Solid-phase synthesis and 1H and 13C high-resolution magic angle spinning NMR of 13C-labeled resin-bound saccharides

Article abstract of DOI:10.1002/mrc.1364

We show how high-resolution magic angle-spinning NMR spectroscopy can be used to characterize ¹³C-labeled saccharides that have been prepared using solid-phase synthesis techniques while they are still bound to a solid-support resin. With the u

Full text of DOI:10.1002/mrc.1364

MAGNETIC RESONANCE IN CHEMISTRY
Magn. Reson. Chem. 2004; 42: 453–458
Published online 27 February 2004 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mrc.1364
Solid-phase synthesis and 1H and 13C high-resolution
magic angle spinning NMR of 13C-labeled resin-bound
saccharides
Nikolaus M. Loening,∗ Takuya Kanemitsu,† Peter H. Seeberger‡ and Robert G. Griffin
Department of Chemistry, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA
Received 27 August 2003; Revised 3 December 2003; Accepted 17 December 2003
We show how high-resolution magic angle-spinning NMR spectroscopy can be used to characterize
13C-labeled saccharides that have been prepared using solid-phase synthesis techniques while they are
still bound to a solid-support resin. With the use of 13C-labeled glucose as the starting material, we have
successfully synthesized mono-, di- and trisaccharides with uniform 13C labeling of the saccharide rings.
1
Using these materials, we have been able to assign the 13C and H spectra and to characterize various
impurities on the resin beads. Copyright  2004 John Wiley & Sons, Ltd.
KEYWORDS: NMR; 1H NMR; 13C NMR; oligosaccharides; high-resolution magic angle spinning; solid support; gel phase;
carbohydrates
techniques along with the application of HR-MAS 1H and 13
C
INTRODUCTION
NMR to these samples. Previous efforts towards the study
of resin-bound saccharides using HR-MAS NMR were only
Solid-phase synthesis techniques have proven to be useful
methods for generating oligopeptides,1,2 oligonucleotides3
and, more recently, oligosaccharides.4,5 However, character-
ization of reaction products by NMR spectroscopy while
they are still resin bound is difficult because the pres-
ence of the solid support reduces the spectral resolution.
Although the use of high-resolution magic angle spinning
(HR-MAS) techniques6–8 can drastically improve the reso-
lution, the resonances in HR-MAS spectra are still broad
relative to conventional solution-state NMR spectra. There-
fore, spectral overlap often complicates the characterization
of resin-bound products; this is particularly relevant in
the analysis of oligosaccharides, where the 1H resonances
are dispersed over a relatively limited range of chemical
shifts. Although high-resolution solution-state NMR spectra
of reaction products can be acquired after cleaving them from
the solid support, the analysis of products while they are still
resin bound is more useful as this allows reaction products
and intermediates to be identified during the course of a
synthesis rather than only at the end.
1
partially successful in assigning the H and 13C spectra9 or
focused on monitoring 13C-labeled10,11 or 19F-containing12–14
protecting groups to track the progression of a synthesis.
By including 13C labels on the saccharide itself, we are able
to use the additional chemical shift dispersion from the 13
spectrum to disentangle and assign the poorly resolved H
spectrum.
