Synthesis of alkylated deoxynojirimycin and 1,5-dideoxy-1,5-iminoxylitol analogues: Polar side-chain modification, sulfonium and selenonium heteroatom variants, conformational analysis, and evaluation as glycosidase inhibitors

Article abstract of DOI:10.1021/ja0482076

The syntheses of N-alkylated deoxynojirimycin and 1,5-dideoxy-1,5- iminoxylitol derivatives having either a D- or an L-erythritol-3-sulfate functionalized N-substituent are reported. The alkylating agent used was a cyclic sulfate derivative, whereby selec

Full text of DOI:10.1021/ja0482076

Published on Web 09/09/2004
Synthesis of Alkylated Deoxynojirimycin and
1,5-Dideoxy-1,5-iminoxylitol Analogues: Polar Side-Chain
Modification, Sulfonium and Selenonium Heteroatom Variants,
Conformational Analysis, and Evaluation as Glycosidase
Inhibitors
Monica G. Szczepina,† Blair D. Johnston,† Yue Yuan,† Birte Svensson,‡,§ and
B. Mario Pinto*,†
Contribution from the Department of Chemistry, Simon Fraser UniVersity,
Burnaby, British Columbia, Canada V5A 1S6, and Department of Chemistry,
Carlsberg Laboratory, Gamle Carlsberg Vej 10, DK-2500 Valby, Denmark
Received March 29, 2004; E-mail: bpinto@sfu.ca
Abstract: The syntheses of N-alkylated deoxynojirimycin and 1,5-dideoxy-1,5-iminoxylitol derivatives having
either a D- or an L-erythritol-3-sulfate functionalized N-substituent are reported. The alkylating agent used
was a cyclic sulfate derivative, whereby selective attack of the nitrogen atom at the least hindered primary
center afforded the desired ammonium salt. In aqueous solution, these salts were configurationally labile
at the ammonium center. Sulfonium and/or selenonium analogues of the ammonium salts were prepared
by analogous reactions. The chalcogen salts were obtained as mixtures of diastereomers, separable in
some cases, differing only in the stereochemistry at the configurationally stable sulfur or selenium atoms.
Proof of configuration and conformation of each compound was obtained by detailed NMR experiments.
The compounds are six-membered ring analogues of salacinol, a known sulfonium-salt glucosidase inhibitor.
Evaluation of the target compounds for enzyme inhibition of the glucosidase enzyme glucoamylase G2
indicated that these compounds were either inactive or, at best, only weak inhibitors of maltose hydrolysis.
Introduction
affinity for glucoamylase, a carbohydrate processing enzyme
from Aspergillus niger (Ki ) 1.1 × 10-12 M).6 This strong
Metabolic processing of oligosaccharides involves a variety
of glycosidase enzymes that catalyze the hydrolytic cleavage
of glycosidic linkages. Inhibition of these enzymes affords
various opportunities to manipulate the rates of glycosidase
reactions, either for enzymology studies or for therapeutic
purposes.1-4 Known inhibitors include certain carbohydrate
mimics that act as substrate surrogates for glycosidase enzymes.5
An example is the naturally occurring glycosidase inhibitor
acarbose 1. This oligosaccharide contains the pseudo-sugar
valienamine at the nonreducing end and has extremely high
binding is postulated to originate from electrostatic interactions
of the positively charged, protonated nitrogen atom with
carboxylate residues in the enzyme active-site. In addition, the
half-chair conformation of the pseudo-sugar resembles the
distorted pyranose ring conformation that is transiently attained
during the generally accepted mechanism for glycosidase-
catalyzed hydrolysis.10
A large number of glycosidase inhibitors are based on simpler
amino anhydroalditol derivatives having the nitrogen atom as
the ring heteroatom in either five-or six-membered ring struc-
tures. Prototypical examples are deoxynojirimycin (2) and its
N-alkylated analogues.7 Two derivatives of deoxynojirimycin,
namely miglitol (3) and N-butyldeoxynojirimycin (4), are
currently in use as drugs for the treatment of Type II diabetes
† Simon Fraser University.
‡ Carlsberg Laboratory.
§ Present address: Biochemistry and Nutrition Group, BioCentrum-DTU,
Denmark Technical University, Søltofts Plads Bldn 224, DK-2800 Kgs.
Lyngby, Denmark.
(1) Varki, A.; Cummings, R.; Esko, J.; Freeze, H.; Hart, G.; Marth, J., Eds.
Essentials of Glycobiology; Cold Spring Harbor Laboratory Press: Cold
Spring Harbor, NY, 1999.
(2) Taylor, M. E.; Drickamer, K. Introduction to Glycobiology; Oxford
University Press: New York, 2003.
(3) Helenius, A.; Aebi, M. Science 2001, 2364-2369.
(4) (a) Dwek, R. A.; Butters, T. D. Platt, F. M.; Zitzmann, N. Nature ReV.
Drug DiscoVery 2002, 1, 65-75. (b) Dwek, R. A. Chem. ReV. 1996, 96,
683-720.
(5) For leading references: (a) Elbein, A. D.; Molyneux, R. J. In ComprehensiVe
Natural Products Chemistry; Pinto, B. M., Ed.; Barton, D. H. R., Nakanishi,
K., Meth-Cohn, O., Ser. Eds.; Elsevier: U.K., 1999; Chapter 7, Vol. 3. (b)
Asano, N,; Nash, R. J.; Molyneux, R. J.; Fleet, G. W. J. Tetrahedron:
Asymmetry 2000, 11, 1645-1680. (c) McCarter, J. D.; Withers, S. G. Curr.
Opin. Struct. Biol. 1994, 4, 885-892. (d) Ly, H. D.; Withers, S. G. Annu.
ReV. Biochem. 1999, 68, 487-522.
(6) Bock, K.; Sigurskjold, B. Stud. Nat. Prod. Chem. 1990, 7, 29-86. Holman,
R. R.; Cull, C. A.; Turner, R. C. Diabetes Care 1999, 22, 960-964. Jacob,
G. S. Curr. Opin. Struct. Biol. 1995, 5, 605-611. Sigurskjold, B. W.;
Berland, C. R.; Svensson, B. Biochemistry 1994, 33, 10191-10199.
(7) Stu¨tz, A. E., Ed. Iminosugars as Glycosidase Inhibitors: Nojirimycin and
Beyond; Wiley-VCH: Weinheim, Federal Republic of Germany, 1999.
(8) Kapit, W.; Macey, R. I.; Meisami, E. The Physiology Coloring Book; Harper
Collins College Publishers: Menlo Park, CA, 1987.
(9) Cox, T.; Lachmann, R.; Hollak, C.; Aerts, J.; van Weely, S.; Hrebicek,
M.; Platt, F.; Butters, T.; Dwek, R.; Moyses, C.; Gow, I.; Elstein, D.;
Zimran, A. Lancet 2000, 355, 1481-1485.
(10) (a) Koshland, D. E. Biol. ReV. 1953, 28, 416-436. (b) Zechel, D. L.;
Withers, S. G. Curr. Opin. Chem. Biol. 2001, 5, 643-649.
9
12458
J. AM. CHEM. SOC. 2004, 126, 12458-12469
10.1021/ja0482076 CCC: $27.50 © 2004 American Chemical Society

