The cycloaddition way to novel deoxy disaccharide analogs

Article abstract of DOI:10.1016/S0040-4020(03)00654-9

A novel heterocycloaddition merges 2-thiono-3-ketolactones with carbohydrate glycals to afford materials which resemble disaccharides with an O-glycosidic linkage at the anomeric center and a thioether linking both C-2 and C-2′, thus creating a third hete

Full text of DOI:10.1016/S0040-4020(03)00654-9

Tetrahedron 59 (2003) 4249–4259
The cycloaddition way to novel deoxy disaccharide analogs
†
Maria M. Tamarez, Richard W. Franck and Aloma Geer
†
*
Department of Chemistry, Hunter College of CUNY, 695 Park Avenue, New York, NY 10021, USA
Received 16 January 2003; revised 17 April 2003; accepted 17 April 2003
Abstract—A novel heterocycloaddition merges 2-thiono-3-ketolactones with carbohydrate glycals to 0afford materials which resemble
disaccharides with an O-glycosidic linkage at the anomeric center and a thioether linking both C-2 and C-2 , thus creating a third heterocyclic
ring. Upon desulfurization, these novel cycloadducts afford materials which are models for 2-deoxydisaccharides. Studies with two keto
lactones and seven glycals are described. q 2003 Elsevier Science Ltd. All rights reserved.
1. Introduction
orthogonal to conventional glycosidation chemistry. At
about the time that we were developing the all-carbon
version of the Bradsher cycloaddition to glycals and
planning the hetero version, LeBlanc and Fitzimmons
elegantly realized the hetero concept in their synthesis of
2-amino glycosides (Eq. (3)).7,8 Our own plans were stalled
until Capozzi reported a key observation.9 Thus, decompo-
sition of phthalimidosulfenyl derivative 9 apparently
produced thione 10 which acted as a dienophile in reaction
with isoprene (Eq. (4)). Our collaborating groups then
showed that 10 and related materials could behave as
oxothiono dienes in reactions with glycals.10 The appli-
cation of this concept to a disaccharide analog synthesis
requires that the thionocarbonyl diene be incorporated into a
pyran or furan ring. The development of such heterodienes
and their cycloaddition behavior is described in this
article.11 A report by the Capozzi group describing a
‘true’ carbohydrate heterodiene has appeared and is nicely
complementary to the research described below.12
The chemistry of glycosyl transfer represents an enormous
research endeavor. The most frequently used methodologies
for forging the glycosidic linkage involve, generation of a
leaving group at the anomeric carbon of the glycosyl donor
which is then displaced by a nucleophilic group from the
glycosyl acceptor. The art and subtlety of this step are found
in a complicated algorithm which includes solvent, leaving
group, Lewis acid catalyst, and the influence of the
neighboring group at C-2.1 The extensive methodology
developed for solution chemistry is now being adapted for
solid supports.2
An important niche in the field is the synthesis of
2-deoxyglycosides, a linkage found in antibiotics.3 For
these materials, the glycosyl transfer algorithm must be
patched to take account of the absence4 of a neighboring
group at C-2 or more commonly, to have a group at C-2 that
can be reductively removed after its control features are no
longer needed.5 The work to be described had its origin in
our Bradsher cycloaddition program where a facile, inverse-
electron-demand cycloaddition with carbohydrate-derived
glycals or with synthetic glycal-like materials to form two
new C–C bonds led smoothly to natural product materials
(Eq. (1)).6 We speculated that an analogous heterocyclo-
addition with a sulfur/oxygen diene could produce glyco-
sidic materials that, upon subsequent desulfurization, would
produce 2-deoxy glycosides (Eq. (2)). The two-step
sequence would thus be a glycosyl transfer that would be
ð1Þ
ð2Þ
Keywords: desulfurization; tetronic acid; cycloaddition; glycosyl transfer.
*
Corresponding author. Tel.: þ1-2127725340; fax: þ1-2127725332;
e-mail: rfranck@hunter.cuny.edu
Taken in part from doctoral theses submitted by M. M. T. and A. G. to the
Graduate School of CUNY in 1999 and 1997, respectively.
†
0040–4020/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0040-4020(03)00654-9

4250
M. M. Tamarez et al. / Tetrahedron 59 (2003) 4249–4259
cycloaddition of heterodiene 15 with excess 12 as its enol 16
(Eq. (5)). A similar dimer has been reported by Cava in
work with dimedone.13
ð3Þ
The desired sulfenylated lactone 14 was obtained by an
inverse addition procedure where the lactone was added to
the sulfenylating agent which was kept at 2408. Although
the reaction appears to be quantitative, recrystallization of
the desired material 14 reduces the yield to about 40%. In
order to obtain a single enantiomer of the heterodiene
precursor, a synthesis of 20, the S-enantiomer of racemic
lactone 12 was developed using a combination of published
methods (Eq. (6)). Hence, Meldrum’s acid 17 was reacted
with diketene to afford the highly enolic 18 in greater than
90% yield.14 Then, a reduction with actively fermenting
yeast afforded the S-alcohol 19 which provided the desired
lactone 20 in about 50% yield for the two steps simply by
refluxing in MeOH.15 Using conditions developed for the
racemic lactone, 20 was phthalimidosulfenylated to produce
21. Similar sulfenylation conditions applied to com-
mercially available tetronic acid 22 afforded the
phthalimidothiofuranolactone 23 (Eq. (7)).
ð4Þ
ð5Þ
ð6Þ
2. Results and discussion
2.1. Synthesis of heterodiene precursors
ð7Þ
Ketolactone 12 is commercially available in racemic form.
Simply reacting it at 08 by addition of phthalimidosulfenyl
chloride 13 led to an incompletely characterized dimeric
material which might be rationalized as the product of a

M. M. Tamarez et al. / Tetrahedron 59 (2003) 4249–4259
4251
2.2. Cycloadditions
noted that the tetronic acid precursor spontaneously
formed heterodiene upon standing in solution at room
temperature. Solvent effects were cursorily examined in
the hexanolactone series, with an inconclusive result
shown in Eq. (9). There is little precedent on face
selectivity in 2-acetoxy glycals and our limited data does
not permit generalization. Adduct 37 was labile to traces
of acid in CDCl3, thus leading to unprotected adduct 38
which was not accessible directly by cycloaddition
(Eq. (13)).
The reactions were carried out simply by mixing
heterodiene precursor and glycal in solvent along with a
limited amount of lutidine and are illustrated in Eqs.
(8)–(18). Reaction rates varied over a wide range from
hours to days. A rough correlation with the HOMO–
LUMO gap can be made with smaller gaps correlating
with faster cycloaddition rates. The HOMO–LUMO gaps
in ev and approximate rates are displayed in each
equation. The MO values were calculated at the semi-
empirical AM1 level using either Spartan or Chem3D
platforms. Methyl ethers were used in place of benzyl in
the calculations. The structures of the products were
routinely assigned using proton NMR data. Observed J
values were consistent with the carbohydrate literature.
