Two-bond 13C-13C spin-coupling constants in carbohydrates: New measurements of coupling signs

Article abstract of DOI:10.1016/S0008-6215(98)00123-2

D-(1,3,6-¹³C₃)Allose (1), (¹³C)methyl α-D-(1,2-¹³C₂)glucopyranoside (2) and (¹³C)methyl β-D-(1,2-¹³C₂)glucopyranoside (3) were synthesized and used to establish the signs of their constituent ²J(CCC) or ²J(COC) values (²J(C1,C3) in the α-pyranose of 1 (15), and ²J(C1,CH₃) in 2 and 3). Compounds 2, 3 and 15 contain three mutually coupled labeled carbons, thus creating a three-spin system from which crosspeak displacements in ¹³C-¹³C COSY-45 spectra were used to determine coupling signs. In all compounds, at least one ³J(CC) value was present as an internal reference: ³J(C2,CH₃ in 2 and 3, and ³J(C1,C6) and ³J(C3,C6) in 15. ²J(C1,CH₃), in 2 and 3, and ²J(C1,C3) in 15, were found to be negative, thus providing experimental confirmation of the sign predictions made via the projection resultant rule described recently.

Full text of DOI:10.1016/S0008-6215(98)00123-2

Carbohydrate Research 309 (1998) 145±152
Two-bond 13C±13C spin-coupling constants in
carbohydrates: New measurements of coupling
signs
Shikai Zhaoa, Gail Bondoa, Jaroslav Zajicekb, Anthony S. Seriannib,*
a Omicron Biochemicals, Inc., 1347 N. Ironwood Drive, South Bend, IN 46615-3566, USA
b Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, IN 46556, USA
Received 19 February 1998; accepted 23 April 1998
Abstract
d-(1,3,6-13C3)Allose (1), (13C)methyl ꢀ-d-(1,2-13C2)glucopyranoside (2) and (13C)methyl ꢁ-d-
(1,2-13C2)glucopyranoside (3) were synthesized and used to establish the signs of their constituent
2JCCC or 2JCOC values (2JC1,C3 in the ꢀ-pyranose of 1 (15), and 2JC1,CH in 2 and 3). Compounds 2,
3 and 15 contain three mutually coupled labeled carbons, thus creating a three-spin system from
3
which crosspeak displacements in 13C±13C COSY-45 spectra were used to determine coupling
3
3
signs. In all compounds, at least one JCC value was present as an internal reference: JC2,CH in 2
2
3
3
3
2
and 3, and JC1,C6 and JC3,C6 in 15. JC1,CH in 2 and 3, and JC1,C3 in 15, were found to be
negative, thus providing experimental con®rmation of the sign predictions made via the projection
3
resultant rule described recently. # 1998 Elsevier Science Ltd. All rights reserved
2
Keywords: JCOC; 2JCCC
;
13C-13C COSY-45; spin-coupling signs; projection resultant method
1. Introduction
key question is the amplitude of this motion, that
is, the extent of the excursions experienced by a
given torsion angle. These questions are frequently
addressed experimentally by exploiting 1H±1H
nuclear Overhauser e€ects (NOE) between resi-
dues, but signi®cant, and often overlooked, pro-
In recent years considerable discussion about the
conformational properties of oligosaccharides has
focused on the ¯exibility of their constituent gly-
cosidic linkages [1]. Much of this discussion, how-
ever, can be misleading, since it is likely that the
¯exibility of a given linkage will depend on its
context, that is, on the speci®c structural environ-
ment in which the linkage is found. The term `¯ex-
ibility' is hierarchical; in reality, all linkages will
experience some degree of motion or libration. The
6
blems accompany this approach. Due to the r
dependence of the NOE, the interpretation of
NOEs in structural terms is complicated when
conformational ¯exibility exists, since the aver-
aging of NOEs is non-linear, as discussed by Jar-
detzky [2]. NOEs will be skewed in favor of those
conformers containing the more potent relaxation
pathways (i.e., the shorter internuclear distances)
even when they are present in minor abundance, as
* Corresponding author. Fax: 001 219 631 6924; e-mail:
serianni.1@nd.edu
0008-6215/98/$Ðsee front matter # 1998 Elsevier Science Ltd. All rights reserved
PII S0008-6215(98)00123-2

