Specific sensing between inositol epimers by a bis(boronate)

Article abstract of DOI:10.1016/j.bmcl.2005.08.112

Bis(boronates) that utilize internal photoinduced electron transfer (PET) quenching mechanisms can specifically signal the binding of chiro-inositol without responding to its epimer, myo-inositol.

Full text of DOI:10.1016/j.bmcl.2005.08.112

Bioorganic & Medicinal Chemistry Letters 15 (2005) 5416–5418
Specific sensing between inositol epimers by a bis(boronate)
Charles W. Gray, Jr.,a,b Leland L. Johnson, Jr.,a,b Brian T. Walker,a Mark C. Sleevi,b,*
A. Stewart Campbell,b Robert Plourdeb and Todd A. Houstona,c,
*
aDepartment of Chemistry, Virginia Commonwealth University, Richmond, VA 23284, USA
bInsmed, Incorporated, Glen Allen, VA 23060, USA
cSchool of Science and Eskitis Institute, Griffith University, Nathan, QLD 4111, Australia
Received 12 August 2005; revised 30 August 2005; accepted 31 August 2005
Available online 6 October 2005
Abstract—Bis(boronates) that utilize internal photoinduced electron transfer (PET) quenching mechanisms can specifically signal
the binding of chiro-inositol without responding to its epimer, myo-inositol.
Ó 2005 Elsevier Ltd. All rights reserved.
One important aspect of drug development, especially
for drugs involved in clinical trials, is the monitoring
of metabolism in vivo. Concentrations of the drug in
serum and measurements of the amount of secreted drug
are parameters that can be used to adjust the dosage of
administrated drug and drug delivery formulations so
that the full potential of the drug may be realized. The
compound D-(+)-chiro-inositol (DCI, 1) has been inves-
tigated in clinical trials for two therapeutic applications:
as an oral sensitizer for type-2 diabetes1 and as a first
line therapy for polycystic ovarian syndrome (PCOS).
Currently, serum concentrations and excretion rates
for DCI are monitored using time-consuming HPLC as-
says. A fluorescence assay that is selective for DCI over
its biologically more-prevalent epimer myo-inositol (2)
would significantly increase the speed with which biolog-
ical samples could be screened for DCI, and work to
develop such a method is described here. To date, a
chemosensor that responds specifically to only one of
two cyclitol epimers has not been reported.
same deficiencies were observed for women diagnosed
with PCOS, which has a known relationship to insulin
resistance.4,5 Recently, a putative insulin mediator
(PIM) containing DCI linked through the 4-OH to a
b-galactosamine residue was reported by Insmed and
collaborators at the University of Virginia.6 It appears
to be the first known b-1,4-linked inositol glycan from
animals that displays insulin mimetic properties.
Cyclitols and their phosphates are abundant in nature
and fulfill many different biochemical roles and this
has inspired chemists to create synthetic receptors for
these systems.7,8 For example, in addition to his seminal
work on inositol triphosphate receptors,7 Anslyn and
co-workers have developed receptors with preorganized
hydrogen bonding sites that accommodate simple cycli-
tol guests.9 The chiro-inositols are two of nine possible
inositol isomers and the only chiral isomers, thus the
prefix indicating this.10 Contributing to its C2-symmetry
is the presence of contiguous cis-1,2-diols, a structural
feature that can be exploited by the affinity of boronic
acids for vicinal diols. Boronic acids have become widely
used as carbohydrate sensors due to their ability to ex-
change covalent bonds with polyols in aqueous solu-
tion.11 To date, these have not previously been
incorporated into the design of receptors for cyclitols.
