Enantioselective synthesis of erythro-4-deoxyglycals as scaffolds for target- And diversity-oriented synthesis: New insights into glycal reactivity

Article abstract of DOI:10.1039/b417429a

An efficient, enantioselective synthesis of eryithro-4-deoxygtycals has been developed using asymmetric aldehyde allylation and tungsten-catalyzed alkynol endo-cycloisomerization as the key steps. These versatile synthetic scaffolds have been elaborated t

Full text of DOI:10.1039/b417429a

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
Enantioselective synthesis of erythro-4-deoxyglycals as scaffolds for
target- and diversity-oriented synthesis: new insights into glycal
reactivity†
Sirkka B. Moilanena and Derek S. Tan*a,b
a Tri-Institutional Training Program in Chemical Biology, Memorial Sloan–Kettering Cancer
Center, 1275 York Ave., Box 422, New York, NY 10021, USA
b Tri-Institutional Research Program and Molecular Pharmacology & Chemistry Program,
Memorial Sloan–Kettering Cancer Center, 1275 York Ave., Box 422, New York, NY 10021,
USA. E-mail: tand@mskcc.org; Fax: +1 212-717-3655; Tel: +1 212-639-7321
Received 18th November 2004, Accepted 14th December 2004
First published as an Advance Article on the web 24th January 2005
An efficient, enantioselective synthesis of erythro-4-deoxyglycals has been developed using asymmetric aldehyde
allylation and tungsten-catalyzed alkynol endo-cycloisomerization as the key steps. These versatile synthetic scaffolds
have been elaborated to a variety of products using stereoselective transformations that are complementary to those
available using the corresponding threo glycals. This work has provided valuable insights into the relationships
between glycal structure and reactivity. In addition, a new diene-forming side reaction during tungsten-catalyzed
alkynol cycloisomerization has been discovered.
Introduction
Results and discussion
Glycals are versatile building blocks with a broad range of
applications in the synthesis of carbohydrates, polyketides, and
other natural products.1 Hence, they are also attractive scaffolds
for diversity-oriented synthesis.2–4 Recently, in the course of a
project aimed at using stereochemical diversity to probe three-
dimensional structure space, we required access to all four
stereoisomeric 4-deoxyglycals 1–4 (Fig. 1). The threo stereoiso-
mers (1 and 2) are readily synthesized by substrate-controlled
stereoselective Luche reduction of the appropriate Danishefsky
diene-derived dihydropyrone.5 Surprisingly, however, the erythro
stereoisomers (3 and 4) pose a much greater synthetic challenge.
Herein we report our investigations of this deceptively simple
problem and the effective solution we ultimately devised to
provide synthetically useful quantities of erythro-4-deoxyglycals.
We also describe stereoselective transformations of these scaf-
folds that are complementary to those available using the threo
stereoisomers and demonstrate their utility as versatile synthons
for stereoselective target- and diversity-oriented synthesis. In the
course of this work, we have compared several alkynol endo-
cycloisomerization reactions, discovered a new side reaction
that can occur in the W(CO)6/Et3N/hm-catalyzed reaction, and
gained valuable insights into the striking differences in glycal
reactivity that arise from seemingly subtle differences in glycal
structure.
Classical approaches
At the outset of these studies, we expected that the erythro-
4-deoxyglycals could be accessed using any of a variety of
well-precedented classical approaches. Thus, our initial efforts
focused on 1,2-reduction of the dihydropyrone 5 (Fig. 2) with
sterically demanding hydride reagents, based on the expectation
that equatorial attack would provide the desired erythro product
7. However, when 5 was treated with any of a variety of
bulky hydride reagents,6 the threo stereoisomer 6 was formed
nearly exclusively. These results are in stark contrast to the
stereoselectivity trends observed with bulky hydride reductions
of cyclohexanones7 and cyclohexenones.8 This unusual stereose-
lectivity for dihydropyrone reduction may be rationalized by the
reduced steric conflict along the axial trajectory (5a),9 increased
torsional strain associated with the transition state for equatorial
attack,10 stereoelectronic effects involving the ring oxygen,11
and/or the availability of a competing reactive conformation
(5b) in which the C5-methyl substituent is oriented axially.
Fig. 1 Four stereoisomeric 4-deoxyglycal scaffolds for target- and
diversity-oriented synthesis.
Fig. 2 Reductions of dihydropyrone 5 with bulky hydride reagents.
Since the threo-4-deoxyglycal 6 is readily available, we next
turned our attention to inversion of its C3-hydroxyl group by
either direct or indirect means. Not surprisingly, Mitsunobu
inversion of 6 was thwarted by competing SN2ꢀ displacement.12
Interestingly, this was also the case for a C1-alkylated con-
gener (not shown), which was synthesized using our recently
† Electronic supplementary information (ESI) available: experimental
procedures and spectral data for compounds 25–35. See http://www.
rsc.org/suppdata/ob/b4/b417429a/
7 9 8
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 7 9 8 – 8 0 3
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
©

View Article Online
reported B-alkyl Suzuki–Miyaura cross coupling approach to
C1-alkylglycals.13 These results indicate that electronic factors
clearly dominate over steric constraints for these reactions of
glycals.
