Ortho-substituted aryl monoboronic acids have improved selectivity for d-glucose relative to d-fructose and l-lactate

Article abstract of DOI:10.1016/j.tet.2010.11.062

Ortho-substituted aryl monoboronic acids have been found to have improved selectivity for d-glucose compared to d-fructose and l-lactate. These findings are supported by computational studies on the B3LYP/6-31G(d) level using Gaussian. This finding is of

Full text of DOI:10.1016/j.tet.2010.11.062

Tetrahedron 67 (2011) 1334e1340
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Ortho-substituted aryl monoboronic acids have improved selectivity for -glucose
D
relative to
D
-fructose and -lactate
L
,
a
a
b
b
J.S. Hansen a , J.B. Christensen , T.I. Solling , P. Jakobsen , T. Hoeg-Jensen
*
a Department of Chemistry, Universitetsparken 5, DK-2100 Copenhagen, Denmark
b Novo Nordisk, 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:
Ortho-substituted aryl monoboronic acids have been found to have improved selectivity for
compared to -fructose and -lactate. These findings are supported by computational studies on the
B3LYP/6-31G(d) level using Gaussian. This finding is of interest for development of boronate based
-glucose sensors.
D-glucose
Received 23 February 2010
Received in revised form 27 October 2010
Accepted 16 November 2010
Available online 24 November 2010
D
L
D
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
exists in the furanose form, which can bind to boronates in tridentate
configuration, while
form, and binds in a bidentate configuration.16,17 Notably
also present in human blood as a metabolite after anaerobic bi-
ological processes, and -hydroxy acids are also capable of forming
boronate esters with aryl boronic acids.18,19 Diboronates20e22 can
show selectivity towards -glucose, but they are synthetically chal-
lenging, compared with the synthesis of aryl monoboronates, and
usually their solubility is low in aqueous media. Carbohydrate af-
finitiesof some boronates can be measured byexploiting their UV/vis
or fluorescence properties, or in cases were such spectroscopic
properties are missing, by competitive binding to spectroscopic
probes.23e29 Another commonly used technique is the pH-de-
pression method,30 where the drop in pH can be correlated with the
binding constant between these two species. However this method
suffers from the fact that it requires a high amount of boronic acid in
solution. Furthermore this method assumes that the boronic acid
diester is fully converted to the tetrahedral anionic form, as a conse-
quence of the lowered pKa value.
In order to overcome the lack of spectroscopic properties of the
given aryl boronic acids, we have adopted the method of UV/vis-
titration experiments with the colored compound alizarin red
sodium (ARS) in a three component competitive binding assay. This
assay has successfully been employed by Wang and
co-workers.11,31,32 As shown in Fig. 1, ARS binds reversibly to the
aryl boronate, forming the corresponding aryl boronate ester,
which displaces the absorption spectrum, and changes the color of
the aqueous solution from clear red towards yellow.
D
-glucose mainly exists in the glucopyranose
Aryl boronic acids are often used in carbohydrate recognition1e6
because they are small and flexible, compared to other classes of
carbohydrate binders, such as lectins7 and artificial macrocycles.8,9
These features make aryl boronic acids easy to incorporate as rec-
ognition motifs in spectral probes or peptides, without dramatically
changing the physical properties of the larger structures. The rec-
ognition event appears since boronic acids and boronates react
with 1,2-cis-diols or 1,3-diols by reversible formation of the cor-
responding boronate esters. It is commonly believed that optimal
binding affinity is achieved when pH is above pKa for a specific
boronic acid. Earlier literature10,11 reports that the binding strength
depends on pKa of both the diol and the boronic acid. The event of
carbohydrate binding is however complicated, since carbohydrates
are polyols, which can exist in different anomeric configurations of
six membered and five membered heterocyclic rings.
L-lactate is
a
D
Glucose monitoring is of great importance in the treatment of
diabetes mellitus. Boronic acids with displacement constants around
15e16 mM for
D
-glucose are desired,12 because blood glucose is
fluctuating between 2 and 30 mM in diabetes patients. The maximum
sensitivity is achieved when Kd is in the middle of the binding curve.
Aryl boronic acids with selective recognition of
other polyol species are therefore of great interest, since
the major carbohydrate present in human blood (z5 mM)13 com-
D-glucose over
D-glucose is
pared to
However
D
-fructose (<0.1 mM, even after a fructose-rich meal).14,15
-fructose generally shows stronger binding affinity to
-glucose. An explanation
might be, that under physiological conditions, -fructose mainly
D
most aryl monoboronates compared to
D
D
Upon addition of the polyol to the ARS-boronate solution, ARS is
released competitively, changing the color of the solution back
towards red.
We used the ARS assay to screen a series of aryl boronic acids
performing the measurements in a physiological environment at
* Correspondingauthor. Tel.:þ4561337866;e-mail address: jonhansen@kemi.ku.dk
(J.S. Hansen).