DCI is a naturally occurring cyclitol found in substan-
tial quantities in a number of plants including the pine
tree and soybean plant. It was first cited in 1993 that
there was a strong correlation between insulin resistance
and DCI deficiencies in the urine and serum of diabet-
ics.2 Furthermore, it was found that DCI excretion is
an index marker for insulin sensitivity.3 Similarly, the
Aminoboronates are useful photoinduced electron
transfer (PET) sensors of polyol binding. Upon esterifi-
cation, the boron generally becomes more electrophilic
and interaction with the nitrogen lone pair is strength-
ened.12 This interrupts the PET quenching pathway
and allows increased fluorescence that can be measured
Keywords: Inositol sensing; Boronate receptor; PET signaling.
*
Corresponding authors. Tel.: +61 7 3735 4115; fax: +61 7 3735
6572; e-mail: t.houston@griffith.edu.au
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.08.112

C. W. Gray et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5416–5418
5417
quantitatively as a signal for binding. James et al. first
described the use of this method with a d-aminoboro-
nate coupled to an anthracene core13 and showed that
a bis(boronate) analog of this system displayed nearly
two orders of magnitude greater affinity for glucose.14
Using a similar strategy with the binaphthol receptor
3, they first reported chiral discrimination in the sensing
of D-glucose with a single enantiomer of this bis(boro-
nate).15 Previously, we showed that ShinkaiÕs glucose
receptor 3 also has high affinity for tartaric acid16 and
sought to apply this compound to the fluorescent detec-
tion of DCI.
Use of a bis(boronate) such as 3 for sensing DCI offers
several advantages. First, the contiguous cis-1,2-diols of
DCI should allow it to bind two boronates quite readi-
ly.17 In addition, while myo-inositol contains two
hydroxyl groups that are in a cis relationship to the lone
axial hydroxyl group, binding of a single boronate at
this point of symmetry would leave a trans relationship
among all of the remaining free hydroxyl groups, virtu-
ally precluding formation of a second boronate ester.
This means a bis(d-aminoboronate) receptor would not
report the binding of a single boronate to myo-inositol
due to PET quenching from the amine adjacent to the
free boronate. Finally, initial molecular modeling results
suggested that 3 could bind DCI while allowing both
amines to coordinate their respective boronate esters.18
In total, this would allow for selective detection of
DCI in the presence of myo-inositol1.
Figure 2. Fluorescent measurement of D-glucose, D-chiro-inositol (1),
and myo-inositol (2) affinities for rac-3.16
affinity and fluorescent response of 3 are much larger
toward glucose, samples could be treated enzymatically
a priori to remove undesired sugars from a biological
sample. Thus, the indicated specificity between inositol
epimers is important as the proposed fluorescence
assays have the potential for higher throughput of
biological/clinical samples relative to time-intensive
HPLC methods. What remains is to develop a bis(bor-
onate) with a higher affinity for DCI.
ShinkaiÕs receptor 3 was synthesized in racemic form
using the modification described previously.16 The
fluorescent response of this compound to both DCI
and myo-inositol was monitored and as the results in
Figure 2 demonstrate, rac-3 responds to the presence
of DCI at millimolar concentrations (Ka = ca.
125 MÀ1) while remaining ÔsilentÕ to myo-inositol at
concentrations as high as 1 M. We propose that the
aforementioned inability of myo-inositol to interact
with both boronates of 3 in a 1:1 complex likely ac-
counts for the lack of fluorescent response. While the
Unlike 3, receptors such as 4 and 5 can distinguish be-
tween 1 and 2 selectively, but not specifically. Once a
single cis-diol binds to either of these receptors, PET
quenching will occur. Monovalent ligands such as malic
acid can generate a fluorescent response from 3 at high
millimolar concentrations, but a-hydroxyacids generally
have a higher affinity for boronates than diols do.16 The
interaction of myo-inositol with a single boronic acid
was confirmed with the simplified receptor 4. While it
lacks some of the steric constraints of the bis(boronate),
it maintains the d-aminoboronate PET sensing arrange-
ment. This compound was synthesized from 9-anthra-
ldehyde using consecutive reductive amination
reactions involving benzylamine (90%) and o-form-
ylphenylboronic acid (79%).16 The affinity of 4 for
myo-inositol is larger than that observed for galactose
and mannose, both of which are documented as having
a higher affinity for benzeneboronic acid than glucose
(Table 1).12 In addition, a-galactose has the same con-
tiguous cis-1,2-diol arrangement as DCI. This further
supports the argument that the bis(boronate) 3, while
possessing the ability to form a 1:1 complex with myo-
inositol by esterification of a single boronate, does not
signal this interaction through an increase in fluorescent
intensity since the second boronate cannot bind the
cyclitol and shut down PET from its proximal amine.