This propensity of glycals to undergo attack at the C1
position brought our attention to an indirect approach to C3
inversion (Fig. 3). A thiophenol Ferrier–sigmatropic rearrange-
ment sequence has been used to convert 3,4,6-tri-O-acetyl-
D-glucal (8) to 3,6-di-O-acetyl-D-allal (11).14,15 Unfortunately,
exposure of our threo-4-deoxyglycal substrate 13 to the reported
Ferrier conditions (BF3·OEt2) resulted in rapid decomposition.
Interestingly, the use of SnCl4 resulted in direct substitution at
the C3 position by thiophenol (12) but allylic substitution at
the C1 position by 4-methoxyphenol (14). Such nucleophile-
dependent regioselectivity has been rationalized by the hard–
soft acid–base principle.16 In this vein, we tested the harder
nucleophile S-(trimethylsilyl)thiophenol, which favors C1 attack
on C4-substituted glycals.17 However, this did not alter the
undesirable C3 regioselectivity for our 4-deoxyglycal substrate
13. These results highlight the striking differences in reactivity
that arise from the absence of a C4-substituent in 4-deoxyglycals.
Fig. 4 Endo- and exo-cycloisomerizations of 3,5-dihydroxyalkynes.
been described by Trost.23,24 Although multistep synthesis of
the alkynol precursor would be required, we recognized that an
enantioselective synthesis could be achieved by drawing upon
the vast arsenal of methodologies that have been developed in
the context of polyketide synthesis.
It should be noted that it was not clear at the outset of
our studies whether or not the key cycloisomerization reaction
would be successful. Trost had reported a single example of 4-
deoxyglycal synthesis as a mixture of threo and erythro diastere-
omers using rhodium-catalyzed cycloisomerization.23 Moreover,
while our work was in progress, Wipf reported a strong influence
of alkynol stereochemistry upon the regiochemical outcome of
W(CO)6/DABCO/hm-catalyzed cycloisomerization.25 Notably,
substrates with our required stereochemical configuration (15)
favored exo-cycloisomerization (16) unless non-coordinating
protecting groups were used at the C3 position. Furthermore,
the practical utility of the reaction was limited by low yields
(15–49%).
Thus, we began by synthesizing the requisite alkynol pre-
cursors 22 (Fig. 5). We designed these substrates to maintain
the spacing of oxygen functionalities found in polyketide
natural products. We also envisioned that the orthogonally
protected C7-hydroxyl might provide an attachment point for
future applications in solid phase synthesis. Enantioselective
allylation of aldehyde 1726,27 with Leighton’s allylsilane reagent28
provided convenient, efficient access to the homoallylic alcohol
18, which was then converted to aldehyde 19.29 We next
had the opportunity to evaluate several recently developed
diastereoselective alkyne addition reactions.30 Unfortunately, in
our hands, attempted couplings of 19 with trimethylsilylacety-
lene using Zn(OTf)2/N-methylephedrine31 led to recovered
starting material, consistent with a recent report involving
related substrates.32 Reactions at higher temperatures or with
Et2Zn/Ti(Oi-Pr)4/BINOL33 led to b-elimination of 19 to give
the corresponding a,b-enone. We thus turned to direct addi-
tion of lithium trimethylsilylacetylide, which afforded modest
substrate-controlled diastereoselectivity favoring the desired
propargylic alcohol 20 (3 : 2 dr). Protecting group manipulations
then provided cycloisomerization substrates 22. Large scale
material throughput was conveniently achieved using this route
(8.0 g of alkynol 22a in 30% overall yield from aldehyde 17).34
Notably, the minor alkynol diastereomer 21 was also readily
separated chromatographically from 20 and carried on to 22 via
standard Mitsunobu inversion protocols (not shown).
Fig. 3 Net C3-inversion of 3,4,6-tri-O-acetyl-D-glucal 8 and related
reactions of threo-4-deoxyglycal 13.
Alkynol cycloisomerization approaches
Having explored the classical approaches, we turned to an
alternative strategy involving endo-cycloisomerization of an
appropriately functionalized linear alkynol precursor (Fig. 4).
