Molecular evolution using intramolecular acyl migration on myo-inositol benzoates with thermodyamic and kinetic selectors

Article abstract of DOI:10.1002/chem.200400234

A molecular evolution model was successfully demonstrated by combining the intramolecular acyl migration on inositol tribenzoates and boron selectors. The addition of boric acid to 12 members of DCL (dynamic combinatorial library) induced the dramatic amp

Full text of DOI:10.1002/chem.200400234

FULL PAPER
Molecular Evolution Using Intramolecular Acyl Migration on myo-Inositol
Benzoates with Thermodynamic and Kinetic Selectors
Young-Hoon Ahn and Young-Tae Chang*[a]
Abstract: A molecular evolution model
was successfully demonstrated by com-
bining the intramolecular acyl migra-
tion on inositol tribenzoates and boron
selectors. The addition of boric acid to
12 members of DCL (dynamic combi-
natorial library) induced the dramatic
amplification of myo-I(2,4,6)Bz3 (1)
with up to 94% under thermodynamic
(see Figure 1c) control while a portion
of phenyl boronic acid caused two sig-
nificant different distributions: under
kinetic control, the pre-equilibrium of
DCL shifted to induce the exclusive
amplification of 1,4,6-tribenzoyl myo-
inositol (7) with decrease of other
members up to 82% from the mixture
(see Figure 2b), and changed gradually
to form 2,4,6-tribenzoyl myo-inositol
(1) with up to 96% under thermody-
namic control (Figure 2c).
Keywords: boronic acid selector ¥
dynamic combinatorial library
intramolecular acyl migration
¥
¥
¥
molecular
amplification
molecular evolution
Introduction
bers.[2,4,6] Scaffold diversification is necessary for a highly ef-
ficient and practical system. One of the challenges in this
field is a successful carbohydrate dynamic library due to its
complexity. There have not been many attempts to generate
dynamic combinatorial libraries (DCL) on a carbohydrate
scaffold, and the selection method was not effective in
giving high yields or selective amplification.[6c,9d] Herein, we
report a highly efficient molecular evolution system utilizing
intramolecular acyl migration on a carbohydrate scaffold
coupled with boric/boronic acid as a selector.
Molecular evolution incorporating a dynamic combinatorial
library (DCL) has emerged as an approach for the identifi-
cation and preparation of new host guest systems.[1] A DCL
is composed of members in a reversible equilibrium in
which all library members are interconverting, thus provid-
ing constant distribution under thermodynamic control.
Consequently, a molecular recognition event in the DCL en-
ables the stabilization of the fittest member that shifts the
equilibrium toward amplification of the selected member
with a consequent decrease in the proportion of poor bind-
ers. It requires two key steps for success: 1) continuous
DCL generation and 2) effective selection. To date, a
number of DCLs have been prepared using such diverse
chemical reactions as intermolecular transesterification,[2]
enzyme catalyzed peptide-bond exchange,[3] imine bond ex-
change of hydrazones or oximes,[4] olefin metathesis,[5] disul-
fide bond exchange,[6] photoisomerization,[7] hydrogen-bond
exchange,[8] and metal ligand coordination.[9] While some
DCLs have successfully demonstrated molecular evolution×s
proof of concept, there is still a limited number of scaffolds.
In fact, most of the efficient DCLs form macrocyclic mem-
Results and Discussion
Our carbohydrate scaffold is hexahydroxyl cyclohexane (in-
ositol), and we have chosen the most naturally abundant in-
ositol isomer, myo-inositol, as the model compound. The
polyhydroxyl nature of inositol allows for diverse regioisom-
ers depending upon the position of the attached functional
group. Under basic conditions, acyl functionalities on an ino-
sitol ring are able to undergo intramolecular migration and
generate various regioisomers, while also giving continuous
interconversion.[10]
One of the inositol benzoate derivatives, 1,4,5-tribenzoyl
myo-inositol [4, (myo-I(1,4,5)Bz3, Figure 1a], was synthe-
sized (for synthetic procedure, see Supporting Information).
