Adenine-based receptor for dicarboxylic acids

Article abstract of DOI:10.1016/j.tetlet.2007.07.110

The adenine-based fluorescent receptor 1 was designed and synthesized for the selective recognition of dicarboxylic acids in CH₃CN. The recognition takes place through the Hoogsteen binding site of adenine with concomitant PET quenching of the anthracene moiety. The carboxylic acid binding to 1 was investigated by ¹H NMR, X-ray, UV-vis, and fluorescence spectroscopic methods. The Hoogsteen (HG) cleft of receptor 1 is found to be selective for glutaric acid.

Full text of DOI:10.1016/j.tetlet.2007.07.110

Tetrahedron Letters 48 (2007) 7022–7026
Adenine-based receptor for dicarboxylic acids
a,
a
b
*
Kumaresh Ghosh, Tanushree Sen and Roland Fr o¨ hlich
a
Department of Chemistry, University of Kalyani, Kalyani 741 235, India
b
Organisch-Chemisches Institut, Universit a¨ t M u¨ nster, Corrensstrabe 40, D-48149, M u¨ nster, Germany
Received 30 May 2007; revised 12 July 2007; accepted 20 July 2007
Available online 27 July 2007
Abstract—The adenine-based fluorescent receptor 1 was designed and synthesized for the selective recognition of dicarboxylic acids
in CH CN. The recognition takes place through the Hoogsteen binding site of adenine with concomitant PET quenching of the
3
1
anthracene moiety. The carboxylic acid binding to 1 was investigated by H NMR, X-ray, UV–vis, and fluorescence spectroscopic
methods. The Hoogsteen (HG) cleft of receptor 1 is found to be selective for glutaric acid.
ꢀ
2007 Elsevier Ltd. All rights reserved.
There is currently great interest in the development of
supramolecular systems that have the ability to selec-
tively bind and sense the presence of dicarboxylic acids.
In relation to this, adenine is known to bind carboxylic
acids at both the Watson–Crick (WC) and Hoogsteen
binding (HG) sites (Fig. 1). Engel reported the exper-
1
10
Sensing and selective recognition of both mono- and
dicarboxylic acids are of considerable attention because
of their important roles in biology. Over the past few
imental differentiation between WC and HG binding
1
1
sites using 6-N-methyl-9-N-ethyladenine. From the
analysis of data from a direct titration between a car-
boxylic acid and 9-butyladenine, Maitra et al. showed
that aromatic carboxylic acids preferred the HG site
for binding 9-butyladenine whereas aliphatic acids pre-
2
years, fluorescent chemosensors for the detection of car-
3
boxylic acids and carboxylates have been developed. In
this respect, one of the recent approaches for the design
of fluorescent signaling systems is to exploit photo-
induced electron transfer (PET) in fluorophore-spacer-
receptor systems where the PET process is suppressed
or enhanced by the introduction of a substrate into the
receptor, exhibiting a fluorescent signal. To date, several
receptors containing different functional groups for
selective binding of both mono- and dicarboxylic acids
1
2
13
ferred the WC site in CDCl . Rebek and Zimmer-
3
1
4
man have exploited these binding features of adenine
in the construction of synthetic receptors for adenine
recognition. The use of adenine in the construction of
a synthetic sensor for carboxylic acids is unknown in
the literature. Our previous report on anthracene-based
1
5
PET sensors for both mono- and dicarboxylic acids
encouraged us to work on new designs where carboxylic
cid recognition takes place at charge neutral recognition
sites with concomitant changes in the photophysical
properties of a fluorophore by modulation of a photo-
induced electron transfer (PET) mechanism. In continu-
4
have been reported in the literature. In this regard, pyr-
1
d,e,5
6
7
idine amine,
oxazole amine, cyclic amide, phos-
8
9
phonamide, the phenolic –OH of a cyclotetramer
etc., are well-known hydrogen bonding motifs for car-
boxylic acids and have been extensively used in develop-
ing carboxylic acid receptors of various architectures.
The search for new hydrogen bonding motifs and their
successful installation onto a fluorophore probe in order
to develop sensors, for selective recognition of carbox-
ylic acids, has been of continuous interest to the supra-
molecular chemistry community.
