Fluorosugars: Synthesis of the 2,3,4-trideoxy-2,3,4-trifluoro hexose analogues of d-glucose and d-altrose and assessment of their erythrocyte transmembrane transport

Article abstract of DOI:10.1039/c0cc01128b

The 2,3,4-trideoxy-2,3,4-trifluoro hexose analogues of d-glucose and d-altrose were prepared and characterised and these novel sugar analogues were explored by 2D-¹⁹F-EXSY NMR for their potential to cross erythrocyte (red blood cell) membranes,

Full text of DOI:10.1039/c0cc01128b

Chemical Communications
www.rsc.org/chemcomm
Volume 46 | Number 30 | 14 August 2010 | Pages 5381–5600
ISSN 1359-7345
COMMUNICATION
David O’Hagan et al.
FEATURE ARTICLE
Fluorosugars: synthesis of
hexose analogues of D-glucose
and D-altrose
Dawei Ma et al.
Total synthesis of antimicrobial and
antitumor cyclic depsipeptides
1359-7345(2010)46:30;1-J

COMMUNICATION
www.rsc.org/chemcomm | ChemComm
View Article Online
Fluorosugars: synthesis of the 2,3,4-trideoxy-2,3,4-trifluoro hexose
analogues of D-glucose and D-altrose and assessment of their
erythrocyte transmembrane transportwz
Stefano Bresciani, Tomas Lebl, Alexandra M. Z. Slawin and David O’Hagan*
Received 26th April 2010, Accepted 24th May 2010
First published as an Advance Article on the web 7th June 2010
DOI: 10.1039/c0cc01128b
The 2,3,4-trideoxy-2,3,4-trifluoro hexose analogues of D-glucose
and D-altrose were prepared and characterised and these novel
sugar analogues were explored by 2D-19F-EXSY NMR for their
potential to cross erythrocyte (red blood cell) membranes, by
comparison with the well known capacity of erythrocytes to
transport D-glucose.
Communication we report the synthesis of hexoses where three
adjacent secondary hydroxyls have been replaced by fluorine,
with a specific stereochemistry. In this series, Sarda et al.11
reported the synthesis of diacetylated 2,3,4-trideoxy-2,3,4-tri-
fluoro analogues of D-glucose 7 and D-galactose 8. These
acetylated compounds were never hydrolysed to explore the
physical or biological properties of the free carbohydrates.
We now report the synthesis and characterisation of the 2,3,4-
Selectively fluorinated carbohydrates have contributed signifi-
cantly to understanding the role of carbohydrate metabolism
and in exploring the mechanism of carbohydrate processing
enzymes.1–3 Typically a single OH group is replaced by
fluorine in a carbohydrate to remove the hydrogen bonding
donor capacity at that site, but to retain the electronegativity,
such that the importance of individual hydrogen bonds for the
kinetics of binding, can be assessed in carbohydrate–protein
interactions. 2-Fluorocarbohydrates have also proven to be
important tools for exploring the mechanism of glycosyl transfer
enzymes where the fluorine suppresses the formation of
oxocarbenium ion intermediates,4 and [18F]-2-deoxy-2-fluoro-
D-glucose is an important compound clinically, for diagnostic
purposes in positron emission tomography (PET).5 Deoxy-
fluoro carbohydrates with more than one fluorine have had
significantly less currency. There are very few examples in the
literature where hexoses with two fluorines replacing secondary
alcohols e.g. 1 and 2 have been reported.6,7 There has been
some recent interest in much more highly fluorinated sugar
analogues. DiMagno et al. have prepared hexafluorohexose
analogues exemplified by 3.8 The high amphiphilic nature of
such a carbohydrate gives it unusual physical properties and 3
was shown to be transported across erythrocyte cells an order
of magnitude faster than D-glucose, presumably as it had a
higher affinity to the relevant transporter protein. Linclau
et al.9 have recently developed robust synthesis methodology
to prepare highly amphiphilic tetrafluorocarbohydrate analogues
exemplified by 4 and 6 and Gouverneur et al.10 have reported
the synthesis of tetrafluoro ribose nucleosides such as 5. In this
trideoxy-2,3,4-trifluoro hexose analogues of D-glucose
9
and D-altrose 10 as the first examples of such carbohydrate
analogues.
The trifluorohexose analogues 9 and 10 were accessed from
the key intermediate (R)-11, which we have recently reported
in three steps from 2,3-O-(R)-isopropylidene-D-glyceraldehyde.12
Epoxidation of (R)-11 generated a mixture of the two diastereo-
isomeric epoxides 12a and 12b which was treated directly with
Et3Nꢀ3HF to give a product containing ring-opened isomers as
shown in Scheme 1. Diastereoisomers 13 and 14 were separated
by chromatography and were then used individually for their
conversion to the D-glucose 9 and D-altrose 10 trifluorohexose
analogues respectively.
Treatment of 13 with DeoxofluorTM generated the trifluoro-
acetal 15. Selective deprotection of the benzyl ether was
accomplished by the method of Adinolfi et al.13 to give 16 and
then Dess–Martin oxidation gave the unstable a-fluoroaldehyde
17 which was immediately subjected to acetonide cleavage and
intramolecular cyclisation, using SnCl2 in CH2Cl214 to generate
the D-glucose analogue 9. The free sugar 9 is a white crystalline
solid and a suitable crystal was subjected to X-ray structure
analysis to confirm the relative and absolute stereochemistry
which was consistent with configurational inversions for each
C–F bond forming reaction. The structure of the resultant
b-anomer of 9 is shown in Fig. 1a. In a similar manner
difluoroalcohol 14 was treated with DeoxofluorTM, debenzylated,
oxidised and cyclised as illustrated in Scheme 1 to give the free
D-altrose analogue 10. Likewise a suitable crystal of hexose
10 was subjected to X-ray structure analysis and the resultant
University of St Andrews, School of Chemistry and Centre for
Biomolecular Sciences, North Haugh, St Andrews, Fife, KY16 9ST,
UK. E-mail: do1@st-andrews.ac.uk; Fax: +44 (0)1334 463800;
Tel: +44 (0)1334 467171
w This publication is part of the web themed issue on fluorine
chemistry
z Electronic supplementary information (ESI) available: Experimental
procedures and compound characterisation data, X-ray structural
data and 19F-EXSY transmembrane studies data. CCDC 775192
and 775193. For ESI and crystallographic data in CIF or other
electronic format see DOI: 10.1039/c0cc01128b
ꢁc
This journal is The Royal Society of Chemistry 2010
5434 | Chem. Commun., 2010, 46, 5434–5436

