Ortho-Substituted fluorescent aryl monoboronic acid displays physiological binding of d-glucose

Article abstract of DOI:10.1016/j.tetlet.2013.01.101

An ortho-fluorinated phenylboronic acid incorporated into the BODIPY-fluorophore, sensor 1, exhibits a significant response in emission intensity upon binding of d-glucose and d-fructose. This sensor displays a desired binding strength toward d-glucose at the physiological level, that is K_d values between 10 and 20 mM. This binding strength of sensor 1 has been shown to be independent of the employed buffer, that is, saline buffer, 50 mM phosphate buffer, and 50 mM phosphate buffer in 52.3 w/w% MeOH. The d-glucose binding strength and the fluorescence response of a BODIPY-based ortho-methylated phenylboronic acid, sensor 2, show a significantly decreased d-fructose binding affinity. This sensor also exhibits buffer-dependent binding strength and fluorescence response upon binding of d-glucose and d-fructose.

Full text of DOI:10.1016/j.tetlet.2013.01.101

Tetrahedron Letters xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
ortho-Substituted fluorescent aryl monoboronic acid displays physiological
binding of D-glucose
J. S. Hansen a, , M. Ficker a, J. F. Petersen a, J. B. Christensen a, T. Hoeg-Jensen b
⇑
a Department of Chemistry, Universitetsparken 5, DK-2100 Copenhagen, Denmark
b Novo Nordisk Park D6.1.142, DK-2760 Maaloev, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
An ortho-fluorinated phenylboronic acid incorporated into the BODIPY-fluorophore, sensor 1, exhibits a
Received 12 October 2012
Revised 10 January 2013
Accepted 24 January 2013
Available online xxxx
significant response in emission intensity upon binding of
a desired binding strength toward -glucose at the physiological level, that is Kd values between 10 and
20 mM. This binding strength of sensor 1 has been shown to be independent of the employed buffer, that
is, saline buffer, 50 mM phosphate buffer, and 50 mM phosphate buffer in 52.3 w/w% MeOH. The -glu-
cose binding strength and the fluorescence response of a BODIPY-based ortho-methylated phenylboronic
acid, sensor 2, show a significantly decreased -fructose binding affinity. This sensor also exhibits buffer-
dependent binding strength and fluorescence response upon binding of -glucose and -fructose.
D-glucose and D-fructose. This sensor displays
D
D
Keywords:
D
ortho-Substituted aryl monoboronic acid
Fluorescent reporter
Binding selectivity
D
D
Ó 2013 Elsevier Ltd. All rights reserved.
Buffer effects
Structural constraints
Arylboronic acids are small and flexible molecules1–6 in com-
parison to lectins7 and artificial macrocycles designed to bind car-
bohydrates.8–11
titrations, using a colorimetric competitive assay, that is, Alizarin
Red Sodium (ARS), and further rationalized by calculations at the
B3LYP/6-31G(d) level with Gaussian.22
The reactions of arylboronic acids with cis-1,2-diols or 1,3-diols
in aqueous media generally afford cyclic esters rapidly and revers-
ibly. This transformation is likely to alter the electronic properties
of an arylboronic acid based sensor, since formation of the corre-
sponding arylboronate is favored in the cyclic ester. An arylboronic
acid based fluorescent sensor can potentially be used as a semi-
These findings have been employed in the engineering of ortho-
substituted aryl monoboronic acids incorporated into the BODIPY
fluorophore.
Sensor 1, which has been synthesized similarly to the procedure
described by DiCesare and Lakowicz,23 contains an o-fluorine sub-
stituent. It exhibits excellent
buffers, with displacement constants, Kd, around 15–20 mM.
Fructose binding of sensor 1 is not reduced significantly as ex-
pected, since the found -fructose displacement constants are
around 0.5–1.0 mM.24 The tight
-fructose binding may be a conse-
quence of a favorable hydrogen bond interaction between a hydro-
xyl group in -fructose and the o-fluorine substituent in sensor 1.
Also the lower pKa of sensor 1 compared to sensor 2 may contrib-
ute positively to the tighter -fructose binding of sensor 1.
