[13C,15N]2-acetamido-2-deoxy-D-aldohexoses and their methyl glycosides: Synthesis and NMR investigations of J-couplings involving1H, 13C, and15N

Article abstract of DOI:10.1021/jo051510k

A series of 2-amino-2-deoxy-D-[1-¹³C]aldohexoses and their methyl glycosides was prepared with use of a simplified cyanohydrin reduction route. Four D-aldopentosylamines (arabino, lyxo, ribo, xylo) were prepared from the corresponding D-aldopentoses by reaction with NH₃(g) in MeOH solvent, isolated in solid form, and characterized by ¹³C and ¹H NMR. Hydrolysis of β-D-xylopyranosylamine was studied using ¹³C-labeled substrates to establish optimal solution conditions for cyanohydrin formation. Major hydrolytic intermediates were observed and identified by time-lapse 1D and 2D NMR analyses of reaction mixtures. The aldopentosylamines were subsequently employed in cyanohydrin reduction reactions with K¹³CN to yield C2-epimeric [1-¹³C]2-aminosugars, which were separated by chromatography on ion-exchange columns. N-Acetylation and methyl glycosklation followed by chromatography gave pure 2-acetamido-2-deoxy-D-[1-¹³C]aldohexopyranosicles, J_CH and J_CC spin-spin coupling constants involving the labeled anomeric carbon were measured and compared to those observed previously in methyl D-[1-¹³C]-aldohexopyranosides. In parallel studies, theoretical J-couplings were calculated in model N-acetylated aldopyranosides using density functional theory (DFT) to predict the effect of OH vs NHCOCH₃ substitution at C2 on J_CH and J_CC values in aldopyranosyl rings. The synthetic method was also modified to accommodate ¹⁵N- and ¹³C-labeling within the N-acetyl side-chain, and some J-couplings involving ¹H, ¹³C, and ¹⁵N atoms in 2-[1,2-¹³C₂;¹⁵N]acetamido-2-deoxy-D-[1- ¹³C]glucose were measured and interpreted.

Full text of DOI:10.1021/jo051510k

[13C,15N]2-Acetamido-2-deoxy-D-aldohexoses and Their Methyl
Glycosides: Synthesis and NMR Investigations of J-Couplings
1
13
15
Involving H, C, and N
Yuping Zhu,† Qingfeng Pan,† Christophe Thibaudeau,† Shikai Zhao,‡ Ian Carmichael,§ and
Anthony S. Serianni*,†
Department of Chemistry and Biochemistry and the Radiation Laboratory, UniVersity of Notre Dame,
Notre Dame, Indiana 46556-5670, and Omicron Biochemicals, Inc., 1347 North Ironwood DriVe,
South Bend, Indiana 46615
serianni.1@nd.edu
ReceiVed July 20, 2005
A series of 2-amino-2-deoxy-D-[1-13C]aldohexoses and their methyl glycosides was prepared with use of
a simplified cyanohydrin reduction route. Four D-aldopentosylamines (arabino, lyxo, ribo, xylo) were
prepared from the corresponding D-aldopentoses by reaction with NH3(g) in MeOH solvent, isolated in
1
solid form, and characterized by 13C and H NMR. Hydrolysis of â-D-xylopyranosylamine was studied
using 13C-labeled substrates to establish optimal solution conditions for cyanohydrin formation. Major
hydrolytic intermediates were observed and identified by time-lapse 1D and 2D NMR analyses of reaction
mixtures. The aldopentosylamines were subsequently employed in cyanohydrin reduction reactions with
K13CN to yield C2-epimeric [1-13C]2-aminosugars, which were separated by chromatography on ion-
exchange columns. N-Acetylation and methyl glycosidation followed by chromatography gave pure 2-acet-
amido-2-deoxy-D-[1-13C]aldohexopyranosides. JCH and JCC spin-spin coupling constants involving the
labeled anomeric carbon were measured and compared to those observed previously in methyl D-[1-13C]-
aldohexopyranosides. In parallel studies, theoretical J-couplings were calculated in model N-acetylated
aldopyranosides using density functional theory (DFT) to predict the effect of OH vs NHCOCH3 substi-
tution at C2 on JCH and JCC values in aldopyranosyl rings. The synthetic method was also modified to
accommodate 15N- and 13C-labeling within the N-acetyl side-chain, and some J-couplings involving 1H,
13C, and 15N atoms in 2-[1,2-13C2;15N]acetamido-2-deoxy-D-[1-13C]glucose were measured and interpreted.
Introduction
colipids.1 For example, the core pentasaccharide (1) of N-linked
glycoproteins such as human IgG contains two R-Manp, one
â-Manp, and two â-GlcNAcp residues, with one of the latter
attached covalently to Asn of the protein.2 Structure 1 is
2-Aminosugars are common constituents of biologically im-
portant oligosaccharides appended to glycoproteins and gly-
* Address correspondence to this author.
† Department of Chemistry and Biochemistry, University of Notre Dame.
‡ Omicron Biochemicals, Inc.
(1) Allen, H. J.; Kisailus, E. C. Glycoconjugates: Composition, Structure
and Function; MarcelDekker: New York, 1992.
§ Radiation Laboratory, University of Notre Dame.
10.1021/jo051510k CCC: $33.50 © 2006 American Chemical Society
Published on Web 12/15/2005
466
J. Org. Chem. 2006, 71, 466-479

Synthesis of 2-Amino-2-deoxy-D-[1-13C]aldohexoses
SCHEME 1
subsequently decorated in vivo with D-galactose, D-mannose,
L-fucose, and/or sialic acid. Ongoing structural studies of 1, and
sufficiently simple, general, and/or reliable for routine use in
isotope labeling of oligosaccharides. In this report, a simple and
general chemical route to 2-amino-2-deoxy-D-aldohexoses is
2
described that allows site-specific labeling with 13C and/or H
at virtually any carbon or hydrogen, and also supports 15N
labeling of the side-chain (Scheme 1). Using this approach,
2-amino-2-deoxy-D-[1-13C]aldohexoses and their 2-acetamido-
2-deoxy methyl glycopyranosides were prepared, and NMR and
computational (density functional theory; DFT) studies were
conducted to extend current knowledge of the structural
compounds derived from it, involve the use of multiple NMR
parameters such as trans-glycoside J-couplings3a-c (2JCC, 3JCC
and 3JCH), dipolar3d-g and quadrupolar4 couplings (1DCC, 1DCH
,
,
1
dependencies of J-couplings involving H, 13C, and/or 15N in
aminosugars.
2H QCC) and spin relaxation to investigate their conformations
and dynamics in solution. These studies are facilitated by
isotopic labeling with 13C, 2H, and/or 15N at one or more sites.5
Synthetic methods are available to prepare a wide range of
D-mannose and D-galactose 13C isotopomers,6 and sialic acid
labeled at C1, C2, and/or C3 is prepared enzymically from
unlabeled ManNAc and labeled pyruvate using sialic acid
aldolase.7 However, current methods8 to label GlcNAc, and to
label sialic acid at C4-C9 using labeled ManNAc, are not
Experimental Section
A. General Procedure for the Synthesis of D-Aldopentopy-
ranosylamines 2-5 (Scheme 1).9a Absolute MeOH (6 mL) was
added to a 250 mL Erlenmeyer flask, and the flask was sealed with
a rubber septum, weighed, and cooled to 4 °C in an ice bath. The
cooled solution was aerated with NH3 gas for ∼3 min and weighed
to determine the amount of dissolved gas (∼1.2 g, ∼71 mmol).
Dry, finely powdered D-arabinose, D-lyxose, D-ribose, or D-xylose
(Sigma) (5.0 g, 33.3 mmol) was added and the flask was sealed
with a rubber septum. After gentle swirling, the pentose slowly
dissolved, yielding a straw-colored solution after ∼15 min. The
sealed reaction vessel was allowed to stand at room temperature
for 2 days, yielding crystals of the lyxo-3, ribo-4, and xylo-5
glycosylamines (Scheme 2). The crystals were removed, washed
with three 5-mL volumes of absolute MeOH, and dried. Isolated
yields ranged from 66% to 95%. For D-arabinosylamine 2, the
reaction solution was concentrated in vacuo, yielding a crude solid
determined by NMR to be >95% pure (Figure 1); unreacted
D-arabinose was the contaminant.
(2) (a) Fujii, S.; Nishiura, T.; Nishikawa, A.; Miura, R.; Taniguchi, N.
J. Biol. Chem. 1990, 265. (b) Dwek, R. A.; Lellouch, A. C.; Wormald, M.
R. J. Anat. 1995, 187, 279-292. (c) Wormald, M. R.; Rudd, P. M.; Harvey,
D. J.; Chang, S.-C.; Scragg, I. G.; Dwek, R. A. Biochemistry 1997, 36,
1370-1380.
(3) (a) Bose, B.; Zhao, S.; Stenutz, R.; Cloran, F.; Bondo, P. B.; Bondo,
G.; Hertz, B.; Carmichael, I.; Serianni, A. S. J. Am. Chem. Soc. 1998, 120,
11158-11173. (b) Cloran, F.; Carmichael, I.; Serianni, A. S. J. Am. Chem.
Soc. 1999, 42, 9843-9851. (c) Cloran, F.; Carmichael, I.; Serianni, A. S.
J. Am. Chem. Soc. 2000, 122, 396-397. (d) Tian, F.; Al-Hashimi, H. M.;
Craighead, J. L.; Prestegard, J. H. J. Am. Chem. Soc. 2001, 123, 485-492.
(e) Yi, X.; Venot, A.; Glushka, J.; Prestegard, J. H. J. Am. Chem. Soc.
2004, 126, 13636-13638. (f) Martin-Pastor, M.; Bush, C. A. Biochemistry
2000, 39, 4674-4683. (g) Kiddle, G. R.; Homans, S. W. FEBS Lett. 1998,
436, 128-130.
1
Compounds 2-5 were characterized by H and 13C NMR by
dissolving them in 0.1 M Na phosphate buffer at pD 10.5. Under
these conditions, hydrolysis rates were very low and spectral
characteristics allowed complete analyses (Figure 1).
(4) (a) Homans, S. W.; Kjellberg, A. J. Magn. Reson. 2001, 151, 90-
93. (b) Bose-Basu, B.; Zajicek, J.; Bondo, G.; Zhao, S.; Kubsch, M.;
Carmichael, I.; Serianni, A. S. J. Magn. Reson. 2000, 144, 207-216.
(5) Milton, M. J.; Harris, R.; Probert, M. A.; Field, R. A.; Homans, S.
W. Glycobiology 1998, 8, 147-153.
B. General Procedure for the Synthesis of 13C-Labeled
2-Amino-2-deoxy-D-[1-13C]aldohexoses. In a well-vented hood,
(6) Serianni, A. S.; Vuorinen, T.; Bondo, P. J. Carbohydr. Chem. 1990,
9, 513-541.
(9) (a) This procedure differs from that reported previously (Isbell, H.
S.; Frush, H. L. J. Org. Chem. 1958, 23, 1309-1319) in that only NH3(g)
is used to effect the conversion of aldose to glycosylamine. (b) Vogel, A.
I. Elementary Practical Organic Chemistry; John Wiley and Sons: New
York, 1958; pp 758-759.
(7) (a) Auge´, C.; David, S.; Gautheron, C.; Malleron, A.; Cavaye´, B.
New J. Chem. 1988, 12, 733-744. (b) Fitz, W.; Schwark, J.-R.; Wong,
C.-H. J. Org. Chem. 1995, 60, 3663-3670.
(8) Walker, T. E.; Barker, R. Carbohydr. Res. 1978, 64, 266-270.
J. Org. Chem, Vol. 71, No. 2, 2006 467