146
S. Zhao et al./Carbohydrate Research 309 (1998) 145±152
clearly explained by Homans [3]. This limitation is
overcome by maximizing the number of NOEs
used in the analysis. Unfortunately, in many oli-
gosaccharides, the number of NOEs that are sensi-
tive to a given linkage geometry is limited. Thus,
despite its widespread use, NOE is not, on its own,
capable of resolving the issue of conformational
¯exibility in oligosaccharides in a unequivocal
fashion.
Fig. 1. Trans-O-glycoside JCH and JCC sensitive to the glyco-
sidic torsional angles, ꢂ and .
Additional information on O-glycoside linkage
conformation can, in principle, be derived from the
analysis of JCH and JCC values across these lin-
kages [4]. For the linkage shown in Fig. 1, two
as described by King-Morris et al. [10], was oxi-
dized with Pb(OAc)4 to give d-(4-13C)erythrose (4)
2
3
3JCOCH, one JCOC and three JCOCC values are
available, and unlike NOE, these scalar couplings
should, in principle, be interpreted reliably in the
presence of conformational averaging [2,5].
Importantly, the ꢂ and torsion angles (Fig. 1)
can each be assessed independently by three cou-
plings, yielding data redundancy which improves
the accuracy of the analysis.
[11]. Compound
4
was converted to d-
(1,5-13C2)arabinose (5) and d-(1,5-13C2)ribose (6)
by cyanohydrin reduction [12,13], and the mixture
was puri®ed by chromatography on Dowex-50Â8
(200±400 mesh) in the Ca+2 form [14,15]; com-
pound 5 eluted ®rst, followed much later by
compound 6 [16]. Compound 5 was converted to
d-(2,5-13C2)ribose (7) by molybdate-catalyzed epi-
merization [17]. Subsequent cyanohydrin reduction
applied to 7 a€orded pure 1 after chromatography
on Dowex-50Â8 (200±400 mesh) in the Ca+2 form
[14,15]; the C-2 epimer, d-(1,3,6-13C3)altrose (8),
eluted ®rst, followed by 1. Compound 1 was iden-
ti®ed by its characteristic 13C chemical shifts [18].
A prerequisite in the use of the trans-O-glycoside
couplings shown in Fig. 1 is a detailed under-
standing of how they are in¯uenced by molecular
structure. Mulloy et al. [6] and Tvaroska et al. [7]
have proposed virtually identical Karplus curves
3
2
for the treatment of JCOCH values, but JCOC and
3JCOCC values in carbohydrates remain poorly
understood. In response to the latter, we proposed
recently a projection resultant method to predict
2
2
the magnitudes and signs of JCCC and JCOC
values in carbohydrates [8], and some experimental
veri®cation of the method has been reported [9a].
In the present study, we provide new veri®cation of
the method through the experimental determina-
2
tion of JCC coupling signs using multiply 13C-
labeled compounds and the 13C±13C COSY-45
method [9b].
(13C)Methyl ꢀ- (2) and b- (3) d-(1,2-13C2)gluco-
pyranosides were prepared from (13C)methanol
and d-(1,2-13C2)glucose [15] by Fischer glycosida-
tion using THF as a cosolvent and Dowex-50 H+
ion-exchange resin as the catalyst [19] according to
the following procedure. HCR-W2 (H+) ion-
exchange resin (Sigma Chemical Co.), washed with
water and dried under vacuum overnight, was
added to a 10-mL round-bottomed ¯ask contain-
ing d-(1,2-13C2)glucose (0.36 g, Omicron Biochem-
icals), and (13C)methanol (0.5 mL, Aldrich
Chemical Co.) and anhydrous THF (3 mL) were
added. The mixture was re¯uxed for 2 h under
nitrogen, cooled, and ®ltered to remove the resin.
The recovered resin was washed with 2-propanol,
and the combined ®ltrate and washings were
2. Experimental
Synthesis of 13C-labeled compounds.Ðd-(1,3,6-13C3)-
Allose (1), (13C)methyl ꢀ-d-(1,2-13C2)glucopyrano-
side (2) and (13C)methyl ꢁ-d-(1,2-13C2)glucopyr-
anoside (3) were prepared and puri®ed by methods
reported previously, and only a brief account of
their synthesis is reported here.
The preparation of 1 is outlined in Scheme 1. d-
(6-13C)Glucose, prepared from 1,2-O-isopropyl-
idene-ꢀ-d-xylo-pentodialdo-1,4-furanose and K13CN
(99 atom% 13C, Cambridge Isotope Laboratories)