The optimized migration condition was sought that would
give the least amount of side products, fast generation of all
members within a reasonable time, and also with applicabili-
ty to the selection system. A series of bases were investigat-
ed including pyridine, DMAP (4-dimethyl-aminopyridine),
[a] Y.-H. Ahn, Prof. Y.-T. Chang
Department of Chemistry
New York University
New York, NY 10003 (USA)
Fax : (+1) 212-995-4203
E-mail: yt.chang@nyu.edu
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2004, 10, 3543 3547
DOI: 10.1002/chem.200400234
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3543

FULL PAPER
DABCO (1,4-diazabicyclo[2.2.2]octane), DIEA (N,N-diiso-
propylethlyamine), BEMP (2-tert-butylimino-2-diethylami-
no-1,3-dimethyl-perhydro-1,3,2-diazaphosphorine) on poly-
styrene, DBU (1,8-diazabicyclo[5.4.0]undec-7-ene) and
NaOH in aprotic solvents with and without water. We found
that increasing amounts of water accelerated benzoyl migra-
tion even with mild bases due possibly to the stabilization of
the tetrahedral intermediate, but it also accelerated benzoyl
group hydrolysis as a side reaction. With the optimized reac-
tion condition (isomer 4 with DBU in anhydrous acetoni-
trile), we generated all 12 of the expected and possible re-
gioisomers within the DCL and without significant side re-
actions. The DCL was analyzed by LC-MS, which allowed
for the identification of 12 peaks in the chromatogram (Fig-
ure 1b); the structure of each isomer was assigned by NMR
by using the literature (Scheme 1).[11]
It is well known that boronic acids are able to bind with
carbohydrates to form five- or six-membered cyclic esters, as
used in carbohydrate recognition.[12] Their binding prefer-
ence with 1,2 or 1,3-cis diols over trans diols in sugar sys-
tems was appropriate for selection amongst a carbohydrate
polyol system. It was also reported that the binding affinity
of boronic acid with a sugar depends on the pH of the
media.[12a] At pH values lower than the pKa of boronic acid
(acidic conditions), coupling between the boronic acid and a
sugar is disfavored, while the coupling is favored at pH
values above the pKa of boronic acid (basic conditions). In
our study, boric acid and phenyl boronic acid were investi-
gated for selective binding and amplification.
Figure 1. Amplification chromatogram using B(OH)3 a) HPLC chromato-
gram (C18 column: 4.6î150 mm, eluted with a solution of water, acetoni-
trile, and methanol in a ratio of 52:18:30) of I(1,4,5)Bz3 (4), b) DCL at
1.5 h after addition of DBU, and c) 72 h after addition of boric acid
(94% of 1).
more available hydroxyl groups. This leads to the isolation
of the product from the equilibrium and gives the amplifica-
tion of 1 (Scheme 1). Once at the final distribution, it re-
mained unchanged over time (Figure 4). This implies that
the selection is driven by thermodynamic control. It has
been also shown that direct migration from isomer 4 to
isomer 1 by adding base and boric acid together without
pre-equilibrium resulted in the same final distribution (94%
of isomer 1) (see Figure 5), which again supports the ther-
modynamic control mechanism.[4a]
Interestingly, when phenyl boronic acid was added to the
equilibrium mixture, two significant different stepwise distri-
butions were observed. At first, the pre-equilibrium of DCL
was shifted to induce an exclusive amplification of another
isomer, myo-I(1,4,6)Bz3 (7) with up to 82% (Figure 2b) as
It was observed that the addition of boric acid to the
DCL significantly affects the distribution of the DCL by in-
ducing the amplification of myo-I(1,4,6)Bz3 (1) with up to
approximately 94% (Figure 1c). As the original amount of
isomer 1 was about 4% in the pre-equilibrium of the DCL,
the enrichment factor is a high 23.5. It was envisioned that
boric acid couples with the 1,3,5-trihydroxyl group of myo-
inositol in a ring-flipped conformation where there are no
Scheme 1. Library members generated by intramolecular acyl migration and two types of selection using boric acid and phenyl boronic acid.