Hoogsteen binding site
Watson-Crick binding site
R
O
O
R
H
N
H
N
N
HO
OH
N
N
Keywords: Anthracene; Adenine; Hoogsteen site; PET sensor; Glutaric
acid.
*
Corresponding author. Tel.: +91 33 25828750; fax: +91 33
5828282; e-mail: ghosh_k2003@yahoo.co.in
2
Figure 1. Hydrogen bonding sites of adenine with carboxylic acids.
0040-4039/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.07.110

K. Ghosh et al. / Tetrahedron Letters 48 (2007) 7022–7026
7023
ation, we herein present our observations on new
anthracene-based tweezer-like receptors 1 and 2 with flat
hydrophobic surfaces bearing 9-butyl adenine as the rec-
ognition determinant, for complexation of both mono-
and dicarboxylic acids of different chain lengths. To
the best of our knowledge, these are the first examples
of anthracene appended adenine-based charge neutral
receptors that show ideal PET behavior for carboxylic
acids involving the HG site as the principal binding do-
main. Receptor 1 can be described as a ‘receptor-spacer-
fluorophore-spacer-receptor’ second generation PET
sensor.
Figure 2. Energy minimized structure of 1.
of C(2)–H results from the binding-promoted p-polari-
zation which induces a partial negative charge on the
adjacent N-1. The downfield shift of the C(8)–H proton
and upfield shift of the C(2)–H proton in 1 during com-
R
N
N
R
R
N
N
N
N
HG- cleft
H
N
H
N
HG- cleft
H
N
N
N
N
N
N
N
1
R = (CH ) CH
2 3 3
2
R = (CH ) CH
2 3 3
Receptor 1 was synthesized according to Scheme 1 and
was obtained in low yield. Adenine was initially alkyl-
ated at the 9-position using butyl bromide in the pres-
ence of K CO in dry DMF to give 3 in 85% yield,
plexation with the diacids in CDCl , suggests a compet-
3
itive mode of complexation between the WC and HG
sites. To substantiate the site selectivity in complexation,
we isolated a single crystal of a 1:1 complex of 1 with
2
3
1
7a
subsequent coupling of which with 9,10-bis(chloro-
methyl)anthracene or 9-chloromethylanthracene affor-
ded 1 and 2, respectively. This method gave 1 in 9%
yield and 2 in 60% yield. All the compounds were char-
adipic acid (Fig. 3a). The disposition of the HG clefts
of 1 in opposite directions led to 1:1 dynamic supramo-
lecular hetero association with adipic acid. The N3 atom
of each adenine is coordinated to methanol, which was
used for crytallization. In the molecular assembly, the
carboxyl moieties are selectively held in the HG clefts
of 1 without showing any proton transfer. This observa-
tion intimates that the binding of the aliphatic acids in
the present case occurs preferentially in the HG site over
the WC site. The HG site prefers to bind aromatic acid.
Figure 3b, in this connection, further substantiated the
1
13
acterized by H NMR, C NMR, UV–vis, and mass
analyses.
A molecular modeling study16 on 1 (Fig. 2; Emin
=
˚
˚
8
6.84 kcal/mol, dNH–NH = 6.48 A, dN7–N7 = 10.85 A)
shows the orientation of adenine with amine hydrogens
into the HG sites around anthracene. The syn form out
of various conformations is capable of forming a stable
hydrogen bonded complex with dimensionally fitted glu-
taric acid (see Supplementary data).
1
7b
binding proposition. The involvement of the WC site
in complexation, although negligible, cannot be ruled
out in solution. Benzoic acid in receptor 2 is complexed
into the HG site over the WC site through more hydro-
gen bonds.