S. Zhao et al./Carbohydrate Research 309 (1998) 145±152
147
concentrated in vacuo to give a colorless syrup
(ꢀ0.39 g). The syrup was dissolved in a minimum
volume of water, and the solution was applied to a
column (2.5 cmÂ60 cm) of Dowex-1Â2 (200±400
mesh) ion-exchange resin in the OH form [20a].
The column was eluted with water, and fractions
(18 mL) were collected and assayed for sugar with
phenol±sulfuric acid [20b]. Fractions 24±28 and
32±35 were pooled and concentrated in vacuo to
give pure (13C)methyl ꢀ-d-(1,2-13C2)glucopyrano-
side (2) (0.15 g) and (13C)methyl ꢁ-d-(1,2-13C2)glu-
copyranoside (3) (0.07 g), respectively, as syrups.
Compounds 2 and 3 were identi®ed by their char-
acteristic 13C chemical shifts [18].
to double Fourier transformation. 13C Chemical
shifts are reported in ppm, and are referenced to the
C-1 chemical shift of ꢀ-d-mannopyranose (95.5 ppm).
3. Results
De®nition of the problem.ÐIn a previous report
[8], we proposed that the magnitude and signs of
2JCCC values could be correlated with an empiri-
cally derived projection resultant, with the latter
determined by inspection of two rotamer projec-
tions along the molecular fragment containing the
coupling constant under consideration. Using a
range of standard compounds, these resultants
NMR spectroscopy.ÐNMR spectra of 2 and 3
2
2
(ꢀ100 mM in H2O, 5 mm tubes) were obtained at
were directly correlated with JCCC magnitude and
ꢁ
30 C on a Varian UnityPlus NMR spectrometer
1
sign, with more positive resultants yielding more
positive couplings. For example, ꢁ-d-glucopyr-
anose and ꢀ-d-allopyranose yield projection resul-
(599.89 MHz H, 150.85 MHz 13C). 2D 13C±13C
COSY data sets were collected in the absolute
intensity mode using pulse sequences described
previously [21]. The WALTZ scheme of broadband
decoupling was used to enhance the intensity of the
detected carbon signals through NOE. The t1
increments (3072) of eight scans each were sampled
in 4096 data points, and zero-®lling to 4 and 8 K
was applied in the F1 and F2 dimensions, respec-
tively. Sinebell weighting functions (determined
empirically) were applied in both dimensions prior
2
tants of +2.5 and 0.5, respectively, and JC1,C3
values of +4.5 and 2.4 Hz are predicted in these
aldohexoses (Fig. 2). This correlation was then
tested experimentally by determining the signs of
2JC1,C3 in ꢀ- (9) and ꢁ- (10) d-(1,3,6-13C3)gluco-
pyranoses and ꢀ- (11) and ꢁ- (12) d-
(1,3,6-13C3)mannopyranoses from crosspeak dis-
placements in 13C±13C COSY-45 spectra [9a].
3
These experiments con®rmed that JC1,C3 in 10
Scheme 1.

148
S. Zhao et al./Carbohydrate Research 309 (1998) 145±152
(+4.5 Hz) and 12 (+4.0 Hz) were positive as pre-
dicted (2JC1,C3=ꢀ0 Hz in 9 and 11). To date,
however, experimental veri®cation of negative
considerably more limited. Projection resultants
2
calculated for JC1,C5 in d-aldopyranoses were
more positive for ꢁ anomers than for ꢀ anomers,
2
2
2
2JCCC values (e.g., JC1,C3 in ꢀ-d-allopyranose,
and by analogy to JCCC behavior, JC1,C5 in ꢀ
anomers was predicted to be more negative (less
positive) than 2JC1,C5 in ꢁ anomers. Since 2JC1,C5 is
ꢀ0 Hz in ꢁ anomers and ꢀ2.0 Hz in ꢀ anomers
[22], the latter couplings were predicted to be
negative, and this prediction has been con®rmed
experimentally via 13C±13C COSY-45 analysis of ꢀ-
(13) and ꢁ- (14) d-(1,5,6-13C3)mannopyranoses
[9a]. Importantly, the extrapolation of these results
Fig. 2) has not been reported, although the method
predicts the existence of such couplings for resul-
2
tants less than ꢀ1. Observation of negative JCCC
values consistent with predictions would provide a
key validation of the method.
2
to JCOC behavior across O-glycosidic linkages
allowed for the prediction of 2JCOC behavior in two
of the three rotamers that relate to ꢂ (rotamers A
and C, Fig. 3), namely, those containing C±O±C
coupling pathways structurally related to those
found in the ꢀ- and ꢁ-d-aldopyranoses; since no
projection is derived from , this torsion does not
appear to in¯uence the value of 2JCOC signi®cantly
[8]. No model compound, however, was available
at that time to evaluate the third rotamer (rotamer
B, Fig. 3). The projection method predicts a nega-
tive coupling in rotamer B, and this prediction is
tested in this report.
2
The treatment of JCCC values was extended to
2JCOC values, where access to model compounds is
2
Fig. 2. Correlation between projection resultants and JC1,C3 in ꢁ-d-glucopyranose and ꢀ-d-allopyranose.