D-glucose binding in three types of
invasive or non-invasive reporter for monitoring of
diabetic patients.12–15 The displacement constant, Kd, must have
a value around 10–20 mM for binding of -glucose in human blood,
D
-glucose in
D-
D
D
since blood glucose varies between 2 and 30 mM in diabetic pa-
tients.16 The maximum sensitivity is achieved, when Kd is in the
middle of the binding curve.
D
D
D
-Glucose is the dominant monosaccharide present in human
blood, that is, [
D
-glc] ꢀ5 mM,17 in comparison to
D
-fructose. [
D
-
D
frc] <0.5 mM, even after a fructose-rich meal.18,19 Thus, selective
Sensor 2 was synthesized according to a recently described pro-
recognition of
accurate glucose readouts.
binding toward aryl monoboronates due to the tridentate binding
mode,20 whereas -glucose exhibits bidentate binding.21 This
stronger -fructose binding leads to errors in -glucose readouts.
A substituent placed ortho to the boronic acid may likely decrease
D
-glucose over other saccharides is important for
cedure.25 This sensor, which contains an o-methyl substituent,
D
-Fructose generally exhibits a stronger
shows a significantly reduced
D-fructose binding affinity, with Kd
around 10 mM. This is a onefold reduction in
D
-fructose binding
D
strength compared to that normally expected for binding aryl
monoboronates.24 An explanation of this phenomenon might be
that sensor 2 is not capable of hydrogen bond formation as sug-
gested for sensor 1, and that the pKa is significantly lower for sen-
sor 1 compared to sensor 2.
D
D
the binding of D-fructose. This has recently been shown by UV/Vis-
Sensors 1 and 2 are shown in Figure 1 (top), along with the
⇑
Corresponding author. Tel.: +45 61337866.
E-mail address: jonhansen@chem.ku.dk (J.S. Hansen).
unsubstituted sensor 3, which has recently been evaluated.25
0040-4039/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2013.01.101
Please cite this article in press as: Hansen, J. S.; et al. Tetrahedron Lett. (2013), http://dx.doi.org/10.1016/j.tetlet.2013.01.101

2
J. S. Hansen et al. / Tetrahedron Letters xxx (2013) xxx–xxx
Increased [D-glc]
1.0
If
0.5
0.0
480
500
520
540
560
580
wavelength (nm)
Figure 2. Increase in the emission intensity of sensor 1, upon augmenting [D-glc]
from 0 to 513 mM in a saline buffer at pH 7.4.
Saline buffer, pH 7.4
1.05
D-frc
D-glc
1.00
Figure 1. Top: ortho-substituted aryl monoboronic acids, 1 and 2, incorporated into
a
BODIPY fluorophore, along with the unsubstituted analogue, 3.25 Bottom:
0.95
0.90
0.85
proposed reductive quenching for the diminished tridentate D-fructose binding of
sensor 2, caused by the presence of an ortho-positioned methyl substituent.26
Research by the groups of Lakowicz and Shinkai revealed a sim-
ilar binding efficiency for arylboronic acids and their correspond-
ing ester analogues.27,28 The similar binding efficiency is due to
rapid ester exchange in boronic acid esters. Shinkai and Lakowicz
have therefore used their boronic acids directly as the correspond-
ing protected esters.
Based on the above, the pinacol ester analogues of boronic acids
have been used throughout the binding studies herein reported,
since these are easier to synthesize and purify. A control binding
experiment with the neopentyl ester analogue of sensor 1 has been
performed. The binding strength of this sensor matched the bind-
ing strength of sensor 1. The neopentyl ester analogue of 3 has pre-
viously been tested by Lakowicz’s group.23
-2
-1
0
1
2
3
log[c]
Figure 3. Curve fit using GraphPad Prism 5.0. The measured emission intensity is
plotted against the logarithm of the concentration data (in mM), using a sigmoidal
dose response (variable slope). Sensor 1 is used at 0.53 lM in the phosphate
buffered saline solution at pH 7.4.