C
1
EXPERIMENTAL
Compounds
The starting material for the solid-phase synthesis ([13C6]dibutyl-3,4-
di-O-benzyl-6-O-levulinoyl-2-O-pivaloyl-D-glucopyranosyl phos-
phate, 7) was synthesized starting from [13C6]-D-glucose as shown
in Fig. 1. The resulting 13C-labeled starting material was then used
in the stepwise solid-phase reaction scheme shown in Fig. 2 to
generate 13C-labeled mono-, di- and trisaccharide samples. The solid
support used for the synthesis was 1% divinylbenzene cross-linked
chloromethyl polystyrene resin (Merrifield resin). Polystyrene resins
have become the standard support for oligosaccharide solid-phase
synthesis owing to their ready availability, synthetic versatility and
compatibility with a wide range of solvents and reagents. Although
other polymeric supports (such as the polyethylene glycol cross-
linked polymeric supports POEPS and POEPOP) facilitate on-resin
analysis, they are not widely used for the routine synthesis of
oligosaccharides. All chemicals were of reagent grade and used as
supplied except as noted. Dichloromethane and tetrahydrofuran
were purified using a J.T. Baker Cycle-Tainer Solvent Delivery
System. Solutions were evaporated under reduced pressure at a
In this paper, we describe the synthesis of 13C-labeled
mono- and oligosaccharides using solid-phase synthesis
ŁCorrespondence to: Nikolaus M. Loening, Department of
Chemistry, Lewis & Clark College, 0615 SW Palatine Hill Road,
Portland, Oregon 97219, USA. E-mail: loening@lclark.edu
†Present address: Division of Organic and Bioorganic Chemistry,
Kyoritsu College of Pharmacy, Shibakoen 1-5-30, Minatoku, Tokyo
105, Japan.
°
bath temperature not exceeding 40 C. Column chromatography
was performed using a forced flow of the indicated solvent on
Sigma H-type silica (10–40µm).
‡Present address: Laboratory for Organic Chemistry, ETH
Ho¨nggerberg, HCI F315, Wolfgang-Pauli-Strasse 10, CH-8093
Zu¨rich, Switzerland.
3,4,6-O-Acetyl-D-arabino-hex-1-enitol (2)
A mixture of [13C6]-D-glucose (1, 1000 mg, 5.73 mmol) and acetic
anhydride (5 ml, 52.9 mmol) in pyridine (10 ml) was stirred at
room temperature for 3 h, after which the reaction mixture was
Contract/grant sponsor: NIH NIBIB; Contract/grant number:
EB002026.
Copyright  2004 John Wiley & Sons, Ltd.

454
N. M. Loening et al.
1) Ac2O, pyridine
2) HBr, AcOH, CH2Cl2
3) Zn, AcOH, AcOEt
BnBr
NaH
DMF
OAc
O
OH
OTIPS
O
OTIPS
O
1) NaOMe, MeOH
2) TIPSCl, imidazole, DMF
O
HO
AcO
AcO
HO
HO
BnO
BnO
OH
HO
84%
95%
OH
89%
3
2
4
1
LevOH
DIPC
DMAP
CH2Cl2
OLev
1) DMDO, CH2Cl2
2) HOP(O)(OBu)2, CH2Cl2
3) PivCl, DMAP
OLev
O
OH
O
P
OBu
O
TBAF
THF
BnO
BnO
O
OBu
O
BnO
BnO
BnO
OPiv
BnO
72%
99%
88%
7
6
5
Figure 1. Synthesis of the 13C-labeled glycosylating agent (7). The starting material (1) was uniformly 13C labeled, as indicated by
the dots in the structure. As a result, 7 was 13C labeled around the glucose ring. See text for further details of the synthesis.
Glycosylation
O
OLev
6
OLev
O
O
O
P
4
5
5
5
BnO
BnO
2 O
O
OBu
BnO
BnO
1
O
O
O
OPiv
HO
OBu
3
OPiv
7
8
9
TMSOTf/CH2Cl2
OLev
Deprotection
6'
3'
4'
5'
NH2NH2•AcOH
CH2Cl2
O
2'O
BnO
BnO
1'
4
O
6
PivO
BnO
BnO
2 O
1
3
OPiv
OLev
10
6''
3''
4''
5''
O
BnO
BnO
2''
1''
4'
O
6'
PivO
BnO
BnO
5'
O
2'O
1'
4
O
6
3'
PivO
BnO
BnO
2 O
1
3
OPiv
11
Figure 2. Solid-phase synthesis of the 13C-labeled mono-, di- and trisaccharides (9, 10 and 11). The labeled sites on the glucose
rings are indicated by dots and by the numbering. Further details of these syntheses are provided in the text.
concentrated. The resulting peracetylglucose was dissolved in
6-O-Triisopropylsilyl-D-arabino-hex-1-enitol (3)
°
CH2Cl2 (20 ml) and cooled to 0 C. A 33% HBr–AcOH mixture
Compound 2 (1220 mg, 4.39 mmol) was dissolved in MeOH (18 ml).