S. Zhao et al./Carbohydrate Research 309 (1998) 145±152
149
The C-1±C-6, C-3±C-6 and C-1±C-3 13C±13C
COSY-45 crosspeaks for 15 are shown in Fig. 4.
The C-1±C-6 and C-3±C-6 crosspeaks show the
same relative displacements as expected, since both
couplings are positive. However, the C-1±C-3
crosspeak shows the opposite displacement, thus
2
2
New JCCC and JCOC sign determinations.ÐThe
13C-labeling pattern in d-(1,3,6-13C3)allose (1)
creates a mutually-coupled three-spin system in
ꢀ-d-(1,3,6-13C3)allopyranose (15) involving three
2
JCC values [22b,23], namely, JC1,C3 (2.4 Hz),
3JC3,C6 (2.8 Hz) and JC1,C6 (3.2 Hz). This three-
3
2
indicating that JC1,C3 is negative, which con®rms
the projection resultant prediction.
spin system is required in order to determine
coupling signs straightforwardly via 13C±13C
COSY-45 [9a]. The two three-bond (vicinal) cou-
plings are expected to be positive in sign [24].
Consequently, the crosspeak displacements
observed for the C-1±C-6 and C-3±C-6 crosspeaks
should have the same relative orientation and can
be correlated with a positive coupling. The sign of
2JC1,C3 can be established by comparing these
standard displacements with that observed for the
C-1±C-3 crosspeak (i.e., same relative displace-
ment, + sign; di€erent relative displacement,
sign).
As in 15, a mutually coupled three-spin system is
present in (13C)methyl ꢀ- (2) and ꢁ- (3) d-
involving
(1,2-13C2)glucopyranosides
1JC1,C2
,
3
2JC1,CH and JC2,CH . The C-2±CH3 and C-1±CH3
3
3
crosspeaks in the 13C±13C COSY-45 spectrum of 3
show opposite displacements (Fig. 5). Since the
relative displacement observed for the three-bond
pathway is correlated with a positive coupling,
2JC1,CH must be negative in sign. Similar data were
3
obtained on 2 (data not shown). These results
con®rm the projection rule predictions made earlier
for rotamer B in Fig. 3 [8].
Fig. 3. Correlation between projection resultants and trans-O-glycoside 2JCOC derived from model studies of 2JC1,C5 in ꢀ- and ꢁ-d-
aldopyranoses.

150
S. Zhao et al./Carbohydrate Research 309 (1998) 145±152
4. Discussion
Fig. 3), and this preferred orientation is supported
2
by the relatively large JC2,CH values observed in
these structures (e.g., 3.0 Hz in 2) and the magni-
3
The results of this investigation provide new
experimental validation of the projection resultant
method for the prediction of JCCC and JCOC
values in carbohydrates. The negative sign of
2JC1,C3 in 15, and the negative signs of 2JC1,CH in 2
and 3, ful®ll the predictions and provide critical
tests of the method. In 2 and 3, the exoanomeric
e€ect [25±27] stabilizes the C-1±O-1 rotamer hav-
ing O-1 gauche to H-1 and O-5 (rotamer B in
3
tudes of JCH ,H1 values reported previously by
3
2
2
Lemieux and Koto in methyl aldopyranosides (3.2±
4.6 Hz) [28]. Thus, while the presence of rotamers
A and C (Fig. 3) cannot be excluded in solutions of
2 and 3, their contributions to the observed cou-
plings are expected to be small.
3
The origin of the displacements observed in
COSY-45 spectra, from which coupling signs are
Fig. 4. Crosspeaks observed in the 13C±13C COSY-45 spectrum of 15. (A) C-1±C-6 crosspeak. (B) C-3±C-6 crosspeak. (C) C-1±C-3
2
crosspeak. The relative displacement observed in (C) di€ers from that observed in (A) and (B), indicating that the sign of JC1,C3
(negative) is opposite to that of 3JC1,C6 and 3JC3,C6 (positive).

