was formed as a single diastereomer; however, their
relative stereochemistries were not determined.
9
J. AM. CHEM. SOC. VOL. 129, NO. 29, 2007 9091

A R T I C L E S
Scheme 8
El-Awa et al.
Scheme 10
Scheme 9
Scheme 11
In contrast to epoxides 8 and 14, reaction of 3,5-dimeth-
ylpyrazole with epoxide 16 gave a mixture of products with
the desired adduct as a minor component.
Hence, Me2NH was tested with 16, giving a quantitative yield
of 1,4-adduct 36 in 15 min at room temperature (Scheme 9).44
Treatment of 36 with methyllithium at -20 °C for 90 min gave
37 in high selectivity. Amine 37 was then converted to
stereotetrad 38 via Cope elimination45 mediated by m-chloro-
perbenzoic acid (Scheme 9).
To determine if the magnesium counterion had any advantage
over lithium, 36 was treated with MeMgBr at 0 °C, surprisingly
resulting in direct formation of 38 in 25:1 diastereoselectivity
and 75% yield.
This may be explained by the stronger energy of the N-Mg
bond relative to that of the N-Li bond,46 rendering the formation
of magnesiumdimethylamide species much more favorable than
lithiumdimethylamide. In addition, the whole sequence starting
from 16 was done in a single pot similar to the case of tetrads
10 and 15 giving tetrad 38 in 75% yield (eq 8).
Unfortunately this led to formation of a 10:1 mixture of 12 and
13 in favor of the undesired anti addition product 12. Switching
to Grignard reagent or varying temperature or solvent did not
offer much advantage. 3,5-Dimethylpyrazole adduct 34 reversed
the selectivity, but gave only a mediocre 2:1 ratio in favor of
13.
The improved result from the 3,5-dimethylpyrazole adduct
encouraged the screening of more nucleophiles for the Lawton
SN2′ reaction of epoxide 9. Trimethylamine and sodium ben-
zylthiolate both added in 1,4 fashion; however, both adducts
39 and 40 reacted with MeMgBr to predominantly give anti
adduct 12 (Scheme 10).
These results suggested that the strong bias favoring formation
of 12 is hard to override. Therefore, we elected to pursue an
epimerization strategy. Tetrad 11 was quantitatively oxidized
to ketone 41 via the Dess-Martin reagent.47 Reduction of 41
after cooling to -70 °C, with diisobutylaluminum hydride in
the same reaction pot, gave the final tetrad 13 in an overall
yield of 96% as a single diastereomer (Scheme 11).
The final task was to make the eighth stereotetrad 13. Since
dimethylamine adduct 35 was in hand, it was subjected to
methyllithium addition followed by oxidative Cope elimination.
(40) Pan, Y.; Hardinger, S. A.; Fuchs, P. L. Synth. Commun. 1989, 19, 403.
(41) (a) Donaldson, R. E.; Saddler, J. C.; McKenzie, A. T.; Byrn, S.; Fuchs, P.
L. J. Org. Chem. 1983, 48, 2167. (b) Hutchinson, D. K.; Fuchs, P. L. J.
Am. Chem. Soc. 1985, 107, 6137. (c) Hutchinson, D. K.; Hardinger, S. A.;
Fuchs, P. L. Tetrahedron Lett. 1986, 27, 1425. (d) Pan, Y.; Hutchinson,
D. K.; Nantz, M. H.; Fuchs, P. L. Tetrahedron 1989, 45, 467.
(43) (a) See ref 16 for a proposition about this selectivity pattern. (b) In ref 16
we report that a 20:1 dr was obtained at room temperature. However, we
later on found out that actually higher temperature (36 oC) is required to
attain such ratio, and the previous result was due to the exotherm resulting
from fast addition of MeMgBr.
(44) 36 was formed as a single diastereomer; however, its relative stereochem-
istry was not confirmed.
(42) With both MeLi and Me4AlLi various solvents and temperature ranges were
tested yet without significant improvement.
(45) Cope, A. C.; Foster, T. T.; Towle, P. H. J. Am. Chem. Soc. 1949, 71, 3929.
9
9092 J. AM. CHEM. SOC. VOL. 129, NO. 29, 2007

Asymmetric Synthesis of Dipropionate Stereotetrads
A R T I C L E S
end, we have subjected tetrads 10, 42,49 and 4350to ozonolytic
cleavage to yield termini-differentiated stereotetrads 44, 45, and
46 (Scheme 12).51
Scheme 12
Conclusion
All eight diastereomeric stereotetrads, seven of which are
present in known polypropionate natural products, have been
synthesized in high stereoselectivities and yields. Both antipodes
of 5, the first chiral intermediate in our synthetic scheme, are
equally easily accessible; therefore, access to all 16 enantiomers
of the eight diastereomeric tetrads is possible. Oxidative
cleavage of the cyclic stereotetrads has been demonstrated, thus
making termini-differentiated acyclic tetrads available for routine
use in target syntheses.
Acknowledgment. We thank Arlene Rothwell and Karl Wood
for providing MS data. We are most grateful to Professor Mark
Lipton for his collaborative efforts in calculating unnatural
targets bearing polypropionates (and related) tetrads.
Note Added after ASAP Publication. After this article was
published ASAP on June 14, 2007, an error was discovered in
the drawing of Jacobsen’s catalyst in Scheme 1. The corrected
version was published ASAP on July 2, 2007.
Because of scalability and economic issues, we also demon-
strated that Swern oxidation48 can replace the Dess-Martin
protocol; however, the latter reagent required aqueous workup
between the oxidation and reduction steps to remove the Et3N‚
HCl.
Oxidative Cleavage of the Stereotetrads. To be viable for
use in synthesis of polypropionate natural products, the above
stereotetrads have to be rendered acyclic and further function-
alized and/or coupled with additional segments. Toward that
Supporting Information Available: Experimental procedures,
1
spectroscopic and analytical data for new compounds, H and
13C NMR spectra of key compounds. X-ray crystallographic
data in CIF format. This material is available free of charge via
the Internet at http://pubs.acs.org.
JA071217X
(46) A recent theoretical study shows that H2N-MgH bond dissociation energy
is 356.3 kJ/mol while that of H2N-Li bond dissociation is 302.4 kJ/mol
which is equal to a difference of 53.9 kJ/mol (12.9 kcal/mol). See: Mo,
O.; Yanez, M.; Eckert-Maksic, M.; Maksic, Z. B.; Alkorta, I.; Elguero, J.
J. Phys. Chem. A 2005, 109, 4359.
(47) Dess, P. B.; Martin, J. C. J. Am. Chem. Soc. 1979, 101, 5294.
(48) (a) Mancuso, A. J.; Huang, S.-L.; Swern, D. J. Org. Chem. 1978, 43, 2480.
(b) Tidwell, T. T. Org. React. 1990, 39, 297.
(49) 42 was obtained from 15 by a protection/selective deprotection operation.
See ref 16.
(50) 43 was obtained from 13 by a protection/selective deprotection operation.
See Supporting Information.
(51) For previous examples of ozonolytic cleavage of vinylsulfones from our
lab see refs 12c, 12d, 12g, and 16.
9
J. AM. CHEM. SOC. VOL. 129, NO. 29, 2007 9093