Sodium methoxide (0.5 M in MeOH, 2 ml) was added to this solution
and stirred at room temperature for 2 h, after which the reaction
mixture was neutralized with Amberlite IR-120 (HC), filtered and
concentrated. This concentrate was dissolved with DMF (10 ml) and
°
(4 ml) was added slowly and the solution was stirred at 0 C for
3 h. After allowing the solution to warm to room temperature,
it was stirred for an additional 3 h. The reaction mixture was
poured into ice-cold aqueous NaHCO3, eluted with CH2Cl2, washed
with H2O, dried over MgSO4, filtered and finally concentrated
to provide glycosyl bromide. The compound was dissolved with
°
cooled to 0 C. Triisopropylsilyl chloride (1.0 ml, 4.67 mmol) and
imidazole (580 mg, 8.52 mmol) were added to this solution, after
°
which the mixture was stirred at 0 C for 2 h and concentrated. The
°
ethyl acetate (10 ml) and cooled to 0 C. A suspension of Zn
residue was purified using silica gel column chromatography with
dust (3.8 g), CuSO4Ð5H2O (50 mg) and AcONa (250 mg) in 60%
toluene–acetone (80 : 20) as eluent to provide 3 (1204 mg, 89%).
AcOH–H2O (26 ml) was added to the reaction solution and
°
stirred at 0 C for 1 h and at room temperature for 1 h. After
the reaction was complete, the suspension was filtered through
a pad of Celite. The resulting filtrate was eluted with ethyl acetate,
washed with aqueous NaHCO3 and H2O, dried over MgSO4,
filtered and concentrated. The residue was purified on a silica
gel column using hexane–ethyl acetate (75 : 25) as eluent to provide
2 (1474 mg, 84%).
3,4-Di-O-benzyl-6-O-triisopropylsilyl-D-arabino-hex-1-enitol (4)
Compound 3 (1200 mg, 3.89 mmol) was dissolved in DMF (20 ml)
°
and cooled to 0 C. NaH (60% oil dispersion, 470 mg, 11.75 mmol)
°
was added and the resulting solution was stirred at 0 C for 1 h.
BnBr (1.86 ml, 15.55 mmol) was added and the mixture was stirred
°
at 0 C for 2 h. After the reaction was complete, the excess of
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 453–458

HR-MAS NMR of 13C-labeled resin-bound saccharides
455
NaH was decomposed by MeOH. The mixture was diluted with
CH2Cl2, washed with H2O twice, dried over MgSO4, filtered and
concentrated. The residue was purified using silica gel column
chromatography with hexane–ethyl acetate (15 : 1) as eluent to
provide 4 (1805 mg, 95%).
resin is crucial for obtaining high-resolution spectra as it is
only in the gel phase that the resin is sufficiently mobile
on a molecular scale to average effects due to dipolar
couplings and chemical shift anisotropy. Even with a swelled
resin, residual dipolar couplings and magnetic susceptibility
gradients due to the solid–liquid interfaces in the sample
substantial broaden the spectrum. Fortunately, these effects
can be greatly attenuated by spinning the sample about
the magic angle. The resulting HR-MAS spectra are very
similar to solution-state NMR spectra in terms of information
content (isotropic chemical shifts and scalar couplings) and,
in favorable cases, resolution. For our experiments, the
samples were spun at a frequency of 6 kHz to ensure that any
sidebands were outside the 0–10 ppm chemical shift range
of the 1H spectra.