S. Zhao et al./Carbohydrate Research 309 (1998) 145±152
151
determined, has been explained previously [5,29].
The structures of the crosspeaks for 15 (Fig. 4)
di€er from those observed in 3 (and 2) (Fig. 5) in
conformational properties of oligosaccharides. In
some instances, torsional motion about these lin-
kages might be signi®cant, leading to the averaging
of the observed coupling. Given the relatively small
expected range of values (ꢀ2 Hz), some limitations
might accrue from the use of these couplings in
these instances. However, recent reports have
shown that JCC values as low as 0.1±0.2 Hz can be
measured reliably in biomolecules [30], including
small proteins, and the application of these meth-
ods to oligosaccharides should permit a meaningful
1
that the large JC1,C2 is absent in the former. This
di€erence makes the detection of the displacement
in the COSY-45 data for 15 more dicult, but
nevertheless observable. Crosspeak displacements
as small as ꢀ2 Hz were readily observed in this
study, and a lower limit for the detection of dis-
placements is estimated to be 1±1.5 Hz when a
1
large JCC is absent (e.g., 15), and 0.5±1Hz when
1JCC splits the crosspeak (e.g., 2 and 3).
interpretation of JCOC values in conformationally
¯exible molecules.
2
The values of 2JCOC across O-glycosidic linkages
range from 0 Hz to ꢀ 2 Hz (Fig. 3), and appear to
depend primarily on the phi (ꢂ) torsion angle [8]
(Fig. 1). Thus, when used in conjunction with other
In conclusion, this study demonstrates the utility
of triply 13C-labeled compounds and the 13C±13C
COSY-45 method in the determination of 13C±13C
spin-coupling signs in saccharides, and has provided
further experimental support of the projection
structural information (e.g., NOE, 3JCOCC
3JCOCH), they may improve the analysis of the
,
Fig. 5. Crosspeaks observed in the 13C±13C COSY-90 and 13C±13C COSY-45 spectra of 3. (A) C-2±CH3 crosspeak, COSY-90. (B)
C-2±CH3 crosspeak, COSY-45. (C) C-1±CH3 crosspeak, COSY-90. (D) C-1±CH3 crosspeak, COSY-45. No displacements are
expected in the COSY-90 data, which serve as controls. The opposite relative displacements observed in (B) and (D) indicate that
3JC2,CH and 2JC1,CH have opposite signs (positive and negative, respectively).
3
3