Tungsten-catalyzed alkynol cycloisomerizations have been de-
veloped extensively by McDonald,18–21 building upon earlier
work with molybdenum-catalyzed reactions.22 More recently,
rhodium- and ruthenium-catalyzed cycloisomerizations have
Fig. 5 Enantioselective synthesis of protected alkynol substrates 22 and endo-cycloisomerization reactions to form erythro-4-deoxyglycals 23.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 7 9 8 – 8 0 3
7 9 9

View Article Online
We were now in position to investigate the key alkynol
cycloisomerization reaction. Unfortunately, in our hands, at-
tempted cycloisomerizations of 22a,e–g using Rh[P(m-FPh)3]3Cl
or Rh(PPh3)3Cl yielded only trace amounts (<10%) of the
corresponding glycals 23a,e–g, despite their close similarity
to the previously reported example.23 However, after extensive
experimentation with various substrates, we were gratified to
find that exposure of 22a to W(CO)6/Et3N/hm20 provided the
desired erythro-4-deoxyglycal 23a in good yield with no evidence
of competing exo-cycloisomerization.25,35
Fig. 7 Half-chair conformations of the erythro-4-deoxyglycal 23a and
steric influences upon double bond reactivity.
Several important observations were made in the course of
these investigations. First, replacement of Et3N with DABCO,
which is now more commonly used as the base,18,20 resulted in
markedly slower reaction rates and no improvement in yields.
Second, attempted tungsten-catalyzed cycloisomerizations of
22b–e resulted in the formation of an unexpected truncated diene
24a. This diene was the exclusive product for 22b and 22c, and a
major product for 22d (1 : 1 with 23d) and 22e (1 : 0.04 : 1.2 with
23e and the corresponding exo-glycal). Similarly, reaction of the
doubly PMB-protected alkynol 22f yielded the corresponding
diene 24b (2.3 : 1 : 1 ratio with 23f and the corresponding exo-
glycal).
A possible mechanism for this side reaction involves normal
cycloisomerization, followed by elimination of the axially-
oriented C3-substituent, alkene isomerization, and [4 + 2]-
cycloreversion (Fig. 6). Formation of this diene has not been
reported previously, perhaps due to the decreased propensity
for elimination of the corresponding threo stereoisomers (see
below), decreased acidity of substrates having substituents at the
C4 position, and volatility of dienes corresponding to substrates
lacking our bulky sidechains.
tion of the a-D-arabino-4-deoxyglycoside 26. Likewise, treatment
of 25 with allylmagnesium chloride provided the corresponding
a-allyl C-glycoside 27 (80% yield). Hydroboration–oxidation of
23a using thexylborane proceeded stereoselectively to give the
b-C2-alcohol 28 (74% isolated yield). In contrast, the use of BH3
led to a 2 : 1 ratio of 28 and its a-C2-epimer while 9-BBN and
dicyclohexylborane were unreactive.
Interestingly, attempted direct methanolysis of the glycal
double bond under mild acidic conditions (PPh3·HBr) led to a 3 :
1 : 3.3 mixture of b-methyl erythro-2,4-dideoxyglycoside 30, its a-
anomer (not shown), and the a-glycoside Ferrier rearrangement
product 31 (51% combined yield of 30 and 31). Notably,
exposure of the corresponding threo-4-deoxyglycal to identical
conditions did not result in any Ferrier rearrangement. These re-
sults are also consistent with axial orientation of the C3-OTIPS
substituent, which would be stereoelectronically predisposed to
Ferrier-type elimination. Furthermore, these observations lend
support to our proposed mechanism for diene formation during
the cycloisomerization reaction (Fig. 6). Similarly, Sc(OTf)3-
catalyzed condensation of 23a with salicylaldehyde38 resulted in
highly regio- and stereoselective formation of tricycle 29, albeit
in low yield, presumably via the intermediacy of a Ferrier-type
elimination.
The erythro-4-deoxyglycal 23a was also converted to C1-
substituted glycals 32a,b via previously described cross coupling-
based routes.13,39 Hydroboration–oxidation proceeded from the
b-face of the glycal double bond with complete stereoselectivity
to provide the corresponding a-D-arabino-4-deoxy-C-glycosides
33a,b in good yields. Interestingly, acid-catalyzed alcoholysis of
the C1-substituted glycal 32a provided the a-methyl ketoglyco-
side 35a (50% yield), corresponding to a double substitution at
both the C1 and C3 positions. We postulate that, in contrast
to the case with the parent unsubstituted glycal 23a, an initial
Ferrier-type elimination is followed by MeOH attack at the
C3 position, leading to the 3-methoxyglycal intermediate 34a.
Indeed, in the C1-phenylglycal series (32b), the 3-methoxyglycal
34b was isolated (24% yield) and found to be in equilibrium with
the doubly substituted product 35b.
Fig. 6 A possible mechanism for formation of truncated diene 24.
Taken together, our results underscore the major influence of
protecting groups and stereochemical configuration upon the
efficiency and outcome of alkynol cycloisomerization reactions.