X-Ray crystallographic structural data for bottom-face
adduct 25 (R-diastereomer at the CH3) obtained from the
R,S mixture formed when racemic diene precursor 14 was
used, and X-ray data for top-face material 45 reinforced
our confidence in the NMR assignments. For the adducts
derived from 2-acetoxyglycal 39, with no proton at C-2,
hence no J1,2 or J2,3 values, the assignments are less
compelling. For example, our assignment of adducts 40
and 41 relies both on their displayed anomeric protons at
6.42 and 6.21, respectively, and the nOe between protons
1 and 3 in top-face isomer 41 (Eq. (14)). In adducts 28
and 29 whose structures are secure, the comparable
resonances are 5.71 and 5.39 (Eq. (9)). Thus, the axial
anomeric proton of the top-face adduct resonates upfield
of the equatorial anomeric proton of the bottom-face
adduct; and this behavior is consistent in other similar
comparisons. The upfield–downfield anomeric resonance
relationship holds in the tetronic acid series as well. Thus
for 44 and 45, the shifts are 5.78 and 5.41 (Eq. (15)). For
46 and 47, the data are 5.83 and 5.58 (Eq. (16)). Our
assignment difficulty arises in the case of the adduct of
glycal 39 with tetronic acid 23, where we have
conflicting data. The adduct’s anomeric proton resonance
is at 6.21, suggesting, by comparison to 40 and 41, a top-
face adduct 50. However, if we rely on a comparison of
its nOe behavior to that of 40 and 41 where the decisive
difference is the absence or presence of an interaction
between H-1 and H-3, we would reach a different
conclusion (Eq. (18)). Thus, the adduct 49 or 50 also
exhibits an absence of interaction between its H-1 and
H-3, so we should assign it as a bottom-face adduct 49
(although it is recognized that absence of an nOe is not as
compelling as presence). We rule out an assignment as
the regioisomeric adduct with anomeric sulfur because
the expected chemical shift for the anomeric H is
between d 5–6. It should be noted that the dominant
stereochemical outcomes are consistent with the literature
on face selectivity of attack of glycal double bonds,
bottom-face for glucals and galactals and top-face for
allals, with one exception. Thus, when the cycloaddition
of glucal triacetate and the tetronic acid heterodiene
precursor 23 was carried out in DMSO, we observed a
predominance of top-face attack producing adduct 47. A
survey of solvent effects was then carried out with the
tetronic acid precursor and tribenzyl glucal. No dramatic
effects were observed. However, the survey was carried
out in the absence of any added lutidine because it was
There are two features of selectivity, S vs O with C-1 and
C-2 and ketone carbonyl vs. lactone carbonyl that were
observed and expected. First, the ketone vs lactone
distinction: the observed product is an enol of a ketone
whereas, if the lactone carbonyl participated in the
addition process, the product would have been a ketene
acetal, a less stable entity. Second, the coefficient of the p
orbital on the ketone oxygen in the LUMO of the diene is
much larger than the corresponding coefficient of the
lactone. In fact, in one cycloaddition reaction pairing 21
and 27 (Eq. (9)), the NMR analysis of the crude
product revealed spectra of two closely related materials.
One set of peaks matched those of the isolated and
characterized material 28. The other set of peaks were
almost identical and suggested the possibility of an
adduct where the lactone carbonyl had participated in
competition with the ketone. Although the mixture
survived florisil chromatography intact and unseparated,
the material, conceivably regioisomeric to 28 did not
survive silica gel chromatography of the reaction product
mixture. The regioisomeric possibility where sulfur
would be attached to C-1 was dismissed because the
anomeric proton chemical shift was not significantly
shielded. With regard to the sulfur selectivity for C-2 of
the glycal, the p-orbital coefficients (and simple ‘electron-
pushing’ considerations) for the sulfur and C-2 of the
glycal are the largest and thus would be predicted to
begin bonding strongly as the reaction progresses. In one
cycloaddition, that of glycal 39 with 21, an additional
cycloadduct was isolated in 13% yield, with its anomeric
proton resonance at 5.71, upfield from the that of adducts
40 and 41. Hence, we tentatively assign it structure
42 (Eq. (14)) where the sulfur is now linked to C-1. This
reversal can be rationalized because the glycal
double bond now has oxygen substitution on both C-1
and C-2.
ð8Þ

4252
M. M. Tamarez et al. / Tetrahedron 59 (2003) 4249–4259
ð13Þ
ð9Þ
ð10Þ
ð14Þ
ð15Þ
ð11Þ
ð16Þ
ð12Þ

M. M. Tamarez et al. / Tetrahedron 59 (2003) 4249–4259
4253
ð17Þ
ð21Þ
ð18Þ
ð22Þ
ð19Þ
2.3. Desulfurizations
The removal of sulfur from our adducts is the ‘Achilles
Heel’ of our synthetic method, in this series and also in
2-thio-substituted glycosides described in earlier work from
these labs.16 Raney Nickel has proven to be the only reagent
that extrudes sulfur from our heterocycles. In addition to the
variability from batch to batch of the nickel catalyst, there is
a problem of elimination competing with reduction. Thus,
whatever the exact details, there is probably a NiH
intermediate depicted as 52 in Eq. (19). To obtain the
desired desulfurization, a reductive elimination of nickel
occurs. A competing elimination of nickel alkoxide would
produce the glycal starting material of our sequence; and
this is observed in every desulfurization reaction. Eqs.
(20)–(22) summarize three representative examples. In
some experiments with adduct 32, reduced and desulfurized
ð20Þ

4254
M. M. Tamarez et al. / Tetrahedron 59 (2003) 4249–4259
glycoside 54 was obtained accompanied by starting
tribenzyl galactal 27 and its reduced form 55 in 21%
combined yield (Eq. (21)). In the tetronic acid example, the
by-products included material 57 where the double bond of
56 is reduced (10% yield), starting glycal 24 in 30% yield
and reduced glycal (not shown) in 20% yield (Eq. (22)).
Other desulfurization reagents examined to no avail were
nickelocene/LAH17 and nickel boride.18
purchased from Aldrich and EM Science. All compounds
isolated by chromatography were sufficiently pure (.95%
by NMR analysis) for use in subsequent preparative
reactions.
4.1.1. 3-(1,3-Dioxoisoindolin-2-ylthio)-6-methyl-3,5,6-
trihydro-2H-pyran-2,4-dione (R/S)-(14). A stirred sol-
ution of phthalimidosulfenyl chloride (170.8 mg, 0.8 mmol)
in THF (8 mL) was cooled at 2408C and the racemic 5,6-
dihydro-4-hydroxy-6-methyl-2H-pyran-2-one 12 (50 mg,
0.4 mmol) dissolved in 2 mL of THF was added dropwise
in a period of 7 min. After 25 min all of the starting material
was consumed. Pentane (10 mL) was added and a white
precipitate was formed, the cooling bath was removed, and
the reaction mixture was allowed to warm to room
temperature without stirring. Filtration of this mixture
provided 122 mg of a white crude solid. Recrystallization
from CH2Cl2:hexane afforded pure compound (R/S)-(14)
(55.4 mg, 46%) as colorless crystals: Rf¼0.05 (9:1 CH2-
3. Conclusion
Novel tricyclic materials derived from glycal hetero-
cycloadditions have been prepared. Desulfurization of
these materials affords model disaccharide analogs. The
two-step sequence of cycloaddition–desulfurization may be
viewed as a glycosyl transfer method which is orthogonal to
conventional glycosidation chemistry.
1
Cl2:MeOH); H NMR d 1.53 (d, J¼6.3 Hz, 3H), 2.86 (dd,
J¼2.36, 16 Hz, 1H), 3.04 (dd, J¼11.33, 16 Hz, 1H), 5.02–
5.06 (m, 1H), 7.77–7.80 (m, 4Harom); 13C NMR d 21.1,
44.2, 72.2, 101.4, 124.7, 132.0, 135.3, 165.3, 167.9, 194.7.
4. Experimental
4.1. General information
1H and 13C NMR spectra were recorded on a 300 MHz
instrument, in most cases using CDCl3 (99.8%, 0.03 v/v
TMS) as solvent. Chemical shifts are reported in d units
with coupling constants reported in Hz. TMS and residual
4.2. (2R,3R,4R,8R)- and (8S,2R,3R,4R)-3,4-Bis-
(phenylmethoxy)-8-methyl-2-[(phenylmethoxy)methyl]-
2,3,4,8,9,4a-hexahydro-10aH,4aH-2H-pyrano[2,3-b]2H-
pyrano[3,4-e]1,4-oxathiin-6-one (25, R,S diastereomers
at CH3–C): general cycloaddition procedure
1
chloroform (d 7.26 for H, d 77.23 for 13C) were used as
internal references. IR spectra were recorded on a Perkin–
Elmer Model 1420 infrared spectrophotometer. High
resolution mass spectra were performed at University of
Illinois Mass Spectrometry Laboratory (Champaign-
Urbana) and medium resolution mass spectra were obtained
at the Hunter College Mass Spectrometry Laboratory.
Optical rotations were measured on a Rudolph Autopol III
polarimeter at the reported temperatures and concentrations.
Melting points were measured on a Fisher–Johns stage
melting point apparatus and are uncorrected. Elemental
analyses were performed by Robertson Microlit
Laboratories Inc., Madison, NJ and by Schwarzkopf
Microanalytical Laboratory, Inc., Woodside, NY.
