Full text of DOI:10.1002/1097-458X(200012)38:12<1041::AID-MRC782>3.0.CO;2-X

MAGNETIC RESONANCE IN CHEMISTRY
Magn. Reson. Chem. 2000; 38: 1041–1042
Reference Data
each carbon and magnetization transfer to protons was observed. The
complete proton and carbon assignments of 1 and 2 are presented in
Tables 1 and 2.
Conformational analysis of
6-deoxy-6-iodo-D-glucose in aqueous solution
∗
Marie-Christine Brochier-Salon and
Christophe Morin
Table 1. 13C and 1H chemical shifts for 1 and 2
13Ca,
1Hb,
13Ca,
1Hb,
Laboratoire d’Etudes Dynamiques et Structurales de la
Se´lectivite´, UMR CNRS 5616, Universite´ Joseph Fourier, B.P. 53,
38041 Grenoble Cedex 9, France
υ (ppm) υ (ppm) Compound υ (ppm) υ (ppm) Compound
1
91.99
95.81
5.2348
4.6390
1˛
1ˇ
92.11
95.72
5.2094
4.7095
2˛
2ˇ
2
71.40
74.06
3.5375
3.2479
1˛
1ˇ
71.44
74.11
3.5456
3.2573
2˛
2ˇ
Received 4 February 2000; revised 21 July 2000; accepted 8 August
2000
3
72.68
75.68
3.7197
3.4929
1˛
1ˇ
72.15
75.01
3.7408
3.5196
2˛
2ˇ
ABSTRACT: Complete 1H and 13C assignments for 6-deoxy-6-iodo-
D-glucose, an analogue of glucose which is transported across the cell
membrane, are given. These allowed rotameric populations around the
C(5)—C(6) bond to be determined. Comparison with those observed
for glucose suggests that the 6-OH group does not play a major role in
glucose transport. Copyright 2000 John Wiley & Sons, Ltd.
4
69.59
69.54
3.4145
3.4059
1˛
1ˇ
73.61
73.27
3.3196
3.3466
2˛
2ˇ
5
71.35
75.84
3.8365
3.4685
1˛
1ˇ
69.49
73.73
3.5444
3.1997
2˛
2ˇ
6R
6S
60.55
60.69
3.7628
3.7248
1˛
1ˇ
7.93
6.70
3.4339
3.4086
2˛
2ˇ
KEYWORDS: NMR; 1H NMR; 13C NMR; conformational analysis;
rotamer population; 6-deoxy-6-iodo-D-glucose
60.55
60.69
3.8452
3.8989
1˛
1ˇ
7.93
6.70
3.5746
3.5896
2˛
2ˇ
a Experimental data.
b Data obtained from simulated spectra.
INTRODUCTION
Medical imaging of glucose-based biological events being of current
interest, various iodinated glucose derivatives have been prepared for
their possible use in SPECT (single photon emitted computed tomogra-
phy) imaging. Among these analogues of glucose (1), 6-deoxy-6-iodo-
D-glucose (2), has shown the required properties of a tracer of D-glucose
transport1 and we have undertaken an NMR study of 22 to determine its
conformation in aqueous solution, which could be of help in understand-
ing on structural grounds why and how it mimics glucose. In order to
compare the results with those for 1, the spectrum of 13–5 was analysed
under identical conditions.
Table 2. J(1H, 1H) coupling constants (Hz) for 1
and 2a
1˛
1ˇ
2˛
2ˇ
J(1,2)
J(2,3)
3.84
9.23
7.50
9.00
3.84
9.65
8.03
9.32
J(3,4)
J(4,5)
9.10
9.90
9.09
9.80
9.28
9.50
9.21
9.21
J(5,6R)
J(5,6S)
J(6R,6S)
5.40
2.20
ꢀ11.80
5.95
1.95
ꢀ12.25
5.80
2.90
ꢀ11.00
5.77
2.72
ꢀ11.24
a Data obtained from simulated spectra.