Conclusion
We have developed an efficient, enantioselective synthesis of
erythro-4-deoxyglycals via a tungsten-catalyzed alkynol endo-
cycloisomerization. We have also carried out a variety of stereo-
selective reactions of these erythro-4-deoxyglycals to probe their
reactivity and to demonstrate their utility as versatile synthetic
intermediates for target- and diversity-oriented synthesis. No-
tably, the products have the appropriate stereochemical config-
uration for applications in the target-oriented synthesis of poly-
ketide natural products such as the apicularens,40 latrunculins,41
spongistatins,42,43 salicylihalimides,44 and compactin.45 Efforts
to use the erythro- and threo-4-deoxyglycals as stereoisomeric
scaffolds for diversity-oriented synthesis are ongoing and will
be reported in due time. In the course of this work, we have
also discovered a new side reaction that occurs during tungsten-
catalyzed cycloisomerization and compared various cycloiso-
merization reactions. Our studies highlight the striking effects
upon glycal reactivity that can result from the absence of a C4-
substituent or the presence of axially-oriented C3-substituents.
Stereoselective reactions of the erythro-4-deoxyglycal
Having secured access to synthetically useful quantities of
erythro-4-deoxyglycal 23a, we were in a position to explore the
reactivity of this novel glycal. In particular, this stereoisomer
provides the opportunity to carry out stereoselective reactions
of the glycal double bond that are complementary to those
available using the corresponding threo stereoisomer. Selective
reactivity at the b-face of the double bond is likely contingent
upon a favored reactive conformation in which the large C3-
OTIPS substituent is axially disposed (23aꢀ, Fig. 7).36
Thus, we carried out a number of exploratory reactions
to probe the reactivity of 23a and to illustrate its potential
synthetic utility in both target- and diversity-oriented synthesis
(Fig. 8). We were gratified to find that epoxidation of 23a with
dimethyldioxirane37 provided the b-epoxide 25 as a single di-
astereomer. Subsequent methanolysis led to quantitative forma-
8 0 0
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 7 9 8 – 8 0 3

View Article Online
Fig. 8 Stereoselective reactions probing the reactivity of the erythro-4-deoxyglycal 23a and illustrating its potential synthetic utility as a versatile
scaffold for target- and diversity-oriented synthesis.
to yield alcohol 18 as a clear oil (14.1 g, 80%). The ee of 18
was determined to be 86% by analysis of its Mosher ester. The
Experimental
Materials and methods
reaction was also completed on a 29 mg scale in 73% yield and
88% ee. Spectral data were in agreement with the literature.29
Reagents were obtained from Aldrich Chemical (www.sigma-
aldrich.com) or Acros Organics (www.fishersci.com) and used
without further purification. Optima grade solvents were ob-
tained from Fisher Scientific (www.fishersci.com), degassed with
Ar, and purified using a solvent drying system as described,46
unless otherwise indicated. Reactions were performed in flame-
(3R)-5-(tert-butyldiphenylsilyloxy)-3-(triethylsilyloxy)pentanal
dried glassware under positive Ar pressure with magnetic stir-
(19)
ring. Cold baths were generated as follows: 0 ◦C, wet ice/water;
−78 ◦C, dry ice/acetone. Photolysis reactions were carried out
The alcohol 18 was converted to aldehyde 19 in two steps
as previously described.29 For analytical scale synthesis, olefin
36 and aldehyde 19 were purified by flash chromatography as
usual. For preparative scale synthesis, 14.1 g of alcohol 18 was
converted to 18.7 g of crude aldehyde 19 (84% overall yield based
on NMR analysis of the crude product), which was carried on
to 20 without further purification (see below).
in a Rayonet photoreactor at 300 nm. Flash chromatography
was performed on E. Merck 60 230–400 mesh silica gel. Optical
rotations were recorded on a JASCO model DIP-370 digital
polarimeter. TLC was performed on 0.25 mm E. Merck silica
gel 60 F254 plates and visualized under UV light (254 nm) or by
using potassium permanganate or cerium ammonium molybde-
nate stains. IR spectra were recorded neat on NaCl plates with
a Perkin–Elmer model 1600 FTIR spectrometer with peaks re-
ported in cm−1. NMR spectra were recorded on Bruker DRX500
or AMX400 instruments at 24 ◦C in CDCl3 unless otherwise
indicated. Chemical shifts are expressed in ppm relative to TMS
(1H, 0 ppm) or solvent signals: CDCl3 (13C, 77.0 ppm), C6D6
(1H, 7.16 ppm; 13C, 128.0 ppm) or d6-acetone (13C, 206.2 ppm).
Coupling constants are expressed in Hz. Mass spectra were ob-
tained at the MSKCC Analytical Core Facility on a PE SCIEX
API 100 mass spectrometer by electrospray (ESI) ionization.