The racemic sulfino lactone (R/S)-(14). (35.5 mg,
0.116 mmol) and tribenzylglucal 24 (48.4 mg,
0.116 mmol) were dissolved in CH2Cl2 (1 mL), and
pyridine (55.1 mg, 0.696 mmol) was added, and the solution
1
was stirred at room temperature for 5 days. H NMR of an
aliquot was taken. The ratio of the anomeric protons of the
product and the glucal was Hp:Hg¼1:1. The reaction was
left stirring 2 more days, but the ratio of glucal to product
remained the same. The reaction was worked up by
dissolving it in CH2Cl2, extracting with saturated aqueous
NH4Cl and back extracting the aqueous layer with CH2Cl2.
The organic extracts were washed twice with brine, dried
(Na2SO4), and concentrated in vacuo to give 95.2 mg of a
crude brown solid. Purification was done by flash chroma-
tography, eluting with petroleum ether:ethyl acetate (3:1),
and finally with pure ethyl acetate in order to remove the
product, 35.6 mg (53.5%, 100% based in the recovered
glucal), as a 1:1 mixture of diastereomers. This mixture was
separated by silica gel P-plate (1000 mm 20£10 cm) eluting
first with ethyl acetate then 5 times with petroleum
ether:ethyl acetate 3:1. This separation afforded two
diastereomers: upper spot 7.3 mg and lower spot 2.8 mg
both as white foamy solids.
All reactions were conducted in oven-dried or flame-dried
glassware under an atmosphere of dry argon or nitrogen. All
solvents were purified before use; THF was distilled from
sodium benzophenone ketyl; dichloromethane and chloro-
form were distilled from P2O5; triethylamine, diethylamine,
DMF, and pyridine were distilled from CaH2; and methanol
was distilled from magnesium turnings. Other solvents were
purified and dried using standard procedures. Phthalimide-
N-sulfenyl chloride,19 3,4,5-tri-O-benzyl-D-galactal,20 tri-
O-benzyl-D-allal,21 4,6-O-isopropylidene-D-glucal22 and
2-acetoxy-3,4,5-tribenzyl-D-glucal23 were prepared as
describe in the literature. 3,4,5-Tri-O-acetyl-D-galactal and
diketene, were purchased from Aldrich. Analytical TLC
was performed with the use of plates coated with a 0.25 mm
thickness of silica gel containing PF254 indicator (EM
Science) and short–long-wave ultraviolet light was used to
visualize the spots. Chromatotron Plates (radial chromato-
graphy) were prepared by using Kielesgel 60 PF254 (EM
Science). Flash chromatography was performed as
described by Still24 with Kieselgel 60 (230–400 mesh)
4.2.1. Compound 25, upper spot, R at CH3. Rf¼0.59 (1:3
1
ethyl acetate:hexane, eluted 5 times); mp 128–1308C; H
NMR d 1.47 (d, J¼6.3 Hz, 3H), 2.40 (dd, J¼3.6, 17.4 Hz,
1H), 2.48 (dd, J¼12.3, 17.4 Hz, 1H), 2.37 (dd, J¼2.8,
10.6 Hz, 1H), 3.67–3.85 (m, 4H), 4.04 (app d, J¼10 Hz,
1H), 4.42–4.50 (m, 1H), 4.57 (d, a part of an AB system,
JAB¼12 Hz, Dn¼34.65 Hz, 1H), 4.59 (d, a part of an AB
system, JAB¼10.2 Hz, Dn¼95.55 Hz, 1H), 4.68 (d,
JAB¼12.3 Hz, Dn¼34.65 Hz, 1H), 4.88 (d, a part of an

M. M. Tamarez et al. / Tetrahedron 59 (2003) 4249–4259
4255
AB system, JAB¼10.2 Hz, Dn¼29.85 Hz, 1H), 4.91 (d, a
part of an AB system, JAB¼9.3 Hz, Dn¼95.55 Hz, 1H),
4.84 (m, 1H); 13C NMR (DMSO-d6) d 20.3, 34.0, 71.5, 90.6,
167.1, 172.7; IR (KBr): 1675 cm21 (n CvO).
4.97 (d,
a
part of an AB system, JAB¼10.5 Hz,
Dn¼29.85 Hz, 1H), 5.69 (d, J¼2.7 Hz, 1H), 7.20–7.50
(m, 15Harom); 13C NMR d 20.5, 35.2, 42.9, 68.1, 72.8, 73.6,
73.8, 75.5, 76.3, 78.6, 94.7, 96.6, 128.1, 128.2, 128.4, 128.6,
128.7, 128.8, 137.6, 137.9, 159.4, 164.6; IR (thin film):
1703 cm21 (n CvO); ESMS m/z 592 (Mþ18).
4.2.5. (6S,12S,17S,18S,1R,10R)-18-Hydroxy-6,15,15-tri-
methyl-5,9,11,14,16-pentaoxa-2-thiatetracyclo[8.8.0.0-
k3,8l.0k12,17l]octadec-3(8)-en-4-one (37). Following
the general stepwise cycloaddition procedure, the enantio-
pure sulfino-lactone (S)-21 (0.339 g, 1.11 mmol), and
isopropylidene glucal 36 (0.206 g, 1.11 mmol) in CHCl3
(4 mL), were combined with 2,6-lutidine (18.9 mg,
1.11 mmol), and after stirring for 6 days, the reaction was
worked up to give 0.352 g of a light brown crude solid.
Purification of this crude by radial chromatography,
eluting with hexane:ethyl acetate (8:1 to 3:1), afforded
98.6 mg (79% based on the reacted glucal) of compound 37:
Rf¼0.12 (1:1 hexane:ethyl acetate); 1H NMR d 1.44 (s, 3H),
1.50 (d, J¼6.3 Hz, 3H), 1.55 (s, 3H), 2.50 (dd, J¼4.2,
17 Hz, 1H), 2.73 (dd, 11.1, 17 Hz, 1H), 3.2 (dd, J¼3, 10 Hz,
1H), 3.8 (d, J¼2.4 Hz, 1H), 3.65–3.93 (m, 5H), 4.11–4.68
(m, 1H), 5.7 (d, J¼3 Hz, 1H). Anal. calcd for C15H20O7S:
C, 52.32; H, 5.85; S, 9.31. Found: C, 52.15; H, 5.80; S,
9.05.
4.2.2. Compound 25, lower spot, S at CH3. Rf¼0.48 (1:3
ethyl acetate:hexane, eluted 5 times); mp 44–478C;
[a]D¼þ128.88 (c¼2.56, CHCl3); 1H NMR d 1.46 (d,
J¼6.3 Hz, 3H), 2.48 (dd, J¼4.5, 17 Hz, 1H), 2.62 (dd,
J¼10.8, 17 Hz, 1H), 3.29 (dd, J¼3, 10.8 Hz, 1H), 3.55 (dd,
J¼9, 10.5 Hz, 1H), 3.68–3.82 (m, 3H), 3.96 (app dt, J¼3,
10.1 Hz, 1H), 4.52–4.69 (m, 4H), 4.79 (d, JAB¼10.2 Hz, a
part of an AB system, 1H), 4.98 (d, J¼10.2 Hz, a part of an
AB system, 1H), 5.72 (d, J¼3 Hz, 1H), 7.15–7.44 (m,
15Harom); 13C NMR d 21.7, 36.0, 43.3, 69.3, 73.2, 74.7,
74.8, 76.5, 78.4, 79.3, 95.1, 98.1, 102.4, 129.0, 129.1, 129.3,
129.6, 138.8, 138.9, 161.3, 164.85; IR (thin film):
1704 cm21 (n CvO); ESMS m/z 592 (Mþ18). Anal.
calcd for C33H3407S: C, 68.97; H, 5.96; S, 5.58. Found: C,
68.96, H, 5.96; S, 5.70.