3,4-Di-O-benzyl-D-arabino-hex-1-enitol (5)
Compound 4 (1780 mg, 3.64 mmol) was dissolved in THF (10 ml) and
°
cooled to 0 C. A 1.0 M solution of TBAF in THF (5.5 ml, 5.5 mmol)
°
was added and the resulting solution was stirred for 1 h at 0 C
and for 3 h at room temperature. The reaction mixture was diluted
with CH2Cl2, washed with H2O twice, dried over MgSO4, filtered
and concentrated. The residue was purified using silica gel column
chromatography with hexane–ethyl acetate (75 : 25) as eluent to
provide 5 (1070 mg, 88%).
3,4-Di-O-benzyl-6-O-levulinoyl-D-arabino-hex-1-enitol (6)
A mixture of 5, levulinic acid (0.64 ml, 6.23 mmol), DIPC (0.98 ml,
6.21 mmol) and DMAP (38 mg, 0.31 mmol) in CH2Cl2 was stirred
°
at 0 C for 2 h. The reaction mixture was purified using silica gel
The composite pulse decoupling sequence WALTZ-
column chromatography with toluene–acetone (90 : 10) as eluent to
1
1615,16 was used for both H decoupling (ω1/2ꢀ D 2.5 kHz)
give 6 (1316 mg, 99%).
and 13C decoupling (ω1/2ꢀ D 4.2 kHz). HSQC spectra17 were
measured using 2048 ꢁt2ꢂ ð 512 ꢁt1ꢂ points and zero-filled
once in each dimension prior to Fourier transformation.
With 16 scans per increment and a recycle delay of
1 s, the experiment time for each HSQC experiment was
2.3 h. INADEQUATE spectra18 were measured using 2048
ꢁt2ꢂ ð 256ꢁt1ꢂ points and zero-filled once in each dimension
prior to Fourier transformation. With 64 scans per increment
and a recycle delay of 1 s, the experiment time used
to acquire each INADEQUATE spectrum was 4.6 h. The
length of the spin-echo period at the beginning of the
Dibutyl 3,4-di-O-benzyl-6-O-levulinoyl-2-O-pivaloyl-D-
glucopyranosyl phosphate (7)
Compound 6 (427 mg, 0.992 mmol) was dissolved in CH2Cl2 (10 ml)
°
and cooled to 0 C. A solution of dimethyldioxirane in acetone
(0.08 M, 19 ml, 1.5 mmol) was added and the reaction mixture was
°
stirred at 0 C for 20 min. The solvent was then removed using a
stream of nitrogen. After the remaining residue had been dried in
°
vacuo at 0 C for 10 min, 10 ml of CH2Cl2 were added to dissolve
°
the residue. The resulting solution was cooled at ꢀ78 C for 15 min.
At this temperature, a solution of dibutyl phosphate (0.235 ml,
1.19 mmol) in CH2Cl2 (10 ml) was added dropwise over a period
°
of 5 min. The reaction mixture was then warmed to 0 C, at which
point DMAP (485 mg, 3.97 mmol) and pivaloyl chloride (0.245 ml,
1.99 mmol) were added to the solution. The solution was then
allowed to warm to room temperature over 1 h and hexane–ethyl
acetate (1 : 1) was added. The reaction mixture was filtered through
a short silica gel column and concentrated. The residue was purified
using silica gel column chromatography with hexane–ethyl acetate
(60 : 40) as eluent to afford 7 (529 mg, 72%).
INADEQUATE sequence was set so that 50 Hz 13C–13
C
scalar couplings resulted in the most efficient generation
of double quantum coherences. INADEQUATE experiments
optimized for weaker 13C–13C couplings (33 Hz) produced
similar results.
°
All spectra were recorded at room temperature (25 C).