To obtain an insight into the binding properties of neu-
1
tral receptor 1, we investigated the change in the H
NMR spectrum of the receptor upon addition of dicar-
With the propensity of the HG site as the preferred
binding site for carboxylic acids established, the selectiv-
ity and sensitivity of receptor 1 toward a series of ali-
phatic dicarboxylic acids, as mentioned in Table 1,
was evaluated by observing the change in their fluores-
boxylic acids in CDCl . The –NH– and C(8)–H protons
3
in 1 moved downfield (Dd = 0.32–0.8 ppm for NH and
0
.01–0.03 ppm for C(8)–H) whereas C(2)–H underwent
an upfield (Dd = 0.01–0.02 ppm) shift upon complexa-
tion with different aliphatic dicarboxylic acids (malonic,
succinic, glutaric, adipic, and pimelic). The upfield shift
cence emission in CH CN. When a solution of 1 (c =
3
ꢀ
6
5.480 · 10 M) in CH CN was excited at 375 nm, the
3
NaH, dry THF
NH2
NH2
2
N
9-chloromethylanthracene
N
K
2
CO
3
, dry DMF
N
N
N
N
H
n-butyl bromide, rt
N
N
R
3
R = (CH ) CH
NaH, dry THF
2
3
3
1
9
,10-bis(chloromethyl)anthracene
Scheme 1. Synthesis of sensors 1and 2.

7024
K. Ghosh et al. / Tetrahedron Letters 48 (2007) 7022–7026
a
b
O22B
C21B
C29
C28
C23
C27
C26
C23B
N11B
N9B
C22B
C23B*
N22
O21B
N20
C21
C5
O33B
C6
C19
N6B
C4
C3
N2B
N24
C7
C8
C32B
N4B
C25
C1B
N18
O4a
C2
O2
C9
C17
C1A
N4A
C32A
C35
C33
O3a
C40a
C10
C11
C1
C5A
C30
N16
C15
C13
C14
C14
C3A
C8A
C36
C41a
N6A
O1
C46a
N2A
C34
C13
O31A
C1
C42a
C12
C31 C3
C4
C45a
C7A
C2
C12
C11
O21A
C44a C43a
N9A
C13A
C32
C5
C23A*
C22A
N11A
C10
C23A
C21A
C10A
C9
C7
C12A
C14A
C15A
O22A
C6
C8
Figure 3. Single crystal X-ray structure of (a) 1 with adipic acid and (b) 2 with benzoic acid.
Table 1. Association constants of 1 with guest acids
1.5
1.4
1.3
1.2
a
ꢀ1 b
ꢀ1 c
Diacids
K
a
(M
)
K
a
(M
)
(b)
(d)
4
2
Malonic
Succinic
Glutaric
Adipic
1.35 · 104
1.31 · 104
3.49 · 10
4.50 · 10
6.52 · 10
8.27 · 10
9.56 · 10
3.25 · 10
(a)
(e)
2
3
2
3
4
1.03 · 104
(
c)
Pimelic
1.49 · 10
a
b
c
All the diacids were dissolved in CH
Determined by the UV method in CH
Determined by the fluorescence method in CH
3
CN containing 0.5% DMSO.
a. malonic
b. succinic
c. glutaric
d. adipic
3
CN.
3
CN.
1
1
.1
.0
e. pimelic
characteristic emissions at 402, 424, and 448 nm for
monomeric anthracene were quenched or ‘switched off’
to different extents in the presence of dicarboxylic acids
without showing any other noticeable changes. Figure 4
shows the changes in fluorescence of 1 upon gradual in-
crease of the concentration of glutaric acid. The extent
of quenching (24–31%) with different diacids of different
chain lengths, though small (Fig. 5), is of great signifi-
cance. The quenching is attributed to the hydrogen
bonding interactions of the adenine moieties with the
carboxylic acid groups of the diacids into the cleft of 1
either in mode A or mode B as shown in Figure 6.
0
10
20
30
40
50
60
70
80
90
-6
[
Guest] x 10
M
0
Figure 5. Stern–Volmer plot of I /I versus concn of carboxylic acid
(M).
The diminished fluorescence emission intensity of 1 is
due to a thermodynamically favored primary PET
between the adenine binding site and the excited chro-
*
mophore ( Anth). Complexation of adenine will prevent
the primary PET and fluorescence of 1, which in princi-
ple, will be switched ‘On’ on the basis of the normal
logic of a fluorescent PET sensor. At the same time,
the opposite situation of ‘On–Off’ switching can be
1
000
1
0
0
.0
.5
.0
*
arranged by a secondary PET process ( Anth to adenine
8
6
4
2
00
00
00
00
0
site). The present example undoubtedly represents a
combination of these two opposite situations resulting
in the quenching of fluorescence leading to an ‘Off’
mode. The same was true for 2 also.