4.2.6. (3S,4S,8S,2R,10aR)-3,4-Dihydroxy-2-(hydroxy-
methyl)-8-methyl-2,3,4,8,9,4a-hexahydro-2H-pyr-
ano[2,3-b]2H-pyrano[3,4-e]1,4-oxathiin-6-one (38). The
isopropylidene adduct 37 was left sitting in the refrigerator
in CDCl3 inside an NMR tube for 3 days, at which time the
original clear solution became cloudy. Filtration of this
cloudy solution afforded the trihydroxy adduct, 38, as a
white powder: Rf¼0.01 (neat ethyl acetate); mp 212–2138C
(decomposed); 1H NMR in DMSO-d6 d 0.47 (d, J¼6.3 Hz,
3H), 1.64 (t, J¼1.8 Hz, 2H), 1.75 (d, J¼6.9 Hz, 2H), 2.25–
2.83 (m, 4H), 3.76–3.84 (m, 2H), 4.4 (d, J¼5.1 Hz, 1H),
4.63 (t, J¼2.4 Hz, 1H), 4.83 (d, J¼2.7 Hz, 1H); 1H NMR in
acetone-d6: d 1.40 (d, J¼6.3 Hz, 3H), 2.60 (dd, J¼4.8,
16.8 Hz, 1H), 2.72 (dd, J¼9.6, 16.8 Hz, 1H), 2.84 (br s, 1H),
3.13 (dd, J¼2.7, 9.9 Hz, 1H), 3.47–3.58 (m, 2H), 3.60–3.84
(m, 3H), 4.49 (br s, 1H), 4.50–4.69 (m, 1H), 4.79 (br s, 1H),
5.69 (d, J¼2.7 Hz, 1H); 13C NMR DMSO-d6: d 20.1, 33.7,
42.3, 60.4, 69.1, 70.6, 71.9, 75.5, 92.3, 96.4, 160.5, 163.5;
IR (KBr): 1672 (n CvO); ESMS m/z 322 (Mþ18, 100),
HRMS calcd for C12H1607S 305.0695 (MþH). Found:
305.0689 (MþH).
4.2.3. 5-((3S)-1,3-Dihydroxybutylidene)-2,2-dimethyl-
1,3-dioxane-4,6-dione [(S)-19]. Dried Fleischman’s yeast
(75 g), sucrose (37.5 g), and ‘tap’ water (150 mL) were
divided into five portions and placed in 5£500 mL
Erlenmeyer flasks and shaken for 24 h, at which time the
yeast suspension expanded to entirely fill the flasks.
Acetoacetylated Meldrum’s acid 18 (1.14 g, 5 mmol),
divided in five 0.228 g portions was placed in each flask
and shaken for 1 day. Most of the water was evaporated.
Celite (10 g) and NaCl (5 g) were placed in each flask and
mixed until became a paste. This paste was extracted 6 times
with CH2Cl2 by vigorous mechanical stirring, followed by
decantation, starting with 150 mL of CH2Cl2 and the
remaining five extractions with 100 mL. The combined
organic extracts were dried (MgSO4), and concentrated in
vacuo to give 1.42 g (100%) of the crude chiral alcohol
(S)-19 as
a
gold oil: Rf¼0.50 (CH2Cl2:MeOH);
[a]D¼þ41.808 (c¼0.55, EtOH), (lit.43 [a]D¼þ50.5,
1
c¼2.05, EtOH); H NMR d 1.3 (d, J¼6.3 Hz, 3H), 1.7 (s,
6H), 3.1 (dd, J¼3.9, 13.5 Hz, 1H), 3.3 (dd, J¼8.4, 13.5 Hz,
1H), 4.3 (m, 1H).
4.2.7. ((8S,2R,3R,4R)-3-Acetyloxy-2-(acetyloxymethyl)-
8-methyl-6-oxo-2,3,4,8,9,4a-hexahydro-10aH,4aH-2H-
pyrano[2,3-b]2H-pyrano[3,4-e]1,4-oxathiin-4-yl acetate
(28). Following the stepwise cycloaddition procedure, a
solution of 0.200 g (0.65 mmol) of thioketolactone 21, tri-
O-acetyl glucal 27 (0.177 g, 0.65 mmol) in chloroform
(2.5 mL), and 2,6-lutidine (0.65 mmol, 69.7 mg) was stirred
for 48 h to provide 0.354 g of a crude solid which after flash
chromatography with florisil, eluting with petroleum
ether:ethyl acetate (4:1 to 2:1) afforded compound 28
(0.139 g, 53%) contaminated with an unstable material that
converted to 28 on silica (and guessed to be an adduct with
4.2.4. (S)-6-Methyl-3,5,6-trihydro-2H-pyran-2,4-dione-
(20). The enantiopure alcohol (S)-19, (1.91 g, 8.29 mmol)
was dissolved in dried MeOH (30 mL) and stirred under
reflux. After 2.5 h, the starting material was consumed as
indicated by TLC, and the reaction mixture was cooled to
room temperature and concentrated in vacuo to provide
1.10 g of crude. Purification of this crude by flash column
chromatography using florisil and eluting with CH2Cl2:
MeOH (20:1) afforded 0.600 g (56%) of the enantiopure
lactone (S)-19 as a white granular solid: Rf¼0.53 (9:1
CH2Cl2:MeOH); mp 127–1298C [a]25¼þ150.748 (c¼2.1,
1
the lactonic rather than the ketonic oxygen). H NMR d
1
EtOH), lit.43 [a]29¼þ153.48, c¼2.1, EtOH); H NMR d
1.48–1.51 (m, 6H), 2.02–2.1 (m, 18H), 2.54–2.67 (m, 4H),
3.34–3.38 (m, 2H), 4.1–4.31 (m, 6H), 4.6–4.8 (m, 2H),
5.11–5.21 (m, 4H), 5.71 (d, J¼2.14 Hz, 1H), 5.79 (d,
J¼2.52 Hz). Anal. calcd for C18H22O10S: C, 50.23; H, 5.15;
S, 7.45. Found: C, 50.35; H, 5.19; S, 7.44.
1.53 (d, J¼6.3 Hz, 3H), 2.46 (dd, J¼11.1, 18.3 Hz, 1H),
2.73 (dd, J¼2.7, 18.3 Hz, 1H), 3.44 (d, a part of an AB
system, JAB¼18.9 Hz, Dn¼38.25 Hz, 1H), 3.57 (d, a part of
an AB system, JAB¼18.9 Hz, Dn¼38.25 Hz, 1H), 4.76–

4256
M. M. Tamarez et al. / Tetrahedron 59 (2003) 4249–4259
4.2.8. ((8S)-3,4-Diacetyloxy-8-methyl-6-oxo-2,3,4,8,9,4a-
hexahydro-10aH-2H-pyrano[2,3-b]2H-pyrano[3,4-e]1,4-
oxathiin-2-yl)methyl acetate [28 (below face)] and [29
(above face)]. Following the stepwise cycloaddition
procedure a solution of 0.879 g (2.88 mmol) of the
thioketolactone 21, tri-O-acetyl glucal 27 (0.784 g,
2.88 mmol) in DMSO (12 mL), and 2,6-lutidine (0.309 g,
2.88 mmol) was stirred for 11 days to provide 0.192 g of
crude. Flash column chromatography of this crude, eluting
with hexane:ethyl acetate (3:1 to 2:1), afforded 63 mg (5%)
of the below face adduct, compound 28: Rf¼0.4 (1:1
hexane:ethyl acetate), and 88.2 mg (7%) of the top face
adduct, compound 29: Rf¼0.34 (1:1 hexane:ethyl acetate);
J¼2.42, 11.12 Hz, 1H), 3.56 (app d, J¼5.6 Hz, 2H), 3.83
(dd, J¼2.83, 11.1 Hz, 1H), 3.96 (app d, J¼1.20 Hz, 1H),
4.42 (d, a part of an AB system, JAB¼11.81 Hz, 1H), 4.45
(d, a part of an AB system, JAB¼11.81 Hz, 1H), 4.57–4.62
(m, 2H), 4.74 (s, 2H), 4.95 (d, a part of an AB system,
JAB¼11.2 Hz, 1H), 5.73 (d, J¼2.79 Hz, 1H), 7.27–7.41 (m
15Harom); 1HNMR (Benzene-d6): d 0.84 (d, J¼6.3 Hz,
3H), 1.57 (dd, J¼4.5, 16.65 Hz, 1H), 1.76 (dd, J¼10.5,
16.65 Hz, 1H), 3.38 (dd, J¼2.7, 11.1 Hz, 1H), 3.63 (dd,
J¼5.4, 9 Hz, 1H), 3.71–3.79 (m, 3H), 3.94 (app d,
J¼0.9 Hz, 1H), 4.09 (t, J¼6.6 Hz, 1H), 4.24 (AB q,
JAB¼12 Hz, Dn/J¼2.4 Hz, 2H), 4.57 (d, a part of an AB
system, JAB¼11.1 Hz, Dn¼135 Hz, 1H), 4.66 (d, J¼12 Hz,
2H), 5.02 (d, a part of an AB system, JAB¼11.1 Hz,
Dn¼135 Hz, 1H), 5.4 (d, J¼2.7 Hz, 1H), 7.05–7.45 (m,
15H); 13C NMR d 20.8, 35.0, 38.6, 68.8, 72.2, 72.6, 73.1,
73.4, 73.7, 73.9, 75.2, 77.4, 77.8, 77.9, 94.2, 97.8, 128.1,
128.2, 128.4, 128.5, 128.6, 128.7, 128.8, 137.8, 138.3,
159.9, 164.0; IR (thin film): 1703 (n CvO) cm21; ESMS m/
z: 592 (Mþ1). Anal. calcd for C33H34O7S: C, 68.97; H,
5.96; S, 5.58. Found: C, 69.08; H, 6.01.