As for glucose, 2 exists as pyranose (˛ : ˇ D 41 : 59) in the stable
4C1 conformation. As the only structural difference between glucose
and its iodinated analogue 2 pies in the C-6 substituent, rotameric pop-
ulations around the C(5)—C(6) bond were determined using 3J.5, 6R/
and 3J.5, 6S/ vicinal coupling constants.5 The experimental coupling
3
constant Jexp is linked to rotamer population by
RESULTS AND DISCUSSION
X
3Jexp
D
pi3Ji
.1/
The 1H resonances of 1 and 2 were assigned through the combined
use of homonuclear COSY and 1D-TOCSY experiments, starting from
the readily identified anomeric protons. The coupling constants J were
determined after iterative analysis and spin simulation. The assign-
ment of the prochiral H-6 protons (H-6R and H-6S) is based on the
observed higher frequency of H-6S than H-6R.3 These assignments are
also consistent with the relative magnitudes of vicinal coupling con-
stants, 3J(5,6R) being larger than 3J(5,6S).4 Heteronuclear correlation
experiments (HMQC, HMBC) were used to assign 13C resonances, with
the exception of C-2˛ and C-5˛ for 1 and C-2ˇ and C-5ˇ for 2, which
are not well enough resolved to allow their unambiguous assignment by
these techniques. Therefore, selective irradiations were performed on
iD1,2,3
where pi represents the population for each rotamer gg, gt and tg and
Ji its theoretical coupling constant.
* Correspondence to: M.-C. Brochier-Salon, Laboratoire d’Etudes Dynamiques
et Structurales de la Se´lectivite´, UMR CNRS 5616, Universite´ Joseph Fourier,
B.P. 53, 38041 Grenoble Cedex 9, France; e-mail: brochier@efpg.inpg.fr
Copyright 2000 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2000; 38: 1041–1042

1042
Reference Data
(1H) and 125.6 MHz (13C). The residual signal of HOD was used
as internal reference (υ D 4.663 ppm at 37 °C), after temperature
calibration using sodium 2,2,3,3-tetradeutero-3-trimethylsilylpropanoate
in D2O.
As electronegative substituents modify the value of theoretical cou-
pling constants J, the general Karplus equation as modified by Colucci
et al.6 was used to calculate the Ji parameter for each of the rotamer:7
1H NMR spectra were acquired using a 1230 Hz spectral width,
32 K data points, 13 s acquisition time, 1 s relaxation delay and 10 µs
for a 90° pulse. Zero-filling was performed without apodization to
give a resolution of 0.037 Hz per point. 13C spectra were obtained
using a 15 kHz spectral width, 20 K data points, 0.7 s acquisition time,
1 s relaxation time and 20 µs for a 60° pulse. Zero-filling and 0.4 Hz
exponential line broadening apodization provided a resolution of 0.9 Hz
per point. T1 relaxation times were obtained by the inversion–recovery
method.
3Ji D A C B cos Âi C C cos 2Âi C cos Âi[.S1 C S4/ cos.Âi ꢀ 120/
C .S2 C S3/ cos.Âi C 120/]
.2/
where Âi is the angle between the two C—H bonds in the rotamer i and
Sn is the contribution of each substituent Xn.
Thus, for glucose (1), S is zero for H, 4.60 for OR, 3.28 for C(OR)
and C(O)R groups and 5.20 for OH with coefficients A, B and C being
8.17, ꢀ1.96 and 6.54 respectively;6 for example, application of these
data to 2 (S being 2.80 for iodine)6 gives the following equation
system from which rotameric populations can be extracted:
1H–1H COSY spectra were recorded on 2K data points for 1K
increments in the F1 dimension and processed using a 4096 ð 4096
transformed matrix with zero-filling in each dimension. Sinusoidal
multiplication was performed before Fourier transformation.
HMQC and HMBC spectra were recorded without proton decou-
pling, on the spectral windows used for 1D spectra acquisition, using
a 2048 ð 256 matrix. Delay was optimized for nJ.H,C/ D 5 Hz and
3J.5, 6R/ D 4.13ptg C 1.04pgg C 11.33pgt
3J.5, 6S/ D 11.33ptg C 2.03pgg C 3.14pgt
ptg C pgg C pgt D 1
.3/
.4/
.5/
1J.C,H/
D 150 Hz. Zero-filling in both dimensions (4096 ð 2048
matrix) and sine multiplication were performed prior to Fourier trans-
formation. A forward linear prediction on 1024K in the 13C dimension
enhanced resolution.