0040-4020/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2010.11.062

J.S. Hansen et al. / Tetrahedron 67 (2011) 1334e1340
1335
HO
B
OH
HO
OH
HO
OH
B
B
O
O
O
S
N
S
S
N
N
O
O
O
3
1
2
HO
OH
HO
OH
B
HO
OH
B
B
O
CF3
O
Fig. 1. Competitive binding of ARS and polyol to aryl monoboronate in a three com-
S
ponent system.
N
5
6
4
O
neutral pH. Comparing our data with data obtained using a non-
physiological phosphate buffer, we report small changes in the
respective binding affinities. Generally decreased selectivity to-
wards D-fructose is achieved, due to the presence of ortho-sub-
stituents at the aryl monoboronic acids.
We have discovered that ortho-substituted aryl monoboronic
acids bind D-fructose with a reduced strength compared to aryl
monoboronic acids with no ortho-substituents. This is a valuable
discovery, because a lot of synthetic effort can be avoided in the
preparation of suitable boronic acid dyes, since aryl monoboronic
acids generally are easier to access, and more soluble.
HO
HO
OH
HO
OH
HO
OH
B
B
B
F
9
F
8
F
7
F
F
O
O
OH
O
HO
OH
HO
HO
OH
HO
B
B
B
F
12
11
10
2. Results and discussion
Screening studies of a series of aryl boronic acids were per-
formed in order to obtain selectivity towards binding of D-glucose.
O
O
HO
O
Fig. 2. Aryl boronic acids tested for binding strength to D-glucose, D-fructose, and
L-lactate.
The measurements were performed in a physiological saline buffer,
containing 10 mM phosphate, 2.7 mM KCl, and 137 mM NaCl in
water, pH 7.4. The structures are shown in Fig. 2, and calculated Kd-
values are shown in Table 1. Compound 1 and 2 are very similar in
structure at the binding site, since they both contain a methyl group
attached in the ortho-position to the boronic acid functionality. Our
first statement is in good agreement with observation of larger Kd
(42 mM) for binding of
because more of 2 is converted to the corresponding boronate at
L
-lactate to 2, compared to 1 (Kd¼29 mM),
studies indicate that binding of
displacement constant Kd is about 10 mM for both boronic acids.
Normally Kd is about 0.5e1 mM for -glucose is
-fructose.10 Kd for
about 60 mM for 1 and 30 mM for 2. The sigmoidal curves for
D
-fructose is disturbed, since the
physiological pH.
D
D
However the complexes formed between L-lactate and 3 and 4 are
remarkably stronger compared to the complexes formed with 1 and
2, which is the opposite of what one would expect, since pKa of 3 and
4 are lower than pKa for 1 and 2. This suggests an effect of the ortho-
binding of 1 and 2 with D-glucose, D-fructose, and L-lactate are
shown in Fig. 6.
The aryl boronic acids 3 and 4 were used as reference compounds
to determine the influence of an ortho-positioned methyl group.
substituent too since decreased binding of L-lactate finds place.
Compound 5 was used as a reference compound to state the
influence of pKa for the binding affinity. Our results showed a little
D
-Fructose is bound remarkably better (one order of magnitude) by 3
and 4 than it is bound with 1 and 2, probably because the more stable
tridentate -fructose-boronate complex is favored, when no ortho-
substituents are present, see Fig. 3.17 The complexes formed
between -glucose and 3 and 4, respectively, are also stronger,
probably due to the decreased pKa of 3 and 4 compared to 1 and 2.
The difference in binding strength of -glucose to 1 and 2, might
stronger binding affinity of 5 to
affinity of 1 and 2. Poor binding affinity of 5 to
observed, and -lactate was bound significantly better in the pres-
D-fructose compared to the binding
D
D
-glucose was
L
D
ence of 5 compared to 1 and 2, respectively. This can be explained
by the much higher pKa value of 5.
D
The binding affinity of 6 to
D-glucose is slightly less compared to
be explained by the difference in acidity. This is also the case for 3 and
4. Pyrrolidine is more electron-donating than dimethylamine, thus
decreasing the acidity of 1 and 3 compared to 2 and 4.33 Our results
2, and -fructose is bound one order of magnitude more strongly by
6. Decreased intramolecular hydrogen bonding from the boronate
OH to oxygen in the trifluoromethoxy group could explain why the
D
suggest that the binding strength towards
the acidity of the boronic acid than the binding strength towards
-glucose. Theproposedstericclashwithanortho-substituent, which
disfavors tridentate binding of -fructose is shown in Fig. 4.
It is commonly believed that an -hydroxy acid reacts with the
trigonal form (boronic acid),34,35 but there are also implications
that -hydroxy acids can react with the tetrahedral boronate at
a slower rate.36 Charge repulsion between the negatively charged
-lactate and the aryl boronate presumably occurs. Therefore the
D
-fructose depend less on
binding of
D
-fructose is not disturbed. The smaller size of O versus
CH3 can also explain the increased binding affinity to
D
-fructose.
D
L-lactate though is tightly bound.