Table 1
Binding constants, Kd, in mM, determined from the change in emission intensities at
pH 7.4
Sensor/buffer
Kd
D
-frc (n)
r
Kd D-glc (n) r
The extinction coefficients of sensors 1 and 2 have been deter-
1, Saline
1.0 (2) 0.3
0.7 (2) 0.2
0.8 (2) 0.1
9.2 (2) 1.3
11.1 (2) 0.3
7.1 (2) 0.1
17 (2) 3.9
18 (2) 1.4
16 (2) 3.3
<500 (2) —
<500 (2) —
97 (2) 0.1
mined in methanol to be e =
1495 nm = 84,000 MÀ1 cmÀ1 and e2495 nm
1, 50 mM Phosphate
1, 52.3 w/w% MeOH/phos
2, Saline
2, 50 mM Phosphate
2, 52.3 w/w% MeOH/phos
80,000 MÀ1 cmÀ1, respectively. Such high extinction coefficients
are as expected for BODIPY dyes.29
The impact of the buffer on the binding properties of sensor 1 is
very similar in the three employed buffers, that is, saline buffer
(10 mM phosphate, 2.7 mM KCl, and 137 mM NaCl), 50 mM phos-
phate buffer, and 52.3 w/w% methanolic and 50 mM phosphate
buffer at physiological pH (7.4). This high assay robustness may
be useful from a biological perspective. The excitation and emis-
sion response intensities are increased significantly upon raising
the monosaccharide concentration. This phenomenon may be
interpreted as oxidative quenching upon saccharide binding, simi-
lar to the proposed binding mechanism of sensor 3.25 The plot of
The number of measurements (n) is shown in brackets, and the standard deviations
) are shown to the right.
(r
drawing effect of fluorine. The tight D-glucose binding of sensor 1
compared to sensor 3 may be due to the relatively lower pKa of 1
(7.3) compared to sensor 3 (8.3–8.8).23 The overall very similar in-
crease in emission intensity of sensor 1 on addition of the two
monosaccharides however, indicates a comparable binding mode
in the boronate ester formation.
Sensor 2 exhibits very weak binding toward
>500 mM). The tightest -glucose binding is found in the 52.3 w/
w% methanolic and 50 mM phosphate buffer with a Kd of 97 mM
being observed. The overall low -glucose binding affinity may
be explained by the very high pKa of sensor 2 (10.7). One would ex-
pect tighter -glucose binding in the 50 mM phosphate buffer
according to the pKa argument, since sensor 2 is slightly more
acidic in this buffer (pKa of 10.4). The response upon -glucose
the slight increase in emission for sensor 1 by addition of
is shown in Figure 2. The binding curves for binding -glucose and
-fructose in saline buffer at pH 7.4 are shown in Figure 3.
Displacement constants, Kd, are listed in Table 1 for sensors 1
and 2, where 1:1 -glucose and -fructose–arylmonoboronate
D-glucose
D
D-glucose (Kd
D
D
D
D
D
complexes are anticipated. The displacement constants are calcu-
lated by plotting the emission intensity increase against the loga-
rithm of the concentration, using a sigmoidal dose response
(variable slope).
D
D
Sensor 1 does not exhibit the same increase in emission inten-
sity as sensor 3.25 This lower emission response increase for 1 may
be a consequence of the relatively high electron affinity of the
meso-aryl substituent in sensor 1, compared to the meso-aryl sub-
stituent in sensor 3. This difference is caused by the electron- with-
binding has been shown to be buffer dependent. A slight decrease
in emission intensity is observed in the saline buffer and in the
50 mM phosphate buffer. Titration of sensor 2 with D-glucose in
a 52.3 w/w% methanolic and 50 mM phosphate buffer, on the other
hand, results in a slight increase in the emission intensity.
