Highly selective disaccharide recognition by a tricyclic octaamide cage

Article abstract of DOI:10.1002/1521-3773(20021104)41:21<4093::AID-ANIE4093>3.0.CO;2-I

Selective oligosaccharide recognition remains a challenge for supramolecular chemistry. Tricyclic receptor 1 was designed to bind the all-equatorial β-cellobiosyl unit in 2. Remarkably, it succeeds while showing no detectable affinity for five other mono-

Full text of DOI:10.1002/1521-3773(20021104)41:21<4093::AID-ANIE4093>3.0.CO;2-I

COMMUNICATIONS
Morin, Jr., E. J. Salaski, J. Am. Chem. Soc. 1985, 107, 8066.
[19] We thank Professor Larry Overman for comparing NMR spectra of
(ꢁ )-1 with those of chiral 1.[4]
OH
H
N
N H
H
H
O
HO
HO
X
O
X
O
N
N
OC8H17
[20] E. J. Corey, A. Venkateswarlu, J. Am. Chem. Soc. 1972, 94, 6190.
[21] CCDC 185370 contains the supplementary crystallographic data for
compound 19. These data can be obtained free of charge via
www.ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge
Crystallographic Data Centre, 12, Union Road, Cambridge CB21EZ,
UK; fax: (þ 44)1223-336-033; or deposit@ccdc.cam.ac.uk).
[22] A. M. Felix, E. P. Heimer, T. J. Lambros, C. Tzougraki, J. Meienhofer,
J. Org. Chem. 1978, 43, 4194.
HO
O
O
O
O
2
O
O
O
O
OH
NH
O
HN
O
O
X
X
HO
HO
N
N
HO
H
H
OC8H17
3
[23] a) G. Lakshmaiah, T. Kawabata, M. Shang, K. Fuji, J. Org. Chem.
1999, 64, 1699; b) R. F. Borch, A. Hassid, J. Org. Chem. 1972, 37, 1673.
X = OC5H11 or NHC(CH2OBn)3
1
H
N
N
H
H
H
Highly Selective Disaccharide Recognition by a
Tricyclic Octaamide Cage**
O
OC5H11
C5H11
O
N
N
O
O
O
O
O
Grÿgory Lecollinet, Andrew P. Dominey,
Trinidad Velasco, and Anthony P. Davis*
O
O
C5H11
O
O
NH
O
HN
O
O
OC5H11
N
N
H
H
Carbohydrate recognition presents a continuing challenge
to supramolecular chemistry,[1] fuelled by the growing aware-
ness of saccharide structures as mediators of biological
events.[2] However, while the biological interest is focussed
largely on oligosaccharides, supramolecular chemists have
concentrated mainly on monosaccharide substrates. A num-
4
NHR
NHR
ber of systems show preferential binding of di- versus
e,3]
monosaccharides,[1c
but only a few boron-based receptors
O
O
O
B
B
I
show good selectivity between disaccharides.[1c,3a 3d] The latter
a
ꢀ
employ covalent B O bond formation, and are thus less
+
relevant to biological carbohydrate recognition.
NHBoc
NHBoc
We recently described the tricyclic cage receptors 1, which
bind monosaccharide derivatives strongly and selectively
even in the presence of a hydroxylic cosolvent.[4] Receptor 1
proved remarkably selective for the all-equatorial b-glucoside
2 as against the a-anomer 3. We now report the synthesis and
binding properties of 4, an ™extended analogue∫ of 1 designed
to accommodate disaccharide substrates. Compound 4 has
proved to be the first receptor capable of distinguishing
clearly between disaccharides through noncovalent interac-
tions.
O
NHR
NHR
R = Boc
R = H
b
5
O
O
C6F5O
C6F5O
CO2C5H11
Receptor 4 was synthesized by two independent routes. In
the first, the tetraaminoterphenyl 5 was prepared as shown in
Scheme 1 and coupled under high dilution with bis(penta-
6
c
[*] Prof. A. P. Davis, Dr. G. Lecollinet, T. Velasco
School of Chemistry
HN
HN
O
University of Bristol
Cantock©s Close
Bristol BS8 1TS (UK)
C5H11O2C
HN
CO2C5H11
O
4
+
NH
O
Fax : (þ 44)117-929-8611
O
HN
O
O
E-mail: Anthony.Davis@bristol.ac.uk
HN
O
O
Dr. A. P. Dominey
Department of Chemistry
Trinity College, Dublin 2 (Ireland)
OC5H11
C5H11O
O
O
NH
HN
[**] Financial support for this work was provided by the European
Commission (TMR contract ERB-FMRX-CT98-0231) and Enterprise
Ireland.