1
mp 88–908C; H NMR d 1.44 (d, J¼6.3 Hz, 3H), 2.05 (s,
3H), 2.085 (s, 3H), 2.09 (s, 3H), 2.49 (dd, J¼4.5, 17.1 Hz,
1H), 2.61 (dd, J¼10.8, 17.1 Hz, 1H), 3.83–3.87 (m, 2H),
4.22–4.23 (m, 2H), 4.24–4.29 (m, 1H), 4.62–4.69 (m, 1H),
5.39 (d, J¼1.5 Hz, 1H), 5.48 (dd¼t, JAøJB¼9.3 Hz, 1H);
13C NMR d 20.8, 21.0, 22.9, 31.9, 34.7, 41.1, 65.3, 71.9,
72.3, 74.4, 77.4, 92.1, 96.8, 156.9, 163.9, 169.45, 170.08; IR
(thin film): 1750 (n CvO, –OC(O)CH3), 1702 (n CvO,
lactone, cm21; ESMS m/z 448 (Mþ18, 100), 431 (Mþ1,
11). Anal. calcd for C18H22O10S: C, 50.23; H, 5.15; S, 7.45.
Found: C, 50.22; H, 5.06; S, 7.20.
4.2.11. (2S,5aS,6R,7R,9aR)-6,7-Bis(phenylmethoxy)-2-
methyl-8-[(phenylmethoxy)methyl]-1,2,3,6,7,8,9,5a,9a-
nonahydro-3-oxaphenoxathiin-4-one (35). Following the
general stepwise cycloaddition procedure, the sulfino-
lactone 21 (155.8 mg, 0.51 mmol), tribenzyl allal 34
(212.16 mg, 0.51 mmol) in CHCl3 (5.0 mL), were combined
with 2,6-lutidine (54.7 mg, 0.0.51 mmol), and stirred for
1.5 h, upon which time, more sulfino-lactone (155.8 mg
0.51 mmol) was added. The stirring was continued for an
additional 0.5 h. The reaction was worked up, and the
phthalimide was removed to provide 275.2 mg of a very
pure crude solid. Filtration of this crude through a short pad
of silica gel, afforded compound 35 (265.2 mg, 91%) as an
off-white solid: Rf¼0.56 (3:1 hexane:ethyl acetate); 1H
NMR d 1.39 (d, J¼6.3 Hz, 3H), 2.26 (dd, J¼4.0, 16.95 Hz,
1H), 2.45 (dd, J¼11.7, 16.95 Hz, 1H), 3.45 (dd, J¼1.5,
4.95 Hz, 1H), 3.61 (app dd, J¼3.6, 4.5 Hz, 2H), 3.74 (dd,
3.0, 4.95 Hz, 1H), 4.02 (dd, J¼3.6, 8.4 Hz, 1H), 4.29 (dt,
2JA¼JB, J¼4.2, 8.4 Hz, 1H), 4.36–4.50 (m, 5H), 4.6 (d,
J¼12 Hz, 2H), 5.6 (d, J¼1.5 Hz, 1H), 7.13–7.34 (m, 15H);
13C NMR d 20.7, 35.1, 39.7, 69.7, 72.0, 72.3, 73.1, 73.7,
76.6, 77.0, 93.5, 97.6, 127.2, 127.9, 128.2, 128.7, 137.7,
4.2.9. (3S,8S,2R,4R)-3-Acetyloxy-2-(acetyloxymethyl)-8-
methyl-6-oxo-2,3,4,8,9,4a-hexahydro-10aH,4aH-2H-
pyrano[2,3-b]2H-pyrano[3,4-e]1,4-oxathiin-4-yl acetate
(31). Following the general ‘one pot’ cycloaddition
procedure, a solution of sulfenyl chloride (0.253 g,
1.2 mmol) and the (S)-lactone 21 (0.101 g, 0.79 mmol) in
THF, was combined with tri-O-acetyl galactal 30 (0.215 g,
0.79 mmol), and 2,6-lutidine (128.6 mg, 1.2 mmol). After
stirring for 6 days, the reaction mixture was worked up to
provide 0.452 g of a dark brown crude solid. Purification of
this crude by radial chromatography, eluting with hexane:
ethyl acetate (5:1 to 3:1) gave 0.136 g (40%) of compound
31 as a clear thick oil: Rf¼0.13 (2:1 hexane:ethyl acetate);
1H NMR d 1.45 (d, J¼6.3 Hz, 3H), 2.01 (s, 3H), 2.04 (s,
3H), 2.14 (s, 3H), 2.53 (dd, J¼4.8, 17.1 Hz, 1H), 2.64 (dd,
J¼9.6, 17.1 Hz, 1H), 3.61 (dd, J¼2.7, 11.7 Hz, 1H), 4.12 (d,
J¼6.6 Hz, 2H), 4.38 (t, J¼6.6 Hz, 1H), 4.61–4.68 (m, 1H),
4.95 (dd, J¼3, 12 Hz, 1H), 5.40 (app d, J¼1.8 Hz, 1H), 5.75
(d, J¼2.7 Hz, 1H); 13C NMR d 20.6, 20.7, 20.8, 34.6, 36.4,
61.6, 65.8, 67.2, 69.3, 72.5, 77.4, 96.4, 124.5, 135.1, 158.9,
169.9, 170.0, 170.5; ESMS m/z 448 (Mþ18, 100), 431
(MþH, 50). Anal. calcd for C18H22O10S: C, 50.23; H, 5.15;
S, 7.45. Found: C, 50.06; H, 5.25.
138.2, 157.9, 164.2; IR (thin film): 1701 (n CvO) cm21
;
ESMS m/z 592 (Mþ18). Anal. calcd for C33H34O7S: C,
66.87; H, 6.12. Found: C, 66.70; H, 6.55.
4.2.12. (8S)-3,4-Bis(phenylmethoxy)-8-methyl-6-oxo-2-
[(phenylmethoxy)methyl]-2,3,4,8,9,4a-hexahydro-10aH-
2H-pyrano[2,3-b]2H-pyrano[3,4-e]1,4-oxathiin-4a-
ylacetate (40 and 41) and (2S,4S,7S,3R,10aR,4aR)-3,4-
bis(phenylmethoxy)-7-methyl-9-oxo-2-[(phenyl-
methoxy)-methyl]-2,3,4,6,7,4a-hexahydro-2H-pyrano-
[3,2-b]2H-pyrano[3,4-e]1,4-oxathiin-4a-yl acetate (42).
Following the general ‘one pot’ cycloaddition procedure,
sulfenyl chloride (1.41 g, 6.58 mmol) in THF (35.8 mL),
was combined with the chiral lactone 21 (0.421 g,
3.29 mmol) in THF (32.9 mL), the glucal 39 (1.56 g,
3.29 mol) in THF (10.9 mL) and 2,6-lutidine (0.762 mL,
6.58 mmol) and after stirring for 10 days at room
temperature, the reaction was worked up to give 2.46 g of
crude solid. Removal of the phthalimide from this crude
4.2.10. (8S)-3,4-Bis(phenylmethoxy)-8-methyl-2-
[(phenylmethoxy)methyl]-2,3,4,8,9,4a-hexahydro-10aH-
2Hpyrano[2,3-b]2H-pyrano[3,4-e]1,4-oxathiin-6-one
(33). Following the general cycloaddition procedure, a
solution of sulfenyl chloride (668.4 mg, 3.13 mmol), and the
(S)-lactone 21 (267.5 mg, 2.09 mmol) in CHCl3 was
combined with tri-O-benzylgalactal 32 (665.6 mg,
1.6 mmol) and 2,6-lutidine (2.09 mmol, 224 mg). The
reaction was stirred 17 h to provide upon work up and
removal of the phthalimide, 0.658 g of a crude oily residue.