Glycosylation reaction on solid phase. General procedure
A mixture of octenediol linker-bound resin 8 (300 mg, ¾0.15 mmol,
1 equiv.) and the glycosylating agent 7 (222 mg, 0.3 mmol, 2 equiv.)
in anhydrous CH2Cl2 (10 ml) was shaken at room temperature for
RESULTS AND DISCUSSION
°
The 1H HR-MAS spectrum of 9 [Fig. 3(a)] highlights the
relatively broad lines that sometimes occur with resin-bound
samples. Resonances due to protons on the saccharide ring
appear between 3 and 5.5 ppm; the strong background
signals that appear between 1–2.5 and 6–8 ppm are
primarily due to the resin. As the ring protons are coupled to
13C nuclei (via both one-bond and longer range couplings),
the use of 13C decoupling improves both the spectral
resolution and the sensitivity [Fig. 3(b)]. Even so, there are
still a number of resonances that overlap the ring protons;
these can be excluded from the spectrum by incorporating
a 13C filter based on two INEPT transfers.19 By using such a
15 min. This suspension was then cooled to ꢀ78 C, at which point
a solution of 0.5 M trimethylsilyltrifluoromethanesulfonic acid in
CH2Cl2 (0.6 ml, 0.3 mmol, 2 equiv.) was added. The suspension was
shaken under a nitrogen atmosphere at a temperature between ꢀ78
°
and ꢀ50 C for 1 h. The resin was then washed with CH2Cl2, MeOH,
THF and CH2Cl2. This procedure was repeated twice.
Deprotection reaction on solid phase. General procedure
A mixture of carbohydrate-bound resin 9 (277 mg, ¾0.13 mmol,
1 equiv.), acetic anhydride (0.05 ml, 0.53 mmol, 4 equiv.) and
DMAP (5 mg, 0.04 mmol, 0.3 equiv.) in pyridine (10 ml) was shaken
at room temperature for 2 h. The resin was then washed with
pyridine–acetic acid (3 : 2). A solution of 0.25 M N2H4ÐHOAc in
pyridine–acetic acid (3 : 2) (10 ml, 2.5 mmol) was added to the resin
and the suspension was shaken at room temperature for 30 min.
The resin was then washed with pyridine–acetic acid (3 : 2), 0.2 M
AcOH in THF, THF and CH2Cl2. This procedure was repeated
twice.
13C filter in the pulse sequence, the H spectrum [Fig. 3(c)]
1
only includes protons that are directly bound to 13C nuclei
and, as a result, the seven ring protons in the monosaccharide
sample (9) can almost be completely resolved. Unfortunately,
the 1H linewidths are ¾50 Hz, even with 13C decoupling, so
the resonances in the much more congested 1H spectra from
the disaccharide [10, Fig. 3(d), 14 resonances are expected]
and trisaccharide [11, Fig. 3(e), 21 resonances are expected]
samples cannot be straightforwardly resolved.
Spectroscopy
The 1H and 13C NMR spectra of the resin-bound mono-, di-,
and trisaccharides (9, 10, and 11 in Fig. 2) were acquired
on a Bruker DRX600 spectrometer operating at 600 MHz for
1
1H and 125 MHz for 13C and equipped with a H/13C HR-
MAS probe. The samples were swollen in CDCl3 (which also
served as the lock solvent and reference) prior to packing
the swelled resin into 4 mm zirconium rotors. Swelling the
The ring protons are difficult to resolve because the peaks
are dispersed by their chemical shifts over a range of only
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 453–458

456
N. M. Loening et al.
(a)
8
7
6
5
4
3
2
1
1H Chemical Shift (ppm)
(b)
(c)
(d)
(e)
6
5
4
3
3
6
2
1
4
3'
4'
5'
6
1
6'
2'
2
6'
5
6
1'
3/3'/3''
4/4'/4''
5''
6/6'
6/6'
6''
1/1'
2/2'
6''
2''
5/5'
1''
5.6 5.4 5.2 5.0 4.8 4.6 4.4 4.2 4.0 3.8 3.6 3.4 3.2
1H Chemical Shift (ppm)
5.8
Figure 3. 1H HR-MAS NMR spectra of the 13C-labeled saccharide samples (9–11). As seen in spectrum (a), resonances arising
from protons on the saccharide rings of 9 lie between 3 and 6 ppm; the large resonances outside of this region are due to the resin.