Deoxynojirimycin/1,5-Dideoxy-1,5-iminoxylitol Analogues
A R T I C L E S
and Gaucher’s disease, respectively. Both drugs act by inhibition
of glycosidase enzymes.
between the strength of inhibition and the extent to which the
bound conformation of the inhibitor matches the conformation
of the transition state for glycoside hydrolysis. A consensus has
emerged, however, that the binding of iminosugar inhibitors is
usually enthalpically driven with a major exothermic contribu-
tion arising from protonation of the amine by an active site
carboxylic acid.14
In the case of Type II diabetes, which generally develops in
adults, insulin secretion from the pancreas may be normal but
the entry of glucose into cells is compromised due to a shortage
of insulin receptors on target cells.8 Blood glucose levels are
abnormally high, and deleterious effects ensue. Treatment with
miglitol (3) before a meal inhibits R-glucosidases such as
sucrase, maltase, and maltose glucoamylase in the intestinal tract
and attenuates the spike in blood glucose levels that follows
ingestion of food, thus leading to a more controlled metabolism
of carbohydrates.
Gaucher’s disease is an inherited metabolic defect that results
in accumulation of glycosphingolipids due to the lack of a
glycosidase enzyme necessary for their degradation. Oral
treatment with N-butyldeoxynojirimycin (4) causes inhibition
of glycolipid synthesis through slowing of Glucosidase I and II
enzymes in the trimming pathways of oligosaccharide synthesis
as well as inhibition of a specific glycosyl transferase involved
in glycosphingolipid synthesis.9 This limits the buildup of
unwanted glycolipids to manageable levels and avoids or
decreases the necessity of administering exogenous cerebro-
cidease enzymes by intravenous methods.
We have begun a program to investigate the synthesis and
properties of sulfonium analogues of known amine glycosidase
inhibitors and have reported the synthesis of a sulfonium
analogue (6) of castanospermine.15 These studies were initiated
under the premise that sulfonium salts should be permanently
charged mimics of the protonated ammonium salts that are
presumed to be the enzyme-bound form of amine glycosidase
inhibitors. Indeed, compound 6 was shown by NMR studies to
bind to glucoamylase G2 with a high energy 1,4B conformation,
and modeling studies provided evidence for electrostatic interac-
tion of the sulfonium center with an active site carboxylate
residue.16 Concurrent with these studies, the importance of
electrostatic binding interactions was reinforced by the isolation
of sulfonium ion glucosidase inhibitors salacinol (7)17 and
kotalanol (8)18 from Salacia reticulata, a plant native to Sri
Lanka and known for its antidiabetic properties. These com-
It has often been postulated that nitrogenous glycosidase
inhibitors bind to their target enzymes at least partially due to
electrostatic interaction of their protonated ammonium salts with
active site carboxylate residues. This is similar to the stabilizing
interactions of these negatively charged carboxylate groups with
the transient charges that develop in the accepted transition state
for glycoside hydrolysis.10 Partial breakage of the glycosidic
bond leads to incremental positive charge development at both
the ring oxygen and the anomeric carbon of the glycoside.
Replacement of either of these two atoms by a protonated
nitrogen center therefore mimics the charge at these centers in
the transition state. Most glycosidase inhibitors result from
replacement of the ring oxygen of sugars with an amine, while
nitrogenous compounds that a mimic positive charge at the
anomeric carbon are found in the form of 1-azasugar analogues
such as isofagomine (5).11
Recently, in the case of an isofagomine derivative bound in
the active site of an endocellulase, direct crystallographic
evidence has been presented to support a protonated imino center
in the bound inhibitor, although the conformation of the six-
membered heterocycle was unchanged from the free ligand.12
Other studies have shown that a variety of distortions in the
conformation of iminosugar inhibitors occur upon binding.13,14
It is unclear at present whether a strong correlation exists
pounds constitute a new class of glycosidase inhibitors in that
they contain a thiosugar sulfonium ion with an internal sulfate
providing the counterion. Presumably the sulfonium center can
act in the same way as the protonated ammonium center of the
nitrogenous glycosidase inhibitors when binding in the active
site of glycosidase enzymes. We have synthesized several
salacinol analogues by either varying the stereochemistry at one
or more chiral centers or by replacing the sulfonium locus by
an ammonium or selenonium center.19-22 Some of these
(14) Zechel, D. L.; Boraston, A. B.; Gloster, T.; Boraston, C. M.; MacDonald,
J. M.; Tilbrook, D. M. G.; Stick, R. V.; Davies, G. J. J. Am. Chem. Soc.
2003, 125, 14313-14323.
(15) Svansson, L.; Johnston, B. D.; Gu, J.-H.; Patrick, B.; Pinto, B. M. J. Am.
Chem. Soc. 2000, 122, 10769-10775.
(16) Johnson, M. A.; Jensen, M. T.; Svensson, B.; Pinto, B. M. J. Am. Chem.
Soc. 2003, 125, 5663-5670.
(17) Yoshikawa, M.; Murakami, T.; Shimada, H.; Matsuda, H.; Yamahara, J.;
Tanabe, G.; Muraoka, O. Tetrahedron Lett. 1997, 38, 8367-8370.
(18) Yoshikawa, M.; Murakami, T.; Yashiro, K.; Matsuda, H. Chem. Pharm.
Bull. 1998, 46, 1339-1340.
(19) (a) Ghavami, A.; Johnston, B. D.; Jensen M. T.; Svensson, B.; Pinto, B.
M. J. Am. Chem. Soc. 2001, 123, 6268-6271. (b) Ghavami, A.; Sadalapure,
K. S.; Johnston, B. D.; Lobera, M.; Snider, B. B.; Pinto, B. M. Synlett
2003, 1259-1262.
(11) Lillelund, V. H.; Jensen, H. H.; Liang, X.; Bols, M. Chem. ReV. 2002,
102, 515-553.
(12) Varrot, A.; Tarling, C. A.; MacDonald, J. M.; Stick, R. V.; Zechel, D. L.;
Withers, S. G.; Davies, G. J. J. Am. Chem. Soc. 2003, 125, 7496-7497.
(13) Varrot, A.; MacDonald, J.; Stick, R. V.; Pell, G.; Gilbert, H. J.; Davies, G.
J. Chem. Commun. 2003, 946-947.
(20) Ghavami, A.; Johnston, B. D.; Pinto, B. M. J. Org. Chem. 2001, 66, 2312-
2317.
(21) Ghavami, A.; Johnston, B. D.; Jensen, M. T.; Svensson, B.; Pinto, B. M.
J. Am. Chem. Soc. 2002, 124, 8245-8250.
(22) Ghavami, A.; Chen, J. J.-W.; Pinto, B. M. Carbohydr. Res. 2004, 339,
401-407.
9
J. AM. CHEM. SOC. VOL. 126, NO. 39, 2004 12459

A R T I C L E S
Szczepina et al.
Scheme 1
derivatives are active glucosidase inhibitors that show interesting
selectivity for glucosidase enzymes from various sources.
The variance in inhibitory power for alkylated deoxynojiri-
mycin and 1,5-dideoxy-1,5-iminoxylitol derivatives with dif-
ferent side chains has been investigated previously.23-25 These
studies were limited to simple alkyl or aromatic derivatives for
the most part, and there seems to have been little investigation
of the effect of including more polar substituents in the alkyl
group. To expand the repertoire of molecules of this class that
could serve as glycosidase inhibitors, we proposed to synthesize
N-alkylated 1,5-dideoxy-1,5-iminoxylitol (9a) and deoxynojiri-
mycin (10a) having the same L-erythritol-derived, sulfated side
chain as salacinol. The advantage of having an internal sulfate
counterion for the ammonium salt was deemed to be worth
pursuing in order to investigate whether such a structural
modification would lead to increased in vivo stability and/or
membrane permeability. In addition, the internal sulfate salt and
polar side chain may provide cationic inhibitors that bind to
glycosidase enzymes without deprotonating the catalytic active-
site carboxylic acid and provide additional insight into the
structural features that are important for inhibition. We report
herein the syntheses of 9a and 10a as well as the corresponding
sulfonium and selenonium analogues 11a, 12a, and 13a. We
report also the syntheses of the corresponding enantiomers or
diastereomers 9b-13b resulting from incorporation of a side
chain derived from D-erythritol. The compounds were evaluated
as inhibitors of glucoamylase G2.
temperature that stereoisomers at the ammonium center were
not observed.
The general synthetic strategy (Scheme 1) involved alkylation
of the piperidine (15 and 16), tetrahydrothiapyran (17 and 18),
or tetrahydroselenapyran (19) heterocycles with either the 2,4-
O-benzylidene-L-1,3-cyclic sulfate (14a),19 derived from L-
glucose, or its enantiomer (14b),19 obtained from D-glucose. In
general, the reactions with the less-expensive 14b were exam-
ined first. These methods are analogous to those that we used
previously to synthesize the five-membered ring analogues,
salacinol and its nitrogen or selenium congeners.19-22
1. Preparation of Starting Materials. In preliminary experi-
ments investigating the reactivity of the cyclic sulfate 14b, we
found that, for complex amine nucleophiles having only
secondary alcohols as additional functional groups, protection
of hydroxyl groups was unnecessary but that any primary
alcohol functional groups may be alkylated in competition with
amines.
Results and Discussion
Accordingly, the unprotected anhydroxylitol imine (15) was
prepared by the literature method,25 while deoxynojirimycin was
prepared as its tetra-O-benzyl derivative (16).23 The tetrahy-
drothiapyran derivative 17 was prepared (Scheme 2) by
deacetylation and benzylation of the known triacetate 20.26 The
benzylated tetrahydrothiapyran 18 was similarly prepared from
the known anhydro-5-thio-D-glucitol tetraacetate (21)15 by
protecting group interchange. Compound 20 was obtained, in
turn, either by reduction of tetra-O-acetyl-5-thio-D-xylopyranose
(22)27 or, more conveniently, from reaction of acetylated 1,5-
dibromoxylitol (23) with sodium sulfide.26 The selenium het-
erocyle 24 was prepared by substituting NaSeB(OEt)3 (obtained
in situ28 by reduction of Se with NaBH4/EtOH) for sodium
sulfide in the reaction with acetylated 1,5-dibromoxylitol (23)
(Scheme 2). Subsequent exchange of the acetates for benzyl
Each target compound was synthesized in two stereoisomeric
forms (a or b) by using either of the enantiomeric forms of the
cyclic sulfate 14a or 14b as the source of the sulfated alkyl
side chain. In the case of compounds 9, 11, and 13, these
stereoisomers are enantiomers, while compounds 10 and 12 were
prepared as either of two diastereomers. For the enantiomeric
sulfonium salts 11a and 11b, R/S isomers at the stereogenic
sulfonium ion center were separated and characterized inde-
pendently. Similar isomers for the sulfonium salts 12 and
selenonium salt 13 were not separable by chromatography, and
the products were characterized as mixtures. In the case of the
ammonium salts 9 and 10, inversion at the nitrogen center, via
the free amine, was sufficiently fast in solution at room
(23) van den Broek, L. A. G. M.; Vermaas, D. J.; Heskamp, B. M.; van Boeckel,
C. A. A.; Tan, M. C. A. A.; Bolscher, J. G. M.; Ploegh, H. L.; van
Kemenade, F. J.; de Goede, R. E. Y.; Miedema, F. Recl. TraV. Chim. Pays-
Bas 1993, 112, 82-94.
(26) Halila, S.; Benazza, M.; Demailly, G. Tetrahedron Lett. 2001, 42, 3307-
3310.
(24) Asano, N.; Kizu, H.; Oseki, K.; Tomioka, E.; Matsui, K.; Okamoto, M.;
Baba, M. J. Med. Chem. 1995, 38, 2349-56.
(27) Bozo, E.; Boros, S.; Kuszmann, J.; Gacs-Baitz, E.; Parkanyi, L. Carbohydr.
Res. 1998, 308, 297-310.
(25) Hausler, H.; Rupitz, K.; Stutz, A. E.; Withers, S. G. Monatsh. Chem. 2002,
133, 555-560.
(28) Klayman, D. L.; Griffin, T. S. J. Am. Chem. Soc. 1973, 95, 197-199.
9
12460 J. AM. CHEM. SOC. VOL. 126, NO. 39, 2004

