Carbohydrate sensing using a fluorescent molecular tweezer

Article abstract of DOI:10.1039/b909230g

A fluorescent molecular tweezer for carbohydrates has been prepared which utilises two boronic acid receptor groups. The Royal Society of Chemistry 2009.

Full text of DOI:10.1039/b909230g

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Carbohydrate sensing using a fluorescent molecular tweezerw
Marcus D. Phillips,a Thomas M. Fyles,b Nicholas P. Barwell*a and Tony D. James*a
Received (in Cambridge, UK) 11th May 2009, Accepted 13th August 2009
First published as an Advance Article on the web 11th September 2009
DOI: 10.1039/b909230g
A fluorescent molecular tweezer for carbohydrates has been
prepared which utilises two boronic acid receptor groups.
by further augmenting the hydrophobicity with the binding
pocket. With the results of receptor 1 and considering the
potential advantages of allowing a degree of conformational
flexibility to be designed into the recognition site, the structure
of the di-boronic molecular tweezer 2 was postulated Fig. 1.
It can be seen from the structure of molecular tweezer 2 that
the o-phenylboronic acid units, tertiary amines, and pyrenyl
fluorophores are positioned so as to permit fluorescence to be
controlled via PET. The dual boronic acid units permit
saccharide selectivity via two point binding with two pyrene
units augmenting the hydrophobicity within the binding
pocket. The two hexamethylene linkers will introduce a degree
of conformational flexibility into the system and allow the
binding cavity to accommodate different saccharides.
Carbohydrates, or saccharides, are the most abundant of the
four major classes of biomolecules, which also include proteins,
lipids, and nucleic acids. Carbohydrates were originally
thought of as food and building materials, however the
realisation that oligosaccharides play an important role in
biological regulation has attracted a great deal of interest.1–3
Receptors for carbohydrates have been designed based on
hydrogen bonding interactions4,5 and covalent ester formation
with boronic acids.6–8 Boronic acids can bind saccharides
via covalent interactions in basic aqueous media through the
formation of cis-1,2- or 1,3-diols which form five- or six-
membered rings, respectively.6–8
Synthesis of 2 was achieved according to Scheme 1 from
readily available starting materials. The selective protection of
diamine 3 gave the mono Boc-protected diamine 4 which was
then reacted with isophthaloyl dichloride17 to yield 5.
Cleavage of the Boc protecting groups was followed by
reductive amination with pyrene-1-carboxaldehyde and a
acidic work-up gave ammonium salt 6. The boronic functionality
was introduced by reacting the neopentyl protected 2-(bromo-
methyl)phenylboronic acid with 6 to give molecular tweezer 2.
Photoinduced electron transfer (PET) saccharide sensors
based on boronic acids were first synthesised over 15 years
ago9–11 which exploit the interactions between o-methyl-
phenylboronic acids (Lewis acids) and proximal tertiary
amines (Lewis bases).
The boronic acid–amine (B–N) interaction provides two
distinct advantages for saccharide binding. The first advantage
of the B–N interaction is that it lowers the pKa of the
boronic acid and thus allows binding to occur at neutral, i.e.
physiological pH.12 The second is related to the contraction of
the O–B–O bond angle upon association with a saccharide
resulting in an increase of acidity at the boron centre. This
increase in acidity of the boron centre strengthens the B–N
interaction and thus disrupts PET. This has the effect of
modulating fluorescence intensity via the amine group and
introduces a digital ‘‘off–on’’ response from the fluorophore,
indicative of the boronic acid being unbound or bound.11,13
Previous work within the group14–16 has shown that the
‘‘inherent stability order’’ for simple boronic acids of
D-fructose 4 D-galactose 4 D-mannose 4 D-glucose can be
perturbed by the use of di-boronic acids. The most potent of
these di-boronic acid receptors was receptor 1, it had an
association constant of 960 MÀ1 for D-glucose, 760 MÀ1
for D-fructose, 660 MÀ1 for D-galactose, and 70 MÀ1 for
D-mannose.15 Constructing a pair of molecular tweezers tipped
with two large aromatic fluorophore units could be of potential
benefit in enhancing the selectivity between specific saccharides
a Department of Chemistry, University of Bath, Claverton Down,
Bath, UK BA2 7AY. E-mail: T.D.James@bath.ac.uk;
Fax: +44 (0)1225 386231; Tel: +44 (0)1225 386444
b Department of Chemistry, University of Victoria, Victoria,
BC V8W 3V6, Canada
Fig. 1 Structure of a previously synthesised di-boronic acid receptor
1 and di-boronic molecular tweezer 2.
w Electronic supplementary information (ESI) available: Experimental
details and fluorescence titrations. See DOI: 10.1039/b909230g
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 6557–6559 | 6557