(3R,5R)-7-(tert-Butyldiphenylsilyloxy)-5-(triethylsilyloxy)-1-
(trimethylsilyl)hept-1-yn-3-ol ((−)-20) and (3S,5R)-7-(tert-
Butyldiphenylsilyloxy)-5-(triethylsilyloxy)-1-(trimethylsilyl)-
hept-1-yn-3-ol ((−)-21)
n-BuLi (2.5 M in hexane, 1.99 mmol, 1.3 equiv.) was added
dropwise to a cooled (−78 ◦C) solution of TMS acetylene
(270 lL, 1.99 mmol, 1.3 equiv.) in anhydrous THF (4.5 mL). The
reaction mixture was stirred at −78 ◦C for 1 h then treated over
5 min with a solution of the aldehyde 19 (722 mg, 1.53 mmol,
1.0 equiv.) in anhydrous THF (3.5 mL). After an additional
20 min at −78 ◦C, the reaction was quenched with H2O. The
aqueous layer was extracted with Et2O (3 ×) and the combined
extracts were washed with brine, dried (MgSO4), filtered, and
concentrated to afford a 59 : 41 mixture of 20 and 21. The crude
material was purified by flash chromatography (elution with 93 :
7 hexanes–EtOAc) to yield propargylic alcohols 20 (454 mg,
52%) and 21 (342 mg, 39%) as clear oils. For preparative scale
synthesis, 18.7 g of crude aldehyde 19 (see above) was converted
to alcohol 20, which was purified by flash chromatography:
(10.4 g, 46% overall yield over 3 steps 18 → 20).
(3S)-1-(tert-Butyldiphenylsilyloxy)hex-5-en-3-ol (18)
A solution of aldehyde 1726,27 (15.6 g, 49.9 mmol, 1.0 equiv.) in
anhydrous toluene (30 mL) was added to a cooled (0 ◦C) so-
lution of (4S,5S)-2-allyl-2-chloro-3,4-dimethyl-5-phenyl-1,3,2-
oxazasilolidine28 (18.6 g, 69.4 mmol, 1.4 equiv.) in anhydrous
toluene (200 mL). The mixture was transferred to a freezer
(−20 ◦C) for 3 h. 1 N aq. HCl (150 mL) and EtOAc (100 mL)
were then added and the mixture was warmed to rt and stirred
for 15 min. The aqueous layer was extracted with EtOAc (2 ×)
then the combined organic layers were washed with brine, dried
(MgSO4), filtered, and concentrated. The residue was purified
by flash chromatography (elution with 95 : 5 hexanes–EtOAc)
20: [a]2D5 = −8.7 (c 0.5, CHCl3). TLC: Rf 0.11 (9 : 1 hexanes–
EtOAc). IR (NaCl, film): 2955, 2871, 1422, 1249, 1111, 1087,
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 7 9 8 – 8 0 3
8 0 1

View Article Online
1009, 841, 739, 703. 1H-NMR (400 MHz): d 7.63 (m, 4H),
7.39 (m, 6H), 4.52 (m, 1H), 4.14 (m, 1H), 3.69 (m, 2H), 2.80
(d, 1H, J = 3.9), 1.89–1.70 (m, 4H), 1.04 (s, 9H), 0.95 (t, 9H, J =
7.9), 0.50 (q, 6H, J = 7.9), 0.16 (s, 9H). 13C-NMR (100 MHz):
d 135.8, 135.0, 133.9, 129.9, 127.9, 106.9, 89.3, 69.2, 62.0, 60.6,
44.2, 40.9, 27.1, 19.3, 7.1, 5.3. ESI-MS m/z: (pos.) 569.4 [M +
H]+, 591.3 [M + Na]+; (neg.) 567.3 [M − H]−.
18.2, 12.4, 7.2, 5.3. ESI-MS m/z: (pos.) 653.6 [M + H]+, 675.3
[M + Na]+; (neg.) 687.4 [M + Cl]−.
(3R,5R)-1-(tert-Butyldiphenylsilyloxy)-5-(triisopropylsilyloxy)-
hept-6-yn-3-ol ((+)-22a)
Acetic acid (1.1 mL) was added to a solution of alkyne 38
(122.9 mg, 0.19 mmol) in THF (0.4 mL, Optima grade) and
water (0.4 mL). After 1.5 h at rt the mixture was quenched with
sat. aq. NaHCO3. The aqueous layer was extractedwith Et2O
(3 ×) and the combined organic layers were dried (MgSO4), fil-
tered, and concentrated. The crude material was purified by flash
chromatography (elution with 95 : 5 hexanes–EtOAc) to yield
alcohol 22a as a clear oil (80.6 mg, 79%). For preparative scale
synthesis, 14.0 g of crude alkyne 38 (see above) was converted to
alcohol 22a, which was purified by flash chromatography (8.0 g,
81% overall yield over 3 steps 20 → 22a).