An expanded view from a 13C decoupled 1H spectrum of 9 is shown in (b). To limit the observed signal to protons attached to
labeled carbon sites, 13C edited and decoupled 1H spectra were acquired for 9, 10 and 11 and are shown as spectra (c), (d) and (e),
respectively. The seven peaks arising from protons directly bound to 13C-labeled sites can be almost resolved in (c). The linewidths
of the 1H resonances are of the order of 50 Hz; this linewidth makes it impossible in spectra (d) and (e) to resolve all the resonances
(14 are expected for 10 and 21 for 11). The assignments were determined from 1H–13C HSQC spectra (see Fig. 6). Degenerate
resonances are indicated by forward slashes.
2.5 ppm. The range of chemical shifts for the 13C nuclei
in the rings is not nearly so limited; peaks are dispersed
over 45 ppm. Consequently, it is much easier to assign and
interpret the 13C spectra and then to use this information
to assign the 1H resonances. Figure 4(a) shows the 1H
decoupled 13C spectrum for 9. Resonances located below
60 and above 110 ppm are due to the natural abundance
concentration of 13C in the resin. The saccharide region of
the 13C spectrum is shown in Fig. 4(b), (c) and (d) for 9, 10
and 11, respectively. The six resonances that are expected for
9 can be clearly seen in Fig. 4(b) even though the resonances
corresponding to carbons 2 and 5 are nearly degenerate.
Some additional low-intensity peaks can be seen in this
spectrum at 66–72 and 98 ppm. These resonances are due
to saccharide molecules that are bound directly to the resin
instead of to the octenediol linker. Although these impurities
complicate the spectra to a certain extent, they do not affect
the purity of the final synthesis product because they are not
removed from the resin by the cleavage reaction.
the INADEQUATE spectra can be analyzed in the same
fashion. In the case of uniformly labeled samples, a number
of additional peaks appear in the INADEQUATE spectra due
to the generation of doubly antiphase coherences between
three or even four spins. The four of these coherences that
are most efficiently generated for 9 are indicated by dotted
circles in Fig. 5(a); several weaker peaks that arise due to this
mechanism appear at contour levels that are not visualized in
the spectra shown in Fig. 5. Although both saccharide rings
of 10 could be independently assigned, the close similarities
between the first and second rings in 11 make most of the
13C resonances from these two rings degenerate (at least
within the limits of the experimental resolution). Resonance
assignments that are degenerate are indicated by forward
slashes in Figs 3, 4 and 6.
Once the 13C resonances have been assigned, it is
1
straightforward to use H–13C HSQC spectra to assign the
1H resonances; such spectra are shown for the three samples
in Fig. 6. The 13C and 1H resonance assignments that resulted
from these experiments are given in Table 1.
To assign the 13C resonances, we used the INADEQUATE
experiment, which is well suited for working with linear spin
systems. The arrows in Fig. 5(a) illustrate how it is possible
to ‘walk through’ the INADEQUATE spectrum to assign
the 13C resonances for 9. For 10 [Fig. 5(b)] and 11 [Fig. 5(c)],
13C–13C NOESY, 1H–1H NOESY and 1H–1H ROESY
spectra of the samples did not reveal any cross peaks; we
interpret this as an indication that the local environment of
the saccharide molecules is dynamic, i.e. that the molecules
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 453–458

HR-MAS NMR of 13C-labeled resin-bound saccharides
457
(a)
140 130 120 110 100
90
80
70
60
50
40
30
13C Chemical Shift (ppm)
3
4
5
2
6
1
(b)
2
5
2'
4'
1'
6'
1
3'
3
4
5'
6
(c)
(d)
1/1'
1''
2/2'
2''
4''
3/3'
3''
5''
6''
4/4'
5/5'
6/6'
100
90
80
13C Chemical Shift (ppm)
70
60
Figure 4. (a) 1H decoupled 13C HR-MAS NMR spectra of the 13C labeled monosaccharide sample (9). The carbohydrate region of
the spectrum for 9 is shown in (b); this same region is also shown for (c) 10 and (d) 11. Spectral resonances above 120 and below
50 ppm are mainly due to the natural abundance concentration of 13C nuclei in the resin. In spectra (b)–(d), the resonances at
98 ppm and between 66 and 72 ppm are due to impurities, as discussed in the text. The numbers give the assignments determined
from the INADEQUATE spectra (see Fig. 5). Note that for 11, several of the resonances are nearly degenerate and could not be
disentangled in the INADEQUATE spectra; these degenerate resonances are indicated by forward slashes.