View Article Online
Scheme 1 Synthesis of 2. Reagents (yields): (a) tert-butyl phenyl
carbonate, EtOH, 78 1C, 45%; (b) isophthaloyl dichloride, triethylamine,
DCM, 96%; (c) HCl, MeOH, EtOAc, 100%; (d) (i) pyrene-1-
carboxaldehyde, caesium carbonate, THF–methanol, reflux; (ii) NaBH4;
(iii) HCl, MeOH, DCM, 68%; (e) 2-(2-(bromomethyl)phenyl)-5,5-
dimethyl-1,3,2-dioxaborinane, potassium carbonate, acetonitrile,
reflux, 14%.
Fig. 2 Fluorescence emission spectra of 2 (0.1 mM) with increasing
amounts of (A) D-glucose (0 to 0.1 M) or (B) D-fructose (0 to 0.1 M) in
an aqueous methanolic buffer [52.1 wt% methanol, KCl (10.0 mM),
KH2PO4 (2.75 mM) and Na2HPO4 (2.75 mM)] at a pH of 8.21.
Relative fluorescence intensity versus carbohydrate concentration profile
of 2 (0.1 mM, lex = 342 nm, (C) lem = 377 nm, (D) lem = 470 nm)
displaying PET with (orange triangle) D-glucose, (red circle) D-fructose,
(blue diamond) D-galactose and (green square) D-mannose.
The binding behaviour of boronic acids and its fluorescence
are pH dependent.9,18 The change in the fluorescence of
tweezers 2, with and without D-fructose, versus pH was
investigated to determine the pH with the maximum fluorescence
response.19 As a result fluorescence titrations of 2 (0.1 mM)
were carried out in an aqueous methanolic buffer at pH 8.21
with various saccharides. Fig. 2 shows the fluorescence spectra
of 2 in the presence of increasing concentrations of either
D-glucose or D-fructose. Spectra obtained by adding increasing
amounts of D-galactose and D-mannose show the same
spectroscopic features as those shown when adding D-glucose,
although the initial quenching of the excimer band at 470 nm
is partially reversed at high galactose concentrations.20
The fluorescence intensity increases with increasing concen-
tration of all four carbohydrates at 377 nm resulting from PET
as found with other aminomethyl boronate complexes. The
fluorescence intensity changes at 470 nm differ among the four
carbohydrates. With increasing concentrations of D-glucose
and D-mannose, the 470 nm band decreases with increasing
carbohydrate concentration. As shown in Fig. 2B, the intensity
of the 470 nm band is apparently invariant with added
D-fructose. Finally, D-galactose shows an initial quenching of
the 470 nm band at low concentrations that is reversed as the
concentration increases (Fig. 2D).
We interpret these diverse observations as follows. All
carbohydrates form a 1 : 1 complex Fig. 3. In the case of
D-glucose, D-galactose, and D-mannose (Fig. 4) this complex
has the cyclic structure envisaged in Fig. 1. This would explain
the quenching of the excimer emission since the binding of the
carbohydrate forces a separation between the pyrene units. We
interpret the more complex behaviour of D-galactose as
indicating the subsequent formation of a 2 : 1 complex.
Fig. 3 Schematic representation showing the two possible ways of
binding for the molecular tweezers with sugars.
ꢀc
This journal is The Royal Society of Chemistry 2009
6558 | Chem. Commun., 2009, 6557–6559