Deoxynojirimycin/1,5-Dideoxy-1,5-iminoxylitol Analogues
A R T I C L E S
Scheme 2
resulting from opening of the cyclic sulfate by the methanol
solvent could be isolated from the early chromatographic
fractions. A similar reaction with the L-cyclic sulfate 14a gave
somewhat less of this side product, and the desired coupled
product 25 was obtained in slightly higher yield (56%). The1H
NMR spectra of compounds 25 and 27 exhibited sharp
resonances for methylene groups R to the amine in D2O (made
basic with K2CO3), but neutral or acidic D2O solutions gave
downfield shifts and much broader resonances for these meth-
ylene resonances. We attribute these observations to exchange,
at an intermediate rate relative to the chemical-shift NMR time
scale, of the conjugate-acid R/S ammonium salts, with nitrogen
inversion taking place via the free amines that exist in
equilibrium with their conjugate acids at acidic pH.
Removal of the benzylidene protecting groups by hydrolysis
in aqueous acetic acid gave the target compounds 9a (73%)
and 9b (72%) after purification by chromatography on silica
gel. These ammonium salts gave severely exchange-broadened
NMR spectra and were more productively characterized by
adding base to the NMR samples to produce the conjugate amine
bases. Prolonged treatment with strong base should be avoided,
however, due to the possibility of sulfate ester hydrolysis, as
noted below. As expected for enantiomers, the NMR data for
9a and 9b were virtually identical, although small differences
in chemical shifts between different samples for both identical
and enantiomeric compounds were noted. These differences
were attributed to the concentration and temperature dependence
of the NMR chemical shifts between samples. The tendency of
zwitterionic compounds to exist as aggregates in solution is the
likely origin of these effects.
Scheme 3
The coupled products 28 and 29, derived from the benzyl-
protected deoxynojirimycin, were obtained by reaction of
compound 16 with the cyclic sulfates 14a and 14b in acetone/
K2CO3 in yields of 80% and 65%, respectively (Scheme 4).
The 1H NMR resonances for compounds 28 and 29 were
extremely broad in CDCl3 but sharpened in CD3OD (made basic
with NaOD), thus indicating that the coupled products were
obtained as an equilibrating mixture of the desired ammonium
salts with the corresponding conjugate bases. Simultaneous
removal of both the benzyl and benzylidene protecting groups
was achieved by hydrogenolysis in aqueous acetic acid to give
the target compounds 10a and 10b.
Analysis by 1H NMR spectroscopy indicated that these
products were contaminated by KOAc. Nevertheless, other than
a resonance at δ 1.8 in the spectrum that was attributed to the
acetate impurity, the target compounds were essentially pure
and all resonances in both the 1H and 13C spectra were assigned
by two-dimensional techniques. Prolonged storage of the NMR
sample of compound 10b in D2O/NaOD at pH > 10 produced
a slow loss of the 3′-sulfate group as evidenced by an upfield
shift of the H-3′ resonance. After 2 days at ambient temperature
the sulfate ester had been completely hydrolyzed to yield cleanly
the tertiary amine compound 30 and inorganic sulfate salts.
protecting groups gave the desired tetrahydroselenapyran deriva-
tive 19, whose preparation was recently reported by an unrelated
method.29
2. Target Ammonium Compounds. Compound 15 was
reacted with the D-cyclic sulfate 14b in MeOH containing K2-
CO3 (Scheme 3). Isolation of the more polar product gave the
ammonium salt 27 in 43% yield. An abundant side product (26)
1
The H NMR data for all of the amine compounds in D2O
(pH > 8) indicated that the predominant conformation of the
piperidine ring was 4C1 (carbohydrate numbering) and that this
conformational preference did not appear to change upon
protonation (pH < 3). Similar conclusions were reached in a
previous conformational study of alkylated deoxynojirimycin
derivatives.24
(29) Lucas, M. A.; Nguyen, O. T. K.; Schiesser, C. H.; Zheng, S.-L. Tetrahedron
2000, 56, 3995-4000.
9
J. AM. CHEM. SOC. VOL. 126, NO. 39, 2004 12461

A R T I C L E S
Scheme 4
Szczepina et al.
gave 31b and 32b in approximately 37% yield. On changing
the solvent to 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP), the
yield improved to 87%. The dramatically beneficial effect of
HFIP solvent on the yields for sulfonium salt formation has
been noted previously.19b The ratio of the major product 31b
to the minor product 32b was 2:1. Pure samples of the two
components were obtained by chromatography and characterized
separately by NMR techniques.
Initially, 1D 1H NMR spectra were obtained which revealed
that the two compounds were isomers, having the same number
of hydrogen atoms. The similarity of the spectra of the two
compounds suggested that the compounds differed in stereo-
chemistry only at the stereogenic sulfur atom. COSY spectra
permitted the assignment of the proton signals for the tetrahy-
drothiapyran ring and for the erythritol side chain in both
compounds. Notably, it was found that all of the ring proton
signals were shifted downfield relative to the parent tetrahy-
drothiapyran 17. This was anticipated, since the positive
sulfonium center is electron withdrawing. Furthermore, although
it was initially expected that the three benzyloxy groups at C-2,
C-3, and C-4 would favor the sterically less-hindered equatorial
positions, analysis of vicinal coupling constants showed that
J2,3 and J3,4 ) 3.5-3.9 Hz. These values are much smaller than
those (J2,3 ≈ J3,4 ≈ 8.9 Hz) observed for the axial-axial vicinal
coupling constants in the precursor 17. Thus, we reasoned that
compounds 31b and 32b preferred a 1C4 conformation, placing
the three benzyloxy groups in axial positions and accounting
for the small vicinal coupling constants. This conformational
preference can be explained by the fact that the axial substituents
at C-2 and C-4 provide stabilizing gauche electrostatic interac-
tions of the polar benzyloxy groups with the sulfonium ion
center; the group at C-3 can also provide stabilizing electrostatic
interactions.15 The results are reminiscent of our previous work
with the sulfonium ion 6.15
3. Target Sulfonium Compounds. The syntheses of sulfo-
nium salts 11 and 12 (Schemes 5 and 6) were achieved in a
fashion similar to those of the ammonium salts. Thus, compound
17 was initially reacted with the D-cyclic sulfate 14b in acetone
at 65 °C. Slow formation of two more-polar products was
observed by TLC analysis. Isolation of the mixture of products
Scheme 5
9
12462 J. AM. CHEM. SOC. VOL. 126, NO. 39, 2004

Deoxynojirimycin/1,5-Dideoxy-1,5-iminoxylitol Analogues
A R T I C L E S
Scheme 6
The configuration at the sulfonium center was next established
by means of a NOESY experiment. The NOESY spectrum for
the major diastereomer showed H-1b′ correlations to H-1ax/
H-1eq/H-5ax as well as H-1a′ and correlations to H-5eq/H-5ax.
This isomer was thus assigned to structure 31b with the
erythritol side chain occupying the equatorial orientation (Figure
1). The absolute configuration at sulfur was thus established as
being R.
The NOESY spectrum for the minor diastereomer showed a
correlation between H-1a′ and the isochronous signal assigned
to H-1ax/H-1eq, as well as a correlation between H-1b′ and
H-5eq. No correlation with H-5ax was observed. This isomer
was thus assigned to structure 32b, the diastereomer with the
erythritol side chain in an axial orientation (Figure 1). The
absolute configuration at sulfur was thus established as being
S. Each of the diastereomers 31b and 32b was deprotected by
hydrogenolysis to give sulfonium salts R-11b and S-11b, which
were obtained in 81% and 95% yields, respectively. Vicinal
coupling constants indicated that deprotection was accompanied
in both cases by a change in the preponderant ring conformation
1
4
from C4 to C1 (R-11b J2,3 ≈ J3,4 ≈ 7.2 Hz, S-11b J2,3 ≈ J3,4
≈ 9.0 Hz) (Figure 1). Transient one-dimensional nuclear
Overhauser enhancement (NOE) difference experiments con-
firmed that there was no configurational inversion at the
sulfonium center upon removal of the benzyl and benzylidene
protecting groups. Thus, the major isomer R-11b, upon irradia-
tion of the H-4′b/H-1′a multiplet, showed no NOE with the ring
axial protons (Figure 2). Irradiation of H-1ax/H-5ax showed
NOEs on the H-1eq/H-5eq/H-3 and H-2/H-4 multiplets only.
No NOEs with the erythritol side chain protons were observed.
These experiments provide evidence for the erythritol side chain
occupying the axial position at sulfur, on the â-face and opposite
to H-1ax, and confirm the R configuration at the sulfonium
center for the major isomer R-11b, as was previously assigned
for the protected precursor 31b.
Figure 2. One-dimensional transient NOE difference spectra of R-11b in
1
D2O. (a) H NMR spectrum (b) with selective irradiation of the H-4′b/H-
Figure 1. Preferred conformations of 31b, 32b, R-11b, and S-11b.
1′a multiplet and (c) with selective irradiation of the H-1ax/H-5ax multiplet.
9
J. AM. CHEM. SOC. VOL. 126, NO. 39, 2004 12463