View Article Online
Fig. 1 X-Ray crystal structure and molecular packing of (a) the
D-glucose analogue 9 (b-anomer) and (b) the D-altrose analogue 10
(b-anomer). In both cases the crystal packing images highlight a
partitioning of the fluorines (green) from the hydrogen bonding
interactions between the oxygens/hydroxyls (red) in the solid state.
similar rates to glucose across erythrocyte membranes, but
that 4- and 6-deoxy-fluoro analogues transport less well. Also
the a-anomer of glucose is more rapidly transported than the
b-anomer, and in general this follows for structural analogues.
The highly fluorinated hexose analogue 3 has a transport
efficiency one order of magnitude faster than 3-deoxy-3-
fluoro-D-glucose, and also the a-anomer showing the better
performance.8 The amphiphilic nature of this molecule with
both a polar and a fluorous hemisphere may account for this
dramatically improved performance. With this background we
have explored the transport ability of the trifluoro D-glucose
analogue 9 across erythrocyte cell membranes, where the three
secondary hydroxyl groups have been replaced in a stereo-
conservative manner with respect to D-glucose.
Scheme 1 Synthesis of D-glucose and D-altrose analogues 9 and 10. i.
mCPBA, CH2Cl2, RT, 5 d; ii. Et3Nꢀ3HF, CH2Cl2, MW, up to 140 1C,
21 h; iii. DeoxofluorTM, CH2Cl2, reflux, 14 h; iv. NaBrO3/Na2S2O4,
EtOAc, H2O, RT, 1 h; v. Dess–Martin periodinane, RT, CH2Cl2, 40
min; vi. SnCl2, CH2Cl2, RT, 1 h.
b-anomer is shown in Fig. 1b. In both cases the crystal packing
reveals a special partitioning of the oxygens/hydroxyls from
the fluoromethylene residues, an intermolecular pattern
presumably dictated by the stronger hydrogen bonding inter-
actions between the oxygen atoms and hydroxyl groups.
The anomeric a : b ratios of both trifluorohexoses 9 and 10
were analysed by 19F{1H}-NMR spectroscopy in a range of
For comparison we have also explored the D-altrose analogue
10 which is no longer stereoconservative relative to D-glucose.
19F-NMR and 19F-2D-EXSY NMR experiments were used to
study the efflux rate in each case from human erythrocytes.
Accordingly cells were suspended in a D2O solution of the
D-hexose analogues in the NMR tube. The resultant 19F-NMR
spectra for each analogue could readily identify signals for the
intra- and extra-cellular anomers. A typical 2D 19F EXSY-NMR
spectrum for 9 is shown in Fig. 2.
The 2D-19F EXSY exchange experiment measures retained
nuclear (19F) polarisation from the intra-cellular to the extra-
cellular molecular population (efflux rate). The intensity of the
cross peaks in this experiment is proportional to the efflux rate
constant (kef). Table 1 reports the resultant rate data for the
D-hexose analogues 9 and 10 by comparison with 3-deoxy-3-
fluoro-D-glucose. This latter analogue is known to have similar
transport ability to D-glucose itself and is a useful reference.18
Thus we have re-examined 3-deoxy-3-fluoro-D-glucose and
find good agreement with the literature in our system.
1
solvents. The individual anomers were assigned by H-NMR,
nOe’s (see ESIz). In chloroform the D-glucose analogue 9
favours the a-anomer (B5 : 1), whereas the D-altrose analogue
10 has an approximate 1 : 1 anomeric ratio. It is interesting to
note that the minor b-anomer of D-glucose 9 in solution was
the anomer that most readily crystallised however the anomeric
ratio of the D-glucose analogue 9 tends towards 1 : 1 in
increasing polar solvents. Erythrocytes (red blood cells) are
recognised to have a transport capability for D-glucose across
their cell membranes.15 The ability of the glucose transporter
(Glut1)15b to recognise and transport D-glucose analogues can
be used as a measure of such analogues to mimic D-glucose.
Selectively fluorinated sugars in particular have been explored
in erythrocyte trans-membrane studies by exploring the intra-
and extra-cellular levels of the sugars using 19F-NMR.16,17
Previous investigations have shown that 2-deoxy-2-fluoro-
and 3-deoxy-3-fluoro-D-glucose analogues are transported at
ꢁc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 5434–5436 | 5435