Please cite this article in press as: Hansen, J. S.; et al. Tetrahedron Lett. (2013), http://dx.doi.org/10.1016/j.tetlet.2013.01.101

J. S. Hansen et al. / Tetrahedron Letters xxx (2013) xxx–xxx
3
The reduced D-fructose binding affinity of sensor 2 in all three
D-frc saline, pH 7.4
buffers is expected based on previous UV/Vis-titrations and calcu-
lations at the B3LYP/6-31G(d) level with Gaussian.22 Titration of
1.0
0.5
sensor 2 with D-fructose in saline buffer and phosphate buffer at
pH 7.4, shows significantly decreased emission intensity on
increasing the sugar concentration. The decreased excitation and
emission intensities of sensor 2 are shown in Figure 4. The binding
If
isotherm is shown in Figure 5. Titration of sensor 2 with D-fructose
in 52.3 w/w% methanolic and 50 mM phosphate buffer, in contrast,
results in increased emission intensity, see Figure 6. This is in good
agreement with the observations from the titration of sensor 2
-1
0
1
2
3
4
with D-glucose. This may suggest a reductive quenching in saline
log[c]
buffer and 50 mM phosphate buffer, while oxidative quenching
might take place in 52.3 w/w% methanolic and 50 mM phosphate
Figure 5. Decreased emission intensity of sensor 2 as a result of increasing [
saline buffer at pH 7.4.
D-frc] in
buffer.26 The proposed oxidative quenching upon
D
-fructose bind-
ing is depicted in Figure 1 (bottom), while the sigmoidal curve
for -fructose binding of sensor 2 is shown in Figure 5. Additional
D
52.3 w/w% MeOH, 50mM phosphate, pH 7.4
excitation measurements mirror the found emission responses for
sensors 1 and 2, and the recorded excitation spectra of 1 and 2
match the absorption spectra.
1.05
D-frc
D-glc
1.00
The binding selectivities have been calculated for sensors 1 and
0.95
If
2 for the binding of
Kd( -frc) value for sensor 1 was found in saline buffer. The overall
lowest Kd( -glc)/Kd( -frc) ratio was, however, found for sensor 2.
D-glucose and D-fructose. The lowest Kd(D-glc)/
D
0.90
D
D
The selectivity ratios are listed in Table 2.
0.85
Determination of the pKa values for the sensors 1 and 2 has
been conducted in 50 mM phosphate buffer and in 52.3 w/w%
methanolic and 50 mM phosphate buffer. The slight blue-shifting
of the absorption and emission spectra upon boronate formation
is exploited to calculate the respective acid–base equilibria. The
titration curve for the determination of the pKa of sensor 1 in
50 mM phosphate buffer is shown in Figure 7, and the found pKa
values are listed in Table 3.
-1
0
1
2
3
4
log[c]
Figure 6. Increased emission intensity of sensor 2 as a result of increasing sugar
concentration.
Table 2
Fluorometric evaluation of sensor 1 reveals excellent binding
Kd(D-glc)/Kd(D-frc) ratios for sensors 1 and 2 in the three employed buffers
affinity toward D-glucose, with displacement constants, Kd, at the
Sensor/buffer
Kd
(D-glc)/Kd(D-frc)
physiological level. Kd is found midway between the extremes of
diabetic blood sugar value, 2–30 mM, that is, the probe should give
good sensitivity over the relevant range.16 The binding strength of
1, Saline
17
25.7
20
1, 50 mM Phosphate
1, 52.3 w/w%MeOH/phos
2, Saline
sensor 1 must still be modulated toward weaker D-fructose bind-
—
ing. This can presumably be achieved by attachment of a larger
electron-withdrawing substituent at the position ortho- to the
boronic acid unit. Attachment of a CF3-group at the ortho-position
2, 50 mM Phosphate
2, 52.3 w/w%MeOH/phos
—
13.7
may result in the desired steric effect causing decreased
D
-fructose
binding, along with a sufficiently low pKa for optimal
D-glucose
Sensor 1 in 50mM phosphate
binding. Alternatively attachment of an ortho-chlorine might also
decrease -fructose binding. However, this may increase the pKa
of the sensor, due to the decreased electron-withdrawing ability
of this halogen. In order not to weaken the -glucose binding, this
latter change of sensor 1 will demand the attachment of electron-
withdrawing substituents elsewhere in the dye to maintain a low
D
1.4
1.2
1.0
0.8
D
pKa. Reduced
D
-fructose selectivity is required in order to obtain
accurate
D-glucose readouts without any significant error.
0.6
0
Increased [D-frc]
2
4
6
8
10
12
14
Increased [D-frc]
pH
1.0
0.5
0.0
1.0
Figure 7. Titration curve for the pKa determination of sensor 1. The ratios of
emission intensities at 500 and 515 nm are plotted against pH.