Zhu et al.
SCHEME 2. Preferred Anomeric Configurations and Ring
Conformation of D-Aldopentosylamines 2-5 in Aqueous
Solution Observed after Dissolution of Solid Samples
synthesis, the gluco product predominated (∼80%), and the elution
profile revealed some overlap between the two products, with the
gluco epimer eluting first. Careful pooling of fractions gave pure
samples of each amine: from 20.1 mmol of 2, GlcNH3Cl, 0.49 g,
2.3 mmol; ManNH3Cl, 0.15 g, 0.7 mmol; mixture, 0.83 g, 3.8 mmol,
total yield, 34%. In the GalNH2/TalNH2 synthesis, galacto eluted
first, followed by talo, with some overlap: from 33.5 mmol of 3,
GalNH3Cl, 1.4 g (6.5 mmol), TalNH3Cl, 1.5 g (7.0 mmol), total
yield, 62% (6.2 g). For the GulNH2/IdoNH2 synthesis, gulo eluted
first, followed by ido, with some overlap: from 35.8 mmol of 5,
GulNH3Cl, 1.0 g (4.6 mmol); IdoNH3Cl, 1.0 g (4.6 mmol), total
yield, 63% (4.9 g). Pooled fractions were concentrated at 30 °C in
vacuo to dryness to yield a crude solid.
C. General Procedure for N-Acetylation. 2-Amino-2-deoxy-
D-aldohexose hydrochloride (0.15 g, 0.6 mmol) was dissolved in a
minimal volume of H2O, and the solution pH was adjusted to 6.6
with use of Dowex 1 × 8 (200-400 mesh) (OH-) anion-exchange
resin. The resin was removed by filtration and washed, the filtrates
were combined (total volume of ∼6 mL), and EtOH (25 mL) was
added. Potassium acetate (0.1 g, 1 mmol, dissolved in a small
amount of H2O) and EEDQ (2-ethoxy-1-ethoxycarbonyl-1,2-
dihydroquinone) (0.18 g, 0.7 mmol dissolved in 10 mL of EtOH)
were added to the aminosugar solution with stirring. The reaction
vessel was sealed with a rubber septum, covered with aluminum
foil, and allowed to stand at 30 °C for 24 h. Reaction progress was
monitored by TLC (ethyl acetate:pyridine:H2O, 10:4:3): Rf 0.31
(free amine), Rf 0.71 (N-acetyl derivative).
K13CN (99 atom % 13C; Cambridge Isotope Laboratories, Inc.) (4.0
g, 60 mmol) was dissolved in 15 mL of water in a 25 mL flask, a
pH electrode was inserted, and the solution was stirred gently with
a magnetic stirrer. The solution pH was adjusted to 8.6 with the
slow addition of glacial acetic acid (∼2 mL), followed by batchwise
addition of the D-aldopentosylamine (3.0 g, 20.1 mmol). After amine
addition, the pH was immediately readjusted to pH 8.6 with acetic
acid, and the reaction was allowed to proceed at room temperature
for ∼6 min while maintaining the pH at 8.6. The pH was then
lowered to 4.2 with acetic acid, and the resulting solution was
purged with N2 through a methanolic-KOH trap to remove excess
H13CN.6
After the reaction was complete, the reaction mixture was
evaporated from 20 mL of H2O to remove ethanol. EEDQ was
extracted with 6 × 15 mL volumes of CH2Cl2, and the resulting
aqueous solution was deionized with batchwise additions of excess
Dowex 50 × 8 (20-50 mesh) (H+) and Dowex 1 × 8 (20-50
mesh) (OAc-) ion-exchange resins. The deionized solution was
concentrated at 30 °C in vacuo to remove acetic acid, and the solid
residue was crystallized from water/MeOH. N-Acetylation yields
ranged from 45% to 80%. The crude N-acetylated aminosugars were
glycosidated without further purification.
D. General Procedure for Methyl Glycosidation. 2-Acetamido-
2-deoxy-D-aldohexose (0.19 g, 0.85 mmol) was dissolved in 20 mL
of absolute MeOH, and Dowex HCR-W2 (H+) resin10 was added
with gentle stirring. The reaction mixture was refluxed for 1 h then
filtered, and the filtrate was concentrated to give crude glycoside
(∼0.2 g, 0.85 mmol). The product was dissolved in a minimal
volume of H2O and loaded on a column (2.5 cm i.d. × 105 cm)
containing Dowex 1 × 8 (200-400 mesh) (OH-) ion-exchange
resin.11a The column was eluted with distilled H2O (1.5 mL/min),
and fractions testing positive for aminosugar (phenol-H2SO4
assay)11b were pooled and concentrated at 30 °C in vacuo. For the
gluco isomer, the R-pyranoside 7 eluted first (60 mg, 0.25 mmol,
29%), followed by the â-pyranoside 8 (65 mg, 0.28 mmol, 33%).
Yields were comparable for the other glycosidation reactions. Eight
methyl 2-acetamido-2-deoxy-D-[1-13C]aldopyranosides 6-13 were
prepared (Scheme 3). For galacto and talo configurations, only one
product (R-pyranosides 6 and 13) was isolated by chromatography.
For the gulo and manno configurations, the R-pyranoside eluted
first, followed by the â-pyranoside.
Pd/BaSO4 catalyst (5%; Aldrich) (1.0 g) was suspended in a
minimal volume of H2O and prereduced at atmospheric pressure
over an H2 atmosphere at room temperature, using a custom-made
catalytic reduction apparatus.9b The HCN-purged reaction mixture
was then added, and aldononitrile reduction was initiated. H2 gas
consumption was monitored via gas displacement in a water ballast,
and the reaction was terminated when consumption ceased (∼18
mmol of H2 was consumed; 3-5 h). The suspension was filtered
to remove the catalyst (glass-fiber filter), and the filtrate was
concentrated at 30 °C in vacuo to a minimal volume to obtain the
crude C2-epimeric aminosugars (∼3.5 g).
The crude aminosugar solution was applied to a column (5.0
cm i.d. × 53 cm) containing Dowex 50 × 8 (200-400 mesh) ion-
exchange resin in the H+ form.8 The column was eluted with a 4
L linear HCl gradient (0-0.5 M) at a flow rate of ∼1.2 mL/min.
Fractions (∼20 mL) were assayed by spotting aliquots (3 µL) from
each fraction onto a Whatman filter paper, drying, and spraying
with a ninhydrin reagent (0.2 g of ninhydrin in 70 mL of absolute
ethanol, 2 mL of H2O, and 8 mL of pyridine; amine-containing
spots appear purple after heating). A broad band of aminosugar
eluted between fractions 267 and 307. In the GlcNH2/ManNH2
E. General Procedure for 15N-Labeling. The synthetic proce-
dure described for 13C-labeling (see above) was followed, but sub-
stituting 15NH3(g) for unlabeled NH3(g) in the protocol. 15NH3(g)
from a small low-pressure gas cylinder (Cambridge Isotope Lab-
oratories, Inc.) was slowly bubbled into the cold methanol solvent
held in a tared flask. After ∼1 min, the gas flow was stopped and
the reaction vessel weighed to determine the weight of the dissolved
FIGURE 1. 1H NMR spectrum (600 MHz) of D-arabinopyranosy-
lamine (2) in aqueous (2H2O) solution at pD 10.5 (0.1 M Na phosphate
buffer) and 20 °C, showing signal assignments. At this pH, sample
stability is high and spectral resolution sufficient to permit the
measurement of J-couplings; significant line broadening is observed
under neutral and acidic pH values, where hydrolysis is more rapid.
(10) Podlasek, C. A.; Wu, J.; Stripe, W. A.; Bondo, P. B.; Serianni, A.
S. J. Am. Chem. Soc. 1995, 117, 8635-8644.
(11) (a) Austin, P. W.; Hardy, F. E.; Buchanan, J. C.; Baddiley, J. J.
Chem. Soc. 1963, 5350-5353. (b) Hodge, J. E.; Hofreiter, B. T. Methods
Carbohydr. Chem. 1962, 1, 380-394.
468 J. Org. Chem., Vol. 71, No. 2, 2006