View Article Online
A sequential 1 : 1–2 : 1 binding model is required to fit the
D-galactose data. The D-fructose data will also accommodate a
2 : 1 complex, but the fit statistics are insignificantly different
from the simpler 1 : 1 binding isotherm. The magnitude of the
binding constants reveal why the D-galactose system uniquely
shows the stepwise behaviour; the 1 : 1 complex is moderately
strong allowing its formation and destruction to be observed
within the concentration range of the experiments. The other
carbohydrates either form complexes that are too weak or too
strong to be manipulated by experimental concentrations.
In conclusion we have developed a molecular tweezer 2
that selectively opens for carbohydrates with the stability
order of D-glucose c D-galactose 4 D-mannose. Although 2
forms a complex with D-fructose with a higher stability
constant than D-glucose, it fails to open the tweezer 2.
We are grateful to the University of Bath and the EPSRC
for support.
Fig. 4 Saccharides used in the study, D-glucose 7, D-fructose 8,
D-galactose 9, and D-mannose 10.
Table 1 Observed 1 : 1 stability constants (Kobs), determination
coefficient (r2), and fluorescence enhancement for 2 (0.1 mM) with
D-glucose, D-fructose, D-galactose, and D-mannose at pH 8.21 in an
aqueous methanolic buffer [52.1 wt% methanol, KCl (10.0 mM),
KH2PO4 (2.75 mM) and Na2HPO4 (2.75 mM)]a
Kobs/MÀ1
r2
I/I0 at 377 nm I/I0 at 460 nm
Notes and references
D-Glucose 7
2660 Æ 200 0.999 6.5
2850 Æ 140 0.999 5.0
440 Æ 10 0.999 4.9
200 Æ 15 0.998 3.9
0.43
0.96
0.66
0.54
1 C. R. Bertozzi and L. L. Kiessling, Science, 2001, 291, 2357–2364.
2 R. A. Dwek and T. D. Butters, Chem. Rev., 2002, 102, 283–284.
3 H. J. Gabius, H. C. Siebert, S. Andre, J. Jimenez-Barbero and
H. Rudiger, ChemBioChem, 2004, 5, 741–764.
4 M. Mazik, Chem. Soc. Rev., 2009, 38, 935–956.
5 A. P. Davis and R. S. Wareham, Angew. Chem., Int. Ed., 1999, 38,
2978–2996.
6 T. D. James, M. D. Phillips and S. Shinkai, Boronic Acids in
Saccharide Recognition, RSC Publishing, Cambridge, UK, 2006.
7 H. Cao and M. D. Heagy, J. Fluoresc., 2004, 14, 569–584.
8 H. Fang, G. Kaur and B. Wang, J. Fluoresc., 2004, 14, 481–489.
9 T. D. James, K. R. A. S. Sandanayake, R. Iguchi and S. Shinkai,
J. Am. Chem. Soc., 1995, 117, 8982–8987.
D-Fructose 8
D-Galactoseb 9
D-Mannose 10
a
The K values were analysed in Scientist using non-linear curve fitting.
The errors reported are the standard errors obtained from the best fit,
b
relative fluorescent enhancements (I/I0) are also reported. Stepwise
formation of a 2 : 1 complex is required for the fit: K21(stepwise from
1 : 1 complex) = 50 Æ 20; I/I0 at 377 nm = 5.2; I/I0 at 470 nm = 0.91.
In such a structure the pyrene excimer can form since the
carbohydrate is no longer forcing the rings apart.
10 T. D. James, K. Sandanayake and S. Shinkai, J. Chem. Soc., Chem.
Commun., 1994, 477–478.
In the case of D-fructose the 1 : 1 complex must form
without disruption of the pyrene excimer, i.e. without
formation of the chelated structure indicated in Fig. 1. The
‘‘inherent stability order’’ favoring D-fructose is based in part
on its ability to coordinate via three hydroxyl groups. This
does not leave any vicinal diol units for a second boronic acid
binding location. It is also possible that 2 : 1 complexes form.
This is known in other systems; carbohydrates such as
D-glucose easily form 1 : 1 cyclic complexes with di-boronic
acids while D-fructose resists forming cyclic complexes and
tends to form 2 : 1 acyclic complexes instead.11,21,22
11 T. D. James, Top. Curr. Chem., 2007, 277, 107–152.
12 G. Wulff, Pure Appl. Chem., 1982, 54, 2093–2102.
13 T. D. James, in Boronic Acids in Organic Synthesis and
Chemical Biology, ed. D. G. Hall, Wiley-VCH, Weinheim, 2005,
pp. 441-480.
14 S. Arimori, M. L. Bell, C. S. Oh and T. D. James, Org. Lett., 2002,
4, 4249–4251.
15 S. Arimori, M. L. Bell, C. S. Oh, K. A. Frimat and T. D. James,
Chem. Commun., 2001, 1836–1837.
16 S. Arimori, M. L. Bell, C. S. Oh, K. A. Frimat and T. D. James,
J. Chem. Soc., Perkin Trans. 1, 2002, 803–808.
17 F. J. Carver, C. A. Hunter, P. S. Jones, D. J. Livingstone,
J. F. McCabe, E. M. Seward, P. Tiger and S. E. Spey, Chem.–Eur.
J., 2001, 7, 4854–4862.
The stability constants (Kobs) of tweezer 2 were calculated
by simultaneous fitting of the emission spectra at two wave-
lengths versus carbohydrate concentration and are shown
in Table 1. As noted above, all carbohydrates form 1 : 1
complexes and, with the exception of D-galactose, the data
could be adequately fit assuming only 1 : 1 complex formation.
18 T. D. James, K. Sandanayake and S. Shinkai, Angew. Chem., Int.
Ed. Engl., 1996, 35, 1911–1922.
19 See Fig. S2 in the ESIw.
20 See Fig. S3 and S4 in the ESIw.
21 J. H. Hartley, T. D. James and C. J. Ward, J. Chem. Soc., Perkin
Trans. 1, 2000, 3155–3184.
22 T. D. James and S. Shinkai, Top. Curr. Chem., 2002, 218, 159–200.
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 6557–6559 | 6559