A R T I C L E S
Szczepina et al.
isomer ratio in 70% yield (Scheme 6). As in the xylitol series,
1
the protected glucitol derivative 33b displayed an unusual C4
conformational preference, as indicated by the coupling con-
stants. This places the three benzyloxy groups at C-2, C-3, and
C-4 as well as the benzyloxymethyl group at C-5 in an axial
orientation.
The stereochemistry at the stereogenic sulfonium center for
the major isomer 33b was established by means of a NOESY
experiment. A strong NOESY correlation was observed between
the H-1b′ proton and the H-5 proton, thus confirming that the
benzylidene-protected erythritol side chain was cis to H-5. NOEs
to H-1ax and to H-6a/H-6b were not observed. Thus, the
absolute configuration at the sulfonium center in the major
isomer was S. Alkylation of the sulfur must occur preferentially
from the R-face of 1,5-anhydro-2,3,4,6-tetra-O-benzyl-5-thio-
D-glucitol 18 due to shielding of the â-face by the adjacent C-5
benzyloxymethyl group.
Figure 3. One-dimensional transient NOE difference spectra of compound
S-11b in D2O. (a) H NMR spectrum (b) with selective irradiation of the
H-4′b/H-1′b multiplet and (c) with selective irradiation of the H-1ax/H-
1
The mixture consisting of compounds 33b and 34b was then
subjected to hydrogenolysis to give primarily 1,5-dideoxy-1,5-
[[(2R,3R)-2,4-dihydroxy-3-(sulfooxy)butyl]-S-episulfonium-
ylidene]-D-glucitol inner salt S-12b in 81% yield (Scheme 6).
Treatment of 1,5-anhydro-2,3,4,6-tetra-O-benzyl-5-thio-D-
glucitol (18) with the L-cyclic sulfate 14a afforded an inseparable
mixture of compounds 33a and 34a with an approximate 3:1
isomer ratio in 68% yield (Scheme 6). Whereas the achiral
anhydro xylitol compound 17 generated enantiomers upon
reaction with the enantiomeric D- and L-cyclic sulfates, this was
not the case for the chiral compound 18. For this reaction, the
products 33a and 33b are diastereomers rather than enantiomers.
The stereochemistry at the stereogenic sulfonium center for
the major isomer 33a was again established by means of a
NOESY experiment. A strong NOE correlation was observed
between the H-1′a proton and H-5. In addition, there was also
an NOE correlation between H-2′ and H-5, confirming that the
benzylidene-protected erythritol side chain was on the same side
as H-5. NOEs to H-1ax and to H-6a/H-6b were not observed.
Thus, the absolute configuration at the sulfonium center for
compound 33a was R, that is, the same stereochemistry at sulfur
previously found for the diastereoisomer 33b. (Note: The
change in R/S descriptor for the sulfonium center between 33a
and 33b derives only from changes in priority between the side
chain and the ring substituents and does not imply a change in
stereochemistry at sulfur in this case.) Therefore, independent
of the configuration (14a or 14b) of the cyclic sulfate reagent,
in both cases, alkylation at sulfur occurred preferentially from
the least-hindered R-face of compound 18.
5ax multiplet.
Figure 4. Cahn-Ingold-Prelog R/S descriptors for the sulfonium centers
in the enantiomeric pair 11a/11b.
The minor isomer S-11b showed, upon irradiation of the
H-4′b/H-1′b multiplet, NOE with the H-1ax/H-5ax protons
(Figure 3). Irradiation of the H-1ax/H-5ax multiplet showed
NOEs with the H-4′b/H-1′b multiplet as well as to the H-2/H-
4/H-4′a/H-1′a multiplet, in addition to NOEs to the ring protons.
These experiments provide evidence for the erythritol side chain
being present on the same face as H-1ax, occupying the
R-equatorial position at sulfur, thus confirming the S configu-
ration of the minor isomer S-11b at the sulfonium center, as
was previously assigned for the protected precursor 32b.
The synthesis of the sulfonium salts from the L-cyclic sulfate
14a was examined next (Scheme 5). Compound 17 was reacted
with 14a at 70 °C in HFIP solvent to give two products 31a
and 32a in a 5:2 ratio (84% yield). The major diastereoisomer
31a, in which the erythritol side chain is cis to the C-3 benzyloxy
group, was separated from the minor diastereoisomer 32a, with
the erythritol side chain trans to the C-3 benzyloxy group. The
1H NMR spectra were virtually identical to those of the
enantiomers 31b and 32b except for small variations due to
concentration, as noted above. Each of the diastereomers 31a
and 32a was deprotected via hydrogenolysis to give the target
compounds R-11a and S-11a, respectively. It is noteworthy that,
for pairs of enantiomers (for example, R-11a/R-11b), the
stereochemical descriptors for the configurations at the sulfo-
nium centers are invariant in these compounds because of
sequence rule changes in priority between the side chain and
the ring substituents (see Figure 4).
The mixture containing 33a and 34a was then subjected to
hydrogenolysis to give primarily 1,5-dideoxy-1,5-[[(2S,3S)-2,4-
dihydroxy-3-(sulfooxy)butyl]-R-episulfonium-ylidene]-D-gluci-
tol inner salt R-12a in 67% yield (Scheme 6).
Upon removal of the protecting groups, compounds R-12a
4
and S-12b adopted a C1 conformation, as indicated by the
vicinal proton coupling constants. This places all of the ring
substituents in an equatorial orientation, as observed for the
xylitol series.
4. Target Selenonium Compounds. The tetrahydroselena-
pyran 19 was coupled to the D-cyclic sulfate 14b in HFIP solvent
and afforded an inseparable mixture of two compounds, 35b
and 36b, in a 1:4 ratio in 96% yield (Scheme 7). These two
compounds are diastereoisomers at the stereogenic selenium
Entry into the 5-thio-D-glucitol analogues began by treatment
of 1,5-anhydro-2,3,4,6-tetra-O-benzyl-5-thio-D-glucitol 18 with
the D-cyclic sulfate 14b. The reaction afforded an inseparable
mixture of compounds 33b and 34b with an approximate 2:1
9
12464 J. AM. CHEM. SOC. VOL. 126, NO. 39, 2004

Deoxynojirimycin/1,5-Dideoxy-1,5-iminoxylitol Analogues
A R T I C L E S
Scheme 7
center. Alkylation can occur on selenium to give, as with sulfur,
the benzylidene-protected erythritol side chain either cis to the
C-3 benzyloxy group or trans to the C-3 benzyloxy group. It
was found by comparison of the NMR data to those of the
sulfonium analogues 31b/32b, and by analysis of the NOESY
spectrum (see below), that the major product, 36b, was that in
which the benzylidene-protected erythritol side chain was trans
to the benzyloxy group at C-3. The minor product, 35b, was
that in which the benzylidene-protected erythritol side chain was
cis to the C-3 benzyloxy group. Curiously, the ratio was opposite
to the results obtained with the tetrahydrothiapyran products
31b and 32b for which the major isomer was the cis isomer.
The predominant conformations observed in both compounds
35b and 36b were, as with the corresponding thio analogues,
those which placed all three benzyloxy groups in an axial
arrangement, thus favoring 1C4 conformations, as evidenced by
the coupling constants. The major isomer 36b in its preferred
1C4 conformation places the selenonium alkyl group in the axial
position. The longer C-Se bonds in compounds 35b/36b
compared to the thio analogues must result in less severe gauche
steric interactions between the selenonium alkyl group and C-2
and C-4.
The configuration at the stereogenic selenonium centers for
the enantiomers 36a and 36b was confirmed by means of
NOESY experiments performed on the mixtures of the com-
pounds containing their minor diastereomers. The major isomer
in each case was found to be that in which the erythritol side
chain occupied the axial position in the preferred 1C4 conforma-
tion. This was evidenced by correlations between H-1b′ and
H-5eq as well as correlations between H-1′a and H-1eq. An
axial preference would imply correlations between H-1′a/H-
1′b and H-5eq, and H-1′a/H-1′b and H-1eq only, since free
rotation about the C-1′-Se bond would not permit the H-1′a and
H-1′b protons to interact with the axial C-1 and C-5 protons as
these are on the opposite side of the selenoether ring. Therefore,
NOEs would not be expected between H-1′a/H-1′b and H-1ax/
H-1eq. On the other hand, an equatorial preference would imply
correlations between H-1′a/H-1′b to H-1ax and H-5ax as well
as possibly to H-1eq and H-5eq. Thus, for compound 36b the
absolute configuration at the selenium center is S and that for
the enantiomeric 36a is also S (see previous discussion on
stereochemical descriptors for 11a/11b). In both cases, the
erythritol side chain is cis to the benzyloxy groups at C-2 and
C-4 and trans to the C-3 benzyloxy group.
The mixture consisting of compounds 35b and 36b was then
deprotected via hydrogenolysis to give mostly one diastereo-
isomer of 13b, in 39% yield (Scheme 7). The low yield was
due to catalyst poisoning by decomposition products and the
reaction could not be brought to completion. This major
compound was characterized by NMR techniques and found to
be 1,5-dideoxy-1,5-[[(2R,3R)-2,4-dihydroxy-3-(sulfooxy)butyl]-
S-episelenoniumylidene]-xylitol inner salt S-13b.
The mixture consisting of 35a and 36a was then deprotected
by hydrogenolysis to afford mostly one diastereoisomer of 13a
in 25% yield (Scheme 7). The major compound was character-
ized by NMR techniques and found to be the desired
1,5-dideoxy-1,5-[[(2S,3S)-2,4-dihydroxy-3-(sulfooxy)butyl]-S-epi-
selenoniumylidene]xylitol inner salt S-13a, the enantiomer of
compound S-13b.
5. Glucoamylase Inhibition Studies. Compounds 9a, 9b,
10a, 10b, S-11a, R-11a, S-11b, R-11b, 12a, 12b, 13a, and 13b
were tested for inhibition of glucoamylase G2 from Aspergillus
niger with maltose as a substrate. The rate of the hydrolysis
reaction was measured by assaying for glucose concentration
Reaction of the selenoether 19 with the L-cyclic sulfate 14a
concluded the synthetic study. The product was an inseparable
mixture of two diastereoisomers at the stereogenic selenium
center, 35a and 36a, in a 1:3 ratio. (Scheme 7).
9
J. AM. CHEM. SOC. VOL. 126, NO. 39, 2004 12465