7
Scheme 1. Synthesis of 4 þ 7: a) [PdCl2(dppf)], Na2CO3 aq. (2m), DMF,
808C; b) trifluoroacetic acid (TFA), CH2Cl2; c) 6, iPr2NEt, THF/DMF, high
dilution. dppf ¼ diphenylphosphanylferrocene.
Supporting information for this article is available on the WWW under
http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2002, 41, No. 21
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4121-4093 $ 20.00+.50/0
4093

COMMUNICATIONS
binding studies on 1[4]). As hop-
ed, we observed changes in the
1H NMR spectrum of just one of
the macrotricycles (later identi-
fied as 4). However, instead of
signal movements with minor
splittings (as observed for 1 þ
NHCbz
NHCbz
Br
Br
I
b,c
a
+
B
O
O
NHBoc
NHBoc
NHBoc
NHBoc
2
[4a]), the spectrum of 4 was
replaced by a new set of reso-
nance signals implying loss of all
symmetry (see Figure S1 in the
Supporting Information). Titra-
tion of 11 into pure 4 confirmed
this result, revealing progressive
loss of receptor and growth of
the new spectrum (Figure 1), and
clearly implying complex forma-
tion with slow exchange. Inte-
gration of bound versus uncom-
plexed receptor proved unwork-
able, but the growth of the new
signals could be monitored by
integration versus an internal
standard. The data was consis-
tent with 1:1 binding (Figure 2)
with a binding constant (Ka) of
NHBoc
NHBoc
I
9
8
NHCbz
NHCbz
CO2C5H11
H
N
H
N
BocHN
BocHN
NHBoc
f,e
O
O
O
O
4
N
H
N
H
NHBoc
10
CO2C5H11
Scheme 2. Synthesis of 4: a) [PdCl2(dppf)], Na2CO3 aq. (2m), DMF, 808C; b) bis(pinacolato)diboron,
[PdCl2(dppf)], KOAc, DMF, 808C; c) 8, [PdCl2(dppf)], Na2CO3 aq. (2m), DMF, 808C; d) H2, 10% Pd/C,
CH3OH/CH2Cl2; e) 6, iPr2NEt, THF/DMF, high dilution; f) TFA, CH2Cl2.
fluorophenyl) ester 6. This direct route seemed viable as, for
steric and geometric reasons, no [1þ1] cyclization is possible
between 5 and 6. The smallest unstrained rings (favored at
high dilution) result from [2þ2] cyclizations, as required for 4.
However, a regioisomeric macrotricycle 7 is also possible, and
two cyclized products were indeed formed. Mass spectromet-
ric and NMR spectroscopic data were consistent with a
mixture of 4 þ 7, but unfortunately the two could not be
separated. Pure 4 was therefore prepared by a longer but
unambiguous sequence (Scheme 2) in which orthogonal
protection in tetracarbamate 9 allowed sequential [2þ2]
cyclizations, leading specifically to 4 via 10.
In common with receptor 1, macrotricycle 4 possesses
extended, parallel apolar surfaces linked through spacers
containing hydrogen-bond donor and acceptor groups. This
geometry should be compatible with equatorially substituted
carbohydrate derivatives.[4a] Accordingly, the all-equatorial n-
octyl-b-d-cellobioside 11[5,6] was employed for initial binding
studies. In the first instance, 11 was added to the mixure of 4
and 7 in CDCl3/CD3OH (92:8) (the solvent used for NMR
OH
O
OH
O
HO
HO
OH
O
OH
O
HO
HO
O
O
OC8H17
OC8H17
HO
HO
OH
OH
OH
OH
11
12
OH
O
OH
O
HO
HO
HO
HO
OH
O
OH
HO
HO
O
O
O
OC8H17
HO
HO
Figure 1. Partial 1H NMR spectra in CDCl3/CD3OH (92:8) of receptor 4
after addition of a) 0, b) 0.4, c) 1, and d) 5 equivalents of 11.
HO
OH
OC12H25
14
13
4094
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4121-4094 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 21

COMMUNICATIONS
Figures S3 and S4 in the Supporting Information). Even under
these conditions, addition of lactoside 12 or b-maltoside 13
caused no change in emission (see Figure S5 in the Supporting
Information). Second, weak but unambiguous induced circu-
lar dichroism was observed when 11 was added to 4 in CHCl3/
MeOH (92:8) (see Figure S6 in the Supporting Information).