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org Chem. Eur. J. 2004, 10, 3543 3547
3544

Molecular Evolution
3543 3547
Figure 5. Direct migration to 1 using boric acid without pre-equilibrium.
Reaction conditions: I(1,4,5)Bz3 (4) (1mg), boric acid (2 mg, 15 equiv) in
acetonitrile (1mL), and then DBU (20 mL, 66 equiv) where t=0 at that
~
*
point, : I(2,4,6)Bz3 (1), : I(1,4,6)Bz3 (7).
the kinetic product, and changed gradually to form the ther-
modynamic isomer (1) up to 96% (Figure 2c). It is envi-
sioned that phenyl boronic acid coupled quickly (kinetically)
with the 1,2-cis diols of the DCL members, whenever availa-
ble during the acyl migration, to form a stable five-mem-
bered ring. Even though there are four possible regioisom-
ers with 1,2-cis coupling (compound 4, 7, 10, and 12), the 5-
benzoyl group seemed to have the most severe steric hin-
drance against phenyl ring of 1,2-cis-boronic ester, and thus
isomer 7, myo-I(1,4,6)Bz3, was kinetically favored and accu-
mulated (Scheme 1). When the reaction proceeds longer,
phenyl boronic acid is able to accept three hydroxyl groups
to form an even more stable complex giving the thermody-
namic isomer 1, as in the case of the boric acid selector
(Figure 6). Also, we investigated the direct migration from
myo-I(1,4,5)Bz3 (4) without pre-equilibrium, in which the
acyl migration occurred in the presence of phenyl boronic
acid. Absolute amplification of the kinetic product, 7, with
up to 98%, was obtained initially (see Figure 3b), with slow
transformation to the thermodynamic product 1 over time
(Figures 3c and 7).
Figure 2. Amplification chromatogram using PhB(OH)2 a) chromatogram
of a 12 member DCL in pre-equilibrium by DBU, b) 5 h after the addi-
tion of PhB(OH)2 (82% of 7), c) 205 h after the addition of PhB(OH)2
(96% of 1).
The practical use of this concept beyond the analytical
scale was performed on a 30 mg scale, purifying the thermo-
dynamic and kinetic products from the reaction mixture, re-
spectively. The purification of the thermodynamic product
Figure 3. Direct migration chromatogram using PhB(OH)2 a) The HPLC
chromatogram of I(1,4,5)Bz3 (4), b) 25 min after addition of DBU and
PhB(OH)2 (98% of 7), c) after an additional 124 h (96% of 1).
Figure 6. Amplification of 1 and 7 using phenyl boronic acid after pre-
equilibrium. Reaction conditions: I(1,4,5)Bz3 (4) (1mg), DBU (20 mL,
66 equiv) in acetonitrile (20 mL) for 1.5 h. Phenyl boronic acid (3.7 mg,
15 equiv) was added and then concentrated to 1 mL where t=0 at that
point. Additional DBU (10 mL, 33 equiv) was added at 55 h to accelerate
Figure 4. Amplification of 1 using boric acid after pre-equilibrium. Reac-
tion conditions: I(1,4,5)Bz3 (4) (1mg), DBU (20 mL, 66 equiv) in acetoni-
trile (20 mL) for 1.5 h, and then boric acid (2 mg, 15 equiv) where t=0 at
~
~
*
*
that point, : I(2,4,6)Bz3 (1), : I(1,4,6)Bz3 (7).
the equilibrium, : I(2,4,6)Bz3 (1), : I(1,4,6)Bz3 (7).