A R T I C L E S
Szczepina et al.
(b) Benzylation: A mixture of 1,5-anhydro-5-thioxylitol (0.520 g,
3.47 mmol) and 60% NaH (0.744 g, 5 equiv) in DMF (50 mL) was
stirred in an ice bath for 1 h. A solution of BnBr (1.4 mL, 4 equiv)
was added, and the solution was stirred at rt overnight. The mixture
was quenched with MeOH (8 mL), H2O (100 mL) was added, and the
solution was extracted with Et2O (3 × 150 mL). The organic solution
was dried over Na2SO4 and concentrated, and the residue was purified
by flash chromatography [hexanes/EtOAc, 20:1] to give 17 as a white
solid (0.928 g, 64%). Mp 46-49 °C. 1H NMR (CDCl3): δ 7.36-7.24
(15H, m, Ar), 4.83 (2H, s, CH2Ph), 4.69 (2H, d, JA,B ) 11.4 Hz, CH2-
Ph), 4.65 (2H, d, JA,B ) 11.6 Hz, CH2Ph), 3.63 (2H, m, J1eq,2 ) J4,5eq
using a linked glucose oxidase method. The tested compounds
were either inactive or, at best, only weak inhibitors of maltose
hydrolysis (compounds 9b and 12a were the only compounds
exhibiting inhibitory activity and were estimated to have Ki
values in excess of 70 mM). It would appear that the five-
membered ring of salacinol (7) is a necessary and key structural
determinant of sulfonium sulfate glycosidase inhibitors and their
analogues19a,19c,21 for this particular enzyme. It is curious,
however, that the nitrogen analogues (10a and 10b) were
inactive given the known inhibitory activities of N-alkylated
deoxynojirimycin derivatives for this enzyme.
) 4.2 Hz, J1ax,2 ) J4,5ax ) 11.0 Hz, H-4 and H-2), 3.31 (1H, t, J2,3
)
J3,4 ) 8.9 Hz, H-3), 2.72 (2H, m, H-5eq and H-1eq), 2.47 (2H, dd,
J5ax,5eq ) J1ax,1eq ) 13.4 Hz, H-5ax and H-1ax). 13C NMR (CDCl3): δ
138.9, 138.37 (3Cipso), 128.42-127.51 (15C, Ar), 86.76 (C-3), 82.26
(2C, C-2 and C-4), 76.33 (CH2Ph), 73.02 (2 CH2Ph), 31.49 (2C, C-1
and C-5). Anal. Calcd for C26H28O3S: C, 74.25; H, 6.71. Found: C,
74.16; H, 6.91.
Experimental Section
General. Optical rotations were measured at 23 °C. Analytical thin-
layer chromatography (TLC) was performed on aluminum plates
precoated with Merck silica gel 60F-254 as the adsorbent. The
developed plates were air-dried, exposed to UV light and/or sprayed
with a solution containing 1% Ce(SO4)2 and 1.5% molybdic acid in
10% aq H2SO4, and heated. Compounds were purified by flash
chromatography on Kieselgel 60 (230-400 mesh). Rexyn 101 was
obtained from Fischer. 1H and 13C NMR spectra were recorded on the
following: Bruker AMX-400 NMR spectrometer at 400.13 MHz,
Bruker AMX-600 NMR spectrometer at 600.13 MHz, and Varian
INOVA 500 NMR spectrometer at 499.97 MHz for 1H. Chemical shifts
are given in ppm downfield from TMS for those measured in CDCl3,
CD3OD, and CD2Cl2 and from 2,2-dimethyl-2-silapentane-5-sulfonate
(DSS) for those spectra measured in D2O. Chemical shifts and coupling
constants were obtained from a first-order analysis of the spectra.
Assignments were fully supported by two-dimensional 1H,1H (COSY),
1H,1H (NOESY), and 1H,13C (HMQC) experiments using standard
Bruker or Varian pulse programs. Processing of the spectra was
performed with standard UXNMR and WINNMR software (Bruker)
or MestReC software (Varian).
The 1D-transient NOE experiments were performed by inverting
the signal of interest with a 80 ms Gaussian selective pulse which was
constructed from 1024 steps. Spectra were collected in difference mode
by alternating the phase of the receiver gain during on- and off-
resonance. The digitized signal was stored in a 32 K data set using a
sweep width of 10 ppm, an acquisition time of 2.72 s, 128 scans, and
8 dummy scans. Processing of the spectra was accomplished by zero
filling to 64 K followed by an exponential multiplication using a line
width of 1 Hz. NOESY spectra were obtained with a mixing time of
500 or 800 ms.
1,5-Anhydro-2,3,4,6-tetra-O-benzyl-5-thio-D-glucitol (18). (a) Ac-
etate Methanolysis: To a solution of 1,5-anhydro-2,3,4,6-tetra-O-acetyl-
5-thio-D-glucitol 21 (0.310 g, 0.89 mmol) in dry MeOH (20 mL) was
added 1 M NaOMe/MeOH (4 mL, 4 equiv), and the mixture was stirred
under N2 overnight. The mixture was neutralized with excess Rexyn
101 ion-exchange resin, the resin was removed by filtration, and the
organic phase was concentrated. The residue was purified by flash
chromatography [CHCl3/MeOH, 5:2] to give 1,5-anhydro-5-thio-D-
glucitol as a white solid (0.125 g, 78%). Mp 110-115 °C; [R]D
+27.4 (c 1.2, MeOH). H NMR (D2O): δ 3.90 (1H, dd, J5,6a ) 3.2
Hz, J6b,6a ) 11.9 Hz, H-6a), 3.75 (1H, dd, J5,6b ) 6.4 Hz, H-6b), 3.64
)
1
(1H, m, H-2), 3.48 (1H, dd, J4,5 ) 10.2 Hz, H-4), 3.19 (1H, t, J2,3
)
J3,4 ) 9.1 Hz, H-3), 2.88 (1H, m, H-5), 2.71 (1H, dd, J1eq,2 ) 4.6 Hz,
J1eq,1ax ) 13.3 Hz, H-1eq), 2.62 (1H, dd, J1ax,2 ) 11.0 Hz, H-1ax).
(b) Benzylation: To a stirred solution of 1,5-anhydro-5-thio-D-
glucitol (0.194 g, 1.08 mmol) in dry DMF (60 mL) was added NaH
(0.5 g, 12.5 mmol) and then BnBr (0.7 mL, 5.9 mmol), and the mixture
was stirred overnight. Excess NaH was destroyed by the addition of
MeOH. The organic phase was concentrated under reduced pressure.
To the residue was added H2O (200 mL), and this was extracted with
CH2Cl2 (5 × 100 mL). The organic phase was dried over Na2SO4 and
concentrated. The product was purified by flash chromatography
[hexanes/EtOAc, 20:1] to give a syrup that was recrystallized from
EtOAc/hexanes to give compound 18 as a white solid (0.276 g, 58%).
Mp 56-59 °C; [R]D ) +15.1 (c 1.1, CHCl3). The 1H NMR spectrum
was consistent with the literature data.30
MALDI mass spectra were obtained on a PerSeptive Biosystems,
Voyager DE time-of-flight spectrometer for samples dispersed in a 2,5-
dihydroxybenzoic acid matrix. High-resolution mass spectra were liquid
secondary ion mass spectrometry (LSIMS) run on a Kratos Concept
double focusing mass spectrometer at 10 000 RP, using a glycerin
matrix or, in the case of compound 31a, with meta-NO2-benzyl alcohol
as the matrix. Solvents were distilled before use and were dried, as
necessary. Solvents were evaporated under reduced pressure and below
50°C.
1,5-Anhydro-2,3,4-tri-O-acetyl-5-selenoxylitol (24). To a stirred
suspension of selenium (1.48 g, 18.7 mmol) in anhydrous EtOH (40
mL) at 0 °C was added NaBH4 (0.9 g, 23.8 mmol). An almost colorless
solution resulted. The ice bath was removed, 2,3,5-tri-O-acetyl-1,5-
dibromo-1,5-dideoxy-xylitol 23 (4.87 g, 12.0 mmol) was added, and
the mixture was stirred at rt overnight. H2O (200 mL) was added, and
the mixture was extracted with Et2O (5 × 100 mL). The solids were
removed by filtration, the solution was concentrated, and the residue
was purified by flash chromatography [hexanes/EtOAc, 1:1] to give
24 as yellow crystals (2.22 g, 57%). Mp 106-111 °C; 1H NMR
1,5-Anhydro-2,3,4-tri-O-benzyl-5-thioxylitol (17). (a) Acetate Meth-
anolysis: A mixture of 1,5-anhydro-2,3,4-tri-O-acetyl-5-thioxylitol 20
(0.125 g, 0.453 mmol) and 1 M NaOMe in MeOH (0.6 mL, 0.6 mmol)
in dry MeOH (10 mL) was stirred under N2 overnight. The mixture
was neutralized with excess Rexyn 101. The resin was removed by
filtration, and the organic phase was concentrated to give 1,5-anhydro-
(CDCl3): δ 5.11 (2H, ddd, J1eq,2 ) J4,5eq ) 4.5 Hz, J1ax,2 ) J4,5ax
)
10.8 Hz, H-2, H-4), 4.96 (1H, t, J2,3 ) J3,4 ) 9.7 Hz, H-3), 2.74 (2H,
dd, H-1eq, H-5eq), 2.67 (2H, t, J5ax,5eq ) J1ax,1eq ) 12.0 Hz, H-1ax,
H-5ax), 2.00 (3H, s, OAc), 1.99 (6H, s, OAc). 13C NMR (CDCl3): δ
169.79 and 169.65 (3CdO), 73.98 (C-3), 73.78 (2C, C-2 and C-4),
21.02 (2 OAc), 20.80 (2C, C-1 and C-5), 20.56 (OAc). Anal. Calcd
for C11H16O6Se: C, 40.88; H, 4.99. Found: C, 40.76; H, 5.02.
1
5-thioxylitol as a solid (59.6 mg, 88%). Mp 137-140 °C. H NMR
(D2O): δ 3.65 (2H, m, J1eq,2 ) J4,5eq ) 4.5 Hz, J1ax,2 ) J4,5ax ) 10.9
Hz, H-2 and H-4), 3.15 (1H, t, J2,3 ) J3,4 ) 9.1 Hz, H-3), 2.66 (2H, m,
H-5eq and H-1eq), 2.56 (2H, dd, J5ax,5eq ) J1ax,1eq ) 13.6 Hz, H-5ax
and H-1ax). 13C NMR (D2O): δ 81.20 (C-3), 75.75 (2C, C-2 and C-4),
34.86 (2C, C-1 and C-5). Anal. Calcd for C5H10O3S: C, 39.99; H, 6.71.
Found: C, 39.68; H, 6.91.
1,5-Anhydro-2,3,4-tri-O-benzyl-5-selenoxylitol (19). (a) Acetate
Methanolysis: A mixture of 1,5-anhydro-2,3,4-tri-O-acetyl-5-selenoxy-
(30) Hashimoto, H.; Masashi, K.; Yuasa, H. Carbohydr. Res. 1996, 282, 207-
222.
9
12466 J. AM. CHEM. SOC. VOL. 126, NO. 39, 2004