Such a feature makes 3, and presumably other
bis(aminoboronates), useful sensors for the detection
of DCI in the presence of myo-inositol even though its
affinity for both boronates in 3 is less than myo-inositolÕs
OH
OH
HO
HO
HO
HO
HO
HO
OH
HO
HO
OH
1
2
B(OH)
2
1
R
N
OMe
OMe
N
B(OH)
2
N
B(OH)
2
1
3
4: R = H
1
5: R = B(OH)
2
Figure 1. Inositol epimers and receptors used herein.

5418
C. W. Gray et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5416–5418
Table 1. Association constants with 4 in 99:1 10 mM aq phosphate
(pH 7.8)/MeOH
9. (a) Huang, C.-Y.; Cabell, L. A.; Lynch, V.; Anslyn, E. V.
J. Am. Chem. Soc. 1992, 114, 1900; (b) Huang, C.-Y.;
Cabell, L. A.; Anslyn, E. V. J. Am. Chem. Soc. 1994, 116,
2778; (c) Diaz, S. G.; Lynch, V.; Anslyn, E. V. J.
Supermol. Chem. 2002, 2, 201.
10. Ostlund, R. E.; Seemayer, R.; Gupta, S.; Kimmel, R.;
Ostlund, E. L.; Sherman, W. R. J. Biol. Chem. 1996, 271,
10073.
Carbohydrate
Ka (102 MÀ1
)
I/I0
myo-Inositol
D-Galactose
D-Mannose
3.5
2.3
1.8
1.5
1.7
1.4
11. (a) Hartley, J. H.; James, T. D.; Ward, C. J. J. Chem. Soc.,
Perkin Trans. 1 2000, 3155; (b) Striegler, S. Curr. Org.
Chem. 2003, 7, 81.
12. Springsteen, G.; Wang, B. Tetrahedron 2002, 58, 5291; An
alternative mechanism for PET quenching has recently
been put forward: Frazen, S.; Ni, W.; Wang, B. J. Phys.
Chem. B 2003, 107, 12942.
affinity for a monoboronate. Here, the mechanism of
signaling takes precedent over raw affinity for particular
guests.19 While further understanding of the exact nat-
ure of the binding event is needed, this is the first report
of epimer specific chemosensing with a synthetic
receptor.20
13. James, T. D.; Sandanayake, K. R. A. S.; Shinkai, S. Chem.
Commun. 1994, 477.
14. (a) James, T. D.; Sandanayake, K. R. A. S.; Iguchi, R.;
Shinkai, S. J. Am. Chem. Soc. 1995, 117, 8982; (b) James,
T. D.; Sandanayake, K. R. A. S.; Shinkai, S. Angew.
Chem., Int. Ed. Engl. 1996, 35, 1910.
15. James, T. D.; Sandanayake, K. R. A. S.; Shinkai, S.
Nature 1995, 374, 345.
Acknowledgment
T.A.H. and C.W.G. thank Insmed, Inc., for funding an
industrial internship for C.W.G. during his doctoral
studies.
16. Gray, C. W., Jr.; Houston, T. A. J. Org. Chem. 2002, 67,
5426; For an in-depth study of the stereochemical basis of
this signaling event, see: Zhao, J.; Fyles, T. M.; James, T.