Chem. Eur. J. 2004, 10, 3543 3547 www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3545

Y.-T. Chang and Y.-H. Ahn
FULL PAPER
ture to a solution of 4 (1mg, 2 mmol) in acetonitrile (1mL). The reaction
mixture was analyzed by HPLC-MS.
Preparation of the thermodynamic product (1): DBU (600 mL, 3.9 mmol)
was added to
a solution of 4 (30 mg, 0.061mmol) in acetonitrile
(600 mL). After the reaction mixture was stirred for 1.5 h, boric acid
(60 mg, 0.96 mmol) in DMSO (0.5 mL) was added to the reaction mix-
ture. To accelerate the rate of amplification, the reaction mixture was
concentrated to about 60 mL, and stirred for 2 d. The reaction mixture
was quenched by acetic acid (1mL), diluted with ethyl acetate, washed
with 1n HCl, aq. NaHCO3, and brine. The organic layer was dried over
MgSO4, concentrated, and purified by chromatography on silica gel
(EtOAc/Hex 1:1) to give 1 (25 mg, 83%). Rf = 0.30 (EtOAc/Hex 1:1);
1H NMR (CD3OD): d
= 3.89 (t, J=9.5 Hz, 1H), 4.00 (dd, J=2.9,
10.0 Hz, 2H), 5.57 (t, J=9.7 Hz, 2H), 5.77 (t, J=2.8 Hz, 1H), 7.26 8.15
Figure 7. Direct migration to 7 and 1 using phenyl boronic acid without
pre-equilibrium. Reaction conditions: I(1,4,5)Bz3 (4) (1mg), phenyl bor-
onic acid (3.7 mg, 15 equiv) in acetonitrile (1 mL), and then DBU (20 mL,
(m, 15H); LC-MS: m/z: 493 [M+H]+, 475 [MÀH2O+H]+.
Preparation of the kinetic product (7): Phenyl boronic acid (113 mg,
0.96 mmol), and DBU (600 mL, 3.9 mmol) was added to a solution of 4
(30 mg, 0.061mmol) in acetonitrile (30 mL). After the solution was stir-
red for 1h, the reaction mixture was quenched by acetic acid (1mL), and
diluted with ethyl acetate, and washed with 1n HCl, aq. NaHCO3, and
brine solution. The resulting organic layer was dried over MgSO4, con-
centrated, and purified by chromatography on silca gel (EtOAc/Hex 1:2)
~
*
66 equiv) where t=0 at that point, : I(2,4,6)Bz3 (1), : I(1,4,6)Bz3 (7).
(1) on silica gel after its amplification of up to 94% afforded
the 83% recovery of 1. The isolation of kinetic product (7)
that is amplified with up to 98% from the direct migration
without pre-equilibrium led to the 92% recovery of 7.
to give
7 (27.5 mg, 92%). Rf =
0.37 (EtOAc/Hex 1:1); 1H NMR
(CD3OD): d = 4.04 (dd, J=2.4, 10.1 Hz, 1H), 4.34 (t, J=2.3 Hz, 1H),
4.48 (t, J=10.0 Hz, 1H), 5.13 (dd, J=2.4, 10.0 Hz, 1H), 5.49 (dd, J=9.3,
10.0 Hz, 1H), 5.80 (t, J=10.0 Hz, 1H), 7.32 8.11 (m, 15H); LC-MS: m/z:
493 [M+H]+.
Conclusion
We demonstrated highly efficient molecular evolution on a
myo-inositol model by combining intramolecular acyl migra-
tion and both kinetic and thermodynamic selectors. We be-
lieve that intramolecular acyl migration is applicable to an
even large number of dynamic carbohydrate libraries, and if
coupled with a broad range of selectors, this method will
open a wide range of applications.
Acknowledgement
This work was supported by the Petroleum Research Fund from the
American Chemical Society.
[1] a) S. Otto, R. L. E. Furlan, J. K. M. Sanders, Curr. Opin. Chem. Biol.