D
The reduced binding affinity of D-fructose to the tested ortho-
a
fluoro aryl monoboronic acids (8 and 10), compared to 11 and 12 can
be explainedbya favorable intramolecularhydrogenbond, see Fig. 5,
which generally favors bidentate complex formation, thus dis-
favoring the tridentate
binding affinity of -fructose to 8 are increased compared to 1 and 2.
a
D-fructoseeboronate complex. However the
L
D

1336
J.S. Hansen et al. / Tetrahedron 67 (2011) 1334e1340
Table 1
1 fit
D-Glc
D-Frc and L-Lac
Displacement constants (Kd) reported in mM, shown for the series of boronic acids in
physiological saline buffer. n is the number of measurements performed for each
0 5
0 4
0 3
0 2
0 1
determination. Standard deviations
s
are shown to the right in bold
D
D
-Frc
-Glc
D
-Glucose Kd(n)
D
-Fructose Kd(n) -Lactate Kd(n)
L
1
2
60 (6) (6.1)
30 (3) (2.6)
13 (7) (2.5)
10 (5) (1.6)
1.5 (3) (0.5)
0.9 (3) (0.1)
5.7 (3) (0.1)
1.6 (3) (0.3)
3.2 (3) (0.5)
2.5 (9) (0.6)
3.4 (5) (0.3)
11 (5) (2.7)
1.0 (3) (0.2)
0.9 (3) (0.3)
29 (7) (2.8)
42 (4) (3.5)
1.9 (3) (0.2)
4.9 (3) (0.1)
5.9 (3) (0.2)
7.5 (3) (0.1)
47 (3) (2.0)
22 (3) (1.0)
49 (3) (3.2)
13 (5) (2.6)
4.8 (3) (0.4)
5.6 (3) (0.7)
L
-Lac
3
4
5
9.1 (3) (0.3)
5.9 (3) (0.2)
d (3) d
6
7
8
9
10
11
12
63 (3) (6.6)
25 (3) (1.7)
26 (9) (3.1)
28 (6) (2.1)
40 (5) (5.2)
10 (3) (2.0)
414 (3) (37.3)
4
3
2
1
0
1
logC
2 fit
D-Glc
D-Frc and L-Lac
0 5
0.4
0 3
0 2
0 1
HO
OH
OH
D
D
-Frc
-Glc
O
O
B
OH
+
O
OH
H
H
O
B
O
H
H
OH
L
-Lac
OH
OH
OH
Fig. 3. Favorable tridentate binding of D-fructose to an aryl monoboronic acid with no
ortho-substituents. The tridentate structure has earlier been elucidated by Norrild and
Eggert.17
4
3
2
1
0
1
R
HO
HO
+
logC
OH
HO
R
O
OH
OH
B
O
OH
H
H
O
B
O
H
Fig. 6. Curve fit using GraphPad Prism 5.0. The measured absorbance at 350 nm, was
plotted against logarithm to concentration data, using sigmoidal dose response (variable
OH
OH
OH
slope). [1, 2]¼100
m
M and [ARS]¼50
mM, here in the physiological saline buffer.
H
OH
Fig. 4. Proposed steric clash for D-fructose induced by an ortho-positioned substituent
in an aryl monoboronic acid.
boronate 7 and 9 display the most decreased selectivity towards
-lactate.
The aryl boronic acids 11 and 12 were used as reference com-
pounds to determine the influence of the ortho-positioned fluorine.
The Kd-values indicate formation of a stronger -fructose complex
between -fructose and 11 and 12, compared to 8 and 10. -lactate is
bound stronger as well, which can be explained by the higher pKa
values of 11 (7.6) and 12 (8.0) compared to 8 and 10.11
-glucose is
L
D
F
HO
+
D
L
OH
HO
H
F
O
OH
OH
O
B
O
OH
H
H
O
B
O
D
H
OH
poorly bound to boronate 12 compared to 8, 10, and 11, which
cannot be explained by pKa arguments.
OH
OH
H
OH
The ratios of Kd for
-glucose/ -lactate Kd(
clearly seen that boronate 1, 2, and 10 displays the greatest reduced
D
-glucose/
-Glc)/Kd(
D
-fructose Kd(
D-Glc)/Kd(D-Frc), and
Fig. 5. Suggested intramolecular hydrogen bonding from boronate OH to fluorine,
preventing D-fructose from binding tridentately.
D
L
D
L-Lac) is given in Table 2. It is
selectivity towards -fructose, while boronate 2, 7, and 9 displays
D
Compounds 7 and 9 are very similar in structure. However both
boronic acids are likely to be converted to either the boronate (7)
and/or carboxylate (9) at pH 7.4. This is consistent with the
Table 2
observed low binding affinity to
affinity to -glucose is around 20 mM
a similar strength (Kd¼3.2 mM and Kd¼3.4 mM) in both species.
-lactate is bound more strongly to 8 and 10, compared to 7 and 9.
L-lactate. In both cases, the binding
Ratios of Kd for
D-glucose/D-fructose and Kd D-glucose/L-lactate for the series of
boronic acids in a physiological saline buffer
D
D
-fructose is bound with
Kd(
D
-Glc)/Kd(
D
-Frc)
Kd(D-Glc)/Kd(L-Lac)
L
1
2
3
4
4.6
3.0
6.1
6.6
d
2.1
0.7
4.8
1.2
d
This is consistent with our expectations since the additional elec-
tron-withdrawing fluorine substituent increase acidity of 7 and 9
compared to 8 and 10.