Iex
0.5
If
Sensor 2 exhibits significantly decreased
also poor -glucose binding in all the three buffers. The poor
-glucose affinity of sensor 2 can be rationalized by the very
D-fructose binding, but
0.0
400
450
500
500
550
600
D
wavelength (nm)
wavelength (nm)
D
high pKa value. Decreasing the pKa of the arylboronic acid by
attachment of electron-withdrawing substituents elsewhere in
the molecule can tune the binding strength of sensor 2 toward
Figure 4. Significant decrease in the excitation and emission intensities of sensor 2
(left and right, respectively) upon addition of
buffer.
D-fructose from 0 to 2.0 M in a saline
Please cite this article in press as: Hansen, J. S.; et al. Tetrahedron Lett. (2013), http://dx.doi.org/10.1016/j.tetlet.2013.01.101

4
J. S. Hansen et al. / Tetrahedron Letters xxx (2013) xxx–xxx
3. Shinkai, S.; Takeuchi, M. Trends Anal. Chem. 1996, 15, 188–193.
Table 3
4. Draffin, S. P.; Duggan, P. J.; Duggan, S. A. M.; Norrild, J. C. Tetrahedron 2003, 59,
9075–9082.
5. Hirata, O.; Kubo, Y.; Takeuchi, M.; Shinkai, S. Tetrahedron 2004, 60, 11211–
11218.
6. Dowlut, M.; Hall, D. G. J. Am. Chem. Soc. 2006, 128, 4226–4227.
7. Goldstein, I. J.; Murphy, L. A.; Ebisu, S. Pure Appl. Chem. 1977, 49, 1095–1103.
8. Anderson, S.; Neidlein, U.; Gramlich, V.; Diederich, F. Angew. Chem., Int. Ed. Engl.
1995, 34, 1596–1600.
pKa Values determined from the change in excitation and emission intensities
Sensor/buffer
pKa (n) r
1, 50 mM Phosphate
1, 52.3 w/w% MeOH/phos
2, 50 mM Phosphate
7.3 (2) 0.1
7.9 (2) 0.1
10.4 (2) 0.1
10.7 (2) 0.5
2, 52.3 w/w% MeOH/phos
9. Barwell, N. P.; Crump, M. P.; Davis, A. P. Angew. Chem., Int. Ed. 2009, 48, 7673–
7676.
The number of measurements (n) is shown in brackets, and the standard deviations
) are shown to the right.
(r
10. Ke, C.; Destecroix, H.; Crump, M. P.; Davis, A. P. Nat. Chem. 2012, 4, 718–723.
11. Sookcharoenpinyo, B.; Klein, E.; Ferrand, Y.; Walker, D. B.; Brotherhood, P. R.;
Ke, C.; Crump, M. P.; Davis, A. P. Angew. Chem., Int. Ed. 2012, 51, 4586–4590.
12. Hiraoka, M.; Firbank, M.; Essenpreis, M.; Cope, M.; Arridge, S. R.; van der Zee,
P.; Delpy, D. T. Phys. Med. Biol. 1993, 38, 1859–1876.
13. Peyser, T.A.; Gamsey, S.; Romey, M.A.; Markle, D.R. US Patent 20100312483,
2010, Chem. Abstr. 2010, 154, 4538.
14. Hansen, J. S.; Christensen, J. B.; Petersen, J. F.; Hoeg-Jensen, T.; Norrild, J. C. Sens.
Actuators, B Chem. 2012, 161, 45–79.
15. Huang, S.; Jia, M.; Xie, Y.; Wang, J.; Xu, W.; Fang, H. Curr. Med. Chem. 2012, 19,
2621–2637.
16. Kamata, K.; Morihiro, M.; Nishimura, T.; Eiki, J.-I.; Nagata, Y. Structure 2004, 12,
429–438.
17. Heise, H. M.; Bittner, A.; Koschinsky, T.; Gries, F. A. Fresenius J. Anal. Chem. 1997,
359, 83–87.
18. Donmoyer, C. M.; Eijofor, J.; Lacy, D. B.; Chen, S.-S.; McGuinness, O. P. Am. J.
Physiol. Endocrinol. Metab. 2001, 280, E703–E711.