View Article Online
Table 1 Efflux rate constants for 9 and 10 across erythrocyte cell
membranes measured by 2D 19F EXSY spectra. The values reported
are averages from different 19F resonances and mixing times. Errors
are standard deviations
kef/sꢂ1
a-Anomer
b-Anomer
3-F-D-glucosea
3-F-D-glucoseb
9 (D-glucose)
10 (D-altrose)
1.35 ꢃ 0.32
1.38 ꢃ 0.02
0.97 ꢃ 0.21
0.33 ꢃ 0.07
1.04 ꢃ 0.23
1.01 ꢃ 0.08
0.22 ꢃ 0.07
0.40 ꢃ 0.05
a
b
Data from ref. 18. Data reproduced in this work.
with the C–F bonds, presumably based to some extent on
shape and dipolar interactions rather than hydrogen bonding.
For the D-glucose analogue 9 the a- and b-anomers are clearly
distinguished by the transmembrane protein in favour of the
a-anomer, similar to D-glucose itself.
We wish to thank EPSRC for financial support, DOH
acknowledges an ERC Advanced Grant and SB thanks the
University of St Andrews for a Scholarship.
Notes and references
1 (a) M. Hein and R. Miethchen, Adv. Org. Synth., 2006, 2, 381–429;
(b) K. Dax, M. Albert, J. Ortner and B. J. Paul, Carbohydr. Res.,
2000, 327, 47–86; (c) I. P. Street, C. R. Armstrong and
S. G. Withers, Biochemistry, 1986, 25, 6021–6027.
2 L. L. Lairson, B. Henrissat, G. J. Davies and S. G. Withers, Annu.
Rev. Biochem., 2008, 77, 521–555.
3 S. A. Allman, H. H. Jensen, B. Vijayakrishnan, J. A. Garnett,
E. Leon, Y. Liu, D. C. Anthony, N. R. Sibson, T. Feizi,
S. Matthews and B. G. Davis, ChemBioChem, 2009, 10, 2522–2529.
4 D. Vocadlo, G. J. Davies, R. Laine and S. G. Withers, Nature,
2001, 412, 835–838.
5 L. Cai, S. Lu and V. W. Pike, Eur. J. Org. Chem., 2008, 2853–2873.
6 B. P. Rempel and S. G. Withers, Aust. J. Chem., 2009, 62, 590–599.
7 (a) J. Pakak, J. Podesva, Z. Tocık and M. Cerny´ , Collect. Czech.
´ ´
Chem. Commun., 1972, 37, 2589–2599; (b) S. G. Withers,
M. D. Percival and I. P. Street, Carbohydr. Res., 1989, 187, 43–66.
8 J. C. Biffinger, H. W. Kim and S. G. DiMagno, ChemBioChem,
2004, 5, 622–627.
9 B. Linclau, A. J. Boydell, R. S. Timofte, K. J. Brown, V. Vinader
and A. C. Weymouth-Wilson, Org. Biomol. Chem., 2009, 7,
803–814.
10 L. Bonnac, S. E. Lee, G. T. Giuffredi, L. M. Elphick,
A. A. Anderson, E. S. Child, D. J. Mann and V. Gouverneur,
Org. Biomol. Chem., 2010, 8, 1445–1454.
11 P. Sarda, F. C. Escribano, R. J. Alves, A. Olesker and G. Lukacs,
J. Carbohydr. Chem., 1989, 8, 115–123.
12 S. Bresciani, A. M. Z. Slawin and D. O’Hagan, J. Fluorine Chem.,
2009, 130, 537–543.
13 M. Adinolfi, G. Barone, L. Guariniello and A. Iadonisi,
Tetrahedron Lett., 1999, 40, 8439–8441.