7, 8, and 9 with a Kd, close to 15e16 mM, which is desired.
-Fructose however is bound more efficiently to 8 than to 10. Since
D-Glucose is bound with a similar strength to
5
6
7
8
9
10
11
12
39.4
7.8
10.4
8.2
3.6
10.0
8.4
0.5
1.2
0.6
3.1
2.1
D
8 and 10 are similar at the recognition site, the diminished selec-
tivity in the latter case cannot be explained by competitive intra-
molecular hydrogen bonding as suggested, see Fig. 5. In accordance
with our measurements, 10 is the boronate where the selectivity
460.0
73.9
towards
D-fructose is decreased most among 7e10, whereas

J.S. Hansen et al. / Tetrahedron 67 (2011) 1334e1340
1337
the greatest reduced selectivity towards
L-lactate. Boronate 12
however shows a remarkable preference of forming the
boronate complex.
D-fructose-
Compounds with larger ortho-substituents, such as o-ethoxy-
methyl, o-aminomethyl, and o-phenyl- have been tested, but they
have not shown any binding affinity to
D
-glucose. Furthermore no
significant binding affinity to
D
-fructose and -lactate has been
L
achieved.
Binding constants were also determined in a phosphate buffer,
with a similar strength as a physiological buffer (100 mM phos-
phate, pH 7.4), and in a methanolic phosphate buffer (53.2 w/w%
MeOH, 52 mM phosphate, 40 mM NaCl). Measurements showed
diminished binding of
D-glucose, improved binding of D-fructose
and -lactate in the 100 mM phosphate buffer with 2, compared to
L
the physiological saline buffer containing 10 mM phosphate,
2.7 mM KCl, and 137 mM NaCl in water, pH 7.4. This suggests that
the ionic strength of the solution influences binding affinity.
Fig. 7. Two energy optimized structures of tridentate complexes formed between
D-fructose and p-methyl phenyl monoboronate (left), and o-methyl phenyl mono-
boronate (right). The left complex is 4.6 kJ/mol more stable than the right. Grey¼C,
white¼H, red¼O and yellow¼B.
Binding of
D
-glucose was remarkably diminished in the methanolic
-fructose unchanged, and binding of -lactate
buffer, binding of
D
L
improved, compared to the physiological buffer. This indicates that
the acidity of 2 is decreased in the methanolic buffer compared to
the physiological buffer solution. The values of Kd are shown in
Table 3.
Table 3
Displacement constants (Kd) reported in mM shown for 2. n is the number of
measurements performed for each determination. Standard deviations are shown to
the right in bold
(2) phosphate (Kd)(n):
(2) phosphate/MeOH.(Kd)(n):
D
D
-Glucose:
-Fructose:
62 (5) (3.2)
4.7 (5)(0.3)
23 (6) (3.6)
160 (6) (24.4)
8.4 (6) (0.5)
3.2 (5) (0.1)
L
-Lactate:
The ratios Kd
(D
-Glc)/Kd(
D-Frc) and Kd
(
D-Glc)/Kd(L-Lac) are cal-
culated in Table 4, showing that the best
D-glucose selectivity is
obtained in the phosphate buffer, compared to the methanolic
phosphate buffer.
Table 4
Ratios Kd(
Fig. 8. The energy optimized structure of the bidentate complex formed between
D-fructose and o-methyl phenyl monoboronate (left). The left complex is 2.1 kJ/mol
more stable than the right, and the stabilization corresponds to free rotation around
the CeB bond. Grey¼C, white¼H, red¼O and yellow¼B.
D-Glc)/Kd(D-Frc) and Kd(D-Glc)/Kd(L-Lac) for 2 in the phosphate buffer and
in the methanolic phosphate buffer, respectively
(2) Phosphate Kd
(2) Phosphate Kd
(2) Phosphate/MeOH Kd
(2) Phosphate/MeOH Kd
(
(
D
D
-Glc)/Kd(
D
-Frc):
13.2
2.7
-Glc)/Kd(
L
-Lac):
In good agreement with our assumptions, the optimized struc-
ture of the bidentate complex formed between o-fluoro phenyl
monoboronate and D-fructose shows a preference for a favorable
(
(
D
D
-Glc)/Kd(
-Glc)/Kd(
D
-Frc):
19.0
50.0
L-Lac):
HeF hydrogen bond (3.3 kJ/mol more stable than the bidentate
complex with no HeF bond formation), see Fig. 9. This is in good
agreement with calculations performed by Razgulin and Mecozzi.37
We performed pKa-titrations of 2 in a 53.2 w/w% methanolic
solution, but realized that the pKa for the boronic acid cannot be
determined under these conditions, because 2 partially form the
methylboronate. The pKa of the methylboronate and the pKa of the
boronic acid cannot be distinguished in this experiment.
To support our experimental investigations, computational
calculations were performed on B3LYP/6-31G(d) level with
Gaussian.