Synthesis of 2-Amino-2-deoxy-D-[1-13C]aldohexoses
SCHEME 3
gas. The process was repeated if necessary. A 4-fold molar excess
of 15NH3(g) was found to be sufficient to convert the pentoses
almost completely into glycosylamines. In the present work, no
effort was made to recover 15NH3(g) lost to the atmosphere,
although this could be achieved by passing the released gas through
an aqueous HCl trap.
F. Mass Spectrometry. Mass spectra (FAB mode) of [1-13C]-
6-13 were obtained on a JEOL JMS-AX505HA mass spectrometer
with 3-nitrobenzyl alcohol as the solvent. All spectra showed (M
+ 1)+ ions at m/z 237 and (M + Na)+ ions at m/z 259 (Figure S1,
Supporting Information). Fragmentation ions were also observed
at m/z 205 (M - 31) and m/z 187 (M - 49) presumably caused by
the loss of OCH3 and OCH3 + H2O, respectively. HRMS data for
the aldopentosylamines 2-5 are given in Table S1 (Supporting
Information). Exact masses are provided for both the monomeric
and dimeric forms of these amines, each of which are detected by
MS analysis (see the Supporting Information for the structure of
the dimer).
D-[5-13C]Xylose was prepared from D-[6-13C]glucose.16 The latter
compound was converted to 1,2-O-isopropylidene-R-D-[6-13C]-
glucofuranose,17 which was treated consecutively with Pb(OAc)4,
NaBH4, and aqueous HCl to give D-[5-13C]xylose after purification
by chromatography on Dowex 50 × 8 (200-400 mesh)(Ca2+).18a
Treatment of D-[5-13C]xylose with NaMoO418b gave D-[5-13C]lyxose,
which was converted to [5-13C]3 and used to determine preferred
anomeric configuration in solution (see below).
For NMR studies of amine hydrolysis, reactions were conducted
at pH 6.5-8.5 for defined time periods, and then quenched by
raising the solution pH to ∼10.5 with NaOH. At pH 10.5, hydrolysis
rates were low enough to permit the characterization of reaction
intermediates by 2D NMR where more extended acquisition times
were required. Characterization of the intermediates observed during
â-D-xylopyranosylamine hydrolysis was assisted by the use of 2D
HMQC12 and HMQC-TOCSY19 spectra and D-[1-13C]- and [5-13C]-
xylopyranosylamines as substrates.
G. NMR Spectroscopy. 1D 1H and 13C{1H} NMR spectra were
obtained in 2H2O on a Varian UnityPlus 600 MHz FT-NMR
spectrometer. Samples were analyzed in 3 mm NMR tubes with a
Nalorac dual 13C/1H microprobe. 13C Signal assignments for 6-13
were made via 2D HMQC spectra12 and interpretation of JCC values
Computations
A. Geometry Optimizations. Ab initio molecular orbital
calculations (geometric optimizations and J-coupling calculations)
were conducted with use of Gaussian98.20 For geometry optimiza-
tion, density functional theory (DFT) was employed with the
B3LYP functional21 and the 6-31G* basis set22 (B3LYP/6-31G*).
2
observed in [1-13C] derivatives. JC1,H2 coupling signs were
determined from analyses of relative cross-peak displacements in
1H-1H TOCSY spectra13a,b of [1-13C] derivatives.
1H-1H and 13C-1H NMR spin-spin coupling constants (J-
1
(15) Serianni, A. S.; Nunez, H. A.; Barker Carbohydr. Res. 1979, 72,
71-78.
couplings) in 6-13 were determined by simulation of H spectra
with MacNUTs (Acorn NMR) (Figure S2, Supporting Informa-
tion).14 Couplings are reported in hertz and are accurate to within
(0.1 Hz unless otherwise indicated.
(16) King-Morris, M. J.; Bondo, P. B.; Mrowca, R. A.; Serianni, A. S.
Carbohydr. Res. 1988, 175, 49-58.
(17) Schmidt, O. T. In Methods in Carbohydrate Chemistry; Whistler,
R. L., Wolfrom, M. L., Eds.; Academic Press: New York, 1963; Vol. 2,
pp 318-325.
H. High-Pressure Liquid Chromatography. HPLC was per-
formed on a ternary HPLC pump connected to an RI detector. The
RI detector was interfaced with desktop PC running ChromPerfect
HPLC software (Justice Laboratory Software, Version 5.1.0) for
data processing. A Rezex RCM (Ca2+) monosaccharide column
(7.8 × 300 mm; Phenomenex) contained in a column heater
equipped with a heater controller was equilibrated at 85 °C, using
deionized/degassed water as the solvent. Solvent flow rate was set
at 0.6 mL/min, and sample retention times were recorded in minutes
using the ChromPerfect software.
(18) (a) Angyal, S. J.; Bethell, G. S.; Beveridge, R. Carbohydr. Res.
1979, 73, 9-18. (b) Hayes, M. L.; Pennings, N. J.; Serianni, A. S.; Barker,
R. J. Am. Chem. Soc. 1982, 104, 6764-6769.
(19) Wijmenga, S. S.; Hallenga, K.; Hilbers, C. W. J. Magn. Reson. 1989,
84, 634-642.
(20) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;
Ortiz, J. V.; Baboul, A. G.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz,
P.; Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-
Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P.
M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.;
Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian98, Revision A.9;
Gaussian, Inc.: Pittsburgh, PA, 1998.
I. Mechanistic Studies of â-D-Xylopyranosylamine Hydrolysis.
â-D-[1-13C]- and [5-13C]Xylopyranosylamines were prepared from
D-[1-13C]xylose and D-[5-13C]Xylose, respectively. D-[1-13C]Xylose
was prepared from D-threose and K13CN as described previously.6,15
(12) Bax, A.; Subramanian, S. J. Magn. Reson. 1986, 67, 565-569.
(13) (a) Wijmenga, S. S.; Hallenga, K.; Hilbers, C. W. J. Magn. Reson.
1989, 84, 634-642. (b) Serianni, A. S.; Podlasaek, C. A. Carbohydr. Res.
1994, 259, 277-282.
(21) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(22) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56,
2257-2261.
(14) MacNUTs Pro; Acorn NMR, Inc.: Livermore, CA.
J. Org. Chem, Vol. 71, No. 2, 2006 469

Zhu et al.
SCHEME 4
1
TABLE 1. 1Ha and 13Cb Chemical Shifts, and H-1H Spin-Coupling
B. J-Coupling Calculations. NMR J-couplings were calculated
by DFT using a modified version of Gaussian9423 and an extended
basis set24 ([5s2p1d|3s1p]) designed to recover the Fermi contact
contribution to the coupling. Actual calculated couplings are
reported (i.e., the reported values were not scaled). Given the
relatively small magnitudes of 2J and 3J, absolute errors are
estimated to be (0.4 Hz.
Eight model compounds 14-21 were chosen for the computa-
tions that mimic the methyl glycosides of R-Glc (14), R-GlcNAc
(15), R-Man (16), R-ManNAc (17), â-Glc (18), â-GlcNAc (19),
â-Man (20), and â-ManNAc (21) (Scheme 4). These structures lack
OH substituents at C4 and C6. Initial exocyclic torsions used for
geometry optimization were as follows: C2-C1-O1-CH3, 180°;
C1-C2-O2-O2H, 180°; C2-C3-O3-O3H, 180°; H2-C2-
N2-N2H, 180°. In all calculations, the amide configuration was
trans (amide hydrogen trans to amide oxygen).
Constants,c in D-Aldopentopyranosylamines 2-5 in H2O (pD 10.5;
2
0.1 M Na Phosphate Buffer)
aldopentosylamine ring configuration
arabino-2
lyxo-3
ribo-4
xylo-5
δH1
4.056
3.446
3.610
3.884
3.793
3.566
90.24
72.95
75.75
71.29
69.63
8.6
4.235
3.828
3.526
3.741
3.845
3.168
86.02
73.73
76.47
68.92
68.97
1.1
3.4
9.7
5.5
10.6
-11.3
1.2
4.383
3.415
4.093
3.780
3.678
3.540
86.15
72.38
72.60
69.53
65.87
8.4
3.990
3.103
3.378
3.561
3.851
3.252
88.54
76.83
δH2
δH3
δH4
δH5
δH5′
δC1
δC2
δC3
79.26
72.13
68.92
δC4
δC5
3JH1,H2
3JH2,H3
3JH3,H4
3JH4,H5
3JH4,H5′
2JH5,H5′
3JC5,H1
8.7 (7.8)e
9.1 (9.3)
9.2 (9.1)
5.4 (5.5)
10.8 (10.5)
-11.4 (-11.6)
9.5 (10.1)d
3.5 (3.5)
2.3 (1.5)
1.4 (2.1)
-13.0 (-12.8)
2.9
2.9
4.9
10.1
Results and Discussion
A. Synthesis of 13C-Labeled Aminosugars. The route shown
in Scheme 1 is a modification of a previous synthetic approach
used to prepare D-[1-13C]GlcNH2 and D-[1-13C]ManNH2 from
N-benzyl-D-arabinosylamine and K13CN.8 The new route has
significant advantages over the prior approach. Simple glyco-
sylamines are employed that are easily prepared on a large scale
and in good yield, and are obtainable in crystalline or amorphous
solid forms. These amines exhibit good chemical stability if
stored in a dry container at 4 °C. Introduction of 15N is also
simplified (see below), requiring relatively inexpensive 15NH3
gas rather than 15N-benzylamine as the isotope source reagent.
Cyanohydrin reduction yielded two C2-epimeric 2-aminoal-
doses, in some cases with one isomer highly favored (e.g.,
GlcNH2/ManNH2, ∼3/1). These products were separated as the
free amines by chromatography on Dowex 50 × 8 (200-400
mesh) (H+) ion-exchange resin. Alternatively, the amine mixture
can be N-acetylated, and the resulting 2-acetamido epimers
separated by chromatography on Dowex 50 × 8 (200-400
mesh) (Ca2+) ion-exchange resin analogous to the analytical
scale separations (Table S1, Supporting Information; see below).
In this work, three of the four aldopentosylamines (2, 3, and 5)
were used as cyanohydrin substrates; amine 4 is expected to
behave similarly.
-11.2
a In ppm, (0.001 ppm relative to internal 3-(trimethylsilyl)-1-propane-
sulfonic acid, 20 °C. In ppm, (0.01 ppm; 13C signal assignments were
b
determined from 2D 13C-1H HMQC spectra. c In Hz, (0.1 Hz. d Values
in parentheses are couplings observed in methyl â-D-arabinopyranoside (ref
10). e Values in parentheses are couplings observed in methyl â-D-
xylopyranoside (ref 10).
rapidly and in good yield. Separation of the anomeric mixtures
was achieved by chromatography on Dowex 1 × 8 (200-400
mesh) (OH-) ion-exchange resin, and the purified products were
analyzed by HPLC (Table S2, Supporting Information). Reten-
tion times varied from ∼11.5-14.1 min, with the galacto
configuration considerably more retained. HPLC of the reducing
sugars showed progressively increasing retention times in the
series ManNAc, GlcNAc, and GalNAc, in agreement with the
glycoside results (Table S2, Supporting Information). Interest-
ingly, HPLC of Man, Glc, and Gal showed Glc eluting first,
followed by the nearly coeluting Man and Gal, indicating that
the N-acetyl group modulates binding interactions on the column
and thus affects the elution order.
B. D-Aldopentosylamine Conformation. Studies of hydroly-
sis rates of D-xylopyranosylamine at different solution pH values
revealed good stability at alkaline pH (e.g., pD 10.5). 13C{1H}
and 1H NMR spectra of 2-5 were obtained under these
conditions, yielding high-resolution spectra (Figure 1, Table 1).
The D-aldopentosylamines crystallize/solidify as pyranoses
Conversion of the 13C-labeled N-acetylated aminosugars to
their methyl pyranosides by Fischer glycosidation proceeded
(23) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G.
A.; Montgomery, J. A.; Raghavachari, K.; AlLaham, M. A.; Zakrzewski,
V. G.; Ortiz, J. V.; Foresman, J. B.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head
Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian94, Gaussian, Inc.:
Pittsburgh, PA, 1995.
1
(initially demonstrated by obtaining H NMR spectra of 2-5
in anhydrous DMSO-d6 solvent; data not shown). On the basis
of 1H-1H spin-coupling analysis in 2H2O solutions at pD 10.5
(0.1 M Na phosphate buffer) (Table 1), the D-lyxo-3, D-ribo-4,
and D-xylo-5 configurations crystallize in the 4C1 ring conforma-
(24) Stenutz, R.; Carmichael, I.; Widmalm, G.; Serianni, A. S. J. Org.
Chem. 2002, 67, 949-958.
470 J. Org. Chem., Vol. 71, No. 2, 2006