19. Arai, T.; Hashimoto, K.; Muzutani, H.; Kawabata, T.; Sako, T.; Washizu, T. Vet.
Res. Commun. 1999, 23, 203–209.
binding of D-glucose at a physiological level. The high assay robust-
ness of sensor 1 in the three employed buffers is promising in the
engineering of a fluorescent sensor, which can be used for semi- or
non-invasive D-glucose monitoring in human blood. The found assay
robustness is quite different compared to the obtained results for the
sensors 2 and 3.25 Here the employed buffer exhibits a much greater
impact on the absorption and emission response, as well as the car-
bohydrate binding affinities and the
Superior selectivity toward
D
-glucose selectivity.
D
-glucose can ultimately be
achieved by the use of fluorescent diboronate derivatives.30–34
However, these usually contain a highly insoluble scaffold, which
is required for appropriate spatial arrangement of the boronate
moieties. Aryl diboronates are also generally more synthetically
challenging than the engineering of aryl monoboronates.
20. Norrild, J. C.; Eggert, H. J. Chem. Soc., Perkin Trans. 2 1996, 2583–2588.
21. Bielecki, M.; Eggert, H.; Norrild, J. C. J. Chem. Soc., Perkin Trans. 2 1999, 449–455.
22. Hansen, J. S.; Christensen, J. B.; Solling, T. I.; Jakobsen, P.; Hoeg-Jensen, T.
Tetrahedron 2011, 67, 1334–1340.
23. DiCesare, N.; Lakowicz, J. R. Tetrahedron Lett. 2001, 42, 9105–9108.
24. Van Duin, M.; Peters, J. A.; Kieboom, A. P. G.; Van Bekkum, H. Tetrahedron 1984,
40, 2901–2911.
25. Hansen, J. S.; Petersen, J. F.; Hoeg-Jensen, T.; Christensen, J. B. Tetrahedron Lett.
2012, 53, 5852–5855.
26. Loudet, A.; Burgess, K. Chem. Rev. 2007, 107, 4891–4932.
27. DiCesare, N.; Lakowicz, J. R. J. Phys. Chem. A 2001, 105, 6834–6840.
28. Mizuno, T.; Fukumatsu, T.; Takeuchi, M.; Shinkai, S. J. Chem. Soc., Perkin Trans. 1
2000, 407–413.
29. Karolin, J.; Johansson, L. B. A.; Strandberg, L.; Ny, T. J. Am. Chem. Soc. 1994, 116,
7801–7806.
30. Yang, W.; He, H.; Drueckhammer, D. G. Angew. Chem., Int. Ed. 2001, 40, 1714–
1718.
31. Karnati, V. R.; Gao, X.; Gao, S.; Yang, W.; Ni, W.; Sankar, S.; Wang, B. Bioorg.
Med. Chem. Lett. 2002, 12, 3373–3377.
32. James, T. D.; Sandanayake, K. R. A. S.; Iguchi, R.; Shinkai, S. J. Am. Chem. Soc.
1995, 117, 8982–8987.
Acknowledgments
Financial support from FTP is gratefully acknowledged, along
with a fruitful collaboration with Novo Nordisk A/S and the Univer-
sity of Copenhagen, Department of Chemistry.
Supplementary data
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.tetlet.2013.01. 101.
References and notes
33. Suri, J. T.; Cordes, D. B.; Cappucio, F. E.; Wessling, R. A.; Singaram, B. Langmuir
2003, 19, 5145–5152.
1. Yang, W.; Yan, J.; Springsteen, G.; Deeter, S.; Wang, B. Bioorg. Med. Chem. Lett.
2003, 13, 1019–1022.
34. Eggert, H.; Frederiksen, J.; Morin, C.; Norrild, J. C. J. Org. Chem. 1999, 64, 3846–
3852.
2. Cao, H.; Diaz, D. I.; DiCesare, N.; Lakowicz, J. R.; Heagy, M. D. Org. Lett. 2002, 4,
1503–1505.
Please cite this article in press as: Hansen, J. S.; et al. Tetrahedron Lett. (2013), http://dx.doi.org/10.1016/j.tetlet.2013.01.101