Flash column purification of this crude afforded 255 mg
(28%) of compound 33 as a single regioisomer: Rf¼0.16
(3:1 petroleum ether:ethyl acetate); 1H NMR d 1.4 (d,
J¼6.4 Hz, 3H), 2.4 (app d, J¼8.1 Hz, 2H), 3.40 (dd,

M. M. Tamarez et al. / Tetrahedron 59 (2003) 4249–4259
4257
using NaOH gave 1.64 g of dark brown crude syrup which
1H NMR showed three different regioisomers, represented
by the anomeric protons at 5.71, 6.21 and 6.41 ppm in a ratio
1.23:1.9:1, respectively. Separation by flash column
chromatography using silica gel, eluting with hexane:
ethyl acetate (2:1) gave 0.91 g of the three diastereomers
(44% combined yield), as a clear oil. This 0.91 g
corresponded to 0.389 g of the anomer at 6.21 ppm,
0.149 g of the anomer at 6.42 ppm, 0.187 g of the anomer
at 5.71 ppm, and 0.184 g of an inseparable mixture of the
three diastereomers.
amounts of 23 of about 80% purity in 67% yield. 1H NMR
(300 MHz, CDCl3) d10.4 (bs, 1H), 8.0–7.77 (m, 4H), 4.76
(s, 2H).
4.3. Typical procedure for cycloaddition of glycals with
23 as heterodiene precursor
To a chloroform (3 mL) solution of the glycal (0.25 mmol)
was added phthalimide sulfenyltetronic acid 24 (0.50 mmol)
and lutidine (29 mL, 0.25 mmol). The reaction mixture was
stirred at room temperature until total consumption of the
starting glycal was indicated by proton NMR of an aliquot
of the reaction mixture. The time for completion of the
reaction was usually not more than 24 h. At this time the
reaction mixture was quenched with an aqueous solution of
ammonium chloride and extracted with dichloromethane
(3£10 mL). The combined organic extracts were dried over
sodium sulfate and concentrated to give a crude material
which was purified by radial chromatography (ethyl
acetate:petroleum ether).
1
4.2.13. Compound 42. (H1 at 5.71 ppm). H NMR d 1.37
(d, J¼6.3 Hz, 3H), 1.93 (dd, J¼11.7, 17.4 Hz, 1H), 2.15 (s,
3H), 2.3 (dd, J¼4.2, 17.4 Hz, 1H), 3.65–3.70 (m, 4H), 3.92
(d, J¼8.4 Hz, 1H), 4.40–4.55 (m, 1H), 4.46 (d, J¼12 Hz,
1H), 4.51 (d, J¼10.8 Hz, 1H), 4.63 (d, J¼12 Hz, 1H), 4.46
(d, J¼12 Hz, 1H), 4.63 (d, J¼12 Hz, 1H), 4.73 (d, J¼12 Hz,
1H), 4.99 (d, J¼10.5 Hz, 1H), 4.99 (d, J¼12.3 Hz, 1H), 5.72
(s, 1H), 7.2–7.36 (m, 15Harom), 13C NMR d 20.6, 21.2,
30.9, 68.1, 73.4, 73.8, 75.5, 75.9, 85.8, 92.6, 104.7, 127.4,
127.9, 128.0, 128.0, 128.1, 128.3, 128.6, 128.7, 137.8,
138.1, 138.3, 161.5, 169.3; ESMS m/z 650 (MþNHþ4 ).
4.3.1. 30,40,6t-Tri-O-benzyl-D-glucosyl-below-face
cycloadduct 44. Yield 61.7%, [a]2D5¼þ120.08 (CDCl3);
1H NMR (300 MHz, CDCl3) 7.35–7.15 (m, 15H), 5.78 (d,
J¼2.54 Hz, 1H), 4.93–4.51 (m, 10H), 4.03–3.98 (m, 1H),
3.85–3.56 (m, 4H), 3.32 (dd, J¼10.91, 2.61 Hz, 1H). 13C
NMR (300 MHz, CDCl3) d 169.3, 166.8, 137.7, 128.7,
128.3, 128.1, 98.8, 93.2, 78.5, 77.3, 75.6, 74.1, 73.8, 68.0,
67.3, 42.0. Anal. calcd for C31H3007S: C, 68.11; H, 5.53; S,
5.87. Found: C, 68.01; H, 5.29; S, 5.90.
4.2.14. Compound 41. (H1 at 6.21 ppm). White foamy
solid; mp 47–498C; 1H NMR d 1.45 (d, J¼6.3 Hz, 3H), 2.08
(s, 3H), 2.50 (d, J¼4.8 Hz, 1H), 2.53 (d, J¼10.5 Hz, 1H),
3.72–3.81 (m, 4H), 4.03 (app d, J¼6.9 Hz, 1H), 4.50 (d,
J¼12.3 Hz, 1H), 4.6 (d, J¼12 Hz, 1H), 4.6–4.67 (m, 1H),
4.78 (d, J¼10.5 Hz, 1H), 4.80 (d, 10.8 Hz, 2H), 5.1 (d,
J¼10.2 Hz, 1H), 6.20 (s, 1H) 7.15–7.40 (m, 15Harom); 13C
NMR d 20.8, 21.8, 34.7, 68.4, 72.7, 73.6, 74.1, 75.4, 76.0,
80.2, 82.7, 95.0, 98.0, 127.6, 127.9, 128.1, 128.5, 128.5,
128.6, 128.8, 137.6, 137.9, 138.1, 157.3, 163.1, 169.6;
ESMS m/z 650 (MþNH4). Anal. calcd for C35H36O9S: C,
66.44; H, 5.73; S, 5.07. Found: C, 66.51; H, 5.87; S, 4.91; IR
4.3.2. 30,40,60-Tri-O-benzyl-D-glucosyl-above-face
1
cycloadduct 45. Yield 10.3%, [a]2D5¼þ76.88 (CDCl3); H
NMR (300 MHz, CDCl3) d 7.36–7.16 (m, 15H, –OBn),
5.41 (d. J¼1.22 Hz, 1H, Cl0 –H), 4.84–4.48 (m,8H,
–OCH2–Ph, –OCH2–), 4.05 (q, J¼8.34, 3.89 Hz, 1H,
C30 –H), 3.98–3.91 (m, 2H, C40 –H, C50 –H), 3.81 (dd,
J¼3.31, 1.12 Hz, 1H, C20 –H), 3.71 (d, J¼2.25 Hz, 2H,
C60 –H, C600 –H). 13C NMR (300 MHz, CDCl3), d 169.7,
164.2, 138.0, 137.9, 137.0, 128.8, 128.6, 128.3, 128.0, 97.6,
94.6, 79.9, 78.0, 77.2, 75.4, 73.8, 73.5, 71.4, 69.0, 67.9,
41.3. Anal. calcd for C31H3007S: C, 68.11; H, 5.53; S, 5.87.
Found: C, 67.96; H, 5.39; S, 6.00.
(thin film): 1760 (n CvO), 1706 (n CvO) cm21
.
1
4.2.15. Compound 40. (H1 at 6.42 ppm): H NMR d 1.38
(d, J¼6.3 Hz, 3H), 2.02 (dd, J¼11.7, 17.3 Hz, 1H), 2.13 (s,
3H), 2.31 (dd, J¼3.9, 17.4 Hz, 1H), 3.63–3.76 (m, 3H),
3.95 (app dd, J¼2.1, 9.9 Hz, 1H), 4.24 (d, J¼9.3 Hz, 1H),
4.48 (d, J¼12.3 Hz, 1H), 4.53 (d, J¼10.8 Hz, 1H), 4.5–4.6
(m, 1H), 4.64 (d, J¼12.3 Hz, 1H), 4.76 (d, 12.3 Hz, 1H),
4.80 (d, J¼10.8 Hz, 1H), 4.97 (d, J¼12 Hz, 1H), 6.42
(s, 1H), 7.13–7.44 (m, 15Harom); 13C NMR d 20.6,
21.2, 31.1, 68.3, 73.6, 73.9, 74.3, 75.7, 75.8, 77.0, 77.4,
81.6, 93.4, 103.6, 127.4, 127.9, 128.0, 128.2, 128.4, 128.7,
137.9, 138.2, 138.5, 161.9, 169.0; ESMS m/z 650
(MþNH4).