Synthesis of 2-Amino-2-deoxy-D-[1-13C]aldohexoses
SCHEME 5
SCHEME 6
1
tion, whereas D-arabino-2 prefers the C4 form (Scheme 2).
Anomeric configurations in the crystalline (or amorphous solid)
state were determined from 3JH1,H2 values observed in aqueous
solution at pD 10.5 in 2 (8.6 Hz; R-pyranose), 4 (8.4 Hz;
â-pyranose), and 5 (8.7 Hz; â-pyranose) (Scheme 2). For 3,
3JH1,H2 could not be used reliably because the coupling is small
in both anomers. In this case, the assignment was made by
CN- predominating at pH 10.5 (Scheme 5). Competing with
cyanide addition is glycosylamine hydrolysis (Scheme 6).
Hydrolysis rates are much greater at pH 7.5 than at pH 10.5,
presumably because the protonated imine is more prone to attack
by water. Optimal pH conditions were examined that promote
cyanide addition while reducing glycosylamine and aldononitrile
hydrolysis, and pH 8.5 was found to be optimal. At this pH,
cyanide addition was rapid and essentially irreversible, and
minimal (<10%) hydrolysis was observed.
3
measuring JC5,H1 in [5-13C]3; the observed coupling (1.2 Hz)
confirmed the â-configuration (a coupling >5 Hz is expected
in the R-anomer).25a
The favored pyranosyl ring conformations of 2-5 orient the
C1-N1 bond equatorial in aqueous solution at pD 10.5. For 2
and 5, preferred anomeric configurations and ring conformations
are identical to those observed in the D-aldopentopyranoses25a
and methyl D-pentopyranosides.10 For 3 and 4, replacement of
OH with NH2 at C1 enhances the stability of the 4C1 form (i.e.,
â-D-lyxopyranose is ∼87% 4C1 and â-D-ribopyranose is ∼74%
4C1,25a whereas 4C1 is almost exclusively favored by the
corresponding glycosylamines). For 3, the favored anomer shifts
from R in the reducing sugar (1C4 (∼38%) and 4C1 (∼62%))25a
to â (almost exclusively 4C1) in the glycosylamine. These
observations suggest that preferred anomeric configuration25b
and ring conformation in 2-5 are correlated and influenced
strongly by the orientational requirements of the C1-N1 bond.
C. D-Xylopyranosylamine Hydrolysis. Solution conditions
for cyanide addition to 2-5 significantly affect the overall yield
of aminosugar. Cyanide addition under mildly basic conditions
(pH 7.5) probably occurs via the acyclic protonated imine, which
is a more potent electrophile than the unprotonated imine that
predominates at higher pH (pH 10.5) (Scheme 5). At higher
pH, cyanohydrin hydrolysis rates increase,25c which reduces the
aminosugar yield. At the same time, pH 7.5 favors the
protonated HCN (pKa ) 9.2), which is less nucleophilic than
13C NMR studies were conducted to monitor the hydrolysis
of â-D-[1-13C]xylopyranosylamine 5 under different solution
conditions. At pH 10.5, hydrolysis was slow, but anomerization
to R-D-[1-13C]xylopyranosylamine was observed early in the
reaction, followed by the slow appearance of putative xylo-
furanosylamines (data not shown). At alkaline pH, anomerization
between xylopyranosylamine anomers appears to be more
kinetically favored than pyranosylamine-furanosylamine and
furanosylamine-furanosylamine anomerization.
A typical time-lapse 13C{1H} NMR spectrum of a â-D-
[1-13C]xylopyranosylamine hydrolysis reaction mixture after
∼10 min at pH 6.5 showed 13C signals arising primarily from
seven labeled species (Figure 2). Three of these signals were
assigned to â-D-[1-13C]xylopyranosylamine (âpA) (1JC1,H1
)
3
154.3 Hz; JH1,H2 ) 8.8 Hz), R-D-[1-13C]xylopyranose (Rp)
3
(1JC1,H1 ) 169.5 Hz; JH1,H2 ) 3.7 Hz), and â-D-[1-13C]-
xylopyranose (âp) (1JC1,H1 ) 161.5 Hz; 3JH1,H2 ) 7.9 Hz) based
1
27
on known C1 chemical shifts26 and JC1,H1
.
The HMQC
spectrum (Figure 3) obtained without 13C-decoupling yielded
characteristic 1H chemical shifts for these three species. HMQC-
TOCSY spectra (Figure 4) extended the proton correlations
beyond H1, confirming the assignments. The 13C signal at
81.03 ppm (Figure 2) was assigned to R-D-[1-13C]xylopyrano-
3
1
(25) (a) Wu, J.; Bondo, P. B.; Vuorinen, T.; Serianni, A. S. J. Am. Chem.
Soc. 1992, 114, 3499-3505. (b) This argument rests on the assumption
that anomeric equilibrium has been reached in aqueous solutions of the
aldopentopyranosylamines at pD 10.5. In contrast to aldopentoses where
equilibration rates are relatively rapid, anomerization rates of the glycosyl-
amines at this pH are considerably slower. However, hydrolysis mechanism
studies of 5 at pH 6.5 (see text) where anomerization rates are more rapid
showed a greater â-pyranose/R-pyranose ratio compared to that observed
in free xylose, supporting the conclusion that the stability of the equatorial
C1-N1 bond is enhanced in the amines. (c) Serianni, A. S.; Nunez, H. A.;
Barker, R. J. Org. Chem. 1980, 45, 3329-3341.
sylamine (RpA) based on JH1,H2 (∼3 Hz) and JC1,H1 (∼158
Hz), and on its rate of formation during the reaction. The R/â
ratio of D-xylopyranosylamines observed throughout the course
of hydrolysis was considerably smaller than that observed for
(26) Bock, K.; Pedersen, C. AdV. Carbohydr. Chem. Biochem. 1983, 41,
27-66.
(27) Drew, K. N.; Zajicek, J.; Bondo, G.; Bose, B.; Serianni, A. S.
Carbohydr. Res. 1998, 307, 199-209.
J. Org. Chem, Vol. 71, No. 2, 2006 471

Zhu et al.
FIGURE 4. 2D 13C-1H HMQC-TOCSY spectrum at 20 °C of the
reaction mixture described in Figure 2 (60 ms mixing time). Note the
significant TOCSY transfer for âpA, CA, âfA, and âp, which contain
3
large JH1,H2 values. In contrast, TOCSY correlations for RpA, RfA,
and Rp are more limited due to inefficient magnetization transfer from
H1 and H2 in these forms (3JH1,H2 is small).
FIGURE 2. 13C NMR spectrum at 20 °C of a â-D-[1-13C]xylopyra-
nosylamine hydrolysis reaction mixture (pH 6.5, ∼10 min reaction
time). Seven enriched signals are detected in the anomeric region: â-D-
[1-13C]xylopyranose (96.94 ppm, âp), R-D-[1-13C]xylopyranose (92.53
ppm, Rp), â-D-[1-13C]xylofuranosylamine (91.55 ppm, âfA), acyclic
[1-13C]carbinolamine (87.66 ppm, CA), R-D-[1-13C]xylofuranosylamine
(86.77 ppm, RfA), â-D-[1-13C]xylopyranosylamine (86.18 ppm, âpA),
and R-D-[1-13C]xylopyranosylamine (81.03 ppm, RpA). The small
signals near 82 and 90 ppm were not identified.
FIGURE 5. 2D 13C-1H HMQC-TOCSY spectrum at 25 °C of a
reaction mixture obtained from the hydrolysis of â-D-[5-13C]xylopy-
ranosylamine (pH 6.5; 80 ms mixing time). Note the significantly
different C5 chemical shifts for the D-xylofuranosylamines (RfA, ∼61
ppm; âfA, ∼66 ppm); similar differences have been observed for
xylofuranoses.28 Note also the detection of H1 for âp, CA, âpA, and
âfA (i.e., completeTOCSY transfer through the structure), whereas H1
(andother) correlations are absent for RpA, RfA, and Rp.
xylopyranosylamine hydrolysis (Figure 5). These data, and those
from the HMQC-TOCSY spectrum (Figure 4), revealed C5
chemical shifts for these species at 61.2 (R) and 65.8 ppm (â),
values consistent with exocyclic hydroxymethyl carbons of
xylofuranosyl anomers.28Assignment of the two C1 signals to
specific furanose anomers was based on the reported effect of
cis-1,2 electronegative substituents on C1 chemical shifts in
furanoses; the trans-N1,O2 anomer (â) is expected to yield a
C1 signal downfield of the cis-N1,O2 anomer (R).29,30
FIGURE 3. 2D 13C-1H HMQC spectrum (13C-coupled) at 20 °C of
the reaction mixture described in Figure 2. Data were used to assign
the H1 resonances of each form, each of which appears as paired signals
1
in the 1H dimension due to JCH. These assignments were used to
3
1
determine JH1,H2 and JC1,H1 values (see text) from the 1H NMR
spectrum of the same mixture (data not shown).
D-xylopyranoses,27 reflecting the increased preference for an
equatorial orientation of the C1-N1 bond in the amines (see
above).
The three remaining signals in Figure 2 were assigned by
using NMR and reaction kinetics data. Two of these signals at
91.55 (1JC1,H1 ) 153.1 Hz) and 86.77 ppm (1JC1,H1 ) ∼160
Hz) were assigned to â-D-xylofuranosylamine (âfA) and R-D-
xylofuranosylamine (RfA), respectively. These assignments were
based partly on HMQC-TOCSY data obtained on D-[5-13C]-
The remaining signal at 87.66 ppm in Figure 2 was assigned
to the acyclic carbinolamine (CA), a presumed hydrolytic
(28) Snyder, J. R.; Serianni, A. S. Carbohydr. Res. 1987, 163, 169-
188.
(29) Ritchie, R. G. S.; Cyr, N.; Korsch, B.; Koch, H. J.; Perlin, A. S.
Can. J. Chem. 1975, 53, 1424-1433.
(30) Serianni, A. S.; Barker, R. J. Org. Chem. 1984, 49, 3292-3300.
472 J. Org. Chem., Vol. 71, No. 2, 2006