The spectra were too noisy for quantitative analysis, but
saturation was clearly observed with increasing [11]. Once
again, negative results were obtained with 12 or 13, the
receptor remaining CD-silent.
In conclusion, these experiments seem to reveal an
interaction of unusual specificity between a synthetic receptor
4 and a biomolecular substrate 11. We cannot be absolutely
certain that 4 does not bind the other glycosides, as complex
formation might not induce a change in spectroscopic proper-
ties in these cases. However, the complete absence of any
signal from any other substrate, on employing three quite
different techniques, provides strong evidence for extreme
selectivity. The substrates, moreover, are only subtly different
from each other. For example, 11 and 12 are related by
inversion at a single asymmetric center, at one end of the
structure. Finally, the design strategy seems to have been
validated. The geometry of 4 does indeed favor the ™flat∫ all-
equatorial substrate, supporting the notion that selective
receptors for at least some complex biomolecules are
accessible through rational design and synthesis.
Figure 2. Experimental ( ) and calculated ( ) values for the 1H NMR
binding study of 4 þ 11 in CDCl3/CD3OH (92:8). [4] ¼ 0.5 mm, [11] ¼ 0.13
3.6 mm. T¼ 278 K. Experimental points were obtained by integration of the
signal at d ¼ 8.8 8.9 (see Figure 1) versus an internal standard (penta-
&
^
fluorobenzaldehyde, d ¼ 10.27). The data were analyzed by using
a
nonlinear least-squares curve-fitting program implemented within Excel
2000.
7000mꢀ1, slightly higher than the value of 980mꢀ1 measured
for 1 þ 2 under these conditions.[4a] In a control experiment,
glycoside 11 was added to macrocycle 10 under the same
conditions; as expected, no change was observed in the
spectrum of 10.To investigate the selectivity of 4, similar
experiments were performed with octyl glucosides 2 and 3,
octyl b-d-lactoside 12,[6,7] octyl b-d-maltoside 13,[6,8] and
dodecyl a-d-maltoside 14 as substrates. Remarkably, none
of these compounds produced any detectable effect on the
NMR spectrum of the receptor. Moreover, addition of 12 to
(4 þ 11) had no effect on the spectrum of the complex. The
implied selectivity was confirmed by two further techniques.
First, irradiation of 4 at l ¼ 285 nm caused fluorescence
emission in the ranges 350 450 nm and 650 750 nm. Addition
of 11 caused substantial increases in the intensities of both
bands (see Figure S2 in the Supporting Information). Analysis
of the data supported a 1:1 binding model, with Ka ¼ 2500mꢀ1
(Figure 3).[9] In contrast, experiments with 2, 3, and 12 14
showed no sign of complex formation. In each case, addition
of 10 equivalents of glycoside (to 0.3 mm) produced no effect
on the fluorescence of 4. The fluorescence titration of 4 versus
11 was repeated in a less competitive solvent system, CHCl3/
MeOH (98:2), yielding a binding constant of 64000mꢀ1 (see
Received: June 6, 2002 [Z19486]
[1] Reviews: a) A. P. Davis, R. S. Wareham, Angew. Chem. 1999, 111, 3160;
Angew. Chem. Int. Ed. 1999, 38, 2978; b) T. D. James, K. R. A. S.
Sandanayake, S. Shinkai, Angew. Chem. 1996, 108, 2038; Angew. Chem.
Int. Ed. Engl. 1996, 35, 1911; recent examples: c) A. Sugasaki, K.
Sugiyasu, M. Ikeda, M. Takeuchi, S. Shinkai, J. Am. Chem. Soc. 2001,
123, 10239; d) A. S. Droz, U. Neidlein, S. Anderson, P. Seiler, F.
Diederich, Helv. Chim. Acta 2001, 84, 2243; e) O. Rusin, K. Lang, V.
Krµl, Chem. Eur. J. 2002, 8, 655; f) R. D. Hubbard, S. R. Horner, B. L.
Miller, J. Am. Chem. Soc. 2001, 123, 5810; g) W. Yang, H. He, D. G.
Drueckhammer, Angew. Chem. 2001, 113, 1764; Angew. Chem. Int. Ed.