The energy optimized structures of the tridentate complexes
formed between D-fructose and p-methyl phenyl monoboronate,
and o-methyl phenyl monoboronate, respectively, were compared,
see Fig. 7. The tridentate complex is in accordance with previous
structural elucidation performed by Norrild and Eggert.17 The latter
complex is 4.6 kJ/mol less stable, indicating that an o-methyl sub-
stituent reduces tridentate complex formation.
The optimized structure of the bidentate complex formed be-
tween D-fructose and o-methyl phenyl monoboronate shown in
Fig. 8, shows that free rotation around the CeB bond can stabilize
the bidentate complex. Thus according to our calculations, the
Fig. 9. The energy optimized structure of a complex formed between D-fructose and
o-fluoro phenyl monoboronate (left). The left complex is 3.3 kJ more stable than the
right. Grey¼C, white¼H, green¼F, red¼O and yellow¼B.
bidentate
D-fructoseeboronate complex is not disturbed due to the
presence of an ortho-methyl substituent.

1338
J.S. Hansen et al. / Tetrahedron 67 (2011) 1334e1340
Thus the ortho-fluoro substituent favors the bidentate
seeboronate complex, which decreases the -fructose selectivity.
The optimized tridentate complexes formed between -fructose
and o-fluoro phenyl monoboronate, and p-fluoro phenyl mono-
boronate, respectively, were compared, and showed a significant
energy difference. The latter complex was calculated to be 7.0 kJ/
mol more stable than the former. This suggests electrostatic effects,
such as lone pairelone pair repulsion, which disfavors the tri-
D-fructo-
Titration of 5
12
10
8
D
D
6
4
dentate D-fructose o-fluoro phenyl monoboronate complex. How-
ever steric repulsion might also contribute, since the covalent
radius of fluorine approximately is twice the covalent radius of
hydrogen; the radii are 60 pm and 30 pm, respectively.38,39 The two
tridentate complexes are shown in Fig. 10.
2
0
0
1
2
3
4
5
V(mL) 0.025 M HCl titrator
pKa values of the monoboronic acids have been determined by
titration of the corresponding boronates, and are shown in Table 5.
The titration curve for 5 is shown in Fig.11. The relative values are in
good agreement with our results for Kd. Besides the Hammett linear
free energy relationship has been used to calculate pKa values for
selected aryl monoboronic acids (7 and 8), see Table 5. The calcu-
lations involves the equation employed by Wang and Springsteen,11
and the found pKa values are in good agreement with our argu-
ments. The much higher pKa values of 1 and 2 compared to 3 and 4,
might be explained sterically by an unfavorable boronate ortho-
Fig. 11. Titration curve for the corresponding base of 5, where pKa is determined to 9.3
for the aryl monoboronic acid.
3. Conclusion
The experimental outcome shows that ortho-substituted aryl
monoboronic acids are capable of binding
creased selectivity relative to -glucose in a physiological buffer,
presumably due to a steric effect of the ortho-positioned methyl
substituent, disfavoring the tridentate boronate- -fructose com-
D-fructose with a de-
D
methyl interaction. The high -glucose affinity for 1 and 2 might
D
D
then be explained by a favorable lipophilic
interaction.
D-glucoseeboronate
plex. This is supported by computational calculations on B3LYP/6-
31G(d) level with Gaussian.
Competitive intramolecular hydrogen bonding, in the case of
ortho-positioned fluorine favors bidentate
D-fructoseeboronate
complex formation, thus -fructose binding is decreased. Compu-
D
tational calculations have also supported these theories. However
calculations have further shown that the disfavored tridentate
D-fructose o-fluoro phenyl monoboronate complex contributes
even more to the decreased -fructose selectivity, possibly due to
D
electrostatic effects, such as lone pairelone pair repulsion. Steric
effects might also contribute since the covalent fluorine radius is
about twice the covalent hydrogen radius. Our experimental results
show that other effects do impact the decreased binding of
D-fructose, as seen when 7 and 8 are compared. 7 does not contain
any ortho-substituents, but two meta-fluoro substituents, while 8
contain an ortho-fluoro substituent.
Furthermore our measurements indicate that L-lactate generally
is weakly bound when the boronate is the dominant configuration,
presumably because of unfavorable coulomb interactions. The
glucose displacement constant for 2, 7, 8, and 9, are closest to
15e16 mM, which is desired, since blood glucose is fluctuating
between 2 and 30 mM in diabetes patients, and the maximum
sensitivity is achieved when Kd matches the middle of the binding
curve. Compounds with larger ortho-substituents, such as
o-ethoxymethyl, o-aminomethyl, and o-phenyl- has been tested,
Fig. 10. The energy optimized structure of a tridentate complex formed between D-
fructose and p-fluoro phenyl monoboronate (left), and o-fluoro phenyl monoboronate.
The left complex is 7.0 kJ/mol more stable than the right. Grey¼C, white¼H, green¼F,
red¼O and yellow¼B.
but they have not shown any binding affinity to
insignificant binding affinity to -fructose and -lactate. This further
outlines how steric repulsion can disturb boronate ester-formation.