Deoxynojirimycin/1,5-Dideoxy-1,5-iminoxylitol Analogues
A R T I C L E S
litol 24 (2.22 g, 6.87 mmol) and 1 M NaOMe in MeOH (10 mL, 10
mmol) in dry MeOH (60 mL) was stirred under an N2 atmosphere
overnight. The mixture was netralized with excess Rexyn 101, the resin
was removed by filtration, and the organic phase was concentrated to
give 1,5-anhydroselenoxylitol as tan crystals (1.19 g, 88%). Mp 98-
105 °C. 1H NMR (D2O): δ 3.75 (2H, m, J1eq,2 ) J4,5eq ) 4.6 Hz, J1ax,2
) J4,5ax ) 10.8 Hz, H-2, H-4), 3.11 (1H, t, J2,3 ) J3,4 ) 9.2 Hz, H-3),
2.66 (2H, t, J5ax,5eq ) J1ax,1eq ) 11.8 Hz, H-5ax, H-1ax), 2.60 (2H, dd,
H-1eq, H-5eq). 13C NMR (D2O): δ 81.40 (C-3), 76.62 (2C, C-2 and
C-4), 25.65 (2C, C-1 and C-5). Anal. Calcd for C5H10O3Se: C, 30.47;
H, 5.11. Found: C, 30.29; H, 5.21.
(b) Benzylation: To 1,5-anhydro-5-selenoxylitol 24 (0.289 g, 1.47
mmol) in dry DMF (20 mL) was added 60% NaH (0.516 g, 6 equiv)
while stirring in an ice bath. The ice bath was removed and BnBr (0.9
mL, 4 equiv) was added. The mixture was stirred under N2 overnight.
The reaction was then quenched with MeOH (5 mL), H2O (100 mL)
was added, and the mixture was extracted with Et2O (3 × 50 mL).
The organic solution was dried over Na2SO4 and concentrated. The
product was purified by flash chromatography [hexanes/EtOAc, 20:1]
to give the title compound 19 as a white solid (0.505 g, 74%). Mp
56-60 °C. 1H NMR (CDCl3): δ 7.32-7.24 (15H, m, ArH), 4.81 (2H,
chromatography (EtOAc/MeOH/H2O, 6:4:1) to give compound 9a
(0.156 g, 73%) as a colorless, hard foam. [R]D +11 (c 0.56, H2O); 1H
and 13C NMR data were virtually identical to those of the enantiomer
9b (see Tables 2 and 4 (Supporting Information)); MALDI MS m/e
399.99 (M+ + Na), 318.28 (M+ +H), 238.12 (M+ + H - SO3).
2,3,4,6-Tetra-O-benzyl-1,5-dideoxy-1,5-[[N-(2R,3R)-2,4-O-benz-
ylidene-2,4-dihydroxy-3-(sulfooxy)butyl]iminoonium]-D-glucitol (29).
Tri-O-benzyldeoxynojirimycin 16 (0.241 g, 0.460 mmol) and 2,4-O-
benzylidene-D-erythritol-1,3-cyclic sulfate 14b (0.143 g, 0.525 mmol)
were dissolved in reagent grade acetone (2 mL). Anhydrous K2CO3
(0.020 g, 0.15 mmol) was added, and the mixture was stirred in a sealed
tube at 70 °C for 20 h. The solvent was removed, and the residue was
purified by column chromatography (CHCl3/MeOH, 5:1) to give the
product 29 as a colorless gum (0.240 g, 65%). [R]D -5.4 (c 0.9, CHCl3);
NMR data in Tables 1 and 3 (Supporting Information).
1,5-Dideoxy-1,5-[[N-(2R,RS)-2,4-dihydroxy-3-(sulfooxy)butyl]imi-
noonium]-D-glucitol (10b). Compound 29 (0.209 g, 0.263 mmol) was
dissolved in 80% aqueous acetic acid (20 mL) and the solution was
stirred with 10% Pd/C catalyst (0.42 g) under 1 atm of H2 for 20 h.
The catalyst was removed by filtration through a small plug of silica
gel and washed with water (50 mL). The filtrate was evaporated, and
the gummy residue was freed of acetic acid by dissolving in water and
reconcentrating (2 × 50 mL). The crude product was purified by column
chromatography (EtOAc/MeOH/H2O, 6:3:1) to give compound 10b
s, CH2Ph), 4.70 (2H, d, JA,B ) 11.6 Hz, CH2Ph), 4.66 (2H, d, JA,B
11.5 Hz, CH2Ph), 3.73 (2H, m, J1eq,2 ) J4,5eq ) 4.2 Hz, J1ax,2 ) J4,5ax
)
)
11.2 Hz, H-2, H-4), 3.27 (1H, t, J2,3 ) J3,4 ) 8.9 Hz, H-3), 2.69 (2H,
dd, J5ax,5eq ) J1ax,1eq ) 12.0 Hz, H-5eq, H-1eq), 2.58 (2H, t, H-5ax,
H-1ax). 13C NMR (CDCl3): 138.89 (Cipso), 138.44 (2Cipso), 128.39-
127.46 (15C, Ar), 86.98 (C-3), 83.17 (2C, C-2 and C-4), 76.34 (CH2-
Ph), 72.97 (2 CH2Ph), 22.11 (2C, C-1 and C-5). Anal. Calcd for
C26H28O3Se: C, 66.80; H, 6.04. Found: C, 66.88; H, 6.22.
1,5-Dideoxy-1,5-[[N-(2R,3R)-2,4-O-benzylidene-2,4-dihydroxy-3-
(sulfooxy)butyl]iminoonium]xylitol (27). 1,5-Dideoxy-1,5-iminoxylitol
15 (0.161 g, 1.21 mmol) and 2,4-O-benzylidene-D-erythritol-1,3-cyclic
sulfate 14b (0.360 g, 1.32 mmol) were dissolved in reagent grade MeOH
(2 mL). Anhydrous K2CO3 (0.015 g, 0.11 mmol) was added, and the
mixture was stirred in a sealed tube at 65 °C for 3.5 h, at which point
TLC showed that the cyclic sulfate had been consumed. The solvent
was removed, and the residue was purified by column chromatography
(EtOAc/MeOH/H2O, 8:2:1) to give the product 27 as a yellow oil (0.209
g, 43%): [R]D -50 (c 0.48, H2O); NMR data in Tables 1 and 3
(Supporting Information).
1,5-Dideoxy-1,5-[[N-(2R,3R)-2,4-dihydroxy-3-(sulfooxy)butyl]imi-
noonium]xylitol (9b). Aqueous 60% HOAc (25 mL) was added to
compound 27 (0.209 g, 0.515 mmol), and the mixture was stirred while
warming in an open flask for 20 h at 70 °C. The mixture was cooled
and concentrated, and the crude product was purified by column
chromatography (EtOAc/MeOH/H2O, 6:4:1) to give compound 9b
(0.118 g, 72%) as a colorless, hard foam: [R]D -9 (c 0.57, H2O); NMR
data in Tables 2 and 4 (Supporting Information); MALDI MS m/e
339.99 (M+ + Na), 238.12 (M+ + H - SO3).
1
(0.096 g, containing 0.56 equiv or 13 wt % of KOAc by H NMR,
91% after correcting for acetate content). NMR data in Tables 2 and 4
(Supporting Information).
2,3,4,6-Tetra-O-benzyl-1,5-dideoxy-1,5-[[N-(2S,3S)-2,4-O-benz-
ylidene-2,4-dihydroxy-3-(sulfooxy)butyl]iminoonium]-D-glucitol (28).
Tri-O-benzyldeoxynojirimycin 16 (0.223 g, 0.426 mmol) and 2,4-O-
benzylidene-L-erythritol-1,3-cyclic sulfate 14a (0.123 g, 0.4535 mmol)
were dissolved in reagent grade acetone (2 mL). Anhydrous K2CO3
(0.020 g, 0.15 mmol) was added, and the mixture was stirred in a sealed
tube at 70 °C for 20 h. The solvent was removed, and the residue was
purified by column chromatography (CHCl3/MeOH, 5:1) to give the
product 28 as a colorless amorphous solid (0.271 g, 80%): [R]D +36
(c 0.8, CHCl3); NMR data in Tables 1 and 3 (Supporting Information).
1,5-Dideoxy-1,5-[[N-(2R,3R)-2,4-dihydroxy-3-(sulfooxy)butyl]imi-
noonium]-D-glucitol (10a). Compound 28 (0.205 g, 0.263 mmol) was
dissolved in 80% aqueous acetic acid (20 mL), and the solution was
stirred with 10% Pd/C catalyst (0.41 g) under 1 atm of H2 for 20 h.
The catalyst was removed by filtration through a small plug of silica
gel and washed with water (50 mL). The filtrate was evaporated, and
the gummy residue was freed of acetic acid by dissolving in water and
reconcentrating (2 × 50 mL). The crude product was purified by column
chromatography (EtOAc/MeOH/H2O, 6:3:1) to give compound 10a
1
(0.094 g, containing 0.77 equiv or 18 wt % of KOAc by H NMR,
1
89% after correcting for acetate content). See Tables 2 and 4 for H
and 13C NMR data (Supporting Information).
2,3,4-Tri-O-benzyl-1,5-dideoxy-1,5-[[(2R,3R)-2,4-O-benzylidene-
2,4-dihydroxy-3-(sulfooxy)butyl]-(R)-episulfoniumylidene]xylitol
Inner Salt (31b) and 2,3,4-Tri-O-benzyl-1,5-dideoxy-1,5-[[(2R,3R)-
2,4-O-benzylidene-2,4-dihydroxy-3-(sulfooxy)butyl]-(S)-episul-
foniumylidene]xylitol Inner Salt (32b). To 1,1,1,3,3,3-hexafluoro-2-
propanol (0.5 mL) were added 2,4-O-benzylidene-D-erythritol-1,3-
cyclic-sulfate 14b (0.565 g, 2.08 mmol), 1,5-anhydro-2,3,4-tri-O-benzyl-
5-thioxylitol 7 (0.677 g, 1.61 mmol), and anhydrous K2CO3 (70 mg).
The mixture was stirred in a sealed tube in a 70 °C oil bath overnight,
after which an extra 40 mg of anhydrous K2CO3 were added. The
solvents were removed, and the residue was chromatographed [CHCl3/
MeOH, 10:1] to give 31b and 32b in a 2:1 ratio (0.975 g, 87%).
Major isomer 31b: Mp 186-189 °C; [R]D +2.1 (c 1.2, CH2Cl2);
NMR data in Tables 1 and 3 (Supporting Information). HRMS Calcd
for C37H40O9S2 (M + H): 693.2192. Found: 693.2209. Anal. Calcd
for C37H40O9S2: C, 64.14; H, 5.82. Found: C, 64.39; H, 5.94.
Minor isomer 32b: mp 169-172 °C; [R]D -49.1 (c 0.8, CH2Cl2);
1,5-Dideoxy-1,5-[[N-(2S,3S)-2,4-O-benzylidene-2,4-dihydroxy-3-
(sulfooxy)butyl]iminoonium]xylitol (25). 1,5-Didexy-1,5-iminoxylitol
15 (0.158 g, 1.19 mmol) and 2,4-O-benzylidene-L-erythritol-1,3-cyclic
sulfate 14a (0.347 g, 1.27 mmol) were dissolved in reagent grade MeOH
(2 mL). Anhydrous K2CO3 (0.018 g, 0.15 mmol) was added, and the
mixture was stirred in a sealed tube at 65 °C for 4 h. The solvent was
removed, and the residue was purified by column chromatography
(EtOAc/MeOH/H2O, 8:2:1) to give the product 25 as a yellow oil (0.273
g, 56%). [R]D +55 (c 0.65, H2O); 1H and 13C NMR data were virtually
identical with those of the enantiomer 27 (see Tables 1 and 3
(Supporting Information)); MALDI MS m/e 428.09 (M+ + Na), 406.11
(M+ + H), 326.15 (M+ + H - SO3).
1,5-Dideoxy-1,5-[[N-(2S,3S)-2,4-dihydroxy-3-(sulfooxy)butyl]imi-
noonium]xylitol (9a). Aqueous 60% HOAc (25 mL) was added to
compound 25 (0.273 g, 0.673 mmol), and the mixture was stirred while
warming in an open flask for 14 h at 75 °C. The mixture was cooled
and concentrated, and the crude product was purified by column
9
J. AM. CHEM. SOC. VOL. 126, NO. 39, 2004 12467