D. Angew. Chem., Int. Ed. 2004, 43, 3461.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2005.08.112.
17. Triboronate esters of chiro-inositol can form in the gas
phase (Wiecko, J.; Sherman, W. R. J. Am. Chem. Soc.
1979, 101, 979) but such a species is not favored in
aqueous solution. In water, DCI (1) must exist predom-
inantly in the conformation represented in Figure 1 as
1
References and notes
the H NMR of 1 in D2O has a downfield resonance of
minimal splitting corresponding to the two equatorial
methine protons (d 3.86 ppm) and more complex splitting
patterns at 3.4 and 3.6 ppm for the four axial protons.
We have not been unable to obtain quantitative binding
data of DCI with boronates using 1H NMR chemical
shifts.
1. For a review of the chemistry and biochemistry involved
in type II diabetes, see: (a) Ross, S. A.; Gulve, E. A.;
Wang, M. Chem. Rev. 2004, 104, 1255; For other
pharmaceutical targets within inositol signaling cascades,
see: (b) Prestwich, G. D. Chem. Biol. 2004, 11, 619.
2. Asplin, I.; Galasko, G.; Larner, J. Proc. Natl. Acad. Sci.
U.S.A. 1993, 90, 5924.
3. Suzuki, S.; Kawasaki, H.; Satoh, Y.; Ohtomo, M.; Hirai,
M.; Hirai, A.; Hirai, S.; Onoda, M.; Matsumoto, M.;
Hinokio, Y.; Akai, H.; Craig, J.; Larner, J.; Toyota, T.
Diabetes Care 1994, 17, 1465.
4. Nestler, J. E.; Jakubowicz, D. J.; Reamer, P.; Gunn, R.
D.; Allan, G. New Engl. J. Med. 1999, 340, 1314.
5. Larner, J. Endocr. J. 1994, 323, 167.
6. Larner, J.; Price, J. D.; Heimark, D.; Smith, L.; Rule, G.;
Piccariello, T.; Fonteles, M. C.; Pontes, C.; Vale, D.;
Huang, L. J. Med. Chem. 2003, 46, 3283.
18. Predictions based on molecular mechanics minimizations
(Merck Molecular Force Field) using Spartan (Wavefunc-
tion, Inc.) with the dative B–N bonds defined as covalent
bonds.
19. Methyl glucosides have previously shown to be ÔsilentÕ
toward glucose-binding bis(boronates) by a mechanism
similar to what is occurring here (i.e., –methyl glucosides
cannot bind two boronates at once although it presumably
also has lower affinity for the receptor as well, unlike the
present case): Takeuchi, M.; Yoda, S.; Imada, T.; Shinkai,
S. Tetrahedron 1997, 53, 8335; Compounds such as 4 are
specific sensors for glucose over glucosamine: Cooper, C.
R.; James, T. D. Chem. Commun. 1997, 1419.
7. (a) Niikura, K.; Metzger, A.; Anslyn, E. V. J. Am. Chem.
Soc. 1998, 120, 8533; (b) Niikura, K.; Anslyn, E. V. J.
Chem. Soc., Perkin Trans. 2 1999, 2769; (c) Niikura, K.;
Anslyn, E. V. J. Org. Chem. 2003, 68, 10156.
20. We define specific sensing as signal response to only one
epimer and selective sensing as differential responses to
each epimer based on enhanced binding affinity of one
epimer over the other. For an epimer selective receptor,
´
`
8. (a) Morey, J.; Orell, M.; Barcelo, M. A.; Deya, P. M.;
Costa, A.; Ballester, P. Tetrahedron Lett. 2004, 45, 1261;
(b) Tanimura, A.; Nezu, A.; Morita, T.; Turner, R. J.;
Tojyo, Y. J. Biol. Chem. 2004, 279, 38095.
´
`
see: Ballester, P.; Capo, M.; Costa, A.; Deya, P. M.;
Gornila, R.; Decken, A.; Deslongchamps, G. Org. Lett.
2001, 3, 267.