A concurrent study of the absorption spectra of 1 in the
presence of the same carboxylic acids show small
changes in the anthracene absorption indicating typical
PET behavior. The Benesi-Hildebrand analysis of the
fluorescence changes observed at 424 nm gave a 1:1
stoichiometry for the complex formed between 1 and
glutaric acid with an association constant of 8.27 ·
250
300
350
400
450
Wavelength (nm)
400
450
500
550
Wavelength (nm)
3
ꢀ1 18
ꢀ
6
1
0 M
.
A similar analysis using absorbance at
Figure 4. Fluorescence spectra of 1 (c = 5.48 · 10 M) in CH
3
CN
ꢀ6
and the change in the UV–vis spectra of 1 (c = 5.48 · 10 M) (inset)
375 nm also gave 1:1 stoichiometry for the complexes
upon addition of glutaric acid.
of all the acids with 1. The association constants are

K. Ghosh et al. / Tetrahedron Letters 48 (2007) 7022–7026
7025
O
O
O
H
H
H
R
R
N
N
O
N
N
O
O
R
N
N
H
H
N
N
N
H
N
N
N
N
N
N
N
N
H
N
N
O
N
R
B
A
R = (CH ) CH
3
2
3
OH
Figure 6. Possible hydrogen bonded complex modes of dicarboxylic acids in the HG cleft of 1.
5
4
3
the cleft in 2 has a marginal preference for an aromatic
acid.
In conclusion, we have shown that the selective recogni-
tion of a dicarboxylic acid by the HG cleft of an adenine-
based receptor is possible. Further optimization of the
HG binding pocket using a combinatorial approach
can now be used to improve both the affinity and selectiv-
ity. Such work is in progress in our laboratory.
Acknowledgments
0
1
2
3
4
5
n
We thank DST (SR/FTP/CS-18/2004) and CSIR, Gov-
ernment of India, for financial support. T.S. thanks
CSIR, Government of India, for a fellowship.
Figure 7. Graphical representation of the association constants logK
of the complexes formed by receptor 1 with HO C–(CH –CO
against the number (n) of atoms connecting the two carboxylic acids
solid cubes: UV titration; solid circles: fluorescence titration).
2
2
)
n
2
H
(
Supplementary data
accumulated in Table 1. Figure 7 shows the calculated
stability constants as a function of the aliphatic chain
lengths connecting the two carboxyl ends for a series
of linear aliphatic dicarboxylic acids HO C–(CH ) –
Job plot of 1 with glutaric, adipic, and pimelic acids,
energy minimized structure of the complex of 1 with
glutaric acid, experimental procedures for the syntheses
of 1 and 2, H NMR, C NMR, and mass spectra of
receptors 1 and 2 are available. Supplementary data
associated with this article can be found, in the online
version, at doi:10.1016/j.tetlet.2007.07.110.
2
2 n
1
13
CO H, with n ranging from 1 to 5 and clearly predicts
2
a preference for glutaric acid over both the short and
long chain diacids. The preference is greater in the
excited state. This is due to the formation of a stable
1
:1 hydrogen bonded complex. The stoichiometry of
the complexes was also confirmed by a Job plot analysis
see Supplementary data). In addition, a plot of emis-
References and notes
(
sion changes in fluorescence at 424 nm versus [G]/[H]
for all the diacids as mentioned in Table 1, gave almost
linear instead of sigmoidal responses. We believe this is
probably due to the interaction of the individual adenine
unit with the carboxylic acid motif in mode B instead of
A as shown in Figure 6. The possibility of formation of
hydrogen bonded complexes A (Fig. 6) with dimension-
ally fitted dicarboxylic acids in solution also cannot be
ruled out.
1
. (a) Konig, B.; Moller, O.; Bubenitschek, P. J. Org. Chem.
995, 60, 4291–4293; (b) Zhao, J.; James, F. D. Chem.