21: [a]2D5 = −6.4 (c 1.0, CHCl3). TLC: Rf 0.34 (9 : 1 hexanes–
EtOAc). IR (NaCl, film): 3436, 2955, 2872, 1467, 1420, 1384,
1
1249, 1096, 1008, 844, 738, 697. H-NMR (400 MHz): d 7.64
(m, 4H), 7.40 (m, 6H), 4.59 (m, 1H), 4.37 (m, 1H), 3.68 (m, 2H),
3.41 (d, 1H, J = 5.5), 1.92–1.63 (m, 4H), 1.06 (s, 9H), 0.95
(t, 9H, J = 7.9), 0.61 (q, 6H, J = 8.0), 0.15 (s, 9H). 13C-NMR
(100 MHz): d 136.3, 135.6, 133.6, 127.8, 127.5, 127.0, 106.7,
88.8. 60.4, 27.1, 26.8, 26.6, 19.1, 5.0. ESI-MS m/z: (pos.) 569.2
[M + H]+, 591.3 [M + Na]+; (neg.) 567.3 [M − H]−, 603.3 [M +
Cl]−.
[a]2D5 = +4.2 (c 1.0, CHCl3). TLC: Rf : 0.35 (9 : 1 hexanes–
EtOAc). IR (NaCl, film): 2934, 2862, 1462, 1426, 1105, 881,
1
820, 736, 699. H-NMR (400 MHz): d 7.66 (m, 4H), 7.41 (m,
6H), 4.72 (m, 1H), 4.18 (m, 1H), 3.85 (m, 2H), 3.27 (d, 1H, J =
2.7), 2.45 (d, 1H, J = 2.0), 1.97 (m, 1H), 1.82 (m, 1H), 1.72 (m,
2H), 1.20–1.13 (m, 30H). 13C-NMR (100 MHz): d 135.8, 133.5,
130.0, 128.0, 85.4, 73.2, 68.7, 62.6, 61.8, 46.1, 39.3, 27.0, 19.3,
18.2, 12.5. ESI-MS m/z: (pos.) 561.4 [M + Na]+; (neg.) 537.3
[M − H]−.
(3R,5R)-7-(tert-Butyldiphenylsilyloxy)-5-(triethylsilyloxy)hept-
1-yn-3-ol (37)
◦
A cooled (0 C) solution of 20 (10.3 g, 18.1 mmol) in MeOH
(180 mL, Optima grade) was treated with H2O (9.8 mL) and
K2CO3 (5.0 g, 36.2 mmol). The mixture was stirred at 0 ◦C
for 10.5 h then the MeOH was evaporated to afford a white
solid. The solid was dissolved in Et2O, washed with H2O (3 ×),
washed with brine, dried (MgSO4), filtered, and concentrated
to yield propargylic alcohol 37 as a white solid (9.41 g, 91%
based on NMR analysis of the crude product) that was carried
on without further purification in preparative scale synthesis.
An analytical sample was purified by flash chromatography for
spectral characterization.
(2R,4R)-2-[2-(tert-Butyldiphenylsilyloxy)ethyl]-4-
(triisopropylsilyloxy)-3,4-dihydro-2H-pyran ((+)-23a)
A solution of the alkynol 22a (116.1 mg, 0.22 mmol, 1.0 equiv.)
in anhydrous THF (0.6 mL) was added to W(CO)6 (19.0 mg,
0.054 mmol, 0.25 equiv.) and the mixture was treated with
distilled Et3N (135 lL, 0.97 mmol, 4.5 equiv.). The mixture was
irradiated without cooling for 8 h. The solvent was evaporated
to give a yellow mixture of oil and solid. The crude material was
purified by flash chromatography (elution with 99 : 1 hexanes–
EtOAc) to yield erythro-4-deoxyglycal 22a as a clear oil (94.3 mg,
81%). For preparative scale synthesis, the reaction was carried
out with three parallel batches of alkynol 22a (158.1, 203.0, and
249.0 mg), which were then combined and purified to afford
479.8 mg of glycal 23a (79%).
TLC: Rf : 0.24 (4 : 1 hexanes–EtOAc). IR (NaCl, film): 3307,
1
2950, 2881, 1461, 1427, 1104, 1006, 816, 735, 695. H-NMR
(400 MHz): d 7.66 (m, 4H), 7.40 (m, 6H), 4.52 (m, 1H), 4.20 (m,
1H), 3.70 (m, 2H), 2.81 (d, 1H, J = 4.0), 2.44 (d, 1H, J = 2.1),
1.91–1.68 (m, 4H), 1.03 (s, 9H), 0.92 (t, 9H, J = 7.9), 0.60 (q,
6H, J = 7.9). 13C-NMR (100 MHz): d 135.8, 135.0, 133.8, 129.9,
127.9, 85.0, 73.1, 69.0, 61.3, 60.6, 44.2, 40.8, 27.1, 19.4, 7.1, 5.3.
ESI-MS m/z: (pos.) 497.3 [M + H]+, 519.3 [M + Na]+; (neg.)
495.3 [M − H]−, 531.0 [M + Cl]−.