(a)
(b)
(c)
2
1
140
150
160
170
180
3
2
5
4
4
3
6
5
100
90
80
70
100
90
80
70
100
90
80
70
13C Chemical Shift (ppm)
Figure 5. 13C–13C INADEQUATE spectra used for assigning the 13C resonances for (a) 9, (b) 10 and (c) 11. In spectrum (a), arrows
illustrate the coupling pathway connecting the resonances and dotted circles indicate strong cross peaks arising from three
quantum spin order (I
because of additional, less efficiently generated, three-quantum spin order terms.
I I
1x 2z 3z). The other peaks seen in the spectra are due to 13C-labeled impurities (as discussed in the text) and
do not have a well-defined structure when bound to
the swelled resin. This interpretation correlates with the
relatively short relaxation times that we observed for the
saccharide resonances (T1 D ¾500 ms for 13C and ¾300 ms
for 1H).
It is often the case in HR-MAS NMR that better
resolution of the desired resonances can be achieved by
inserting a CPMG pulse train immediately before signal
acquisition.20,21 This takes advantage of the fast rate at
which broad resonances dephase during the CPMG period.
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 453–458

458
N. M. Loening et al.
(a)
(b)
(c)
60
60
70
80
90
100
6
60
6/6'
6''
6'
6
70
70
2''
5''
4''
2'
5'
4'
2
5
2/2'
2
5
5/5'
4
80
4/4'
4
80
3
3/3'
3
3''
3'
90
90
100
1
1/1'
100
1
1''
4.5
1H Chemical Shift (ppm)
1'
5.5
5.0
4.5
4.0
3.5
5.5
5.0
4.5
4.0
3.5
5.5
5.0
4.0
3.5
1H Chemical Shift (ppm)
1H Chemical Shift (ppm)
Figure 6. 1H–13C HSQC spectra for (a) 9, (b) 10 and (c) 11. The increase in spectral congestion with the addition of more saccharide
rings is apparent going from spectrum (a) to (c). Degenerate peaks that cannot be resolved are indicated by forward slashes.
solid-phase syntheses of 13C labeled oligosaccharides, an
application that is important in the synthesis of both uni-
formly and selectively labeled saccharides by solid-phase
techniques. Such labeled materials are becoming increas-
ingly important for studying the structure and function of
oligosaccharides and glycoproteins in biological processes.
Narrow resonances are only slightly attenuated by the CPMG
pulse train and, consequently, are enhanced relative to the
broad signals in the resulting spectrum. The end result is
that broad resonances are removed (or at least attenuated)
in the spectrum and therefore the apparent resolution of
the remaining resonances in the spectrum is improved.
Unfortunately, the interesting resonances in our samples
had relatively fast transverse relaxation times (i.e. they were
the broad components of the signal), so this technique was
not applicable.
Acknowledgments
The authors thank Dr Anthony Bielecki for assistance. N.M.L. thanks
the National Institutes of Health for support via a National Research
Service Award post-doctoral fellowship (F32 NS42425-01). P.H.S.
gratefully acknowledges financial support from GlaxoSmithKline
(Research Scholar Award), the Alfred P. Sloan Foundation (Scholar
Award) and Merck. This work was supported in part by NIH NIBIB
grant EB002026.