Synthesis of 2-Amino-2-deoxy-D-[1-13C]aldohexoses
SCHEME 7
SCHEME 8
and the same may be true for R-D-xylofuranosylamine. These
conclusions are consistent with conformational predictions
reported recently for aldofuranosylamines based on DFT
calculations.31
D. 1H and 13C NMR Analyses of 6-13. 1H and 13C chemical
shifts and JHH, JCH, and JCC values in 6-13 are reported in
Tables 2-6. Substitution of an NHCOCH3 for an OH substituent
at C2 uniformly shifts δH2 downfield by 0.2-0.4 ppm; the
magnitude of the effect appears to depend on relative config-
uration at C1 and C2 (Table 2). The same substitution shifts
δC2 upfield by ∼18 ppm; in this case, the effect of rela-
tive configuration at C1 and C2 appears smaller (Table 3). In
both cases, however, a relatively wide range of chemical shifts
is observed: for δH2, 3.7-4.5 ppm and for δC2, 47-56 ppm,
depending on the configuration of the pyranosyl ring. Reso-
nances for the amide carbonyl carbons were observed at ∼175
ppm.
intermediate (Scheme 6). Formation of the acyclic carbinolamine
is rapid and its steady-state concentration remains relatively high
during the course of hydrolysis. An overall reaction scheme for
â-D-xylopyranosylamine hydrolysis at pH 6.5 is summarized
in Scheme 7.
Several interesting structural observations emerged from these
hydrolysis studies. The H1 signal of the acyclic xylocarbinol-
3JHH values in 6-13 are similar to those observed in the cor-
responding methyl aldohexopyranosides in most cases, although
3JH2,H3 is uniformly larger in the aminosugars by 0.7-1.6 Hz
(Table 4). The latter change is attributed mainly to local
electronegative substituent effects on the coupling rather than
3
amine (CA) appears as a doublet at ∼4.15 ppm with a JH1,H2
value of 8.8 Hz. When the same species contains 13C-enrichment
at C1, however, the H1 signal appears as a pseudoquartet split
by the directly bonded 13C. These results suggest that, in the
absence of 13C-enrichment, the C1-epimeric carbinolamines have
identical δH1. Upon 13C-enrichment, however, this degeneracy
3
to conformational change, since the adjacent JHH are less
3
affected, especially JH3,H4. Couplings involving the hydroxy-
methyl protons of 12 also differ from those in methyl â-D-
mannopyranoside, suggesting possible differences in C5-C6
1
is eliminated due to different JC1,H1 values in the two diaster-
1
eomers (150 and 158 Hz).
bond rotamer populations. Long-range (4-bond) H-1H coup-
3JH1,H2 values in the D-xylofuranosylamines also deserve
lings were observed between H1 and H3, and between H2 and
H4, in nearly all of the aminosugars studied (Table 4); the
presence of these couplings and their negative signs were
confirmed via spectral simulation. 4JH2,H4 is especially large in
13 (-1.2 Hz) in which the di-equatorial orientation of the
coupled protons creates a planar W-shaped arrangement along
the coupling pathway;32 similar couplings have been observed
previously in talopyranoses.33
3
comment. In the R-configuration, JH1,H2 ) 3.3 Hz, which is
similar to 3JH1,H2 values observed in 5-deoxy-R-D-xylofuranose
(3.9 Hz) and 5-O-methyl-R-D-xylofuranose (4.3 Hz).28 In
contrast, 3JH1,H2 in â-D-xylofuranosylamine (8.8 Hz) is signifi-
cantly larger than that observed in 5-deoxy-â-D-xylofuranose
(1.7 Hz) and 5-O-methyl-â-D-xylofuranose (1.4 Hz), suggesting
a major change in conformational preference. The strong
equatorial preference of the C1-N1 bond observed in 2-5 may
explain these observations. C1-N1 bond orientation dictates
ring conformation in xylofuranosylamines, in both cases leading
to preferred furanose conformations that orient this bond quasi-
equatorial. In the R-furanose, this conformation is near E2,
1JC1,H1 values show the expected dependence on anomeric
configuration, with axial C1-H1 bonds (â-pyranoses) yielding
smaller couplings by ∼10 Hz than equatorial C1-H1 bonds
(R-pyranoses) (Table 5).34 These couplings increase by
0.5-2.7 Hz in the aminosugars compared to related values in
the corresponding methyl aldohexopyranosides. Interestingly,
2
whereas in the â-furanose it is near E (Scheme 8). In the
xylofuranoses, the C1-O1 bond prefers a quasi-axial orientation
due to the anomeric effect (Scheme 8).30 Thus, 3JH1,H2 will not
change significantly when O1 is substituted by NH2 in the R-xylo
configuration, since the H1-C1-C2-H2 torsion angle remains
approximately gauche in both structures. In contrast, this torsion
angle changes from gauche to trans in the â-xylo configuration,
(31) Cloran, F.; Zhu, Y.; Osborn, J.; Carmichael, I.; Serianni, A. S. J.
Am. Chem. Soc. 2000, 122, 6435-6448.
(32) Barfield, M.; Dean, A. M.; Fallick, C. J.; Spear, R. J.; Sternwell,
S.; Westerman, P. W. J. Am. Chem. Soc. 1975, 97, 1482-1492.
(33) Snyder, J. R.; Johnston, E. R.; Serianni, A. S. J. Am. Chem. Soc.
1989, 111, 2681-2687.
(34) (a) Bock, K.; Lundt, I.; Pedersen, C. Tetrahedron Lett. 1973, 1037-
1040. (b) Bock, K.; Pedersen, C. J. Chem. Soc., Perkin Trans. 2 1974,
293-297. (c) Bock, K.; Pedersen, C. Acta Chem. Scand. 1975, B29, 258-
264.
3
causing a significant increase in JH1,H2. The unusually large
3JH1,H2 in â-D-xylofuranosylamine suggests that pseudorotation
is more restricted in this structure compared to â-D-xylofuranose,
J. Org. Chem, Vol. 71, No. 2, 2006 473

Zhu et al.
TABLE 2. 1H Chemical Shiftsa in Methyl 2-Acetamido-2-deoxy-D-[1-13C]Aldohexopyranosides
1H chemical shift (ppm)
compd
H1
H2
H3
H4
H5
H6
H6′
OCH3
CH3
R-GalNAc 6
4.824
4.935
4.787
4.904
4.494
4.472
4.819
4.719
4.658
4.776
4.969
4.197
3.917
3.943
3.655
3.738
3.358
4.302
4.511
4.072
4.151
3.943
3.910
3.906
3.741
3.761
3.584
3.588
3.924
3.831
3.721
4.064
3.933
4.015
4.063
3.508
3.495
3.488
3.477
3.860
3.542
3.652
4.011
4.009
3.953
3.997
3.700
3.741
3.510
3.560
4.157
3.433
3.459
3.919
3.973
3.813
3.855
3.907
3.965
3.987
4.024
3.816
3.943
4.021
3.852
3.933
3.790
3.832
3.811
3.852
3.801
3.822
3.801
3.842
3.825
3.797
3.869
3.419
3.511
3.415
3.515
3.557
3.670
3.435
3.533
3.630
3.430
3.519
2.075
R-GlcNAc 7
â-GlcNAc 8
2.067
2.088
R-GulNAc 9
â-ManNAc 12
2.095
2.082
R-TalNAc 13
2.075
2
1
a In H2O at 25 °C; ppm ((0.001 ppm) relative to the residual H2O signal (4.800 ppm); determined by spectral simulation. Values in italics are for the
corresponding methyl aldopyranosides (ref 10).
TABLE 3. 13C Chemical Shiftsa in Methyl 2-Acetamido-2-deoxy-D-[1-13C]Aldohexopyranosides
13C chemical shift (ppm)
compd
C1
C2
C3
C4
C5
C6
OMe
CO
MeAc
22.87
R-GalNAc 6
99.09
[99.1]
100.1
98.51
[98.6]
100.0
102.36
[102.3]
104.0
99.05
100.4
100.96
102.6
100.33
101.9
100.43
101.3
100.21
102.2
50.80
[50.8]
69.2
54.05
[54.3]
72.2
55.88
[56.1]
74.1
47.24
65.5
51.00
69.1
52.90
71.2
53.02
70.6
51.61
70.7
68.73b
[68.7]
70.5
71.57
[71.9]
74.1
74.40
[74.6]
76.8
69.43b
[69.4]
70.2
70.41
[70.4]
70.6
70.38
[70.9]
70.6
71.66
[71.6]
71.6
72.09
[72.2]
72.5
76.32
[76.3]
76.8
67.19
67.3
74.88
74.9
72.58
73.7
76.42
76.6
71.16
72.1
62.21
[62.1]
62.2
61.01
[61.4]
61.6
61.18
[61.5]
61.8
62.19
62.0
61.95
62.1
60.78
62.1
60.39
61.4
61.57
62.3
56.03
56.0
175.53
R-GlcNAc 7
â-GlcNAc 8
55.56
[55.6]
55.9
57.49
[57.2]
58.1
56.37
56.3
57.64
58.1
55.19
55.9
57.00
56.9
54.88
55.6
174.87
175.10
22.33
22.60
R-GulNAc 9
â-GulNAc 10
R-ManNAc 11
â-ManNAc 12
R-TalNAc 13
69.82b
71.4
69.52b
70.4
174.94
175.02
175.21
22.79
22.87
22.33
21.99
22.49
69.29b
72.3
70.87b
70.5
69.52b
71.8
67.08b
68.0
72.02
73.3
63.85
66.2
66.75
67.1
68.48
70.3
174.28
2
a In H2O at 25 °C; ppm ((0.01 ppm) relative to the C1 signal of R-D-[1-13C]mannopyranose (95.5 ppm). Values in brackets were taken from ref 26.
Values in italics are for the corresponding methyl aldopyranosides (ref 26). b Assignments may be reversed.
TABLE 4. 1H-1H Spin-Coupling Constantsa in Methyl
TABLE 5. 13C-1H Spin-Coupling Constantsa in Methyl
2-Acetamido-2-deoxy-D-[1-13C]Aldohexopyranosides
2-Acetamido-2-deoxy-D-[1-13C]Aldohexopyranosides
coupled nuclei
coupling constant (Hz)
compd
1,2
2,3
3,4
3.2
3.4
4,5 5,6 5,6′ 6,6′ 1,3 2,4
compd
1JC1,H1
2JC1,H2
3JC1,H3
3JC1,H5
3JC1,CH
3
R-GalNAc 6
3.8 11.0
3.9 10.3
3.6 10.7
3.8 9.8
8.5 10.4
8.0 9.5
4.1 3.7
1.5 4.5
0.9 3.2
1.2 5.0
0.7 8.1 4.2 -11.7 0.4 0.4
1.2 8.2 4.2 -11.7
R-GalNAc 6
R-GlcNAc 7
171.9
171.8
170.1
161.8
161.3
171.0
162.2
159.5
173.1
-2.5b
-2.7
0.4
0.7
br
1.1
1.2
5.4
∼0
0
1.4
1.4
4.4
4.4
4.4
4.7
4.6
4.6
4.6
4.5
4.4
R-GlcNAc 7
â-GlcNAc 8
9.0 10.1 2.3 5.6 -12.3 0.3 0.5
9.2 10.0 2.3 5.6 -12.3
8.7 10.0 2.2 6.1 -12.3
9.2 10.0 2.3 6.2 -12.4
3.9
9.8 10.0 2.3 5.3 -12.4
9.6
3.2
+1.0
∼2.0
â-GlcNAc 8
-7.5
2.4
0.6
-6.3
2.3
1.7
obs
∼2.2
1.8
R-GulNAc 9
â-ManNAc 12
-3.3
R-GulNAc 9
â-ManNAc 12
1.2 8.2 4.3 -11.8 0.4 0.6
-3.7
(∼1.5
9.7 2.4 6.5 -12.2
1.5 8.0 4.1 -11.8 0.2 1.2
R-TalNAc 13
-2.7
0.1
R-TalNAc 13
a In 2H2O at 25 °C; in Hz ((0.1 Hz); br ) broadened resonance; obs )
obscured resonance. Values in italics are for the corresponding methyl
aldopyranosides (ref 10).
to more negative couplings upon conversion of an OH to a NHCOCH3
substituent at C2 (ref 37), DFT calculations (see Table 8), spectral simulation
(see Figure S2, Supporting Information), and/or 1H-1H TOCSY cross-peak
analysis (see Figure S3, Supporting Information).
∼1.9 3.4 ∼3.5 ∼1.3 7.7 3.9 -11.2
a In 2H2O at 25 °C; in Hz ((0.1 Hz). Values in italics are for the
corresponding methyl aldopyranosides (ref 44 for 6-8; ref 10 for 12 and
13).
J
signs determined from the expected shift
b 2
C1,H2
1JC1,H1 in 7 and 8 differ slightly from 1JC1,H1 in the corresponding
reducing sugars (R-pyranose, 171.1 Hz; â-pyranose, 162.0 Hz)
(Scheme 9). These changes may be caused, in part, by slightly
different C1-O1 torsion angles for the OH and OCH3 substit-
uents at C1 (i.e., vicinal oxygen lone-pair effects on C-H bond
length, modulated by the C1-O1 torsion angle, influence
1JCH35).
3JC1,H3 and JC1,H5 values in the aminosugars show the
expected dependence on the C1-C2-C3-H3 and C1-O5-
3
(35) Serianni, A. S.; Wu, J.; Carmichael, I. J. Am. Chem. Soc. 1995,
117, 8645-8650.
474 J. Org. Chem., Vol. 71, No. 2, 2006