4.3.3. 30,40,60-Tri-O-acetyl-D-glucosyl-below-face
cycloadduct 46. Yield 9.7%, [a]2D5¼þ186.08 (CDCl3); H
1
NMR (300 MHz CDCl3) d 5.83 (d, J¼2.55 Hz, 1H), 5.18–
5.14 (m, 2H), 4.79, (dd, J¼15.97, 11.28 Hz, 2H), 4.35–4.16
(abq, J¼12.23, 4.42 Hz, 2H), 4.27–4.42 (m, 1H), 3.42 (dd,
J¼2.55 Hz, 1H), 2.12, 2.08, 2.03 (s£3, 9H). 13C NMR
(300 MHz, CDCl3) 170.7, 169.9, 169.6, 168.7, 166.1, 134.5,
97.6, 93.5, 71.0, 68.9, 67.8, 61.6, 39.9, 20.9, 20.7. Anal.
calcd for C16H18010S: C, 47.76; H,4.47; S, 7.97. Found: C,
47.78; H, 4.61; S, 7.64.
4.2.16. Preparation of phthalimidesulfenyltetronic acid,
23. To a solution of phthalimide-N-sulfenyl chloride
(1.0 mmol) in THF (5 mL) under a nitrogen atmosphere at
2208C (by internal thermometer), was added over a 15 min
period via syringe a solution of tetronic acid 22 (1.0 mmol)
in THF (5 mL) also prepared under a nitrogen atmosphere.
After 20 min at 2208C the cooling bath was removed and
pentane (20 mL) was added. The reaction mixture was
warmed to room temperature and the solid material was
removed by vacuum filtration and was confirmed to be the
desired 23 by NMR. This pure sample was obtained in 22%
yield. The filtrate was concentrated to provide additional
4.3.4. 30,40,60-Tri-O-acetyl-gluco-above-face cycloadduct
47. Yield 58.3%, [a]2D5¼þ29.88 (CDC13); 1H NMR
(300 MHz, CDCl3) d 5.58 (d, J¼1.22 Hz, 1H), 5.46 (t,
J¼9.47 Hz, 1H), 5.34 (dd, J¼9.28, 4.45 Hz, 1H), 4.78 (s,
2H), 4.24 (d, J¼3.95 Hz, 2H), 3.95–3.91 (t, J¼3.95 Hz,
1H), 3.88 (dd, J¼4.32, 1.06 Hz, 1H), 2.11, 2.08, 2.05 (s £3,
9H, –OAc). 13C NMR (300 MHz, CDCl3) d 170.8, 169.9,
169.4, 169.0, 163.9, 95.4, 93.7, 74.8, 71.6, 67.8, 65.0, 62.2,

4258
M. M. Tamarez et al. / Tetrahedron 59 (2003) 4249–4259
41.0, 20.9, 20.8, 20.7. Anal. calcd for C16H18010S: C, 47.76;
H, 4.47; S, 7.97. Found: C, 47.97; H, 4.27; S, 8.16.
3.65 (app t, J¼9 Hz, 1H), 3.84 (app t, J¼6.3 Hz, 1H), 3.92
(ddd, J¼2.1, 4.5, 13.95 Hz, 1H), 3.99 (s, 1H), 4.39 (d, a part
of an AB system, JAB¼11.4 Hz, Dn¼19.8 Hz, 1H), 4.46 (d,
a part of an AB system, JAB¼11.7 Hz, Dn¼19.8 Hz, 1H),
4.46–4.58 (m, 1H), 4.61 (d, a part of an AB system,
JAB¼11.1 Hz, Dn¼97.65 Hz, 1H), 4.66 (s, JAB¼0, 2H),
4.93 (d, a part of an AB system, J¼11.4 Hz, Dn¼97.65 Hz,
1H), 5.49 (d, J¼1.8 Hz, 1H), 5.63 (d, J¼2.7 Hz, 1H), 7.22–
7.38 (m, 15H); 13C NMR d 20.7, 30.6, 34.7, 69.0, 70.8, 72.1,
72.3, 72.6, 73.8, 74.1, 74.8, 94.4, 97.4, 127.5, 127.9, 129.0,
128.0, 128.4, 128.5, 128.6, 128.7, 138.0, 138.31, 138.7,
4.3.5. 30,40,60-Tri-O-acetyl-D-galactosyl-below-face
1
cycloadduct 48. Yield 61%, H NMR (300 MHz, CDCl3)
d 5.88 (d, J¼2.46 Hz, 1H), 5.47 (d, J¼2.15 Hz, 1H), 5.01
(dd, J¼11.85, 1.16 Hz, 1H), 4.78 (d, J¼5.34 Hz, 2H), 4.45
(t, J¼6.46 Hz, 1H), 4.16 (d, J¼6.5 Hz, 2H), 3.70 (dd,
J¼9.32, 2.52 Hz, 1H), 2.18, 2.08, 2.05 (s£3, 9H, –OAc).
Anal. calcd for C16H18010S: C, 47.76; H, 4.47; S, 7.97.
Found: C, 47.57; H, 4.46; S, 7.96.
167.5, 169.3; IR (thin film): 1706 (n CvO) cm21
.
4.3.6. 20-Acetoxy-30,40,60-tri-O-benzyl-D-glucosyl cyclo-
adduct 49 or 50. Yield 51%, [a]2D5¼þ5.458 (CDCl3); H
4.3.9. 4-(20-Deoxy-30,40,60-tri-O-benzyl-a-D-glucosyl)-O-
tetronic acid glycoside 56. Yield 40%, [a]2D5¼þ139.88
1
NMR (300 MHz, CDCl3) d 7.34–7.16 (m, 15H), 6.21 (s,
1H), 4.99–4.49 (m, 8H), 4.07 (d, J¼5.66 Hz, 2H), 3.85–
3.82 (m, 1H), 3.77 (dd, J¼10.86, 3.09 Hz, 1H), 3.71 (d,
J¼10.66 Hz, 1H), 2.09 (s, 3H). 13C NMR (300 MHz,
CDCl3) d 168 6, 168.4, 163.9, 138.0–127.8, 96.9, 96.5,
82.6, 79.4, 76.9, 76.8, 75.7, 75.5, 74.5, 68.1, 67.2, 21.8.
Anal. calcd for C33H3209S: C, 65.55; H, 5.33; S, 5.30.
Found: C, 65.44; H, 5.31; S, 5.06.
1
(CDCl3): H NMR (300 MHz, CDCI3) d 7.35–7.15 (m,
15H), 5.61 (d, J¼2.10 Hz, 1H), 5.40 (s, 1H), 4.91–4.45 (m,
8H), 4.02–3.93 (m,1H), 3.81–3.75 (m, 2H), 3.62 (t,
J¼9.79 Hz, 2H), 2.47–2.41 (dd, J¼4.84, 0.95 Hz, 1H),
1.96–1.86 (dd, J¼10.96, 3.41 Hz, 1H). 13C NMR
(300 MHz, CDCl3) d 176.0, 173.6, 138.3, 138.3, 137.8,
128.7, 128.6, 128.1, 128.0, 127.8, 99.8, 92.8, 76.6, 75.4,
73.8, 73.2, 72.4, 68.3, 68.0, 34.6. Anal. calcd for C31H3207:
C, 72.08; H, 6.24. Found: C, 71.83; H, 6.31.
4.3.7. 4-{3,4,6-Tribenzyl-2-deoxy-D-a-glucosyl}-6-[S]-
methyl-5,6-dihydro-2H-pyran-2-one (53). Desulfuriza-
tion with W-2 Raney nickel. General procedure. A slurry
of W-2 RaNi25 (3 teaspoons, 9.0 g) in absolute ethanol was
transferred to a flask provided with a stirring bar and a
nitrogen inlet, and rinsed with absolute ethanol, followed by
benzene until it became pH neutral. Toluene (2 mL) and
benzene (2 mL) were added to this neutralized slurry and
this slurry was cooled to 08C, followed by the addition of
compound 25S (500 mg, 0.871 mmol) in benzene (6 mL).