Our results are promising for developing the -glucose moni-
D-glucose, and only
D
L
D
toring dyes based on aryl monoboronic acids, which are attractive
compared to more synthetically challenging, notoriously larger and
thus usually less soluble aryl diboronic acid dyes.
Table 5
pKa values of selected aryl monoboronic acids, determined by titration of the cor-
responding base. For 7 and 8, the calculated Hammett-values are shown in the
brackets to the right
Aryl monoboronic acid
pKa (titration)
4. Experimental
4.1. General
1
2
3
4
5
6
7
8
8.4
8.2
7.6
7.2
9.3
8.3
6.3 (6.2)
6.8 (6.7)
The computational calculations were performed on B3LYP/6-
31G(d) level with Gaussian.
The aryl boronic acids 1e6, 11, and 12 were purchased from
Combi Blocks and used as received. Compounds 7 and 8 were

J.S. Hansen et al. / Tetrahedron 67 (2011) 1334e1340
1339
purchased from SigmaeAldrich and used as received. 9 and 10 were
prepared by oxidation of the corresponding aldehydes with
KMnO4.40 Alizarin red sodium,
-glucose, -fructose, -lactate,
P4417 (phosphate buffered saline pellets), and P7994 (phosphate
buffer pellets) were purchased from SigmaeAldrich and used as
received. Deionized water was used for the binding studies.
All data were fitted in GraphPad Prism 5.0, where the measured
absorbances at 340 nm, 350 nm, and 360 nm were plotted against
the logarithm to concentration data, using sigmoidal dose response
phosphate, [ARS]¼50
m
M, [boron]¼100
m
M, and [polyol]¼2 M. The
L
-lactate solution was prepared as in the previous experiment,
D
D
L
resulting in 100 mM phosphate, [ARS]¼50
mM, [boron]¼100
mM,
and [
L-lactate]¼2 M.
(3b). Phosphate buffer containing ARS and aryl boronic acid: 6 mL
ARS buffer containing aryl boronic acid was diluted with 6 mL of
deionized water, to obtain [ARS]¼50
m
M and [boron]¼100
mM in
a 100 mM phosphate buffer (pH¼7.4).
fit (variable slope). Kd¼100
m
M (boronate) and [Ligand]¼50
mM
(ARS), or [Ligand]¼52
m
M (ARS). UV/vis-measurements were all
(4b). Sample preparation: sample preparation was carried out as
described in previous experiment.
performed on a PerkineElmer apparatus.
4.2. pKa determination of aryl monoboronic acids via
acidebase titration
4.5. UV/vis-titration with ARS in 53.2 w/w% MeOH, 52 mM
phosphate, and 40 mM NaCl buffer
Aryl monoboronic acid (1e8) (0.125 mmol) was dissolved in 5 mL
of 0.025 M NaOH and 5 mL of deionized water, creating the corre-
sponding base (1.25 x10ꢀ2 M), which was titrated with 0.025 M HCl.
(1c). Phosphate/MeOH buffer containing ARS and aryl boronic acid:
480 mL 104 mM methanolic phosphate buffer containing 53.2 w/w
% MeOH was prepared by mixing 200 mL deionized water and
300 mL MeOH, and 8.5 g phosphate buffer pellets. Afterwards
18 mg ARS (0.05 mmol) was dissolved in 480 mL buffer solution,
4.3. UV/vis-titration with ARS in saline buffer
stirring for 4 h at room temperature, which gave [ARS]¼104
0.048 mmol boronic acid was dissolved in 240 mL buffer solution,
which gave [boron]¼200 M.
mM.
(1a). Buffer containing ARS and aryl boronic acid: 1.0 L buffer
containing 20 mM phosphate, 5.4 mM potassium chloride, and
274 mM sodium chloride was prepared. Afterwards 36 mg ARS
(0.1 mmol) was dissolved with stirring for 5 h while slightly heating
m
(2c). MeOH (53.2 w/w%): 53.2 g MeOH (67.3 mL) and 46.8 g
deionized water (46.8 mL) was mixed, to obtain 53.2 w/w% MeOH.
(40 ꢁCe50 ꢁC), [ARS]¼100
mM. 0.05 mmol aryl boronic acid was
dissolved in 250 mL of the prepared ARS buffer solution, [boron]¼
200
mM.
(3c). Polyol solutions: D-glucose: 10.0 mmol D-glucose (1.803 g)
was dissolved in 5 mL of the solution prepared under procedure
(1c). Afterwards 0.1 mL 4 M HCl was added, and the total volume
was adjusted with 53.2 w/w% MeOH (prepared under 2c) to 10 mL.
(2a). Polyol solution: A polyol solution was prepared by mixing
20.0 mmol ( -fructose or -glucose, 3.61 g) and 5 mL ARS buffer
D
D
containing aryl boronic acid. pH was adjusted with NaOH (7.4), and
addition of water to a total volume of 10 mL. This gave a buffered
This gave [
D
-glucose]¼1 M, [phosphate]¼52 mM, and [NaCl]¼
40 mM, pH¼7.4.