Commun. 2005, 1889; (c) Yang, N.; Yan, J.; Fang, H.;
Wung, B. Chem. Commun. 2003, 792–793; (d) Etter, M. C.;
Adsmond, D. A. J. Chem. Soc., Chem. Commun. 1990,
1
5
89–592; (e) Karle, I. L.; Ranganathan, D.; Haridas, V. J.
Am. Chem. Soc. 1997, 119, 2777–2783; (f) Goodman, M.
S.; Hamilton, A. D.; Weiss, J. J. Am. Chem. Soc. 1995,
117, 8447–8455.
2
. (a) Dugas, H. Bioorganic Chemistry; Springer: New York,
1996; (b) Stryer, L. Biochemistry; W. H. Freeman,
Company: New York, 1998.
ꢀ
5
The fluorescence sensitivity of 2 (c = 1.26 · 10 M)
with the monocarboxylic acids was also observed and
3
. (a) Jadhav, V. D.; Schmidtchen, F. P. Org. Lett. 2006, 8,
2
exhibited the same ‘On–Off’ switching behavior as 1
2
329–2332; (b) Fan, E.; van Arman, S. A.; Kincaid, S.;
in CH CN. It was found that receptor 2 binds benzoic
3
Hamilton, A. D. J. Am. Chem. Soc. 1993, 115, 369–370; (c)
Mei, M.; Wu, S. New J. Chem. 2001, 25, 471–475; (d) Liu,
S.-Y.; Fang, L.; He, Y.-B.; Chan, W.-H.; Yeung, K.-T.;
Chen, Y.-K.; Yang, R.-H. Org. Lett. 2005, 7, 5825–5828;
(e) Gunnlaugssson, T.; Davis, A. P.; O’Brien, J.-E.; Glynn,
and acetic acids with association constants of 4.29 ·
3
ꢀ1
3
ꢀ1
1
0 M and 2.71 · 10 M , respectively. The values
are slightly less in magnitude than those obtained
for dicarboxylic acids with 1 and also indicate that

7026
K. Ghosh et al. / Tetrahedron Letters 48 (2007) 7022–7026
M. Org. Lett. 2002, 4, 2449–2452; (f) Kim, S. K.; Kang,
16. Energy minimization was carried out using MMX (PC
Model Serena Software 1993). Molecular modeling was
performed using standard constants, and the dielectric
constant was maintained at 1.5.
B.-G.; Koh, H.-S.; Yoon, Y. J.; Jung, S. J.; Jeong, B.; Lee,
K.-D.; Yoon, J. Org. Lett. 2004, 6, 4655–4658; (g) Kim,
D.-S.; Miyaji, H.; Chang, B. Y.; Park, S.-M.; Ahn, K.-H.
Chem. Commun 2006, 3314–3316; (h) Liu, S.-Y.; He,
Y.-B.; Nu, J.-L.; Wei, L.-H.; Qin, H.-J.; Meng, L.-Z.; Hu,
L. Org. Biomol. Chem. 2004, 2, 1582–1586.
17. (a) Crystal data of the 1:1 complex of adipic acid with 1:
C
34
H
36
N
10Æ2/2C
6 10 4 3 1
H O Æ2/2CH OH, monoclinic, P2 /n
˚
(No. 14), a = 14.321(1), b = 14.814(1), c = 18.512(1) A,
b = 98.80(1) , V = 3881.1(4) A , T = 198(2) K, Z = 4,
o
˚
3
4
. Fitzmaurice, R. J.; Kyne, G. M.; Douheret, D.; Kilburn, J.
D. J. Chem. Soc., Perkin Trans 1 2002, 841–864.
ꢀ3
qcalcd = 1.306 mg m , 25,333 total reflections of which
5
. (a) Goswami, S.; Ghosh, K.; Dasgupta, S. J. Org. Chem.
6811 were independent, 3690 observed [I > 2r (I)]. Struc-
2
000, 65, 1907–1914, and references cited therein; (b)
Halter, I. P.; Smith, T. J.; Weiss, J. J. Org. Chem. 1997, 62,
186–2192; (c) Moriuchi, T.; Yoshida, K.; Hirao, T. Org.
ture solution and refinement with SHELXS-97 and SHELXL-
2
97, final refinement against F with 601 parameters, R
1
2
2
[I > 2r (I)] = 0.114, wR = 0.268.; (b) Crystal data of the
complex of benzoic acid with 2: C H N ÆC H O Æ
Lett. 2003, 35, 4285–4288.