[a]2D5 = +68.8 (c 1.0, CHCl3). TLC: Rf : 0.33 (9 : 1 hexanes–
EtOAc). IR (NaCl, film): 2941, 2864, 2360, 1636, 1241, 1111,
1088, 997, 882, 735, 701. 1H-NMR (400 MHz): d 7.64 (d, 4H, J =
6.8), 7.39 (m, 6H), 6.39 (d, 1H, J = 6.1), 4.88 (t, 1H, J = 4.9),
4.28–4.17 (m, 2H), 3.81 (m, 2H), 1.81 (m, 3H), 1.60 (m, 1H),
1.03 (m, 30H). 13C-NMR (100 MHz): d 145.9, 135.8, 134.3,
134.1, 129.8, 127.8, 104.1, 68.4, 60.6, 60.3, 38.7, 38.5, 27.1, 19.4,
18.4, 12.6. ESI-MS m/z: (pos.) 539.3 [M + H]+, 561.4 [M + Na]+;
(neg.) 573.5 [M + Cl]−.
(3R,5R)-7-(tert-Butyldiphenylsilyloxy)-5-(triethylsilyloxy)-3-
(triisopropylsilyloxy)hept-1-yne (38)
Triisopropylsilyl chloride (73 lL, 0.34 mmol, 1.5 equiv.) was
added dropwise to a solution of the propargylic alcohol 37
(113.1 mg, 0.23 mmol, 1.0 equiv.) and imidazole (31.3 mg,
0.46 mmol, 2.0 equiv.) in anhydrous DMF (2.3 mL, Sureseal
bottle). After 22 h at rt the reaction mixture was quenched with
sat. aq. NaHCO3. The aqueous layer was extracted with Et2O
(3 ×) and the combined extracts were washed with H2O and
brine, dried (MgSO4), filtered, and concentrated. The residue
was purified by flash chromatography (elution with 98 : 2
hexanes–EtOAc) to yield alkyne 38 as a clear oil (125.5 mg,
84%). For preparative scale synthesis, 9.4 g of crude alcohol 37
was converted to 14.0 g of crude alkyne 38, which was carried
on without further purification (see below).
Acknowledgements
We thank Professor Geoffrey W. Coates, Dr Christine M. Di-
Blasi, and Dr Young Ho Rhee for helpful discussions and
Dr George Sukenick, Anna Dudkina, Hui Fang, and Sylvia
Rusli (MSKCC Analytical Core Facility) for mass spectral
analyses. Financial support from the Tri-Institutional Training
Program in Chemical Biology, the William Randolph Hearst
Fund in Experimental Therapeutics, Mr William H. Goodwin
and Mrs Alice Goodwin and the Commonwealth Foundation
for Cancer Research, and the Experimental Therapeutics Center
of MSKCC is gratefully acknowledged.
TLC: Rf : 0.59 (9 : 1 hexanes–EtOAc). IR (NaCl, film): 3296,
2950, 2869, 1461, 1427, 1237, 1104, 1058, 1006, 885, 822, 735,
References and notes
1 S. J. Danishefsky and M. T. Bilodeau, Angew. Chem., Int. Ed. Engl.,
1996, 35, 1380–1419.
2 D. S. Tan, Comb. Chem. High Throughput Screening, 2004, 7, 631–
643.
3 D. S. Tan and S. L. Schreiber, Tetrahedron Lett., 2000, 41, 9509–9513.
1
695. H-NMR (400 MHz): d 7.67 (d, 4H, J = 6.9), 7.40 (m,
6H), 4.61 (m, 1H), 4.19 (m, 1H), 3.70 (m, 2H), 2.40 (d, 1H, J =
1.8), 1.91–1.69 (m, 4H), 1.13–1.02 (m, 30H), 0.92 (t, 9H, J =
7.9), 0.59 (q, 6H, J = 7.9). 13C-NMR (100 MHz): d 135.8, 134.1,
129.7, 127.8, 85.7, 73.2, 67.1, 61.3, 60.6, 46.5, 41.1, 27.0, 19.3,
8 0 2
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 7 9 8 – 8 0 3

View Article Online
4 H. Kubota, J. Lim, K. M. Depew and S. L. Schreiber, Chem. Biol.,
2002, 9, 265–276.
26 R. K. Boeckman, Jr., A. B. Charette, T. Asberom and B. H. Johnston,
J. Am. Chem. Soc., 1991, 113, 5337–5353.
5 S. Danishefsky and C. Maring, J. Am. Chem. Soc., 1985, 107, 1269–
27 V. Caprio, M. A. Brimble and D. P. Furkert, Tetrahedron, 2001, 57,
1274.
4023–4034.
6 DIBAL, 9-BBN, or LiPyrrBH3 (lithium pyrroloborohydride: J. C.
Fuller, E. L. Stangeland, C. T. Goralski and B. Singaram, Tetra-
hedron Lett., 1993, 34, 257–260); treatment with L-Selectride led to
decomposition.
7 B. W. Gung, Chem. Rev., 1999, 99, 1377–1386.
8 H. L. Goering and J. P. Blanchard, J. Am. Chem. Soc., 1954, 76,
5405–5408.