D
-Fructose: 20.0 mmol -fructose (3.608 g) was
D
solution(10mMphosphate, 2.7mMKCl,137mMNaCl), [ARS]¼50
[boron]¼100 M, and [polyol]¼2 M. The -lactate solution was pre-
pared by mixing 20.0 mmol of -lactate (2.24 g) and 5 mL ARS buffer
mM,
dissolved in 5 mL of the solution prepared under (1c). Afterwards
m
L
0.1 mL 4 M HCl was added, and the total volume was adjusted with
L
53.2 w/w% MeOH (prepared under 2c) to 10 mL. This gave [
fructose]¼2 M, [phosphate]¼52 mM, and [NaCl]¼40 mM, pH¼7.4.
-lactate: 20.0 mmol -lactate was dissolved in 5 mL of the solution
D-
containing aryl boronic acid. Addition of deionized water to a total
volumeof10mLwasperformed, whileadjustingpHto7.4, byaddition
L
L
of 50
m
L 0.01 M HCl. This gave buffer (10 mM phosphate, 2.7 mM KCl,
prepared under procedure (1c). Afterwards the volume of the so-
lution was adjusted with 53.2 w/w% MeOH, while addition of 4 M
HCl dropwise until pH¼7.4. The total volume was 10 mL. This gave
137 mM NaCl), [ARS]¼50
m
M, [boron]¼100 M, and [ -lactate]¼2 M.
m
L
(3a). Physiological buffer containing ARS and aryl boronic acid:
6 mL ARS buffer containing aryl boronic acid was diluted with 6 mL
[L
-lactate]¼2 M, [phosphate]¼52 mM, and [NaCl]¼40 mM.
of deionized water, to obtain [ARS]¼50
m
M and [boron]¼100
m
M in
(4c). Buffer used in sample preparation: 6 mL of solution prepared
under procedure (1c), was diluted with 5.9 mL 53.2 w/w% MeOH,
and pH was adjusted with 0.12 mL 4 M HCl to 7.4. This gave
a physiological buffer (10 mM phosphate, 2.7 mM KCl, 137 mM
NaCl, pH¼7.4).
[boron]¼100
m
M, [ARS ]¼52 M, and [NaCl]¼40 mM. Total volume
m
(4a). Sample preparation: ARS-boronate-polyol solution mixed
12 mL.
with ARS-boronate-solution, gave the following 14 polyol concen-
trations: [polyol]¼200
m
M, 500
m
M, 1 mM, 2 mM, 5 mM, 10 mM,
(5c). Sample preparation.
glucose concentrations were prepared by mixing the ARS-boro-
nate- -glucose solution with the ARS-boronate-solution:
glucose]¼200 M, 500 M, 1 mM, 2 mM, 5 mM, 10 mM, 20 mM,
50 mM, 100 mM, 200 mM, 500 mM, and 1.0 M. Sample consument
(3c) about 1.9 mL. Sample consument (4c) about 10.1 mL -fructose
and -lactate: ARS-boronate- -fructose or ARS-boronate- -lactate
solution mixed with ARS-boronate-solution, gave the following 14
M, 500 M, 1 mM, 2 mM,
D-Glucose: solutions with 12 different D-
20 mM, 50 mM, 100 mM, 200 mM, 500 mM, 1.0 M, 1.5 M, and 2.0 M.
Sample consument (2a) about 2.7 mL. Sample consument (3a)
about 11.3 mL Samples with
solutions.
D
[D-
L
-lactate was prepared as the polyol
m
m
D
4.4. UV/vis-titration with ARS in phosphate buffer
L
D
L
(1b). Buffer containing ARS and boronic acid: 250 mL 200 mM
phosphate buffer was prepared. Afterwards 9 mg ARS (0.025 mmol)
polyol concentrations: [polyol]¼200
m
m
5 mM, 10 mM, 20 mM, 50 mM, 100 mM, 200 mM, 500 mM, 1.0 M,
1.5 M, and 2.0 M.
was dissolved with stirring for
5
h
while slightly heating
M. 0.05 mmol aryl boronic acid was
M.
(40 ꢁCe50 ꢁC), [ARS]¼100
m
dissolved in the prepared ARS buffer solution, [boron]¼200
m
Acknowledgements
(2b). Polyol solution: preparation was performed as in the previous
experiment with -glucose and -fructose, resulting in 100 mM
Financialsupport from the UniversityofCopenhagen, Department
of Chemistry, and from Novo Nordisk A/S is gratefully acknowledged.
D
D

1340
J.S. Hansen et al. / Tetrahedron 67 (2011) 1334e1340
20. Arimori, S.; Ward, C. J.; James, T. D. Tetrahedron Lett. 2002, 43, 303e305.
References and notes
21. Suri, J. T.; Cordes, D. B.; Cappucio, F. E.; Wessling, R. A.; Singaram, B. Langmuir
2003, 19, 5145e5152.
22. Eggert, H.; Frederiksen, J.; Morin, C.; Norrild, J. C. J. Org. Chem. 1999, 64,
3846e3852.
23. Cabell, L. A.; Monahan, M.-K.; Anslyn, E. V. Tetrahedron. Lett. 1999, 40,
7753e7756.