Synthesis of 2-Amino-2-deoxy-D-[1-13C]aldohexoses
SCHEME 9. J-Couplings in 22 and 23a
-2.5 Hz. Recent DFT calculations in 2-deoxy-â-D-ribofurano-
sylamine have shown that 2JC2,H1 shifts to more negative values
by ∼3.5 Hz in the amine relative to that in the corresponding
deoxyfuranose, in agreement with the nucleoside results.31 The
2
behavior of JC1,H2 in 6-13 is consistent with these prior
2
findings; for 7, JC1,H2 changes sign upon N-substitution, and
for 12, a similar sign change may occur.
1JC1,C2 values in 6-13 range from 42.5 to 46.9 Hz (Table 6).
1
The change from O- to N-substitution at C2 reduces JC1,C2 in
aldopyranosyl rings by 1.7 ( 0.3 Hz based on data available
1
for 7, 8, 11, and 12. These data support prior claims that JCC
magnitude depends on the number and nature of electronegative
substituents appended to the C-C fragment, with more elec-
tronegative substituents increasing the coupling.38 1JCC values
also increase slightly upon methyl glycosidation. For example,
1JC1,C2 in the R- (44.6 Hz) and â- (44.1 Hz) pyranose forms of
N-acetyl-D-glucosamine (Scheme 9) are 0.3 and 1.0 Hz smaller,
respectively, than related values in methyl glycosides 7 and 8.
1
Likewise, JC1,C2 in the R- (46.2) and â- (46.1) pyranoses of
D-glucose are 0.5 and 0.9 Hz smaller, respectively, than related
values in their methyl glycosides. By comparison, 1JC1,C2 in the
R- and â-pyranose forms of D-glucosamine-HCl are 44.3 and
43.8 Hz, respectively.39
2JC1,C3 in 6-13 depends mainly on the orientation of OH
substituents at the coupled carbons (Table 6). The di-equatorial
orientation (8, 12) gives the largest (most positive) coupling of
∼4 Hz. The equatorial/axial arrangement (6, 7, 10, 11, 13) yields
little or no coupling. Coupling is observed in the di-axial
orientation (9), and based on the projection resultant rule,40 it
is expected to be negative (2JC1,C3 in R-D-allopyranose is -2.4
a Values in Hz; negative signs are indicated in parentheses.
C5-H5 dihedral angles, respectively (Table 5). Both couplings
appear minimally affected by the change of substituent from
OH to NHCOCH3 at C2. Configuration at C1 appears to affect
both couplings, with equatorial C1-O1 bonds slightly enhancing
2
Hz).40 JC1,C3 appears minimally affected by a change in
substitution at C2 from OH to NHCOCH3.
Dual pathway 3+3JC1,C4 values are rarely observed in aldopy-
ranosyl rings. A small coupling (0.9 Hz) was observed in 9; a
similar coupling (0.7 Hz) has been reported in ethyl R-D-
mannopyranoside.3a
3
3
the coupling (e.g., JC1,H3 and JC1,H5 are slightly larger in 8
than in 6 and 7) (Table 5). The presence of oxygen along the
pathway also enhances the coupling (e.g., the average gauche
2
2JC1,C5 depends on the orientation of OH
3
3JC1,H5 is ∼1 Hz larger than the average gauche JC1,H3). The
Like JC1,C3
,
large 3JC1,H3 for 9 reflects the ∼180° torsion angle between the
coupled nuclei (H3 is equatorial). These results track those
observed previously in the methyl aldopyranosides.10
substituents at the coupled carbons. In this case, however, only
configuration at C1 is relevant. 2JC1,C5 values in the R-pyranoses
(6, 7, 9, 11, 13) range from -1.8 to -2.1 Hz and are negative
in sign.41 Little or no coupling is observed in â-pyranoses. Like
2JC1,C3, 2JC1,C5 appears to be minimally affected by a change of
substitution at C2 from OH to NHCOCH3.
3
2
In contrast to JC1,H3/H5, JC1,H2 values change significantly
when an NHCOCH3 group replaces an OH group at C2. The
available data indicate a shift to more negative values in the
aminosugars, as confirmed via analysis of cross-peak displace-
3JC1,C6 depends mainly on the C1-O5-C5-C6 dihedral
angle.3a In all cases examined, this angle is ∼180°. Couplings
in 6-13 range from 3.2 to 4.0 Hz, with larger values generally
observed in the â-configuration. Other factors influence this
coupling, however, notably the C5-C6 torsion angle, since an
1
ments in representative H-1H TOCSY spectra (Figure S3,
Supporting Information); absolute changes range from 1 to 4
Hz depending on ring configuration. Importantly, trends ob-
served in these couplings are conserved in the aminosugars and
the simple methyl aldohexopyranosides. For example, 8 exhibits
3
“in-plane” O6 enhances JC1,C6 as does an “in-plane” O1.3a
3JC1,C6 appears to be minimally affected by a change of
2
a much different JC1,H2 (-7.5 Hz) than the remaining com-
substitution at C2 from OH to NHCOCH3.
pounds (-2.5 to -3.7 Hz), as is true for methyl â-D-
glucopyranoside. This difference is predicted in the latter case
by the projection rule proposed by Bock and Pedersen.36
13C-13C couplings are also observed between C1 and the
carbonyl carbon of the N-acetyl side-chain in 6-13 (Table 6).
Couplings vary with ring configuration and range from ∼0.6
to 2.2 Hz. These couplings are expected to depend on the
Prior studies of nucleosides have suggested that a change from
O- to N-substitution at the coupled carbon affects 2JCCH values
much less than a similar change made at the carbon bearing
the coupled proton.37 2JCCH values were reported to shift to more
negative values in the latter case; the shift was estimated at ca.
(38) Duker, J. M.; Serianni, A. S. Carbohydr. Res. 1993, 249, 281-
303.
(39) Walker, T. E.; London, R. E.; Barker, R.; Matwiyoff, N. A.
Carbohydr. Res. 1978, 60, 9-18.
(40) Church, T.; Serianni, A. S. Carbohydr. Res. 1996, 280, 177-186.
(41) Serianni, A. S.; Bondo, P. B.; Zajicek, J. J. Magn. Reson. 1996,
112B, 69-74.
(36) Bock, K.; Pedersen, C. Acta Chem. Scand. 1977, B31, 354-358.
(37) Bandyopadhyay, T.; Wu, J.; Stripe, W. A.; Carmichael, I.; Serianni,
A. S. J. Am. Chem. Soc. 1997, 119, 1737-1744.
J. Org. Chem, Vol. 71, No. 2, 2006 475

Zhu et al.
TABLE 6. 13C-13C Spin-Coupling Constantsa in Methyl 2-Acetamido-2-deoxy-D-[1-13C]Aldohexopyranosides
coupling constant (Hz)
3+3
compd
1JC1,C2
44.8
46.0
44.9
46.2 (46.7)
45.1
46.0 (46.9)
46.1
45.9
46.9
47.7
45.3
46.7 (47.2)
42.5
42.7 (43.8)
45.4
2JC1,C3
nc
nc
nc
nc (nc)
+4.4
+4.5 (4.5)
2.5b
2.0
nc
J
2JC1,C5
3JC1,C6
3.4
3.6
3.2
3.3 (3.3)
4.0
4.1 (4.1)
3.1
3.2
3.6
3.7
3.2
3.3 (3.2)
4.0
3.9 (4.0)
3.2
2JC1,Me
3JC1,CO
C1,C4
R-GalNAc 6
nc
nc
nc
nc (nc)
nc
nc (nc)
0.9b
obs
nc
nc
nc
nc (0.7)
nc
nc
-2.1
-1.9
0.9
-1.9
-1.9
-1.8 (-2.0)
nc
R-GlcNAc 7
â-GlcNAc 8
R-GulNAc 9
â-GulNAc 10
R-ManNAc 11
â-ManNAc 12
R-TalNAc 13
-1.9
-1.8
nc
1.3
1.6
2.2
obs
nc (nc)
-1.8
-1.9
nc
-2.0
-2.0
nc
nc
nc
-1.8
-2.0 (-2.0)
nc
-∼1.7
nc (nc)
+4.0
+4.0 (3.9)
nc
nc
nc
nc
-1.9
-1.8
-1.8
∼0.6
46.5
nc
3.3
a In Hz ((0.1 Hz); in 2H2O at 25 °C; nc denotes couplings <0.6 Hz; obs denotes obscured signal. Values in italics are for the corresponding aldohexopyranoses
2
(refs 16 and 25). Values in parentheses are for the corresponding ethyl aldopyranosides (ref 3a). Indicated J signs are predicted based on the projection
resultant rule (ref 40) and experimental sign determinations in related systems (refs 41 and 45). b Assignments may be reversed.
1
1
TABLE 7. Computed JCH and JCC Valuesa in Compounds 14-21
coupling constant (Hz)
compd
1JC1,H1
1JC2,H2
1JC3,H3
1JC1,C2
1JC2,C3
R-Glc 14ad
R-Glc 14be
R-GlcNAc 15
R-Man 16af
R-Man 16bg
R-ManNAc 17
â-Glc 18
â-GlcNAc 19
â-Man 20
â-ManNAc 21
171.2 (170.1)b
170.4
150.2 (∼145.9)
143.0 (146.5)
147.2
143.3
143.0
147.9
145.2
140.7 (142.9)
136.5
143.3 (142.0)
139.1
45.5 (46.7)c
52.0
46.9 [44.9]
46.8 (47.2)c
52.7
48.7 [45.3]
52.9 (46.9)c
49.5 [45.1]
50.1 (43.8)c
45.6 [42.5]
42.7 (38.2)c
40.0
42.5 [36.7]
41.6 (37.8)c
38.6
146.0
172.3 [171.8]h
171.5 (171.0)
171.9
144.1
156.7 (148.5)
153.3
149.4
150.2 (145.0)
147.5
172.3
41.0
158.2 (161.3)
153.6 [161.8]
151.7 (159.5)
158.2 [162.2]
40.4 (38.9)c
42.1 [37.0]
38.8 (38.2)c
42.7
150.6 (147.9)
148.3
a In Hz. b Values in parentheses are couplings observed experimentally in methyl pyranosides (ref 10). c Couplings observed in ethyl glycosides (ref 3a).
d C1-C2-O2-O2H torsion, -47.0°. e C1-C2-O2-O2H torsion, -168.8°. f C1-C2-O2-O2H torsion, 43.3°. g C1-C2-O2-O2H torsion, 164.6°.
h Experimental couplings in square brackets are from this study.
C1-C2-N2-CO torsion angle and thus may be useful in
assessing side-chain conformation (see below).
C2-H2 bond, which lengthens the bond and reduces the
coupling. 1,3-Lone-pair effects are also operating on 1JC1,H1, as
supported by the essentially identical computed JC1,H1 in 16a
1
E. DFT Calculations of J-Couplings. JCH and JCC values
were calculated in model compounds 14-21 (Scheme 4;
Scheme S1, Supporting Information) to further study the effect
of OH vs NHCOCH3 substitution at C2 on their magnitudes.
Only the Fermi contact component was recovered; based on
prior studies,3a-c,24 the non-Fermi contact terms are expected
to be negligible. All model glycosides lack hydroxyl groups
at C4 and C6 to simplify the calculations, but their absence
is not expected to significantly affect coupling pathways
involving C1, C2, C3, and their attached protons. Only one
C2-N2 torsion was examined in each structure, namely, that
having H2 and the amide proton approximately trans. Also, only
the trans configuration of the amide bond (carbonyl oxygen and
amide proton trans) was considered (Scheme S1, Supporting
Information).
and 16b (Schemes S1 and S2, Supporting Information) (no 1,3
effects are possible in the O1,O2-trans arrangement). In cases
where experimental 1JCH couplings are available, their agreement
with computed values is only moderate presumably because the
experimental couplings reflect the presence of C-O populational
averaging in solution. Notwithstanding the C-O torsional
1
complications, computed JC1,H1 generally decrease in the
aminosugars, in agreement with the experimental findings.
1
1
1
Computed JC2,H2 and JC3,H3 are smaller than JC1,H1 in all
compounds, also in agreement with experiment. These data show
1
that predictions of absolute JCH in saccharides must take into
account C-O torsional averaging as well as possible solvation
effects.
1JC1,C2 is influenced by the C2-O2 torsion angle due to O2
1
1
Calculated JCH values were found to be very dependent on
lone-pair effects (Table 7). JC1,C2 is considerably smaller in
1
C-O torsion angles. For example, JC1,H1 values of 171.2 and
14a (45.5 Hz) than in 14b (52.0 Hz) due to the presence of an
170.4 Hz were calculated when the C1-C2-O2-H torsion
angle is -47.0° (14a) and -168.8° (14b) (Schemes S1 and S2,
Supporting Information), respectively (Table 7). Likewise,
2JC2,H2 is 150.2 Hz in 14a and 146.0 Hz in 14b. Oxygen lone-
pair effects on nearby bond lengths and angles are probably
responsible for these effects. The smaller 2JC2,H2 in 14b is partly
attributed to the presence of a vicinal O2 lone-pair anti to the
O2 lone-pair anti to the C1-C2 bond. 1JC2,C3 is predicted to be
1
smaller than JC1,C2, in agreement with experimental data,
presumably due to the reduced number of electronegative
1
substituents appended to the coupled carbons (two for JC2,C3
1
1
vs three for JC1,C2). JCC appears to decrease by 3-7 Hz for
each loss of an oxygen substituent.38 C-O torsional issues
notwithstanding, JC1,C2 appears to be slightly smaller in 15,
1
476 J. Org. Chem., Vol. 71, No. 2, 2006