28 J. W. A. Kinnaird, P. Y. Ng, K. Kubota, X. Wang and J. L. Leighton,
J. Am. Chem. Soc., 2002, 124, 7920–7921.
29 A. B. Smith III, I. G. Safonov and R. M. Corbett, J. Am. Chem. Soc.,
2001, 123, 12426–12427.
30 L. Pu, Tetrahedron, 2003, 59, 9873–9886.
31 D. E. Frantz, R. Faessler and E. M. Carreira, J. Am. Chem. Soc.,
2000, 122, 1806–1807.
32 J. A. Marshall and M. P. Bourbeau, Org. Lett., 2003, 5, 3197–3199.
33 G. Gao, D. Moore, R.-G. Xie and L. Pu, Org. Lett., 2002, 4, 4143–
4146.
34 Large scale synthesis was facilitated by using crude products without
further purification when feasible. In the example cited, 50 mmol of
17 was converted to 14.8 mmol of 22a and column chromatography
was necessary only for compounds 18, 20, and 22a.
35 The reaction was most efficient when carried out in multiple parallel
200 mg batches. Increasing the batch size to 750 mg decreased the
yield to 58%.
9 W. G. Dauben, G. J. Fonken and D. S. Noyce, J. Am. Chem. Soc.,
1956, 78, 2579–2582.
10 Y. D. Wu, K. N. Houk and B. M. Trost, J. Am. Chem. Soc., 1987,
109, 5560–5561.
11 S. Danishefsky and M. E. Langer, J. Org. Chem., 1985, 50, 3672–
3674.
12 B. K. Shull, T. Sakai, J. B. Nichols and M. Koreeda, J. Org. Chem.,
1997, 62, 8294–8303.
13 J. S. Potuzak and D. S. Tan, Tetrahedron Lett., 2004, 45, 1797–1801.
14 M. D. Wittman, R. L. Halcomb, S. J. Danishefsky, J. Golik and D.
Vyas, J. Org. Chem., 1990, 55, 1979–1981.
15 Although deoxgenation of 11 at the C4-position would provide a
desired erythro-4-deoxyglycal, we did not pursue this strategy since
it would only provide access to the D-enantiomer.
16 W. Priebe and A. Zamojski, Tetrahedron, 1980, 36, 287–292.
17 L. V. Dunkerton, N. K. Adair, J. M. Euske, K. T. Brady and P. D.
Robinson, J. Org. Chem., 1988, 53, 845–850.
36 Rudimentary molecular modeling experiments indicated that the 23aꢀ
(4H5) conformation is, indeed, favored by approximately 1.6 kcal
mol−1 (MM2 minimization using CS Chem3D Pro v5.0).
37 R. L. Halcomb and S. J. Danishefsky, J. Am. Chem. Soc., 1989, 111,
6661–6666.
38 J. S. Yadav, B. V. S. Reddy, L. Chandraiah, B. Jagannadh, S. K. Kumar
and A. C. Kumar, Tetrahedron Lett., 2002, 43, 4527–4530.
39 R. W. Friesen and C. F. Sturino, J. Org. Chem., 1990, 55, 2572–
2574.
18 F. E. McDonald and K. S. Reddy, J. Organomet. Chem., 2001, 617–
618, 444–452.
19 F. E. McDonald, Chem. Eur. J., 1999, 5, 3103–3106.
20 F. E. McDonald, K. S. Reddy and Y. Diaz, J. Am. Chem. Soc., 2000,
122, 4304–4309.
40 A. Bhattacharjee and J. K. De Brabander, Tetrahedron Lett., 2000,
41, 8069–8073.
41 R. Zibuck, N. J. Liverton and A. B. Smith III, J. Am. Chem. Soc.,
1986, 108, 2451–2453.
21 F. E. McDonald and J. L. Bowman, Tetrahedron Lett., 1996, 37,
4675–4678.
42 E. B. Holson and W. R. Roush, Org. Lett., 2002, 4, 3719–3722.
43 E. B. Holson and W. R. Roush, Org. Lett., 2002, 4, 3723–3725.
44 B. B. Snider and F. Song, Org. Lett., 2001, 3, 1817–1820.
45 G. E. Keck and D. F. Kachensky, J. Org. Chem., 1986, 51, 2487–2493.
46 A. B. Pangborn, M. A. Giardello, R. H. Grubbs, R. K. Rosen and
F. J. Timmers, Organometallics, 1996, 15, 1518–1520.
22 F. E. McDonald, C. B. Connolly, M. M. Gleason, T. B. Towne and
K. D. Treiber, J. Org. Chem., 1993, 58, 6952–6953.
23 B. M. Trost and Y. H. Rhee, J. Am. Chem. Soc., 2003, 125, 7482–7483.
24 B. M. Trost and Y. H. Rhee, J. Am. Chem. Soc., 2002, 124, 2528–2533.
25 P. Wipf and T. H. Graham, J. Org. Chem., 2003, 68, 8798–8807.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 7 9 8 – 8 0 3
8 0 3