A R T I C L E S
Szczepina et al.
NMR data in Tables 1 and 3 (Supporting Information); Anal. Calcd
for C37H40O9S2: C, 64.14; H, 5.82. Found: C, 63.84: H, 5.96.
1,5-Dideoxy-1,5-[[(2R,3R)-2,4-dihydroxy-3-(sulfooxy)butyl]-(R)-
episulfoniumylidene]xylitol Inner Salt (R-11b). To compound 31b
(0.33 g, 0.48 mmol) dissolved in 80% AcOH (12 mL) was added Pd-
(OH)2 (0.2 g). The mixture was stirred under H2 (110 psi) for 48 h and
then filtered through Celite with MeOH. The solvent was evaporated,
and the residue was purified by column chromatography [EtOAc/
MeOH/H2O, 7:3:1]. Compound R-11b was obtained as a syrup (0.13
g, 81%); [R]D -21.8 (c 1.1, H2O); NMR data in Tables 2 and 4
(Supporting Information). HRMS Calcd for C9H19O9S2 (M + H):
335.0470. Found: 335.0454. Anal. Calcd for C9H18O9S2: C, 32.33;
H, 5.43. Found: C, 32.03; H, 5.59.
1,5-Dideoxy-1,5-[[(2R,3R)-2,4-dihydroxy-3-(sulfooxy)butyl]-(S)-
episulfoniumylidene]xylitol Inner Salt (S-11b). Compound 32b (0.249
g, 0.36 mmol) was deprotected by hydrogenolysis using the procedure
described above for R-11b to give the title compound as a syrup (0.13
g, 95%); [R]D -16.2 (c 0.9, H2O); NMR data in Tables 2 and 4
(Supporting Information). HRMS Calcd for C9H19O9S2 (M + H):
335.0470. Found: 335.0478. Anal. Calcd for C9H18O9S2: C, 32.33;
H, 5.43. Found: C, 31.88; H, 5.21.
2,3,4-Tri-O-benzyl-1,5-dideoxy-1,5-[[(2S,3S)-2,4-O-benzylidene-
2,4-dihydroxy-3-(sulfooxy)butyl]-(R)-episulfoniumylidene]xylitol
Inner Salt (31a) and 2,3,4-Tri-O-benzyl-1,5-dideoxy-1,5-[[(2S,3S)-
2,4-O-benzylidene-2,4-dihydroxy-3-(sulfooxy)butyl]-(S)-episul-
foniumylidene]xylitol Inner Salt (32a). To 1,1,1,3,3,3-hexafluoro-2-
propanol (0.5 mL) were added 2,4-O-benzylidene-L-erythritol-1,3-
cyclic-sulfate 14a (0.265 g, 0.97 mmol), 1,5-anhydro-2,3,4-tri-O-benzyl-
5-thioxylitol 17 (0.328 g, 0.78 mmol), and anhydrous K2CO3 (24 mg).
The mixture was stirred in a sealed tube in a 70 °C oil bath for 5 days.
The solvent was evaporated, and the residue was purified by column
chromatography [CHCl3/MeOH, 10:1] to give 31a and 32a in a 5:2
ratio as a white solid (0.465 g, 86%). Pure samples were obtained by
rechromatography.
citol Inner Salts (33b and 34b). To 1,1,1,3,3,3-hexafluoro-2-propanol
(0.5 mL) were added 2,4-O-benzylidene-D-erythritol-1,3-cyclic-sulfate
14b (0.115 g, 0.42 mmol), 1,5-anhydro-2,3,4,6-tetra-O-benzyl-5-thio-
D-glucitol 18 (0.174 g, 0.32 mmol), and anhydrous K2CO3 (30 mg).
The mixture was stirred in a sealed tube in a 70 °C oil bath for 5 days.
The solvent was removed, and the residue was purified by column
chromatography [CHCl3/MeOH, 10:1] to give an inseparable mixture
of 33b and 34b in a 2:1 ratio as a white solid (0.182 g, 70%); [R]D
)
+2.1 (c 1.3, CH2Cl2). Major isomer 33b: See Tables 1 and 2
1
(Supporting Information) for H and 13C NMR data. Anal. Calcd for
C45H48O10S2: C, 66.48; H, 5.96. Found: C, 66.36; H, 6.08.
1,5-Dideoxy-1,5-[[(2R,3R)-2,4-dihydroxy-3-(sulfooxy)butyl]-(S)-
episulfoniumylidene]-D-glucitol Inner Salt (12b). To a mixture of
compounds 33b and 34b (0.1639 g, 0.20 mmol) dissolved in 80%
AcOH (10 mL) was added Pd(OH)2 (0.17 g). The mixture was stirred
under 120 psi of H2 for 48 h. The mixture was filtered through Celite
with MeOH, the solvent was removed, and the residue was purified
by column chromatography [EtOAc/MeOH/H2O, 7:3:1]. Compound 12b
was obtained as a syrup (0.06 g, 81%); [R]D ) -20.4 (c 0.8, H2O).
1
See Tables 2 and 4 (Supporting Information) for H and 13C NMR
data. HRMS Calcd for C10H21O10S2 (M + H): 365.0576. Found:
365.0574.
2,3,4,6-Tetra-O-benzyl-1,5-dideoxy-1,5-[[(2S,3S)-2,4-O-benzylidene-
2,4-dihydroxy-3-(sulfooxy)butyl]-(R/S)-episulfoniumylidene]-D-glu-
citol Inner Salts (33a and 34a). To 1,1,1,3,3,3-hexafluoro-2-propanol
(0.5 mL) were added 2,4-O-benzylidene-L-erythritol-1,3-cyclic-sufate
14a (0.148 g, 0.54 mmol), 1,5-anhydro-2,3,4,6-tetra-O-benzyl-5-thio-
D-glucitol 18 (0.240 g, 0.44 mmol), and anhydrous K2CO3 (33 mg).
The mixture was stirred in a sealed tube in a 69-70 °C oil bath for 84
h. The solvent was evaporated, and the residue was purified by column
chromatography [CHCl3/MeOH, 10:1] to give an inseparable 3:1
mixture of 33a and 34a as a white solid (0.25 g, 68%); [R]D ) +48.8
(c 1.6, CH2Cl2). Major isomer 33a: See Tables 1 and 2 (Supporting
1
Information) for H and 13C NMR data. Anal. Calcd for C45H48O10S2:
C, 66.48; H, 5.95. Found: C, 66.19; H, 6.07.
Major isomer 31a: Mp 175-180 °C; [R]D -3.7 (c 0.9, CH2Cl2);
1H and 13C NMR data were virtually identical to those of the enantiomer
31b. Anal. Calcd for C37H40O9S2: C, 64.14; H, 5.82. Found: C, 63.81;
H, 5.68.
Minor isomer 32a: Mp 163-170 °C; [R]D +41.8 (c 1.1, CH2Cl2);
1H and 13C NMR data were virtually identical to those of the enantiomer
32b. Anal. Calcd for C37H40O9S2: C, 64.14; H, 5.82. Found: C, 64.42;
H, 5.75.
1,5-Dideoxy-1,5-[[(2S,3S)-2,4-dihydroxy-3-(sulfooxy)butyl]-(R)-
episulfoniumylidene]-D-glucitol Inner Salt (12a). To a mixture of
compounds 33a and 34a (0.180 g, 0.22 mmol) dissolved in 80% AcOH
(10 mL) was added Pd(OH)2 (0.20 g), and the mixture was stirred under
120 psi of H2 for 6 days. The mixture was filtered through Celite with
MeOH, the solvent was removed and the residue was purified by
column chromatography [EtOAc/MeOH/H2O, 7:3:1]. Compound 12a
was obtained as a syrup (0.05 g, 67%); [R]D ) +10.3 (c 0.6, H2O).
1,5-Dideoxy-1,5-[[(2S,3S)-2,4-dihydroxy-3-(sulfooxy)butyl]-(R)-
episulfoniumylidene]xylitol Inner Salt (R-11a). To compound 31a
(0.304 g, 0.44 mmol) dissolved in 80% AcOH (10 mL) was added
Pd/C (0.5 g). The mixture was stirred under 120 psi of H2 for 96 h.
The mixture was filtered through Celite with MeOH, and the solvent
was removed. The residue was then redissolved in 80% AcOH (10
mL). To the solution was added Pd(OH)2 (0.2 g) and the solution was
stirred under 120 psi of H2 for 48 h. The mixture was filtered through
Celite with MeOH, the solvent was evaporated, and the residue was
purified by column chromatography [EtOAc/MeOH/H2O, 7:3:1] to give
the title compound as a syrup (0.08 g, 55%); [R]D +21.7 (c 0.8, H2O).
1H and 13C NMR data were virtually identical to those of the enantiomer
R-11b (see Tables 1 and 3 (Supporting Information)). HRMS Calcd
for C9H18O9S2Na (M + Na): 357.0290. Found: 357.0284.
1,5-Dideoxy-1,5-[[(2S,3S)-2,4-dihydroxy-3-(sulfooxy)butyl]-(S)-
episulfoniumylidene]xylitol Inner Salt (S-11a). Compound 32a (0.240
g, 0.35 mmol) was deprotected by hydrogenolysis using the procedure
described above for S-11b to give the title compound as a syrup (0.08
g, 67%); [R]D +19.5 (c 0.7, H2O). 1H and 13C NMR data were virtually
identical to those of the enantiomer S-11b (see Tables 2 and 4
(Supporting Information)). HRMS Calcd for C9H19O9S2 (M + H):
335.0470. Found: 335.0477.
See Tables 2 and 4 (Supporting Information) for H and 13C NMR
data. HRMS Calcd for C10H21O10S2 (M + H): 365.0576. Found:
365.0577.
1
2,3,4-Tri-O-benzyl-1,5-dideoxy-1,5-[[(2R,3R)-2,4-O-benzylidene-
2,4-dihydroxy-3-(sulfooxy)butyl]-(S/R)-episelenoniumylidene]xyli-
tol Inner Salt (35b and 36b). To 1,1,1,3,3,3-hexafluoro-2-propanol
(0.5 mL) were added 2,4-O-benzylidene-D-erythritol-1,3-cyclic-sufate
14b (0.272 g, 1.00 mmol), 1,5-anhydro-2,3,4-tri-O-benzyl-5-selenoxy-
litol 19 (0.362 g, 0.78 mmol), and anhydrous K2CO3 (50 mg). The
mixture was stirred in a sealed tube in a 70 °C oil bath for 48 h. The
solvent was concentrated, and the residue was purified by column
chromatography [CHCl3/MeOH, 10:1] to give an inseparable mixture
of 35b and 36b in a 1:4 ratio (0.20 g, 96%). [R]D -45.7 (c 1.1, CH2-
Cl2). For the major isomer 36b: See Tables 1 and 2 (Supporting
Information) for 1H and 13C NMR data. Anal. Calcd for C37H40O9SSe:
C, 59.99; H, 5.45. Found: C, 59.73; H, 5.36.
1,5-Dideoxy-1,5-[[(2R,3R)-2,4-dihydroxy-3-(sulfooxy)butyl]-(S)-
episelenoniumylidene]xylitol Inner Salt (13b). To the mixture of
compounds 35b and 36b (0.295 g, 0.40 mmol) dissolved in 80% AcOH
(10 mL) was added Pd(OH)2 (0.29 g), and the mixture was stirred under
120 psi of H2 for 5 days. TLC revealed one major product and two
minor products. The mixture was filtered through Celite and concen-
trated, and the residue was purified by column chromatography [EtOAc/
2,3,4,6-Tetra-O-benzyl-1,5-dideoxy-1,5-[[(2R,3R)-2,4-O-benzylidene-
2,4-dihydroxy-3-(sulfooxy)butyl]-(S/R)-episulfoniumylidene]-D-glu-
9
12468 J. AM. CHEM. SOC. VOL. 126, NO. 39, 2004