14 X.-H. Xu, Z.-W. You, X. Zhang and F.-L. Qing, J. Fluorine
Chem., 2007, 128, 535–539.
15 (a) G. J. Riley and N. F. Taylor, Biochem. J., 1973, 135, 773–777;
(b) A. Salas-Burgos, P. Iserovich, F. Zuniga, J. C. Vera and
J. Fischbarg, Biophys. J., 2004, 87, 2990–2999.
16 J. R. Potts, A. M. Hounslow and P. W. Kuchel, Biochem. J., 1990,
266, 925–928.
17 R. E. London and S. A. Gabel, Biophys. J., 1995, 69, 1814–1818.
18 T. M. O’Connell, S. A. Gabel and R. E. London, Biochemistry,
1994, 33, 10985–10992.
Fig. 2 (a) Control, 19F{1H}NMR spectrum (470.4 MHz) of trifluoro-
D-glucose analogue 9 in buffer (123 mM NaCl, 15 mM Tris/HEPES,
5 mM ascorbate) recorded at 310 K shows six resonances corresponding
to the a- and b-anomers. (b) 2D 19F EXSY-NMR spectrum (mixing
time 500 ms) of trifluoro-D-glucose analogue 9 in the presence of
erythrocytes suspended in the buffer at 310 K now shows twelve
resonances corresponding to the intra- and extra-cellular populations
of the a- and b-anomers. The projection traces show the corresponding
19F{1H}-NMR spectrum. All six resonances of the trifluoro-D-glucose
anomers of 9 are accompanied by broad downfield shifted peaks which
can be assigned to intracellular trifluoro-D-glucose 9 resonances. The
cross-peaks indicate exchange between intra- and extra-cellular pools
where the cross-peak intensities are proportional to the exchange rates.
Comparison of cross-peaks associated with F2a (ꢂ199.3 ppm) and
F3b (ꢂ195.7 ppm) implies a difference of exchange rate between the
a- and b-anomers. There are no cross-peaks corresponding to a/b
anomer interconversion, suggesting that mutarotation is negligible
w.r.t. the mixing time.
For the D-hexose analogues, both are transported less well,
however the D-glucose analogue 9 performs better (B70%
efficiency) and there is a clear preference for the a-anomer,
with a greater partitioning between a- and b- than any of the
other sugars examined, suggesting a level of stereochemical
recognition by the Glut1 protein with a similar sense to that of
D-glucose. Indeed for the D-altrose analogue 10, then there is
no clear preference and indeed the b-anomer appears to be
transported a little faster.
In summary we have presented the first synthesis and
structural analyses of trideoxy-trifluoro D-hexose analogues.
Transmembrane studies suggest that the Glut1 transmembrane
protein can distinguish the D-glucose from the D-altrose
analogue, and thus recognise the stereogenicity associated
ꢁc
This journal is The Royal Society of Chemistry 2010
5436 | Chem. Commun., 2010, 46, 5434–5436