2002, 6, 321 327; b) S. J. Rowan, S. J. Cantrill, G. R. L. Cousins,
J. K. M. Sanders, J. F. Stoddart, Angew. Chem. 2002, 114, 938 993;
Angew. Chem. Int. Ed. 2002, 41, 898 952; c) J. M. Lehn, A. V. Eli-
seev, Science 2001, 291, 2331 2332; d) C. Karan, B. L. Miller, Drug
Discovery Today 2000, 5, 67 75.
[2] a) P. A. Brady, R. P. Bonar-Law, S. J. Rowan, C. J. Suckling, J. K. M.
Sanders, Chem. Commun. 1996, 319 320; b) P. A. Brady, J. K. M.
Sanders, J. Chem. Soc. Perkin Trans. 1 1997, 3237 3253; c) G.
Kaiser, J. K. M. Sanders, Chem. Commun. 2000, 1763 1764.
[3] R. J. Lins, S. L. Flitsch, N. J. Turner, E. Irving, S. A. Brown, Angew.
Chem. 2002, 114, 3555 3557; Angew. Chem. Int. Ed. 2002, 41, 3405
3407.
[4] a) S. L. Roberts, R. L. E. Furlan, S. Otto, J. K. M. Sanders, Org.
Biomol. Chem. 2003, 1, 1625 1633; b) R. L. E. Furlan, Y. F. Ng, S.
Otto, J. K. M. Sanders, J. Am. Chem. Soc. 2001, 123, 8876 8877;
c) G. R. L. Cousins, R. L. E. Furlan, Y. F. Ng, J. E. Redman, J. K. M.
Sanders, Angew. Chem. 2001, 113, 437 442; Angew. Chem. Int. Ed.
2001, 40, 423 428; d) A. Star, I. Goldberg. B. Fuchs, Angew. Chem.
2000, 112, 2797 2801; Angew. Chem. Int. Ed. 2000, 39, 2685 2689;
e) O. Storm, U. L¸ning, Chem. Eur. J. 2002, 8, 793 798.
[5] T. Giger, M. Wigger, S. Audetat, S. A. Benner, Synlett 1998, 688
692.
[6] a) S. Otto, R. L. E. Furlan, J. K. M. Sanders, Science 2002, 297, 590
593; b) S. Otto, R. L. E. Rurlan, J. K. M. Sanders, J. Am. Chem. Soc.
2000, 122, 12063 12064; c) O. Ramstrˆm, J. M. Lehn, ChemBio-
Chem 2000, 1, 41 48.
[7] a) A. V. Eliseev, M. I. Nelen, J. Am. Chem. Soc. 1997, 119, 1147
1148; b) A. V. Eliseev, M. I. Nelen, Chem. Eur. J. 1998, 4, 825 834.
[8] a) M. Crego-Calama, P. Timmerman, D. N. Reinhoudt, Angew.
Chem. 2000, 112, 771 774; Angew. Chem. Int. Ed. 2000, 39, 755
758; b) F. Hof, C. Nuckolls, J. Rebek, J. Am. Chem. Soc. 2000, 122,
4251 4252.
Experimental Section
Materials and general methods: All reactions were performed in oven-
dried glassware under positive nitrogen pressure. Unless otherwise noted,
starting materials and solvents were purchased from Aldrich and Acros
organics and used without purification. Analytical TLC was carried out
on Merck 60 F254 silica gel plate (0.25 mm layer thickness) and visualiza-
tion was done with UV light, and/or by spaying with a 5% solution of
phosphomolybdic acid followed by charring with a heat gun. Column
chromatography was performed on Merck 60 silica gel (230 400 mesh).
1H NMR (200 MHz) spectra were determined on Varian Genini 200 spec-
trometer. Chemical shifts were reported in parts per million (ppm) rela-
tive to internal standard as tetramethylsilane, and coupling constants (J)
are in Hertz (Hz). All compounds were identified by LC-MS (Agilent
Technologies) using a C18 column (4.6î150 mm) with 65 minutes of elu-
tion time using a solution of H2O/CH3CN/MeOH in a ratio of 52:18:30
(containing 0.1% acetic acid) with a UV detector at l=250 and 230 nm
and an electrospray ionization source.