24. Yang, W.; He, H.; Drueckhammer, D. G. Angew. Chem., Int. Ed. 2001, 40,
1714e1718.
1. Yang, W.; Yan, J.; Springsteen, G.; Deeter, S.; Wang, B. Bioorg. Med. Chem. Lett.
2003, 13, 1019e1022.
2. Cao, H.; Diaz, D. I.; DiCesare, N.; Lakowicz, J. R.; Heagy, M. D. Org. Lett. 2002, 4,
1503e1505.
3. Shinkai, S.; Takeuchi, M. Trends Anal. Chem. 1996, 15, 188e193.
4. Draffin, S. P.; Duggan, P. J.; Duggan, S. A. M.; Norrild, J. C. Tetrahedron 2003, 59,
9075e9082.
5. Hirata, O.; Kubo, Y.; Takeuchi, M.; Shinkai, S. Tetrahedron 2004, 60, 11211e11218.
6. Dowlut, M.; Hall, D. G. J. Am. Chem. Soc. 2006, 128, 4226e4227.
7. Goldstein, I. J.; Murphy, L. A.; Ebisu, S. Pure Appl. Chem. 1977, 49, 1095e1103.
8. Anderson, S.; Neidlein, U.; Gramlich, V.; Diederich, F. Angew. Chem., Int. Ed. Engl.
1995, 34, 1596e1600.
9. Barwell, N. P.; Crump, M. P.; Davis, A. P. Angew. Chem., Int. Ed. 2009, 48, 7673e7676.
10. Van Duin, M.; Peters, J. A.; Kieboom, A. P. G.; Van Bekkum, H. Tetrahedron 1984,
40, 2901e2911.
11. Wang, B.; Springsteen, G. Tetrahedron 2002, 58, 5291e5300.
12. Kamata, K.; Morihiro, M.; Nishimura, T.; Eiki, J.-I.; Nagata, Y. Structure 2004, 12,
429e438.
25. Soundararajan, S.; Badawi, M.; Kohlrust, C. M.; Hageman, J. H. Anal. Biochem.
1989, 178, 125e134.
26. Tong, A. J.; Yamauchi, A.; Hayashita, T.; Zhang, Z. Y.; Smith, B. D.; Teramae, N.
Anal. Chem. 2001, 73, 1530e1536.
27. Yamamoto, M.; Takeuchi, M.; Shinkai, S. Tetrahedron 1998, 54, 3125e3140.
28. Kijima, H.; Takeuchi, M.; Shinkai, S. Chem. Lett. 1998, 781e782.
29. James, T. D.; Linnane, P.; Shinkai, S. Chem. Commun. 1996, 281e288.
30. Chapelle, S.; Verchere, J. Tetrahedron 1988, 44, 4469e4482.
31. Springsteen, G.; Wang, B. Chem. Commun. 2001, 1608e1609.
32. Yan, J.; Sprinsteen, G.; Deeter, S.; Wang, B. Tetrahedron 2004, 60,
11205e11209.
33. Hammerich, O.; Hansen, T.; Thorvildsen, A.; Christensen, J. B. ChemPhysChem
2009, 10, 1805e1824.
13. Heise, H. M.; Bittner, A.; Koschinsky, T.; Gries, F. A. Fresenius’ J. Anal. Chem. 1997,
359, 83e87.
14. Donmoyer, C. M.; Eijofor, J.; Lacy, D. B.; Chen, S.-S.; McGuinness, O. P. Am. J.
Physiol.Endocrinol. Metab. 2001, 280, E703eE711.
34. Pizer, R.; Selzer, R. Inorg. Chem. 1984, 23, 3023e3026.
35. Friedman, S.; Pace, B.; Pizer, R. J. Am. Chem. Soc. 1974, 96, 5381e5384.
36. Babcock, L.; Pizer, R. Inorg. Chem. 1980, 19, 56e61.
37. Razgulin, A. V.; Mecozzi, S. J. Med. Chem. 2006, 49, 7902e7906.
38. Robinson, E.; Johnson, S.; Tang, T.-H.; Gillespie, R. Inorg. Chem. 1997, 36,
3022e3030.
15. Arai, T.; Hashimoto, K.; Muzutani, H.; Kawabata, T.; Sako, T.; Washizu, T. Vet. Res.
Commun. 1999, 23, 203e209.
16. Angyal, S. J. Adv. Carbohydr. Chem. Biochem. 1984, 42, 15e68.
17. Norrild, J. C.; Eggert, H. J. Chem. Soc., Perkin Trans. 2 1996, 2583e2588.
18. Chi, L.; Zhao, J.; James, T. D. J. Org. Chem. 2008, 73, 4684e4687.
19. Sartain, F. K.; Yang, X.; Lowe, C. R. Chem.dEur. J. 2008, 14, 4060e4067.
39. Schomaker, V.; Stevenson, D. P. J. Am. Chem. Soc. 1941, 63, 37e40.
40. de Filipps, A.; Morin, C.; Thimon, C. Synth. Commun. 2002, 32, 2669e2676.