2001, 40, 1714; h) S. Tamaru, M. Yamamoto, S. Shinkai, A. B. Khasanov,
T. W. Bell, Chem. Eur. J. 2001, 7, 5270; i) M. Mazik, H. Bandmann, W.
Sicking, Angew. Chem. 2000, 112, 562; Angew. Chem. Int. Ed. 2000, 39,
551; j) J. Bitta, S. Kubik, Org. Lett. 2001, 3, 2637; k) S. Arimori, M. L.
Bell, C. S. Oh, K. A. Frimat, T. D. James, Chem. Commun. 2001, 1836;
l) N. Sugimoto, D. Miyoshi, J. Zou, Chem. Commun. 2000, 2295; m) M.
Inouye, J. Chiba, H. Nakazumi, J. Org. Chem. 1999, 64, 8170.
[2] R. A. Dwek, T. D. Butters, Chem. Rev. 2002, 102, 283 (and succeeding
articles); I. Capila, R. J. Linhardt, Angew. Chem. 2002, 114, 426;
Angew. Chem. Int. Ed. 2002, 41, 390; T. Feizi, B. Mulloy, Curr. Opin.
Struct. Biol. 2001, 11, 585 (and succeeding articles); C. R. Bertozzi, L. L.
Kiessling, Science 2001, 291, 2357; J. K. Bashkin, Chem. Rev. 2000, 100,
4265 (and succeeding articles).
[3] a) A. Sugasaki, M. Ikeda, M. Takeuchi, S. Shinkai, Angew. Chem. 2000,
112, 3997; Angew. Chem. Int. Ed. 2000, 39, 3839; b) M. Ikeda, S.
Shinkai, A. Osuka, Chem. Commun. 2000, 1047; c) H. Kijima, M.
Takeuchi, S. Shinkai, Chem. Lett. 1998, 781; d) K. Sandanayake, K.
Nakashima, S. Shinkai, J. Chem. Soc. Chem. Commun. 1994, 1621;
e) A. S. Droz, F. Diederich, J. Chem. Soc. Perkin Trans. 1 2000, 4224;
f) J. Otsuki, K. Kobayashi, H. Toi, Y. Aoyama, Tetrahedron Lett. 1993,
34, 1945.
[4] a) A. P. Davis, R. S. Wareham, Angew. Chem. 1998, 110, 2397; Angew.
Chem. Int. Ed. 1998, 37, 2270; b) T. J. Ryan, G. Lecollinet, T. Velasco,
A. P. Davis, Proc. Natl. Acad. Sci. USA 2002, 99, 4863.
[5] Y. D. Ma, A. Takada, M. Sugiura, T. Fukuda, T. Miyamoto, J. Watanabe,
Bull. Chem. Soc. Jpn. 1994, 67, 346.
&
^
Figure 3. Experimental ( ) and calculated ( ) values for the fluorescence
binding study of 4 þ 11 in CHCl3/CH3OH (92:8). [4] ¼ 0.03 mm, [11] ¼
0.065 2.3 mm.
Angew. Chem. Int. Ed. 2002, 41, No. 21
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4121-4095 $ 20.00+.50/0
4095

COMMUNICATIONS
[6] Glycosides 11 13 were synthesized by using the method of Vill et al.; V.
Vill, T. Bocker, J. Thiem, F. Fischer, Liq. Cryst. 1989,6, 349.
[7] G. Pohlenz, M. Trimborn, H. Egge, Glycobiology 1994, 4, 625.
[8] H. Alpes, K. Allman, H. Plattner, J. Reichert, R. Riek, S. Schulz,
Biochim. Biophys. Acta 1986, 862, 294.
H3C
A
X
C
C
C
O
O
O
N
N
H
N
N
H
N
N
H
[9] This value is somewhat lower than that obtained by 1H NMR
spectroscopy in CDCl3/CD3OH (92:8). The discrepancy may be due
to a slight difference in experimental conditions; for example, different
amounts of adventitious water may have been present in the two
solvent systems.