Molecular amplification experiment with pre-equilibrium: DBU (20 mL,
0.13 mmol) was added at room temperature to a solution of 4 (1mg,
2.0 mmol) in acetonitrile (20 mL). After the reaction mixture was stirred
for 1.5 h, the full migration of 12 isomers was confirmed by HPLC-MS. A
solution of boric acid (2 mg, 0.032 mmol) in DMSO (50 mL) or phenyl
boronic acid (3.75 mg, 0.032 mmol) was added to the reaction mixture.
The change of the distribution was analyzed by HPLC-MS.[13]
Molecular amplification experiment without pre-equilibrium: Boric acid
(2 mg, 0.032 mmol) in DMSO (50 mL) or phenyl boronic acid (3.75 mg,
0.032 mmol), and DBU (20 mL, 0.13 mmol) were added at room tempera-
[9] a) S. Choudhary, J. R. Morrow, Angew. Chem. 2002, 114, 4270 4272;
Angew. Chem. Int. Ed. 2002, 41, 4096 4098; b) E. Stulz, Y. F. Ng,
3546
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 3543 3547

Molecular Evolution
3543 3547
S. M. Scott, J. K. M. Sanders, Chem. Commun. 2002, 524 525; c) M.
Ziegler, J. J. Miranda, U. N. Andersen, D. W. Johnson, J. A. Leary,
K. N. Raymond, Angew. Chem. 2001, 113, 755 758; Angew. Chem.
Int. Ed. 2001, 40, 733 736; d) S. Sakai, Y. Shigemasa, T. Sasaki, Tet-
rahedron Lett. 1997, 38, 8145 8148; e) C. Karan, B. L. Miller, J. Am.
Chem. Soc. 2001, 123, 7455 7456; f) M. A. Case, G. L. McLendon, J.
Am. Chem. Soc. 2000, 122, 8089 8090; g) V. Goral, M. I. Nelen,
A. V. Eliseev, J. M. Lehn, Proc. Natl. Acad. Sci. USA 2001, 98,
1347 1352.
Chem. Soc. 1979, 101, 979 983; c) H. Eggert, J. Frederiksen, C.
Morin, J. C. Norrild, J. Org. Chem. 1999, 64, 3846 3852; d) W.
Yang, H. He, D. G. Drueckhammer, Angew. Chem. 2001, 113, 1764
1768; Angew. Chem. Int. Ed. 2001, 40, 1714 1718; e) T. D. James, S.
Sandanayake, R. Iguchi, S. Shinkai, J. Am. Chem. Soc. 1995, 117,
8982 8987.
[13] The concentrated reaction mixture led to intermolecular benzoyl mi-
gration as a side reaction, which could be suppressed by dilution
(20 times of direct migration). However, in the case of phenyl bor-
onic acid, the reaction mixture was concentrated to 1mL in order to
accelerate the rate of the amplification since it was found that the
addition of boric acid or phenyl boronic acid stopped intermolecular
benzoyl migration from going further.
[10] S. K. Chung, Y. T. Chang, J. Chem. Soc. Chem. Commun. 1995, 1 3
14.
[11] S. K. Chung, Y. T. Chang, K. H. Sohn, J. Chem. Soc. Chem.
Commun. 1996, 163 164; b) S. K. Chung, Y. T. Chang, Y. Ryu, Pure
Appl. Chem. 1996, 68, 931 935.
[12] a) T. D. James, S. Sandanayake, S. Shinkai, Angew. Chem. Int. Ed.
Engl. 1996, 35, 1910 1922; b) J. Wiecko. W. R. Sherman, J. Am.
Received: March 11, 2004
Published online: May 27, 2004
Chem. Eur. J. 2004, 10, 3543 3547
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3547