CONCLUSION
By incorporating 13C labels, we have succeeded in char-
acterizing and assigning the 1H and 13C NMR spectra of
resin-bound oligosaccharide samples. This highlights an
important application of HR-MAS to the monitoring of
REFERENCES
1. Merrifield RB. J. Am Chem. Soc. 1964; 86: 304.
2. Merrifield RB, Singer J, Chait B. Anal. Biochem. 1988; 174: 399.
3. Caruthers MH. Science 1985; 230: 281.
4. Seeberger PH, Haase W-C. Chem. Rev. 2000; 100: 4349.
5. Seeberger PH (ed), Solid Support Oligosaccharide Synthesis and
Combinatorial Carbohydrate Libraries. Wiley-VCH: New York,
2001.
Table 1. Resonance assignments (in ppm) for the ring carbons
and protons for the resin-bound saccharides 9, 10 and 11. The
resins were swollen with CDCl3, which was also used as the
lock reference
6. Keifer PA, Baltusis L, Rice DM, Tymiac AA, Shoolery JN. J.
Magn. Reson. A 1996; 119: 65.
1H (ppm)
13C (ppm)
Atom
9
10
11
9
10
11
7. Sarkar SS, Gargipati RS, Adams JL, Keifer PA. J. Am. Chem. Soc.
1996; 118: 2035.
8. Fitch WL, Detre G, Holmes CP, Shoolery JN, Keifer PA. J. Org.
Chem. 1994; 59: 7955.
9. Seeberger PH, Beebe X, Sukenick GD, Pochapsky S, Danishef-
sky SJ. Angew. Chem., Int. Ed. Engl. 1997; 36: 491.
10. Kanemitsu T, Kanie O, Wong C-H. Angew. Chem., Int. Ed. Engl.
1998; 37: 3415.
11. Kanemitsu T, Wong C-H, Kanie O. J. Am. Chem. Soc. 2002; 124:
3591.
12. Mogemark M, Elofsson M, Kihlberg J. Org. Lett. 2001; 3: 1463.
13. Mogemark M, Elofsson M, Kihlberg J. ChemBioChem 2002; 3:
1266.
14. Mogemark M, Elofsson M, Kihlberg J. J. Org. Chem. 2003; 68:
7281.
15. Shaka AJ, Keeler J, Frenkiel T, Freeman R. J. Magn. Reson. 1983;
52: 335.
16. Shaka AJ, Keeler J, Freeman R. J. Magn. Reson. 1983; 53: 313.
17. Bodenhausen G, Ruben DJ. Chem. Phys. Lett. 1980; 69: 185.
18. Bax A, Freeman R, Kempsell SP. J. Am. Chem. Soc. 1980; 102:
4849.
1
2
3
4
5
4.47
5.11
3.77
3.66
3.59
4.52
5.16
3.81
3.78
3.65
4.51
5.15
3.80
3.77
3.64
101.2 100.9 100.8
73.0
83.2
77.4
73.2
63.1
73.2
83.2
77.6
75.4
61.9
73.2
83.3
77.7
75.8
62.0
6
4.43/4.30 3.97/3.85 3.95/3.83
10
20
30
40
50
60
100
200
300
400
500
600
4.65
5.20
3.81
3.76
3.66
4.51
5.15
3.80
3.77
3.64
101.4 101.8
72.8
83.3
77.4
73.2
63.2
73.2
83.3
77.7
75.8
62.0
101.2
72.9
83.5
77.5
73.2
62.9
4.48/4.36 3.95/3.83
4.64
5.19
3.80
3.75
3.65
19. Morris GA, Freeman R. J. Am. Chem. Soc. 1979; 101: 760.
20. Carr HY, Purcell EM. Phys. Rev. 1954; 94: 630.
21. Meiboom S, Gill D. Rev. Sci. Instrum. 1958; 29: 688.
4.49/4.34
Copyright  2004 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2004; 42: 453–458