Synthesis of 2-Amino-2-deoxy-D-[1-13C]aldohexoses
2
2
TABLE 8. Computed JCH and JCC Valuesa in Compounds 14-21
coupling constant (Hz)
compd
2JC1,H2
2JC2,H1
2JC2,H3
2JC3,H2
2JC3,H4 (ax)
2JC1,C3
2JC1,C5
R-Glc 14ad
R-Glc 14be
R-GlcNAc 15
R-Man 16af
R-Man 16bg
R-ManNAc 17
â-Glc 18
â-GlcNAc 19
â-Man 20
â-ManNAc 21
-0.5 ((1.0)b
0.0
-0.7 (-)
-1.3
-4.4 (-)
-4.4
-4.4 (-6.2)
-3.3
-5.1 (-4.3)
-5.4
-1.0 (nc)c
-0.9
0.4 [nc]
0.7 (nc)c
-0.8
0.3 [nc]
3.9 (4.5)c
6.7 [4.4]
3.5 (3.9)c
6.0
-2.2 (-2.0)c
-2.2
-2.3 [-1.8]
-2.0 (-2.0)c
-2.2
-2.1 [-1.8]
-0.7 (nc)c
-0.6 [nc]
-0.4 (nc)c
-0.4
-2.0 [-2.7]h
-0.9 (-1.2)
-0.2
-0.1
-3.6
-5.7
-5.1
-1.4 (∼-1.8)
1.5 (∼+1.4)
-3.1 (∼-3.7)
-5.2
-2.1
2.0
-1.9
-5.4
-2.7
-0.7
0.6
-4.6
-5.1
-5.9 (-6.3)
-7.2 [-7.5]
-0.5 ((1.5)
-3.0 [-3.7]
0.4 (0)
1.4
-4.3 (-4.3)
-2.6 (-4.2)
-5.5
-5.3 (-4.2)
-5.2
-3.3
7.7 (+7.1)
6.4
2.5 (+1.5)
1.5
-2.4 (-4.6)
-4.8
-5.5 (-5.3)
-5.2
a In Hz. b Values in parentheses are couplings observed experimentally in methyl pyranosides (ref 10). c Couplings observed in ethyl glycosides (ref 3a);
nc ) no coupling was observed. d C1-C2-O2-O2H torsion, -47.0°. e C1-C2-O2-O2H torsion, -168.8°. f C1-C2-O2-O2H torsion, 43.3°. g C1-
C2-O2-O2H torsion, 164.6°. h Experimental couplings in brackets from this study.
3
3
TABLE 9. Computed JCH and JCC Coupling Constantsa in Compounds 14-21
coupling constant (Hz)
compd
3JC1,H3
3JC1,H5
3JC2,H4 (ax)
3JC2,H4 (eq)
3JC3,H1
3JC1,C6
3JC2,CH
3
R-Glc 14ad
0.7 (br)b
[70°]i
1.4
1.6 (∼2.0)
[-62°]
1.5
[-61°]
1.5 [1.4]
[-62°]
1.5 (∼2.0)
[-61°]
1.4
2.2 (nc)h
[-67°]
2.3
[-67°]
2.3
[-67°]
1.9 (nc)
[-68°]
1.8
[-65°]
1.7
[-69°]
2.0 (nc)
[-67°]
2.0
[-68°]
1.6 (nc)
[-66°]
1.8
7.4 (5.1)c
[176°]
7.4
[175°]
7.3
[176°]
5.5
[175°]
5.2
[176°]
5.3
[174°]
7.2 (5.6)c
[175°]
7.0
[175°]
5.1
[176°]
5.4
4.8 (5.2)
[169°]
5.1
[173°]
4.8
[170°]
4.3 (4.6)
[173°]
4.7
[166°]
4.3
[171°]
1.5 (1.1)
[-65°]
1.4
[-64°]
1.2 (nc)
[-62°]
1.1
4.2 (3.3)
[-180°]
4.1
[-179°]
4.2 [3.2]
[-180°]
3.7 (3.2)
[-178°]
3.9
[-179°]
3.8 [3.2]
[-178°]
4.7 (4.1)
[-175°]
4.9 [4.0]
[-175°]
4.7 (4.0)
[-177°]
4.6 [4.0]
[-176°]
3.4
[-167°]
3.7
[168°]
3.3
[-167°]
4.2
[-170°]
5.0
[-170°]
4.7
[-170°]
4.3
[171°]
4.3
[173°]
4.4
[173°]
4.2
R-Glc 14be
R-GlcNAc 15
R-Man 16af
R-Man 16bg
R-ManNAc 17
â-Glc 18
[65°]
0.9 [0.7]j
[67°]
0.3 (nc)
[62°]
0.2
[70°]
0.2
[66°]
1.6 (1.2)
[67°]
1.4 [1.1]
[67°]
0.5 (nc)
[67°]
[-61°]
1.6
[-60°]
2.3 (2.3)
[-57°]
2.6 [2.4]
[-57°]
2.3 (∼2.2)
[-59°]
2.5
â-GlcNAc 19
â-Man 20
â-ManNAc 21
1.1 [nc]
[63°]
[-58°]
[-69°]
[174°]
[-58°]
[173°]
a In Hz. b Values in parentheses are couplings observed experimentally in methyl or ethyl pyranosides (refs 3a and 10). c Couplings observed in methyl
galactopyranosides (ref 44). d C1-C2-O2-O2H torsion, -47.0°. e C1-C2-O2-O2H torsion, -168.8°. f C1-C2-O2-O2H torsion, 43.3°. g C1-C2-
O2-O2H torsion, 164.6°. h nc denotes no coupling (J < 0.6 Hz); br denotes broadened resonance. i Torsion angles in brackets were obtained from the
DFT-optimized structures (see Schemes S1 and S2, Supporting Information). j Experimental couplings in brackets from this study.
17, 19, and 21 relative to their respective methyl glycosides, in
agreement with the experimental findings.
experiment. A significant deviation between computed and
2
experimental JC1,C3 values was found in 19, for reasons that
2
2
Computed 2JCH and 2JCC are generally in good agreement with
experimental data (Table 8). The effect of C-O bond rotation
are unknown. In contrast to JC1,C3, JC1,H5 is larger in 14-17
(ca. -2.1 Hz) than in 18-21 (ca. -0.5 Hz), in good agreement
with experiment. These two-bond couplings appear unaffected
by C2-O2 torsion and by C2 substituent structure.
on these two-bond couplings is smaller than observed for 1JCH
,
2
and sign predictions appear robust. Computed JC1,H2 shift to
more negative values by 1.5-2.5 Hz when the C2 substituent
is changed from OH to NHCOCH3, in agreement with experi-
mental findings. More significant differences between calculated
3
3
Computed JC1,H3 and JC1,H5 are in good agreement with
experimental data (Table 9) for gluco and manno configurations.
Interestingly, 3JC1,H3 appears on average to be slightly larger in
â-pyranosides 18-21 than in R-pyranosides 14-17, although
the difference is small and the effects of C-O torsions cannot
be completely excluded. Likewise, 3JC1,H5 appears larger in the
â-pyranosides; in this case, the conclusion is less ambiguous
since C-O torsions affect this pathway minimally. Similar
observations have been reported based on trends observed in
2
2
and experimental JCCH are observed for JC3,H2 (less negative
in the calculations), presumably due to the lack of an O4
2
substituent in the model compounds, which can affect JCCH
magnitudes via remote effects10 and to C2-O2 torsional effects.
2
The data suggest a shift to more negative JC3,H2 in the
2
aminosugars, analogous to observations for JC1,H2. Computed
2JC3,H4 are more negative than corresponding experimental
values, again presumably due to the lack of an OH group on
the carbon bearing the coupled proton. These couplings appear
unaffected by substituent structure at C2.
experimental 3JC1,H3 and 3JC1,H5
.
10 Gauche 3JC1,H5 are predicted
3
to be larger than gauche JC1,H3, again in agreement with
experiment.10 The computed couplings between C2 and H4ax/eq
deviate from experimental findings due to the lack of an O4
substituent in the model structures. Interestingly, the relatively
wide range of anti couplings observed for 3JC2,H4(eq) suggest that
O2/N2 orientation influences this coupling in a manner analo-
2
Small JC1,C3 values are computed in R-pyranosides 14-17
(+0.7 to -1.0 Hz), and large values are computed in â-pyra-
nosides 18-21 (+3.5 to +6.7 Hz), in general agreement with
J. Org. Chem, Vol. 71, No. 2, 2006 477