Deoxynojirimycin/1,5-Dideoxy-1,5-iminoxylitol Analogues
A R T I C L E S
MeOH/H2O, 7:3:1] to give the major product, compound 13b as a syrup
(0.06 g, 39%); [R]D -16.6 (c 0.9, H2O). See Tables 2 and 4 (Supporting
Information) for 1H and 13C NMR data. HRMS Calcd for C9H19O9SSe
(M + H): 382.9915. Found: 382.9916. Anal. Calcd for C9H18O9SSe:
C, 28.35; H, 4.76. Found: C, 28.44; H, 4.71.
Bagsvaerd, Denmark) as described.31,32 The initial rates of glucoamylase
G2 catalyzed hydrolysis were tested with 1.2 mM maltose as substrate,
in 0.1 M sodium acetate, pH 4.5, at 45 °C. The enzyme concentration
was 8 × 10-7 M, and seven inhibitor concentrations in the range 0.2-
4.2 mM were tested. The glucose release was analyzed in aliquots
removed at appropriate time intervals, using a glucose oxidase assay
adapted to microtiter plate reading,33 and with a total reaction volume
for the enzyme reactions of 150 or 300 µL. The Ki values were
calculated assuming competitive inhibition, from 1/ν ) (1/Vmax) +
[(Km)/(Vmax[S]Ki)][I] where ν is the rate measured in the presence and
absence of inhibitor, [I] and [S] are the concentrations of inhibitor and
substrate, Km ) 1.6 mM, and kcat ) 11.3 s-1 (kcat is the catalytic
constant), using ENZFITTER.34
2,3,4-Tri-O-benzyl-1,5-dideoxy-1,5-[[(2S,3S)-2,4-O-benzylidene-
2,4-dihydroxy-3-(sulfooxy)butyl]-(S/R)-episelenoniumylidene]xyli-
tol Inner Salts (35a and 36a). To 1,1,1,3,3,3-hexafluoro-2-propanol
(0.5 mL) were added 2,4-O-benzylidene-L-erythritol-1,3-cyclic-sufate
14a (0.226 g, 0.83 mmol), 1,5-anhydro-2,3,4-tri-O-benzyl-5-seleno-
xylitol 19 (0.308 g, 0.66 mmol), and anhydrous K2CO3 (20 mg). The
mixture was stirred in a sealed tube in a 70 °C oil bath for 72 h. The
solvent was removed and the residue was purified by column chro-
matography [CHCl3/MeOH, 10:1] to give an inseparable 1:3 mixture
of 35a and 36a as a white solid (0.42 g, 85%). [R]D -44.0 (c 0.9,
Acknowledgment. We are grateful to Dr. A. Ghavami for
providing the L-cyclic sulfate 14a, to Sidsel Ehlers for expert
technical assistance with the glucoamylase assays, and to
University Medical Discoveries, Inc. for financial support.
1
CH2Cl2). For the major isomer 36a, the H and 13C NMR data were
virtually identical to those of the enantiomeric compound 36b (see
Tables 1 and 2 (Supporting Information)) except for small chemical
shift differences due to concentation effects. Anal. Calcd for C37H40O9-
SSe: C, 59.99; H, 5.45. Found: C, 59.85; H, 5.58.
Supporting Information Available: NMR data tables consist-
ing of assigned chemical shifts and coupling constants for
isomers of compounds 9, 10, 11, 12, and 13 and their protected
precursors as well as selected NMR spectra illustrating typical
1H and 13C resonances (61 pages). This material is available
free of charge via the Internet at http://pubs.acs.org.
1,5-Dideoxy-1,5-[[(2S,3S)-2,4-dihydroxy-3-(sulfooxy)butyl]-(S)-
episelenoniumylidene]xylitol Inner Salt (13a). To a mixture of
compounds 35a and 36a (0.406 g, 0.55 mmol) dissolved in 80% AcOH
(10 mL) was added Pd(OH)2 (0.50 g), and the mixture was stirred under
120 psi of H2 for 8 days. TLC revealed one major product and two
minor products. The mixture was filtered through Celite with MeOH,
the solvent was removed, and the residue was purified by column
chromatography [EtOAc/MeOH/H2O, 7:3:1]. Compound 13a was
obtained as a syrup (0.05 g, 25%); [R]D +14.1 (c 0.4, H2O). For
JA0482076
1
compound 13a, the H and 13C NMR data were virtually identical to
(31) Svensson, B.; Pedersen, T.; Svendsen, I.; Sakai, T.; Ottesen, M. Carlsberg
Res. Commun. 1982, 53, 423-442.
(32) Stoffer, B.; Frandsen, T. P.; Busk, P. K.; Schneider, P.; Svendsen, I.;
Svensson, B. Biochem. J. 1993, 292, 197-202.
the enantiomeric compound 13b (see Tables 1 and 2 (Supporting
Information)) except for small chemical shift differences due to
concentration effects. HRMS Calcd for C9H18O9SSeNa (M + Na):
404.9734. Found: 404.9735. Anal. Calcd for C9H18O9SSe: C, 28.35;
H, 4.76. Found: C, 28.56; H, 4.54.
(33) Frandsen, T.; Dupont, C.; Lehmbeck, J.; Stoffer, B.; Sierks, M. R.;
Honzatko, R. B.; Svensson, B. Biochemistry 1994, 33, 13808-13816.
(34) Leatherbarrow, R. J. Enzfitter, a Nonlinear Regression Data Analysis
Program for IBM PC; Elsevier Science Publishers BV: Amsterdam, The
Netherlands, 1987.
Enzyme Inhibition Assays. Glucoamylase G2 from Aspergillus
niger was purified from a commercial enzyme (Novo Nordisk,
(35) Svensson, B.; Sierks, M. R. Carbohydr. Res. 1992, 227, 29-44.
9
J. AM. CHEM. SOC. VOL. 126, NO. 39, 2004 12469