The calculated relative populations ptg, pgg and pgt of 1 and 2 are dis-
played in Table 3. It can be seen that the tg population, which reflects the
syn diaxial interaction between the C-6 substituent and the 4-OH, is neg-
ative but such negative values have frequently been observed for glucose
derivatives5 and, although not being of physical significance, they can be
used to assay relative populations. Variations in rotameric populations
upon comparison of 1 and 2 (Table 3) are found to be minimal, which
suggests that, as the transport of carbohydrates across the cell membrane
is controlled by hydrogen bonding, the 6-substituent (hydroxyl for glu-
cose and iodine for 2) does not play a significant role in their entry into
the cell. Whether bulkier and/or more lipophilic groups (such as affinity
markers) can be introduced at position 6 remains to be investigated.
1D TOCSY experiments were recorded using a selective ‘e-burp’-
shaped pulse (pulse width D 240 ms) on each anomeric proton; the
mixing time was incremented from 10 to 100 ms.
Selective irradiations of individual 13C sites followed by polarization
transfer to proton were performed using the reverse detection mode
pulse sequence proposed by Blechta et al.8a and Nishida et al.,8b in
which some modifications were introduced: broadband WALTZ homo-
decoupling was applied during the preparation delay (3 s), then a
half-Gaussian selective excitation (94 ms) was applied to a selected
13C site under decoupled conditions. The polarization transfer used a
3.4 ms D 0.5/J.C,H/ delay. A 1230 Hz spectral width was retained and
a slight exponential broadening (0.6 Hz) was applied prior to Fourier
transformation.
EXPERIMENTAL
Materials
Acknowledgement
D-Glucose was from a commercial source and 6-deoxy-6-iodo-D-glucose
was prepared according to the literature procedure.2a In aqueous solu-
tion, 2 equilibrates to a mixture of anomers which, as for glucose, can
be separated by HPLC; however, each isolated anomer mutarotated to
produce the initial mixture;2a therefore, all NMR experiments on 1 and
2 were performed on pre-equilibrated aqueous solutions of mixture of
anomers.
We are grateful to C. Bougault for critical reading of the manuscript.
REFERENCES
1. Henry C, Tanti J-F, Gre´meaux T, Morin C, Van Obberghen E,
Comet M, LeMarchand-Brustel Y. Nucl. Med. Biol. 1997; 24:
99.
2. (a) Charronneau E, Mathieu J-P, Morin C. Appl. Radiat. Isot. 1998;
49: 1605. (b) Garcia Fernandez JM, Gadelle A, Defaye J. Carbohydr.
Res. 1994; 265: 249.
3. Rockwell GD, Grindley TB. J. Am. Chem Soc. 1998; 120: 10953.
4. Haasnoot CAG, DeLeeuw FAAM, Altona C. Tetrahedron 1980; 36:
2783.
5. Bock K, Duus JF. J. Carbohydr. Chem. 1994; 13: 513.
6. Colucci WJ, Jung SJ, Gandour RD. Magn. Reson. Chem. 1985; 23:
335.
Spectra
NMR spectra were recorded in 5 mm tubes (c D 0.17 mol lꢀ1 in D2O) at
37 °C using a Varian UNITYC spectrometer operating at 499.89 MHz
Table 3. Rotamer populations about C5—C6 cal-
culated for 1 and 2 in solution
ptg(%)
pgg(%)
pgt(%)
Compound
7. Altona C, Ippel JH, Westra Hoekzema AJA, Erkelens C, Groes-
beek M, Donders L. Magn. Reson. Chem. 1989; 27: 564.
8. (a) Blechta F, Del Rio-Portilla R, Freeman R. Magn. Reson. Chem.
1994; 32: 134. (b) Nishida G, Widdmalm P, Sandor P. Magn. Reson.
Chem. 1995; 33: 596.
˛
ˇ
ꢀ1
4
47
51
45
53
54
45
62
45
1
2
1
2
ꢀ7
2
Copyright 2000 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2000; 38: 1041–1042