Zhu et al.
gous to 3JC1,H3, with an in-plane O2/N2 increasing the coupling.
3JC3,H1 is larger in 14-17 than in 18-21 due to the significantly
different C3-C2-C1-H1 dihedral angles (∼60° vs ∼180°).
3JC1,C6 shows a dependence on anomeric configuration, with
an “in-plane” O1 enhancing the coupling.3a For example, 3JC1,C6
TABLE 10. Structure-Spin Coupling Correlations in 2-Acetamido
Sugars
J-coupling
structural correlations
3JH2,H3
increases in C2-NAc derivatives (0.9-1.6 Hz)a
increases in C2-NAc derivatives (0.5-2.7 Hz)a
decreases in C2-NAc derivatives (∼1.7 Hz)a
shifts to more negative (-) values in C2-NAc
derivatives (1-4 Hz)a
1JC1,H1
1JC1,C2
2JC1,H2
3
is smaller in 14/15 than in 18/19, and JC1,C6 is smaller in
16/17 than in 20/21. Absolute values of the computed couplings
are larger than the experimental values presumably because of
the lack of O6 in the model compounds.
3JC1,H5, 3JC1,H3
,
shifted to more positive (+) values when “in-plane”
terminal oxygen(s) are present in the pathway
2JC1,C3, 3JC1,C6
F. Side-Chain 13C/15N Labeling of 2-Aminosugars. 2-
Acetamido-2-deoxy-D-glucose was prepared with 13C-labeling
at C1 and both acetyl carbons, and with 15N-labeling (Figure
S4, Supporting Information). Analysis of the 13C NMR spectrum
revealed multiple JCC and JCN couplings within the R- (22) and
a Relative to methyl aldopyranosides.
configuration at C1, C2, and C3 may modulate one or both of
these preferences.
â- (23) pyranoses which are summarized in Scheme 9.
1
Intra-ring 13C-13C couplings involving C1 include JC1,C2
,
Conclusions
3
2JC1,C3
,
2JC1,C5, and JC1,C6. Most of these couplings are
A simple and general chemical route to prepare 2-amino-2-
deoxy-D-aldohexoses has been developed that is applicable to
13C, 2H, and/or 15N labeling with use of inexpensive and readily
available labeled precursors. Simple D-aldopentosylamines can
be prepared in good yield, are chemically stable, and convert
rapidly to 2-amino-2-deoxyaldononitriles, which are subse-
quently reduced catalytically to give C2-epimeric 2-amino-2-
deoxyaldohexoses. Solution conditions were determined to
minimize pentosylamine hydrolysis during cyanide addition, and
separation of the epimeric amines can be achieved either as their
free amines or their N-acetylated derivatives. This simple
methodology provides access to the biologically relevant
GlcNAc and GalNAc structures, but importantly to other ring
configurations that may be valuable in studies of biological
structure-function properties.
comparable in magnitude ((0.1 Hz) to those observed in the
1
methyl glycosides 7 and 8 (Table 6). The exception is JC1,C2
,
which differs by 0.3 Hz between 7 and 22, and by 1 Hz between
8 and 23. This trend is observed in other reducing sugar/methyl
glycoside comparisons (data not shown). In general, methyl
1
glycosidation increases JC1,C2 by 0.3-0.4 Hz in R-aldopyra-
nosyl rings and by 0.9-1.0 Hz in â-aldopyranosyl rings. In
contrast, 1JC1,H1 increases by ∼0.7 Hz upon methyl glycosidation
of 22, whereas 1JC1,H1 decreases slightly (0.2 Hz) upon methyl
glycosidation of 23. A small change (0.3 Hz) is also observed
3
in JC1,C6 between 8 and 23.
Numerous J-couplings were observed within the N-acetyl
side-chain (Figure S4, Supporting Information). These include
the vicinal couplings 3JC1,CO, 3JC3,CO, 3JC1,NH, 3JC3,NH, and 3JH2,NH
,
all of which presumably report on the torsional preferences of
the C2-N2 bond, although Karplus curves for these pathways
remain to be formulated. Vicinal coupling between C2 and the
side-chain methyl group was also observed in 22 and 23 (∼1.5
Hz), but it is unclear how this J-coupling may be influenced
by the cis-trans configuration of the amide bond.
13C and 1H NMR data were obtained on the D-aldopentosyl-
amines in aqueous solution at pD 10.5. At this pD, hydrolysis
is slow, and well-resolved spectra are obtained. Solution studies
suggest that orientation of the C1-N1 bond plays a dominant
role in dictating both preferred anomeric configuration and ring
conformation in these structures, with an equatorial orientation
preferred in all cases. Time-lapse 13C NMR studies of the
hydrolysis of 13C-labeled â-D-xylopyranosylamines revealed four
major intermediates that were assigned to the two anomeric
D-xylofuranosylamines, R-D-xylopyranosylamine, and an acyclic
carbinolamine. Conformational preferences of xylofuranosy-
lamines appear to differ significantly from those of simple
xylofuranoses based on 3JH1,H2 analyses, indicating that substitu-
tion of NH2 for OH at C1 significantly changes preferred
furanose ring conformation.
Vicinal coupling of the 15N to C4 (∼180° dihedral angle)
was observed in both anomers (1.5-1.7 Hz; sign unknown).
There is no apparent 2JNC involving C1 and C3 in 22 and 23. A
2JNC is, however, observed for the methyl carbon of the side-
1
chain (7.6-7.8 Hz; sign unknown). JC2,N (ca. -11 Hz) and
1JCO,N (ca. -15 Hz) are comparable in magnitude and prob-
1
ably negative in sign.42 JNH appears to differ in 22 (-92.9
Hz) and 23 (-91.7 Hz). A number of geminal couplings were
also observed (e.g., 2JCO,NH, 2JCH3,N, 2JC2,CO), but their depend-
encies on side-chain structure/conformation are presently un-
known.
Fischer glycosidation of the free HexNAc sugars was rapid
and yielded pure methyl glycosides after column purification.
Access to pure glycoside anomers simplified subsequent NMR
analyses by reducing the spectral complexity inherent to
reducing sugars. N-Acetyl substitution at C2 caused minimal
change in intra-ring JCH and JCC values in most cases, and thus
prior empirical rules governing the effect of pyranosyl ring
structure on J-coupling magnitude and sign based on observa-
tions in simple aldohexopyranosyl rings appear applicable to
the 2-acetamido-2-deoxy ring system. There are, however, some
Taken collectively, the ensemble of J-couplings involving
N-acetyl side-chain atoms should provide useful, redundant
information about the torsional preferences of the C2-N2 bond,
and possibly about the cis-trans equilibrium of the amide bond.
At present, the preferred H2-C2-N2-H torsion in structures
such as 22 and 23 is generally believed to be trans,43 while
amide configuration is assumed to be trans. However, relative
(42) (a) Levitt, M. H. Spin Dynamics. Basics of Nuclear Magnetic
Resonance; John Wiley & Sons: New York, 2001; p 419. (b) Ando, I.;
Webb, G. A. Theory of NMR Parameters; Academic Press: New York,
1983; pp 83-113.
(43) (a) Cowman, M. K.; Cozart, D.; Nakanishi, K.; Balazs, E. A. Arch.
Biochem. Biophys. 1984, 230, 203-212. (b) Holmbeck, S. M. A.; Petillo, P.
A.; Lerner, L. E. Biochemistry 1994, 33, 14246-14255. (c) Scott, J. E.;
Heatley, F. Proc. Natl. Acad. Sci. 1999, 96, 4850-4855.
1
1
2
exceptions, specifically for JC1,C2, JC1,H1, and JC1,H2 (Table
2
10). For JC1,H2, O- to N-substitution at C2 consistently shifts
these couplings to more negative values. Overall, these results
are important for anticipated NMR studies of GlcNAc- and
GalNAc-containing oligosaccharides. A key question that
remains unanswered is how long-range JCC values involving
478 J. Org. Chem., Vol. 71, No. 2, 2006

Synthesis of 2-Amino-2-deoxy-D-[1-13C]aldohexoses
C2 as the coupled carbon are affected by N-substitution at C2.
parametrization should prove useful in this context. Likewise,
while it is generally assumed that amide bond configuration in
2-acetamido-2-deoxy-D-aldohexoses is trans, it is not unreason-
able to expect the presence of some cis configuration in solution,
albeit in minor amounts. Further studies of chemical shifts and/
or J-couplings within the N-acetyl side-chain may shed further
light on this question.
This question is particularly relevant for trans-glycoside 3JC2,Cx
,
which are useful structural constraints for the φ torsion angle
in O-glycosidic linkages.3a This problem is currently under
investigation.
1
1
DFT studies showed that JCH and JCC are significantly
3
3
affected by exocyclic C-O torsions, whereas JCH and JCC
appear much less influenced by them. This observation suggests
that efforts to interpret the former J-couplings in solution will
need to take into account the averaging of exocyclic C-O
rotamer populations and possibly other factors.
Acknowledgment. This work was supported by grants from
the National Institutes of Health (GM) (to A.S.). The Notre
Dame Radiation Laboratory is supported by the Office of Basic
Energy Sciences of the United States Department of Energy.
This is Document No. NDRL-4607 from the Notre Dame
Radiation Laboratory.
Exocyclic N-acetyl side-chain conformation in 2-acetamido-
2-deoxyaldohexoses has been implicated in determining their
biological properties, most notably in heteropolysaccharides such
as hyaluronic acid.43 13C- and 15N-labeling of this side-chain
makes available a large ensemble of scalar couplings that are
sensitive to the C2-N2 torsion angle, to amide bond configu-
ration, and/or to other presently unrecognized structural features.
Supporting Information Available: Table S1 showing HRMS
data obtained on aldopentosylamines 2-5; Table S2 showing HPLC
retention times of aminosugars; Scheme S1 showing DFT-computed
molecular structures of 14a, 15, 16a, and 17-21; Scheme S2
showing DFT-computed molecular structures of 14 and 16; Fig-
ure S1 showing MS data for 7; Figure S2 showing real and
simulated 600-MHz 1H NMR spectra of 7; Figure S3 showing the
assignment of the negative sign of 2JC1,H2 in 7 from 1H-1H TOCSY
data; Figure S4 showing partial 13C NMR spectra of 22 and 23;
Cartesian coordinates for DFT structures 14-21; 13C{1H} NMR
Prior studies have utilized only one of these couplings, 3JH2,NH
,
to evaluate C2-N2 torsional preferences.43 As shown in recent
studies of exocyclic hydroxymethyl (CH2OH) conformation,44
the use of multiple, redundant J-couplings sensitive to the same
conformational domain can significantly improve determinations
of rotamer populations in solution. In the exocyclic N-acetyl
1
spectra (150 MHz) of 2-5; and H NMR spectra (600 MHz) of
3
3
2-5. This material is available free of charge via the Internet at
case, at least two additional J-couplings, JC1,CO and JC3,CO
,
http://pubs.acs.org.
are likely to depend on the C2-N2 torsion angle, and thus their
JO051510K
(44) Thibaudeau, C.; Stenutz, R.; Hertz, B.; Klepach, T.; Zhao, S.; Wu,
Q.; Carmichael, I.; Serianni, A. S. J. Am. Chem. Soc. 2004, 126, 15668-
15685.
(45) Zhao, S.; Bondo, G.; Zajicek, J.; Serianni, A. S. Carbohydr. Res.
1998, 309, 145-152.
J. Org. Chem, Vol. 71, No. 2, 2006 479