Y
N
N
N
X
H
1
2
X = Y = H
X = CH3, Y = H
3 X = H, Y = CH3
4
5 X = H, A = CH3
6 X = CH3, A = CH3
7 X = H, A = Cl
8 X = CH3, A = Cl
Z
Dynamic Acylhydrazone Metal Ion Complex
Libraries: A Mixed Ligand Approach to
Increased Selectivity in Extraction**
CH2Q
9 X = H, Z = Cl
10 X = CH3, Z = Cl
11 X = H, Z = CH3
C
O
N
N
H
C
O
N
H
12 X = CH3, Z = CH3
13 X = H, Z = tBu
Seema Choudhary and Janet R. Morrow*
N
N
14 X = CH3, Z = tBu
15 X = H, Z = OH
16 X = CH3, Z = OH
17 X = H, Z = OCH3
18 X = CH3, Z = OCH3
19 X = H, Z = phenyl
20 X = CH3, Z = phenyl
X
N
Metal ion coordination chemistry is well suited to dynamic
X
3]
combinatorial chemistry approaches.[1 Diverse metal ion
21 X = H, Q = H
complex libraries can be created simply by mixing a labile
metal ion with different ligands.[4] If ligand exchange is rapid,
these metal ion complexes form the equilibrating components
of a dynamic library. The equilibrium between complexes may
be shifted upon addition of molecular targets. For example,
small molecule or biopolymer targets perturb the equilibrium
between metal ion complexes to increase the concentration of
22 X = CH3, Q = H
23 X = H, Q = phenyl
24 X = CH3, Q = phenyl
25 X = CH3, Q = (CH2)2CH3
Scheme 1. Acylhydrazone ligands used in this study.
Table 1. Extraction of Zn2þ or Cd2þ into chloroform by acylhydrazone
ligands.
7]
the metal ion complex that best binds the target.[5 We
Metal Ligand Extracted [%][a] Metal Ligand Extracted [%][a]
previously reported on dynamic combinatorial libraries of
metal ion Schiff-base complexes of limited diversity.[8] Our
selection protocol utilizes extraction of metal ion complex
libraries from aqueous into organic solvent. In this method,
the organic solvent is the ultimate target and complexes with
the greatest stability and solubility in organic solvent versus
aqueous solution are selected.[9] Here we present studies using
acylhydrazone ligand libraries and show that these libraries
lead to improvements in efficiency and selectivity of metal ion
extraction in comparison to single ligand systems.
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Cd2þ
Cd2þ
Cd2þ
Cd2þ
1
3
5
7
25
12
3.0
9.0
54
23
20
0
21
44
0
1.8
0
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Zn2þ
Cd2þ
Cd2þ
Cd2þ
Cd2þ
2
4
6
8
10
12
14
16
18
20
22
24
1
62
26
28
43
62
56
54
0
54
64
1.2
2.0
15
2.1
4.0
13
9
11
13
15
17
19
21
23
25
2
Acylhydrazone ligands (Scheme 1) were chosen to explore
the possibility of forming a double-orthogonal library[4] by
[10
exchange at metal ligand (M L) and C N bonds. 13] Initial
studies were carried out to establish the feasibility of using
these exchange processes. ZnII extracts into chloroform from a
buffered aqueous solution in the presence of two equivalents
of acylhydrazone ligand 1 (Table 1, Scheme 1). The predom-
inant complex that extracts into chloroform is the neutral
48
3.7
14
32
5
7
13
ꢀ
¼
6
8
14
[a] Extractions were carried out in 5.0 mm Mes buffer (10 mL), 0.100m
NaCl, and CHCl3 (1 mL) at 238C at pH 5.50 for ZnII (0.0500 mm) or
pH 6.50 for CdII (0.0500 mm) with 0.100 mm ligand.
complex [Zn(1ꢀ)2] as determined by comparison of H NMR
1
and mass spectral data to that of an authentic sample.
After two hours, the extent of extraction of ZnII from a
[*] Prof. J. R. Morrow, Dr. S. Choudhary
Department of Chemistry
water/chloroform mixture containing
0.0500 mm [Zn(1ꢀ)2] is identical within
University at Buffalo
R2
State University of New York
Amherst, NY 14260 (USA)
Fax : (þ 1)716-645-6963
H
N
experimental error to a second experi-
ment containing 0.0500 mm Zn(NO3)2
and 0.100 mm 1, confirming that our
system is at equilibrium. Addition of a
different acylhydrazone ligand (13) to
either of the solutions above followed
by stirring for 2 h increases the amount
O
N
N
N
N
Zn
H
E-mail: jmorrow@acsu.buffalo.edu
N
O
[**] This work was supported by the American Chemical Society
Petroleum Research Fund (34358-AC3) and by the Environment
and Society Institute of the University at Buffalo.
R1
[Zn(1-)2, R = C6H5 ]
Supporting information for this article is available on the WWW under
http://www.angewandte.org or from the author.
4096
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4121-4096 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 21