. (a) Hernandez, J. V.; Oliva, A. I.; Simon, L.; Muniz, F.
M.; Mateos, A. A.; Moran, J. R. J. Org. Chem. 2003, 68,
2
4
23
5
7
6
2
ꢀ
˚
6
1/2C
7 6 2
H O , triclinic, P1 (No. 2), a = 10.492(1), b =
11.247(1), c = 13.288(1) A, a = 76.45(1), b = 82.04(1),
o
˚
3
7513–7516; (b) Oliva, A. I.; Simon, L.; Muniz, F. M.;
c = 74.42(1) , V = 1463.6(2) A , T = 223(2) K, Z = 2,
3
ꢀ
Sanz, F.; Moran, J. R. Org. Lett. 2004, 6, 1155–1157.
. Moore, G.; Papomicael, C.; Levacher, V.; Bourguignon,
J.; Dupas, G. Tetrahedron 2004, 60, 4197–4204.
. Mussons, L.; Raposo, C.; Anaya, J.; Grando, M.; Moran,
J. R.; Caballero, C. J. Chem. Soc., Perkin Trans. 1 1992,
qcalcd = 1.281 mg m , 20,123 total reflections of which
5151 were independent, 4487 observed [I > 2r (I)]. Struc-
ture solution and refinement with SHELXS-97 and SHELXL-
7
8
2
97, final refinement against F with 418 parameters, R1
2
[I > 2r (I)] = 0.069, wR = 0.182. CCDC Nos. 648928 and
2
2, 3125–3127.
. Tanaka, Y.; Kato, Y.; Aoyama, Y. J. Am. Chem. Soc.
990, 112, 2807–2808.
0. Adrian, J. C.; Wilcox, C. S. J. Am. Chem. Soc. 1989, 111,
055–8057.
1. Engel, J. D.; von Hippel, P. H. Biochemistry 1974, 13,
143–4158.
2. Rao, P.; Ghosh, S.; Maitra, U. J. Phys. Chem. B 1999,
03, 4528–4533.
648929.
9
18. Binding constants were determined using the expression
ꢀ
1
ꢀ1
1
A /(A ꢀ A) = [e /(e ꢀ e )] (K ÆC þ 1Þ where e and
0
0
M
M
C
a
g
M
1
1
1
e
C
are the molar extinction coefficients of receptor 1 and
hydrogen bonding complex, respectively, at a selected
wavelength. A denotes the absorbance of free receptor 1
at that specific wavelength and C is the concentration of
8
0
4
g
carboxylic acid guest. The measured absorbance [A
(A
0
/
1
0
ꢀ A)] as a function of the inverse of carboxylic acid
1
1
3. Rebek, J., Jr. Acc. Chem. Res. 1990, 23, 399–404.
4. Zimmerman, S. C.; Wu, W. J. Am. Chem. Soc. 1989, 111,
guest concentrations fits a linear relationship, indicating
1:1 stoichiometry of the receptor/carboxylic acid complex.
The ratio for the intercept versus slope determines the
8
054–8055.
1
5. (a) Ghosh, K.; Masanta, G. Tetrahedron Lett. 2006, 47,
binding constant (K
a
) and the binding constant values
2
365–2369; (b) Ghosh, K.; Adhikari, S. Tetrahedron Lett.
using fluorescence intensity were calculated using the
ꢀ1
ꢀ1
2006, 47, 3577–3581; (c) Ghosh, K.; Masanta, G. Chem.
expression F
0
/(F
0
ꢀ F) = [/
M
/(/
M
ꢀ /
C
)] ðKa ÆC þ 1).
g
Lett. 2006, 34, 414–415; (d) Ghosh, K.; Masanta, G.
Supramol. Chem. 2005, 17, 331–334; (e) Ghosh, K.;
Masanta, G. Tetrahedron Lett. 2006, 47, 9233–9237.
Chou, P. T.; Wu, G. R.; Wei, C. Y.; Cheng, C. C.; Chang,
C. P.; Hung, F. T. J. Phys. Chem. B 2000, 104, 7818–
7829.