152
S. Zhao et al./Carbohydrate Research 309 (1998) 145±152
resultant method for predicting the magnitudes
and signs of 2JCC values in these molecules. Future
work will extend the use of these coupling con-
stants, and 3JCOCC values (Fig. 1), to studies of the
conformational properties of the O-glycosidic lin-
kages in 13C-labeled oligosaccharides.
[10] M.J. King-Morris, P.B. Bondo, R.A. Mrowca, and
A.S. Serianni, Carbohydr. Res., 175 (1988) 49±58.
[11] A.S. Perlin, Methods Carbohydr. Chem., 1 (1962)
64±66.
[12] A.S. Serianni, H.A. Nunez, and R. Barker, Carbo-
hydr. Res., 72 (1979) 71±78.
[13] A.S. Serianni and P.B. Bondo, J. Biomol. Struct.
Dyn., 11 (1994) 1133±1148.
[14] S.J. Angyal, G.S. Bethell, and R. Beveridge, Car-
bohydr. Res., 73 (1979) 9±18.
Acknowledgement
[15] A.S. Serianni, T. Vuorinen, and P.B. Bondo, J.
Carbohydr. Chem., 9 (1990) 513±541.
[16] P.C. Kline and A.S. Serianni, J. Am. Chem. Soc.,
112 (1990) 7373±7381.
Financial support of this work by Omicron Bio-
chemicals, Inc. of South Bend, IN, is gratefully
acknowledged.
[17] M.L. Hayes, N.J. Pennings, A.S. Serianni, and R.
Barker, J. Am. Chem. Soc., 104 (1982) 6764±6769.
[18] K. Bock and C. Pedersen, Adv. Carbohydr. Chem.
Biochem., 41 (1983) 27±66.
References
[1] (a) R.U. Lemieux, K. Bock, L.T.J. Delbaere, S.
Koto, and V.S. Rao, Can. J. Chem., 58 (1980) 631±
653; (b) H. Paulsen, T. Peters, V. Sinnwell, M.
Heume, and B. Meyer, Carbohydr. Res., 156 (1986)
87±106; (c) D.A. Cumming and J.P. Carver, Bio-
chemistry, 26 (1987) 6664±6676. (d) Z.-Y. Yan and
C.A. Bush, Biopolymers, 29 (1990) 799±811.
[2] O. Jardetzky, Biochem. Biophys. Acta, 621 (1980)
227±232.
[19] C.A. Podlasek, J. Wu, W.A. Stripe, P.B. Bondo,
and A.S. Serianni, J. Am. Chem. Soc., 117 (1995)
8635±8644.
[20] (a) P.W. Austin, F.E. Hardy, J.C. Buchanan, and
J. Baddiley, J. Chem. Soc., 1963, 5350±5353. (b)
J.E. Hodge and B.T. Hofreiter, Methods Carbo-
hydr. Chem., 1 (1962) 380±394.
[21] (a) G. Bodenhausen, R. Freeman, R. Niedermeyer,
and D.L. Turner, J. Magn. Reson., 26 (1977) 133±
164. (b) P. Bachman, W.P. Aue, L. Mueller, and
R.R. Ernst, J. Magn. Reson., 28 (1977) 28±39.
[22] (a) R. Barker and T.E. Walker, Methods Carbo-
hydr. Chem., 8 (1980) 151±165. (b) M.J. King-
Morris and A.S. Serianni, J. Am. Chem. Soc., 109
(1987) 3501±3508.
[23] J. Wu, P.B. Bondo, T. Vuorinen, and A.S. Ser-
ianni, J. Am. Chem. Soc., 114 (1992) 3499±3505.
[24] J.L. Marshall, Carbon±Carbon and Carbon±Proton
NMR Couplings: Applications to Organic Stereo-
chemistry and Conformational Analysis, Verlag
Chemie International, Deer®eld Beach, FL, 1983,
pp. 65±122.
[3] S.W. Homans, Prog. NMR Spectrosc., 22 (1990)
55±81.
[4] (a) T.J. Rutherford, J. Partridge, C.T. Weller, and
S.W. Homans, Biochemistry, 32 (1993) 12715±
12724; (b) D.Y. Gagnaire, R. Nardin, F.R. Tar-
avel, and M.R. Vignon, Nouv. J. Chem., 1 (1977)
423±430. (c) H.A. Nunez and R. Barker, Biochem-
istry, 19 (1980) 489±495. (d) M.L. Hayes, A.S.
Serianni, and R. Barker, Carbohydr. Res., 100
(1982) 87±101. (e) C. Landersjo, R. Stenutz, and G.
Widmalm, J. Am. Chem. Soc., 119 (1997) 8695±
8698.
[5] M. Eberstadt, G. Gemmecker, D.F. Mierke, and
H. Kessler, Angew. Chem., Int. Ed. Engl., 34 (1995)
1671±1695.
[25] R.U. Lemieux, Pure Appl. Chem., 25 (1971) 527±
548.
[6] B. Mulloy, T.A. Frenkiel, and D.B. Davies, Car-
bohydr. Res., 184 (1988) 39±46.
[26] R.U. Lemieux, S. Koto, and D. Voisin, ACS Symp.
Ser., 87 (1979) 17±29.
[7] I. Tvaroska, M. Hricovini, and E. Petrakova, Car-
bohydr. Res., 189 (1989) 359±362.
[8] T. Church, I. Carmichael, and A.S. Serianni, Car-
bohydr. Res., 280 (1996) 177±186.
[27] H. Thogersen, R. Lemieux, K. Bock, and B.
Meyer, Can. J. Chem., 60 (1982) 44±57.
[28] R.U. Lemieux and S. Koto, Tetrahedron, 30 (1974)
1933±1944.
[9] (a) A.S. Serianni, P.B. Bondo, and J. Zajicek, J.
Magn. Reson. Ser. B, 112 (1996) 69±74. (b) A. Bax
and R. Freeman, J. Magn. Reson., 44 (1981) 542±
561.
[29] C. Biamonti, C.B. Rios, B.A. Lyons, and G.T.
Montelione, Adv. Biophys. Chem., 4 (1994) 51±120.
[30] J.-S. Hu and A. Bax, J. Am. Chem. Soc., 118 (1996)
8170±8171.