After stirring this cooled reaction mixture for 45 min, TLC
indicated disappearance of all of the starting material. The
reaction mixture was filtered through a celite pad, which
was rinsed exhaustively with ethyl acetate. The filtrate was
concentrated in vacuo to provide 409.6 mg of an oily
residue. Radial chromatography of this residue, eluting with
hexane:ethyl acetate (4:1 to 1:1) afforded 270.4 mg (57%)
of the disaccharide 53 as a clear oil: Rf¼0.33 (1:1
Acknowledgements
This research has been supported by (i) NIH: GM51216
(R. W. F), MBRS/GM56945 (M. M. T.) and RCMI/
RR03037 (chemistry department infrastructure); (ii)
PSC/CUNY research awards.
References
1. (a) Zhang, Z.; Ollmann, I. R.; Ye, X.-S.; Wischnat, R.; Baasov,
T.; Wong, C.-H. J. Am. Chem. Soc. 1999, 121, 734–753.
(b) Barresi, F.; Hindsgaul, O. J. Carbohydr. Chem. 1995, 14,
1043–1087. (c) In Carbohydrates in Chemistry and Biology,
Part I: Chemistry of Saccharides. Chemical Syntheses of
Glycosides and Glycomimetics; Ernst, B., Hart, G. W., Sinay,
P., Eds.; Wiley-VCH: Weinheim, 2000; Vol. 1.
1
hexane:ethyl acetate); H NMR d 1.3 (d, J¼6.3 Hz, 3H),
1.79–1.83 (m, 1H), 2.16 (dd, J¼2.83, 16.9 Hz, 1H), 2.26–
2.36 (m, 2H), 3.52–3.59 (m, 2H), 3.69–3.72 (m, 2H), 3.91–
3.93 (m, 1H), 4.36–4.60 (m, 5H), 4.82 (d, J¼10.8 Hz, 1H),
7.11–7.33 (m, 15H); 13C NMR d 20.7, 34 6, 34.7, 68.5,
72.2, 72.4, 72.8, 73.8, 75.3, 76.9, 77.4, 94.4, 96.8, 127.9,
127.9, 128.0, 128.1, 128.6, 138.0, 138.4, 138.4, 167.4,
168.9; IR (thin film): 1710 (n CvO) cm21. Anal. calcd for
C33H3607: C, 72.78; H, 6.66. Found: C, 72.60; H, 6.78.
2. Solid Support Oligosaccharide Synthesis and Combinatorial
Carbohydrate Libraries; Seeberger, P. H., Ed.; Wiley: New
York, 2001.
3. (a) Marzabadi, C. H.; Franck, R. W. Tetrahedron 2000, 56,
8385–8417. (b) Veyrieres A. Ref. 1(c), Chapter 15, 367–406.
4. Guo, Y.; Sulikowski, G. A. J. Am. Chem. Soc. 1998, 120,
1392–1397.
4.3.8. 4-{3,4,6-Tribenzyl-2-deoxy-a-D-galactosyl}-6-[S]-
methyl-3,4,5,6-tetrahydrohydro-2H-pyran-2-one (54).
Following the general desulfurization procedure, compound
33 (50 mg 0.087 mmol), W-2 RaNi (1/2 teaspoon, 1.5 g), in
benzene (2 mL), and toluene (1 mL) were stirred for 3 h to
provide 56 mg of an oily residue. Purification of this crude
by flash column chromatography, eluting with hexane:ethyl
acetate (9:1 to 5:1) afforded 14.7 mg (31%) of compound
54: Rf¼0.39 (3:1 petroleum ether:ethyl acetate); 1H NMR d
1.42 (d, J¼6.3 Hz, 3H), 2.09 (dd, J¼4.5, 13.2 Hz, 1H), 2.25
(dd, J¼3.9, 17.4 Hz, 1H), 2.32–2.42 (m, 1H), 2.44 (ddd,
J¼1.8, 11.7, 17.4 Hz, 1H), 3.50 (dd, J¼5.4, 9.15 Hz, 1H),
5. (a) Roush, W. R.; Bennett, C. E. J. Am. Chem. Soc. 1999, 121,
3541–3542. (b) Roush, W. R.; Bennett, C. E. J. Am. Chem.
Soc. 2000, 122, 6124–6125.
6. (a) Yin, H.; Franck, R. W.; Chen, S.-l.; Quigley, G. J.; Todaro,
L. J. Org. Chem. 1992, 57, 644–651. (b) Gupta, R. B.; Franck,
R. W. J. Am. Chem. Soc. 1989, 111, 7668–7669.
7. (a) Leblanc, Y.; Labelle, M. Am. Chem. Soc. Symp. Ser. 494
1992, 80–96. (b) Leblanc, Y.; Fitzsimmons, B. J.; Springer,
J. P.; Rokach, J. J. Am. Chem. Soc. 1989, 111, 2995–3000.
8. For a unique glycal functionalization via a non-cycloaddition
route: (a) Liu, J.; Gin, D. Y. J. Am. Chem. Soc. 2002, 124,

M. M. Tamarez et al. / Tetrahedron 59 (2003) 4249–4259
4259
9789–9797. (b) Honda, E.; Gin, D. Y. J. Am. Chem. Soc. 2002,
124, 7343–7352.
16. Franck, R. W.; Marzabadi, C. H. J. Org. Chem. 1998, 63,
2197–2208.
9. Capozzi, G.; Menichetti, S.; Nativi, C.; Rosi, A.; Valle, G.
Tetrahedron 1992, 48, 9023–9032.
17. Chan, M. C.; Cheng, K. M.; Ho, K. M.; Ng, C. T.; Yam, T. M.;
Wang, B. S. L.; Luh, T. Y. J. Org. Chem. 1988, 53, 4466–4471.
18. Back, T. G.; Yang, K.; Krouse, H. R. J. Org. Chem. 1992, 57,
2923–2925.
10. For two recent papers describing this approach (a) Bartolozzi,
A.; Capozzi, G.; Falciani, C.; Menichetti, S.; Nativi, C.;
Bacialli, A. P. J. Org. Chem. 1999, 64, 6490–6494.
(b) Marzabadi, C. H.; Franck, R. W.; Schinazi, R. F. Bioorg.
Med. Chem. 2002, 10, 273–281.
19. (a) Bombala, M. U.; Ley, S. V. J. Chem. Soc. Perkin Trans. 1
1979, 3013–3016. (b) Win, W. W.; Franck, R. W. J. Org.
Chem. 1997, 4510, 4512.
11. A preliminary report of a part of these results: Capozzi, G.;
Dios, A.; Franck, R. W.; Geer, A.; Marzabadi, C.; Menichetti,
S.; Nativi, C.; Tamarez, M. Angew. Chem., Int. Ed. 1996, 35,
777–779.
20. Kozikowski, A. P.; Lee, J. J. J. Org. Chem. 1990, 55,
863–870.
21. Witman, M. D.; Halcomb, R. L.; Danishefsky, S. J.; Golik, J.;
Vyas, D. J. J. Org. Chem. 1990, 55, 1979–1981.
22. Fraser-Reid, B.; Walker, D. L.; Tam, S. Y.; Holder, N. L. Can.
J. Chem. 1973, 51, 3950–3954.
12. For related work on cycloaddition of ketolactones: Bartolozzi,
A.; Capozzi, G.; Menichetti, S.; Nativi, C. Org. Lett. 2000, 2,
251–253.
23. Lichtenthaler, F. W.; Schneider-Adams, T. J. Org. Chem.
1994, 59, 6728–6734.
13. Huang, N. Z.; Lakshmikantham, M. V.; Cava, M. P. J. Org.
Chem. 1987, 52, 169–172.
24. Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43,
2923–2925.
14. Hausler, J. Monatsh. Chem. 1982, 113, 1213–1216.
15. Sato, M.; Sakaki, J.; Sugita, Y.; Nakano, T.; Kaneko, C.
Tetrahedron Lett. 1990, 31, 7463–7466.
25. Mozingo, R. Organic Syntheses; 1955; Collect. Vol. 3, pp
181–183.